This commit is contained in:
2018-10-27 01:56:42 -04:00
commit 78e38cbca6
352 changed files with 106022 additions and 0 deletions

7
.idea/misc.xml generated Normal file
View File

@@ -0,0 +1,7 @@
<?xml version="1.0" encoding="UTF-8"?>
<project version="4">
<component name="JavaScriptSettings">
<option name="languageLevel" value="ES6" />
</component>
<component name="ProjectRootManager" version="2" project-jdk-name="Python 3.6 (thue)" project-jdk-type="Python SDK" />
</project>

8
.idea/modules.xml generated Normal file
View File

@@ -0,0 +1,8 @@
<?xml version="1.0" encoding="UTF-8"?>
<project version="4">
<component name="ProjectModuleManager">
<modules>
<module fileurl="file://$PROJECT_DIR$/.idea/thue.iml" filepath="$PROJECT_DIR$/.idea/thue.iml" />
</modules>
</component>
</project>

13
.idea/thue.iml generated Normal file
View File

@@ -0,0 +1,13 @@
<?xml version="1.0" encoding="UTF-8"?>
<module type="PYTHON_MODULE" version="4">
<component name="NewModuleRootManager">
<content url="file://$MODULE_DIR$">
<excludeFolder url="file://$MODULE_DIR$/venv" />
</content>
<orderEntry type="inheritedJdk" />
<orderEntry type="sourceFolder" forTests="false" />
</component>
<component name="TestRunnerService">
<option name="PROJECT_TEST_RUNNER" value="Unittests" />
</component>
</module>

1
requirements.txt Normal file
View File

@@ -0,0 +1 @@
lark-parser

49
rethue.py Normal file
View File

@@ -0,0 +1,49 @@
from lark import Lark
grammar = r'''
suite: prod*
prod: _named_prod
_named_prod: _direct_prod | alias
_direct_prod: _compound_prod | _rule_prod | _literal_prod | _abstract_prod
_abstract_prod: output
_compound_prod: cont | sing
_rule_prod: full | part
_literal_prod: lit | ref
cont: "{" suite "}" // continual production
sing: "[" suite "]" // singular production
alias: ref "=" prod // production alias
full: _direct_prod "=>" prod // full application
part: ctx "::=" prod // partial application
output: "~" prod
lit: STRING
ref: CNAME
ctx: REGEX
CNAME: /[a-z_][a-z0-9_]*/i
STRING: /(r)?(['"])(.*?)(?<!\\)\2/
REGEX: /(r)?([\/])(.*?)(?<!\\)\2(i)?/
%import common.WS
%ignore WS
'''
def main():
lrk = Lark(grammar, start='suite')
with open('sample.ret') as f:
src = f.read()
tree = lrk.parse(src)
print(tree.pretty())
if __name__ == '__main__':
main()

22
sample.ret Normal file
View File

@@ -0,0 +1,22 @@
UNARY = ';1;\0' => {
~ 'Conversion step: \0'
/(.*);(.*);(.*)0/ ::= '\1;\2\2;\3'
/(.*);(.*);(.*)1/ ::= '\1\2;\2\2;\3'
/(.*);(.*);$/ ::= '\1'
}
FIZZ = UNARY => /(.{3})*/ ::= ~ 'fizz'
BUZZ = UNARY => /(.{5})*/ ::= ~ 'buzz'
FIZZBUZZ = UNARY => /(.{15})*/ ::= ~ 'fizzbuzz'
FIZZ_BUZZ = [
FIZZBUZZ
FIZZ
BUZZ
~ '\0'
]
':::' => FIZZ_BUZZ

76
venv/bin/activate Normal file
View File

@@ -0,0 +1,76 @@
# This file must be used with "source bin/activate" *from bash*
# you cannot run it directly
deactivate () {
# reset old environment variables
if [ -n "${_OLD_VIRTUAL_PATH:-}" ] ; then
PATH="${_OLD_VIRTUAL_PATH:-}"
export PATH
unset _OLD_VIRTUAL_PATH
fi
if [ -n "${_OLD_VIRTUAL_PYTHONHOME:-}" ] ; then
PYTHONHOME="${_OLD_VIRTUAL_PYTHONHOME:-}"
export PYTHONHOME
unset _OLD_VIRTUAL_PYTHONHOME
fi
# This should detect bash and zsh, which have a hash command that must
# be called to get it to forget past commands. Without forgetting
# past commands the $PATH changes we made may not be respected
if [ -n "${BASH:-}" -o -n "${ZSH_VERSION:-}" ] ; then
hash -r
fi
if [ -n "${_OLD_VIRTUAL_PS1:-}" ] ; then
PS1="${_OLD_VIRTUAL_PS1:-}"
export PS1
unset _OLD_VIRTUAL_PS1
fi
unset VIRTUAL_ENV
if [ ! "$1" = "nondestructive" ] ; then
# Self destruct!
unset -f deactivate
fi
}
# unset irrelevant variables
deactivate nondestructive
VIRTUAL_ENV="/home/allem/PycharmProjects/thue/venv"
export VIRTUAL_ENV
_OLD_VIRTUAL_PATH="$PATH"
PATH="$VIRTUAL_ENV/bin:$PATH"
export PATH
# unset PYTHONHOME if set
# this will fail if PYTHONHOME is set to the empty string (which is bad anyway)
# could use `if (set -u; : $PYTHONHOME) ;` in bash
if [ -n "${PYTHONHOME:-}" ] ; then
_OLD_VIRTUAL_PYTHONHOME="${PYTHONHOME:-}"
unset PYTHONHOME
fi
if [ -z "${VIRTUAL_ENV_DISABLE_PROMPT:-}" ] ; then
_OLD_VIRTUAL_PS1="${PS1:-}"
if [ "x(venv) " != x ] ; then
PS1="(venv) ${PS1:-}"
else
if [ "`basename \"$VIRTUAL_ENV\"`" = "__" ] ; then
# special case for Aspen magic directories
# see http://www.zetadev.com/software/aspen/
PS1="[`basename \`dirname \"$VIRTUAL_ENV\"\``] $PS1"
else
PS1="(`basename \"$VIRTUAL_ENV\"`)$PS1"
fi
fi
export PS1
fi
# This should detect bash and zsh, which have a hash command that must
# be called to get it to forget past commands. Without forgetting
# past commands the $PATH changes we made may not be respected
if [ -n "${BASH:-}" -o -n "${ZSH_VERSION:-}" ] ; then
hash -r
fi

37
venv/bin/activate.csh Normal file
View File

@@ -0,0 +1,37 @@
# This file must be used with "source bin/activate.csh" *from csh*.
# You cannot run it directly.
# Created by Davide Di Blasi <davidedb@gmail.com>.
# Ported to Python 3.3 venv by Andrew Svetlov <andrew.svetlov@gmail.com>
alias deactivate 'test $?_OLD_VIRTUAL_PATH != 0 && setenv PATH "$_OLD_VIRTUAL_PATH" && unset _OLD_VIRTUAL_PATH; rehash; test $?_OLD_VIRTUAL_PROMPT != 0 && set prompt="$_OLD_VIRTUAL_PROMPT" && unset _OLD_VIRTUAL_PROMPT; unsetenv VIRTUAL_ENV; test "\!:*" != "nondestructive" && unalias deactivate'
# Unset irrelevant variables.
deactivate nondestructive
setenv VIRTUAL_ENV "/home/allem/PycharmProjects/thue/venv"
set _OLD_VIRTUAL_PATH="$PATH"
setenv PATH "$VIRTUAL_ENV/bin:$PATH"
set _OLD_VIRTUAL_PROMPT="$prompt"
if (! "$?VIRTUAL_ENV_DISABLE_PROMPT") then
if ("venv" != "") then
set env_name = "venv"
else
if (`basename "VIRTUAL_ENV"` == "__") then
# special case for Aspen magic directories
# see http://www.zetadev.com/software/aspen/
set env_name = `basename \`dirname "$VIRTUAL_ENV"\``
else
set env_name = `basename "$VIRTUAL_ENV"`
endif
endif
set prompt = "[$env_name] $prompt"
unset env_name
endif
alias pydoc python -m pydoc
rehash

75
venv/bin/activate.fish Normal file
View File

@@ -0,0 +1,75 @@
# This file must be used with ". bin/activate.fish" *from fish* (http://fishshell.org)
# you cannot run it directly
function deactivate -d "Exit virtualenv and return to normal shell environment"
# reset old environment variables
if test -n "$_OLD_VIRTUAL_PATH"
set -gx PATH $_OLD_VIRTUAL_PATH
set -e _OLD_VIRTUAL_PATH
end
if test -n "$_OLD_VIRTUAL_PYTHONHOME"
set -gx PYTHONHOME $_OLD_VIRTUAL_PYTHONHOME
set -e _OLD_VIRTUAL_PYTHONHOME
end
if test -n "$_OLD_FISH_PROMPT_OVERRIDE"
functions -e fish_prompt
set -e _OLD_FISH_PROMPT_OVERRIDE
functions -c _old_fish_prompt fish_prompt
functions -e _old_fish_prompt
end
set -e VIRTUAL_ENV
if test "$argv[1]" != "nondestructive"
# Self destruct!
functions -e deactivate
end
end
# unset irrelevant variables
deactivate nondestructive
set -gx VIRTUAL_ENV "/home/allem/PycharmProjects/thue/venv"
set -gx _OLD_VIRTUAL_PATH $PATH
set -gx PATH "$VIRTUAL_ENV/bin" $PATH
# unset PYTHONHOME if set
if set -q PYTHONHOME
set -gx _OLD_VIRTUAL_PYTHONHOME $PYTHONHOME
set -e PYTHONHOME
end
if test -z "$VIRTUAL_ENV_DISABLE_PROMPT"
# fish uses a function instead of an env var to generate the prompt.
# save the current fish_prompt function as the function _old_fish_prompt
functions -c fish_prompt _old_fish_prompt
# with the original prompt function renamed, we can override with our own.
function fish_prompt
# Save the return status of the last command
set -l old_status $status
# Prompt override?
if test -n "(venv) "
printf "%s%s" "(venv) " (set_color normal)
else
# ...Otherwise, prepend env
set -l _checkbase (basename "$VIRTUAL_ENV")
if test $_checkbase = "__"
# special case for Aspen magic directories
# see http://www.zetadev.com/software/aspen/
printf "%s[%s]%s " (set_color -b blue white) (basename (dirname "$VIRTUAL_ENV")) (set_color normal)
else
printf "%s(%s)%s" (set_color -b blue white) (basename "$VIRTUAL_ENV") (set_color normal)
end
end
# Restore the return status of the previous command.
echo "exit $old_status" | .
_old_fish_prompt
end
set -gx _OLD_FISH_PROMPT_OVERRIDE "$VIRTUAL_ENV"
end

12
venv/bin/easy_install Executable file
View File

@@ -0,0 +1,12 @@
#!/home/allem/PycharmProjects/thue/venv/bin/python
# EASY-INSTALL-ENTRY-SCRIPT: 'setuptools==39.1.0','console_scripts','easy_install'
__requires__ = 'setuptools==39.1.0'
import re
import sys
from pkg_resources import load_entry_point
if __name__ == '__main__':
sys.argv[0] = re.sub(r'(-script\.pyw?|\.exe)?$', '', sys.argv[0])
sys.exit(
load_entry_point('setuptools==39.1.0', 'console_scripts', 'easy_install')()
)

12
venv/bin/easy_install-3.6 Executable file
View File

@@ -0,0 +1,12 @@
#!/home/allem/PycharmProjects/thue/venv/bin/python
# EASY-INSTALL-ENTRY-SCRIPT: 'setuptools==39.1.0','console_scripts','easy_install-3.6'
__requires__ = 'setuptools==39.1.0'
import re
import sys
from pkg_resources import load_entry_point
if __name__ == '__main__':
sys.argv[0] = re.sub(r'(-script\.pyw?|\.exe)?$', '', sys.argv[0])
sys.exit(
load_entry_point('setuptools==39.1.0', 'console_scripts', 'easy_install-3.6')()
)

12
venv/bin/pip Executable file
View File

@@ -0,0 +1,12 @@
#!/home/allem/PycharmProjects/thue/venv/bin/python
# EASY-INSTALL-ENTRY-SCRIPT: 'pip==10.0.1','console_scripts','pip'
__requires__ = 'pip==10.0.1'
import re
import sys
from pkg_resources import load_entry_point
if __name__ == '__main__':
sys.argv[0] = re.sub(r'(-script\.pyw?|\.exe)?$', '', sys.argv[0])
sys.exit(
load_entry_point('pip==10.0.1', 'console_scripts', 'pip')()
)

12
venv/bin/pip3 Executable file
View File

@@ -0,0 +1,12 @@
#!/home/allem/PycharmProjects/thue/venv/bin/python
# EASY-INSTALL-ENTRY-SCRIPT: 'pip==10.0.1','console_scripts','pip3'
__requires__ = 'pip==10.0.1'
import re
import sys
from pkg_resources import load_entry_point
if __name__ == '__main__':
sys.argv[0] = re.sub(r'(-script\.pyw?|\.exe)?$', '', sys.argv[0])
sys.exit(
load_entry_point('pip==10.0.1', 'console_scripts', 'pip3')()
)

12
venv/bin/pip3.6 Executable file
View File

@@ -0,0 +1,12 @@
#!/home/allem/PycharmProjects/thue/venv/bin/python
# EASY-INSTALL-ENTRY-SCRIPT: 'pip==10.0.1','console_scripts','pip3.6'
__requires__ = 'pip==10.0.1'
import re
import sys
from pkg_resources import load_entry_point
if __name__ == '__main__':
sys.argv[0] = re.sub(r'(-script\.pyw?|\.exe)?$', '', sys.argv[0])
sys.exit(
load_entry_point('pip==10.0.1', 'console_scripts', 'pip3.6')()
)

BIN
venv/bin/python Executable file

Binary file not shown.

BIN
venv/bin/python3 Executable file

Binary file not shown.

BIN
venv/bin/python3.6 Executable file

Binary file not shown.

View File

@@ -0,0 +1,2 @@
./setuptools-39.1.0-py3.6.egg
./pip-10.0.1-py3.6.egg

View File

@@ -0,0 +1,8 @@
from .tree import Tree
from .visitors import Transformer, Visitor, v_args, Discard
from .visitors import InlineTransformer, inline_args # XXX Deprecated
from .exceptions import ParseError, LexError, GrammarError, UnexpectedToken, UnexpectedInput, UnexpectedCharacters
from .lexer import Token
from .lark import Lark
__version__ = "0.6.5"

View File

@@ -0,0 +1,16 @@
class LexerConf:
def __init__(self, tokens, ignore=(), postlex=None, callbacks=None):
self.tokens = tokens
self.ignore = ignore
self.postlex = postlex
self.callbacks = callbacks or {}
class ParserConf:
def __init__(self, rules, callback, start):
self.rules = rules
self.callback = callback
self.start = start

View File

@@ -0,0 +1,89 @@
from .utils import STRING_TYPE
###{standalone
class LarkError(Exception):
pass
class GrammarError(LarkError):
pass
class ParseError(LarkError):
pass
class LexError(LarkError):
pass
class UnexpectedInput(LarkError):
pos_in_stream = None
def get_context(self, text, span=40):
pos = self.pos_in_stream
start = max(pos - span, 0)
end = pos + span
before = text[start:pos].rsplit('\n', 1)[-1]
after = text[pos:end].split('\n', 1)[0]
return before + after + '\n' + ' ' * len(before) + '^\n'
def match_examples(self, parse_fn, examples):
""" Given a parser instance and a dictionary mapping some label with
some malformed syntax examples, it'll return the label for the
example that bests matches the current error.
"""
assert self.state is not None, "Not supported for this exception"
candidate = None
for label, example in examples.items():
assert not isinstance(example, STRING_TYPE)
for malformed in example:
try:
parse_fn(malformed)
except UnexpectedInput as ut:
if ut.state == self.state:
try:
if ut.token == self.token: # Try exact match first
return label
except AttributeError:
pass
if not candidate:
candidate = label
return candidate
class UnexpectedCharacters(LexError, UnexpectedInput):
def __init__(self, seq, lex_pos, line, column, allowed=None, considered_tokens=None, state=None):
message = "No terminal defined for '%s' at line %d col %d" % (seq[lex_pos], line, column)
self.line = line
self.column = column
self.allowed = allowed
self.considered_tokens = considered_tokens
self.pos_in_stream = lex_pos
self.state = state
message += '\n\n' + self.get_context(seq)
if allowed:
message += '\nExpecting: %s\n' % allowed
super(UnexpectedCharacters, self).__init__(message)
class UnexpectedToken(ParseError, UnexpectedInput):
def __init__(self, token, expected, considered_rules=None, state=None):
self.token = token
self.expected = expected # XXX str shouldn't necessary
self.line = getattr(token, 'line', '?')
self.column = getattr(token, 'column', '?')
self.considered_rules = considered_rules
self.state = state
self.pos_in_stream = getattr(token, 'pos_in_stream', None)
message = ("Unexpected token %r at line %s, column %s.\n"
"Expected one of: \n\t* %s\n"
% (token, self.line, self.column, '\n\t* '.join(self.expected)))
super(UnexpectedToken, self).__init__(message)
###}

View File

@@ -0,0 +1,66 @@
class Symbol(object):
is_term = NotImplemented
def __init__(self, name):
self.name = name
def __eq__(self, other):
assert isinstance(other, Symbol), other
return self.is_term == other.is_term and self.name == other.name
def __ne__(self, other):
return not (self == other)
def __hash__(self):
return hash(self.name)
def __repr__(self):
return '%s(%r)' % (type(self).__name__, self.name)
fullrepr = property(__repr__)
class Terminal(Symbol):
is_term = True
def __init__(self, name, filter_out=False):
self.name = name
self.filter_out = filter_out
@property
def fullrepr(self):
return '%s(%r, %r)' % (type(self).__name__, self.name, self.filter_out)
class NonTerminal(Symbol):
is_term = False
class Rule(object):
"""
origin : a symbol
expansion : a list of symbols
"""
def __init__(self, origin, expansion, alias=None, options=None):
self.origin = origin
self.expansion = expansion
self.alias = alias
self.options = options
def __str__(self):
return '<%s : %s>' % (self.origin, ' '.join(map(str,self.expansion)))
def __repr__(self):
return 'Rule(%r, %r, %r, %r)' % (self.origin, self.expansion, self.alias, self.options)
class RuleOptions:
def __init__(self, keep_all_tokens=False, expand1=False, priority=None):
self.keep_all_tokens = keep_all_tokens
self.expand1 = expand1
self.priority = priority
def __repr__(self):
return 'RuleOptions(%r, %r, %r)' % (
self.keep_all_tokens,
self.expand1,
self.priority,
)

View File

@@ -0,0 +1,49 @@
//
// Numbers
//
DIGIT: "0".."9"
HEXDIGIT: "a".."f"|"A".."F"|DIGIT
INT: DIGIT+
SIGNED_INT: ["+"|"-"] INT
DECIMAL: INT "." INT? | "." INT
// float = /-?\d+(\.\d+)?([eE][+-]?\d+)?/
_EXP: ("e"|"E") SIGNED_INT
FLOAT: INT _EXP | DECIMAL _EXP?
SIGNED_FLOAT: ["+"|"-"] FLOAT
NUMBER: FLOAT | INT
SIGNED_NUMBER: ["+"|"-"] NUMBER
//
// Strings
//
//STRING: /"(\\\"|\\\\|[^"\n])*?"i?/
STRING_INNER: ("\\\""|/[^"]/)
ESCAPED_STRING: "\"" STRING_INNER* "\""
//
// Names (Variables)
//
LCASE_LETTER: "a".."z"
UCASE_LETTER: "A".."Z"
LETTER: UCASE_LETTER | LCASE_LETTER
WORD: LETTER+
CNAME: ("_"|LETTER) ("_"|LETTER|DIGIT)*
//
// Whitespace
//
WS_INLINE: (" "|/\t/)+
WS: /[ \t\f\r\n]/+
CR : /\r/
LF : /\n/
NEWLINE: (CR? LF)+

View File

@@ -0,0 +1,55 @@
"Provides Indentation services for languages with indentation similar to Python"
from .lexer import Token
###{standalone
class Indenter:
def __init__(self):
self.paren_level = 0
self.indent_level = [0]
def handle_NL(self, token):
if self.paren_level > 0:
return
yield token
indent_str = token.rsplit('\n', 1)[1] # Tabs and spaces
indent = indent_str.count(' ') + indent_str.count('\t') * self.tab_len
if indent > self.indent_level[-1]:
self.indent_level.append(indent)
yield Token.new_borrow_pos(self.INDENT_type, indent_str, token)
else:
while indent < self.indent_level[-1]:
self.indent_level.pop()
yield Token.new_borrow_pos(self.DEDENT_type, indent_str, token)
assert indent == self.indent_level[-1], '%s != %s' % (indent, self.indent_level[-1])
def process(self, stream):
for token in stream:
if token.type == self.NL_type:
for t in self.handle_NL(token):
yield t
else:
yield token
if token.type in self.OPEN_PAREN_types:
self.paren_level += 1
elif token.type in self.CLOSE_PAREN_types:
self.paren_level -= 1
assert self.paren_level >= 0
while len(self.indent_level) > 1:
self.indent_level.pop()
yield Token(self.DEDENT_type, '')
assert self.indent_level == [0], self.indent_level
# XXX Hack for ContextualLexer. Maybe there's a more elegant solution?
@property
def always_accept(self):
return (self.NL_type,)
###}

View File

@@ -0,0 +1,235 @@
from __future__ import absolute_import
import os
import time
from collections import defaultdict
from io import open
from .utils import STRING_TYPE
from .load_grammar import load_grammar
from .tree import Tree
from .common import LexerConf, ParserConf
from .lexer import Lexer, TraditionalLexer
from .parse_tree_builder import ParseTreeBuilder
from .parser_frontends import get_frontend
class LarkOptions(object):
"""Specifies the options for Lark
"""
OPTIONS_DOC = """
parser - Decides which parser engine to use, "earley" or "lalr". (Default: "earley")
Note: "lalr" requires a lexer
lexer - Decides whether or not to use a lexer stage
"standard": Use a standard lexer
"contextual": Stronger lexer (only works with parser="lalr")
"dynamic": Flexible and powerful (only with parser="earley")
"dynamic_complete": Same as dynamic, but tries *every* variation
of tokenizing possible. (only with parser="earley")
"auto" (default): Choose for me based on grammar and parser
ambiguity - Decides how to handle ambiguity in the parse. Only relevant if parser="earley"
"resolve": The parser will automatically choose the simplest derivation
(it chooses consistently: greedy for tokens, non-greedy for rules)
"explicit": The parser will return all derivations wrapped in "_ambig" tree nodes (i.e. a forest).
transformer - Applies the transformer to every parse tree
debug - Affects verbosity (default: False)
keep_all_tokens - Don't automagically remove "punctuation" tokens (default: False)
cache_grammar - Cache the Lark grammar (Default: False)
postlex - Lexer post-processing (Default: None) Only works with the standard and contextual lexers.
start - The start symbol (Default: start)
profile - Measure run-time usage in Lark. Read results from the profiler proprety (Default: False)
propagate_positions - Propagates [line, column, end_line, end_column] attributes into all tree branches.
lexer_callbacks - Dictionary of callbacks for the lexer. May alter tokens during lexing. Use with caution.
"""
__doc__ += OPTIONS_DOC
def __init__(self, options_dict):
o = dict(options_dict)
self.debug = bool(o.pop('debug', False))
self.keep_all_tokens = bool(o.pop('keep_all_tokens', False))
self.tree_class = o.pop('tree_class', Tree)
self.cache_grammar = o.pop('cache_grammar', False)
self.postlex = o.pop('postlex', None)
self.parser = o.pop('parser', 'earley')
self.lexer = o.pop('lexer', 'auto')
self.transformer = o.pop('transformer', None)
self.start = o.pop('start', 'start')
self.profile = o.pop('profile', False)
self.ambiguity = o.pop('ambiguity', 'auto')
self.propagate_positions = o.pop('propagate_positions', False)
self.earley__predict_all = o.pop('earley__predict_all', False)
self.lexer_callbacks = o.pop('lexer_callbacks', {})
assert self.parser in ('earley', 'lalr', 'cyk', None)
if self.parser == 'earley' and self.transformer:
raise ValueError('Cannot specify an embedded transformer when using the Earley algorithm.'
'Please use your transformer on the resulting parse tree, or use a different algorithm (i.e. lalr)')
if o:
raise ValueError("Unknown options: %s" % o.keys())
class Profiler:
def __init__(self):
self.total_time = defaultdict(float)
self.cur_section = '__init__'
self.last_enter_time = time.time()
def enter_section(self, name):
cur_time = time.time()
self.total_time[self.cur_section] += cur_time - self.last_enter_time
self.last_enter_time = cur_time
self.cur_section = name
def make_wrapper(self, name, f):
def wrapper(*args, **kwargs):
last_section = self.cur_section
self.enter_section(name)
try:
return f(*args, **kwargs)
finally:
self.enter_section(last_section)
return wrapper
class Lark:
def __init__(self, grammar, **options):
"""
grammar : a string or file-object containing the grammar spec (using Lark's ebnf syntax)
options : a dictionary controlling various aspects of Lark.
"""
self.options = LarkOptions(options)
# Some, but not all file-like objects have a 'name' attribute
try:
self.source = grammar.name
except AttributeError:
self.source = '<string>'
cache_file = "larkcache_%s" % str(hash(grammar)%(2**32))
else:
cache_file = "larkcache_%s" % os.path.basename(self.source)
# Drain file-like objects to get their contents
try:
read = grammar.read
except AttributeError:
pass
else:
grammar = read()
assert isinstance(grammar, STRING_TYPE)
if self.options.cache_grammar:
raise NotImplementedError("Not available yet")
assert not self.options.profile, "Feature temporarily disabled"
self.profiler = Profiler() if self.options.profile else None
if self.options.lexer == 'auto':
if self.options.parser == 'lalr':
self.options.lexer = 'contextual'
elif self.options.parser == 'earley':
self.options.lexer = 'dynamic'
elif self.options.parser == 'cyk':
self.options.lexer = 'standard'
else:
assert False, self.options.parser
lexer = self.options.lexer
assert lexer in ('standard', 'contextual', 'dynamic', 'dynamic_complete') or issubclass(lexer, Lexer)
if self.options.ambiguity == 'auto':
if self.options.parser == 'earley':
self.options.ambiguity = 'resolve'
else:
disambig_parsers = ['earley', 'cyk']
assert self.options.parser in disambig_parsers, (
'Only %s supports disambiguation right now') % ', '.join(disambig_parsers)
assert self.options.ambiguity in ('resolve', 'explicit', 'auto', 'resolve__antiscore_sum')
# Parse the grammar file and compose the grammars (TODO)
self.grammar = load_grammar(grammar, self.source)
# Compile the EBNF grammar into BNF
self.terminals, self.rules, self.ignore_tokens = self.grammar.compile()
self.lexer_conf = LexerConf(self.terminals, self.ignore_tokens, self.options.postlex, self.options.lexer_callbacks)
if self.options.parser:
self.parser = self._build_parser()
elif lexer:
self.lexer = self._build_lexer()
if self.profiler: self.profiler.enter_section('outside_lark')
__init__.__doc__ += "\nOPTIONS:" + LarkOptions.OPTIONS_DOC
def _build_lexer(self):
return TraditionalLexer(self.lexer_conf.tokens, ignore=self.lexer_conf.ignore, user_callbacks=self.lexer_conf.callbacks)
def _build_parser(self):
self.parser_class = get_frontend(self.options.parser, self.options.lexer)
self._parse_tree_builder = ParseTreeBuilder(self.rules, self.options.tree_class, self.options.propagate_positions, self.options.keep_all_tokens, self.options.parser!='lalr')
callback = self._parse_tree_builder.create_callback(self.options.transformer)
if self.profiler:
for f in dir(callback):
if not (f.startswith('__') and f.endswith('__')):
setattr(callback, f, self.profiler.make_wrapper('transformer', getattr(callback, f)))
parser_conf = ParserConf(self.rules, callback, self.options.start)
return self.parser_class(self.lexer_conf, parser_conf, options=self.options)
@classmethod
def open(cls, grammar_filename, rel_to=None, **options):
"""Create an instance of Lark with the grammar given by its filename
If rel_to is provided, the function will find the grammar filename in relation to it.
Example:
>>> Lark.open("grammar_file.lark", rel_to=__file__, parser="lalr")
Lark(...)
"""
if rel_to:
basepath = os.path.dirname(rel_to)
grammar_filename = os.path.join(basepath, grammar_filename)
with open(grammar_filename, encoding='utf8') as f:
return cls(f, **options)
def __repr__(self):
return 'Lark(open(%r), parser=%r, lexer=%r, ...)' % (self.source, self.options.parser, self.options.lexer)
def lex(self, text):
"Only lex (and postlex) the text, without parsing it. Only relevant when lexer='standard'"
if not hasattr(self, 'lexer'):
self.lexer = self._build_lexer()
stream = self.lexer.lex(text)
if self.options.postlex:
return self.options.postlex.process(stream)
return stream
def parse(self, text):
"Parse the given text, according to the options provided. Returns a tree, unless specified otherwise."
return self.parser.parse(text)
# if self.profiler:
# self.profiler.enter_section('lex')
# l = list(self.lex(text))
# self.profiler.enter_section('parse')
# try:
# return self.parser.parse(l)
# finally:
# self.profiler.enter_section('outside_lark')
# else:
# l = list(self.lex(text))
# return self.parser.parse(l)

View File

@@ -0,0 +1,324 @@
## Lexer Implementation
import re
from .utils import Str, classify, get_regexp_width, Py36
from .exceptions import UnexpectedCharacters, LexError
class Pattern(object):
def __init__(self, value, flags=()):
self.value = value
self.flags = frozenset(flags)
def __repr__(self):
return repr(self.to_regexp())
# Pattern Hashing assumes all subclasses have a different priority!
def __hash__(self):
return hash((type(self), self.value, self.flags))
def __eq__(self, other):
return type(self) == type(other) and self.value == other.value and self.flags == other.flags
def to_regexp(self):
raise NotImplementedError()
if Py36:
# Python 3.6 changed syntax for flags in regular expression
def _get_flags(self, value):
for f in self.flags:
value = ('(?%s:%s)' % (f, value))
return value
else:
def _get_flags(self, value):
for f in self.flags:
value = ('(?%s)' % f) + value
return value
class PatternStr(Pattern):
def to_regexp(self):
return self._get_flags(re.escape(self.value))
@property
def min_width(self):
return len(self.value)
max_width = min_width
class PatternRE(Pattern):
def to_regexp(self):
return self._get_flags(self.value)
@property
def min_width(self):
return get_regexp_width(self.to_regexp())[0]
@property
def max_width(self):
return get_regexp_width(self.to_regexp())[1]
class TerminalDef(object):
def __init__(self, name, pattern, priority=1):
assert isinstance(pattern, Pattern), pattern
self.name = name
self.pattern = pattern
self.priority = priority
def __repr__(self):
return '%s(%r, %r)' % (type(self).__name__, self.name, self.pattern)
###{standalone
class Token(Str):
__slots__ = ('type', 'pos_in_stream', 'value', 'line', 'column', 'end_line', 'end_column')
def __new__(cls, type_, value, pos_in_stream=None, line=None, column=None):
self = super(Token, cls).__new__(cls, value)
self.type = type_
self.pos_in_stream = pos_in_stream
self.value = value
self.line = line
self.column = column
self.end_line = None
self.end_column = None
return self
@classmethod
def new_borrow_pos(cls, type_, value, borrow_t):
return cls(type_, value, borrow_t.pos_in_stream, line=borrow_t.line, column=borrow_t.column)
def __reduce__(self):
return (self.__class__, (self.type, self.value, self.pos_in_stream, self.line, self.column, ))
def __repr__(self):
return 'Token(%s, %r)' % (self.type, self.value)
def __deepcopy__(self, memo):
return Token(self.type, self.value, self.pos_in_stream, self.line, self.column)
def __eq__(self, other):
if isinstance(other, Token) and self.type != other.type:
return False
return Str.__eq__(self, other)
__hash__ = Str.__hash__
class LineCounter:
def __init__(self):
self.newline_char = '\n'
self.char_pos = 0
self.line = 1
self.column = 1
self.line_start_pos = 0
def feed(self, token, test_newline=True):
"""Consume a token and calculate the new line & column.
As an optional optimization, set test_newline=False is token doesn't contain a newline.
"""
if test_newline:
newlines = token.count(self.newline_char)
if newlines:
self.line += newlines
self.line_start_pos = self.char_pos + token.rindex(self.newline_char) + 1
self.char_pos += len(token)
self.column = self.char_pos - self.line_start_pos + 1
class _Lex:
"Built to serve both Lexer and ContextualLexer"
def __init__(self, lexer, state=None):
self.lexer = lexer
self.state = state
def lex(self, stream, newline_types, ignore_types):
newline_types = frozenset(newline_types)
ignore_types = frozenset(ignore_types)
line_ctr = LineCounter()
while line_ctr.char_pos < len(stream):
lexer = self.lexer
for mre, type_from_index in lexer.mres:
m = mre.match(stream, line_ctr.char_pos)
if not m:
continue
t = None
value = m.group(0)
type_ = type_from_index[m.lastindex]
if type_ not in ignore_types:
t = Token(type_, value, line_ctr.char_pos, line_ctr.line, line_ctr.column)
if t.type in lexer.callback:
t = lexer.callback[t.type](t)
yield t
else:
if type_ in lexer.callback:
t = Token(type_, value, line_ctr.char_pos, line_ctr.line, line_ctr.column)
lexer.callback[type_](t)
line_ctr.feed(value, type_ in newline_types)
if t:
t.end_line = line_ctr.line
t.end_column = line_ctr.column
break
else:
raise UnexpectedCharacters(stream, line_ctr.char_pos, line_ctr.line, line_ctr.column, state=self.state)
class UnlessCallback:
def __init__(self, mres):
self.mres = mres
def __call__(self, t):
for mre, type_from_index in self.mres:
m = mre.match(t.value)
if m:
t.type = type_from_index[m.lastindex]
break
return t
###}
def _create_unless(terminals):
tokens_by_type = classify(terminals, lambda t: type(t.pattern))
assert len(tokens_by_type) <= 2, tokens_by_type.keys()
embedded_strs = set()
callback = {}
for retok in tokens_by_type.get(PatternRE, []):
unless = [] # {}
for strtok in tokens_by_type.get(PatternStr, []):
if strtok.priority > retok.priority:
continue
s = strtok.pattern.value
m = re.match(retok.pattern.to_regexp(), s)
if m and m.group(0) == s:
unless.append(strtok)
if strtok.pattern.flags <= retok.pattern.flags:
embedded_strs.add(strtok)
if unless:
callback[retok.name] = UnlessCallback(build_mres(unless, match_whole=True))
terminals = [t for t in terminals if t not in embedded_strs]
return terminals, callback
def _build_mres(terminals, max_size, match_whole):
# Python sets an unreasonable group limit (currently 100) in its re module
# Worse, the only way to know we reached it is by catching an AssertionError!
# This function recursively tries less and less groups until it's successful.
postfix = '$' if match_whole else ''
mres = []
while terminals:
try:
mre = re.compile(u'|'.join(u'(?P<%s>%s)'%(t.name, t.pattern.to_regexp()+postfix) for t in terminals[:max_size]))
except AssertionError: # Yes, this is what Python provides us.. :/
return _build_mres(terminals, max_size//2, match_whole)
# terms_from_name = {t.name: t for t in terminals[:max_size]}
mres.append((mre, {i:n for n,i in mre.groupindex.items()} ))
terminals = terminals[max_size:]
return mres
def build_mres(terminals, match_whole=False):
return _build_mres(terminals, len(terminals), match_whole)
def _regexp_has_newline(r):
"""Expressions that may indicate newlines in a regexp:
- newlines (\n)
- escaped newline (\\n)
- anything but ([^...])
- any-char (.) when the flag (?s) exists
"""
return '\n' in r or '\\n' in r or '[^' in r or ('(?s' in r and '.' in r)
class Lexer:
"""Lexer interface
Method Signatures:
lex(self, stream) -> Iterator[Token]
set_parser_state(self, state) # Optional
"""
set_parser_state = NotImplemented
lex = NotImplemented
class TraditionalLexer(Lexer):
def __init__(self, terminals, ignore=(), user_callbacks={}):
assert all(isinstance(t, TerminalDef) for t in terminals), terminals
terminals = list(terminals)
# Sanitization
for t in terminals:
try:
re.compile(t.pattern.to_regexp())
except:
raise LexError("Cannot compile token %s: %s" % (t.name, t.pattern))
if t.pattern.min_width == 0:
raise LexError("Lexer does not allow zero-width terminals. (%s: %s)" % (t.name, t.pattern))
assert set(ignore) <= {t.name for t in terminals}
# Init
self.newline_types = [t.name for t in terminals if _regexp_has_newline(t.pattern.to_regexp())]
self.ignore_types = list(ignore)
terminals.sort(key=lambda x:(-x.priority, -x.pattern.max_width, -len(x.pattern.value), x.name))
terminals, self.callback = _create_unless(terminals)
assert all(self.callback.values())
for type_, f in user_callbacks.items():
assert type_ not in self.callback
self.callback[type_] = f
self.terminals = terminals
self.mres = build_mres(terminals)
def lex(self, stream):
return _Lex(self).lex(stream, self.newline_types, self.ignore_types)
class ContextualLexer(Lexer):
def __init__(self, terminals, states, ignore=(), always_accept=(), user_callbacks={}):
tokens_by_name = {}
for t in terminals:
assert t.name not in tokens_by_name, t
tokens_by_name[t.name] = t
lexer_by_tokens = {}
self.lexers = {}
for state, accepts in states.items():
key = frozenset(accepts)
try:
lexer = lexer_by_tokens[key]
except KeyError:
accepts = set(accepts) | set(ignore) | set(always_accept)
state_tokens = [tokens_by_name[n] for n in accepts if n and n in tokens_by_name]
lexer = TraditionalLexer(state_tokens, ignore=ignore, user_callbacks=user_callbacks)
lexer_by_tokens[key] = lexer
self.lexers[state] = lexer
self.root_lexer = TraditionalLexer(terminals, ignore=ignore, user_callbacks=user_callbacks)
self.set_parser_state(None) # Needs to be set on the outside
def set_parser_state(self, state):
self.parser_state = state
def lex(self, stream):
l = _Lex(self.lexers[self.parser_state], self.parser_state)
for x in l.lex(stream, self.root_lexer.newline_types, self.root_lexer.ignore_types):
yield x
l.lexer = self.lexers[self.parser_state]
l.state = self.parser_state

View File

@@ -0,0 +1,779 @@
"Parses and creates Grammar objects"
import os.path
import sys
from ast import literal_eval
from copy import deepcopy
from .utils import bfs
from .lexer import Token, TerminalDef, PatternStr, PatternRE
from .parse_tree_builder import ParseTreeBuilder
from .parser_frontends import LALR_TraditionalLexer
from .common import LexerConf, ParserConf
from .grammar import RuleOptions, Rule, Terminal, NonTerminal, Symbol
from .utils import classify, suppress
from .exceptions import GrammarError, UnexpectedCharacters, UnexpectedToken
from .tree import Tree, SlottedTree as ST
from .visitors import Transformer, Visitor, v_args, Transformer_InPlace
inline_args = v_args(inline=True)
__path__ = os.path.dirname(__file__)
IMPORT_PATHS = [os.path.join(__path__, 'grammars')]
EXT = '.lark'
_RE_FLAGS = 'imslux'
_TERMINAL_NAMES = {
'.' : 'DOT',
',' : 'COMMA',
':' : 'COLON',
';' : 'SEMICOLON',
'+' : 'PLUS',
'-' : 'MINUS',
'*' : 'STAR',
'/' : 'SLASH',
'\\' : 'BACKSLASH',
'|' : 'VBAR',
'?' : 'QMARK',
'!' : 'BANG',
'@' : 'AT',
'#' : 'HASH',
'$' : 'DOLLAR',
'%' : 'PERCENT',
'^' : 'CIRCUMFLEX',
'&' : 'AMPERSAND',
'_' : 'UNDERSCORE',
'<' : 'LESSTHAN',
'>' : 'MORETHAN',
'=' : 'EQUAL',
'"' : 'DBLQUOTE',
'\'' : 'QUOTE',
'`' : 'BACKQUOTE',
'~' : 'TILDE',
'(' : 'LPAR',
')' : 'RPAR',
'{' : 'LBRACE',
'}' : 'RBRACE',
'[' : 'LSQB',
']' : 'RSQB',
'\n' : 'NEWLINE',
'\r\n' : 'CRLF',
'\t' : 'TAB',
' ' : 'SPACE',
}
# Grammar Parser
TERMINALS = {
'_LPAR': r'\(',
'_RPAR': r'\)',
'_LBRA': r'\[',
'_RBRA': r'\]',
'OP': '[+*][?]?|[?](?![a-z])',
'_COLON': ':',
'_COMMA': ',',
'_OR': r'\|',
'_DOT': r'\.',
'TILDE': '~',
'RULE': '!?[_?]?[a-z][_a-z0-9]*',
'TERMINAL': '_?[A-Z][_A-Z0-9]*',
'STRING': r'"(\\"|\\\\|[^"\n])*?"i?',
'REGEXP': r'/(?!/)(\\/|\\\\|[^/\n])*?/[%s]*' % _RE_FLAGS,
'_NL': r'(\r?\n)+\s*',
'WS': r'[ \t]+',
'COMMENT': r'//[^\n]*',
'_TO': '->',
'_IGNORE': r'%ignore',
'_DECLARE': r'%declare',
'_IMPORT': r'%import',
'NUMBER': r'\d+',
}
RULES = {
'start': ['_list'],
'_list': ['_item', '_list _item'],
'_item': ['rule', 'term', 'statement', '_NL'],
'rule': ['RULE _COLON expansions _NL',
'RULE _DOT NUMBER _COLON expansions _NL'],
'expansions': ['alias',
'expansions _OR alias',
'expansions _NL _OR alias'],
'?alias': ['expansion _TO RULE', 'expansion'],
'expansion': ['_expansion'],
'_expansion': ['', '_expansion expr'],
'?expr': ['atom',
'atom OP',
'atom TILDE NUMBER',
'atom TILDE NUMBER _DOT _DOT NUMBER',
],
'?atom': ['_LPAR expansions _RPAR',
'maybe',
'value'],
'value': ['terminal',
'nonterminal',
'literal',
'range'],
'terminal': ['TERMINAL'],
'nonterminal': ['RULE'],
'?name': ['RULE', 'TERMINAL'],
'maybe': ['_LBRA expansions _RBRA'],
'range': ['STRING _DOT _DOT STRING'],
'term': ['TERMINAL _COLON expansions _NL',
'TERMINAL _DOT NUMBER _COLON expansions _NL'],
'statement': ['ignore', 'import', 'declare'],
'ignore': ['_IGNORE expansions _NL'],
'declare': ['_DECLARE _declare_args _NL'],
'import': ['_IMPORT _import_path _NL',
'_IMPORT _import_path _LPAR name_list _RPAR _NL',
'_IMPORT _import_path _TO TERMINAL _NL'],
'_import_path': ['import_lib', 'import_rel'],
'import_lib': ['_import_args'],
'import_rel': ['_DOT _import_args'],
'_import_args': ['name', '_import_args _DOT name'],
'name_list': ['_name_list'],
'_name_list': ['name', '_name_list _COMMA name'],
'_declare_args': ['name', '_declare_args name'],
'literal': ['REGEXP', 'STRING'],
}
@inline_args
class EBNF_to_BNF(Transformer_InPlace):
def __init__(self):
self.new_rules = []
self.rules_by_expr = {}
self.prefix = 'anon'
self.i = 0
self.rule_options = None
def _add_recurse_rule(self, type_, expr):
if expr in self.rules_by_expr:
return self.rules_by_expr[expr]
new_name = '__%s_%s_%d' % (self.prefix, type_, self.i)
self.i += 1
t = NonTerminal(new_name)
tree = ST('expansions', [ST('expansion', [expr]), ST('expansion', [t, expr])])
self.new_rules.append((new_name, tree, self.rule_options))
self.rules_by_expr[expr] = t
return t
def expr(self, rule, op, *args):
if op.value == '?':
return ST('expansions', [rule, ST('expansion', [])])
elif op.value == '+':
# a : b c+ d
# -->
# a : b _c d
# _c : _c c | c;
return self._add_recurse_rule('plus', rule)
elif op.value == '*':
# a : b c* d
# -->
# a : b _c? d
# _c : _c c | c;
new_name = self._add_recurse_rule('star', rule)
return ST('expansions', [new_name, ST('expansion', [])])
elif op.value == '~':
if len(args) == 1:
mn = mx = int(args[0])
else:
mn, mx = map(int, args)
if mx < mn:
raise GrammarError("Bad Range for %s (%d..%d isn't allowed)" % (rule, mn, mx))
return ST('expansions', [ST('expansion', [rule] * n) for n in range(mn, mx+1)])
assert False, op
class SimplifyRule_Visitor(Visitor):
@staticmethod
def _flatten(tree):
while True:
to_expand = [i for i, child in enumerate(tree.children)
if isinstance(child, Tree) and child.data == tree.data]
if not to_expand:
break
tree.expand_kids_by_index(*to_expand)
def expansion(self, tree):
# rules_list unpacking
# a : b (c|d) e
# -->
# a : b c e | b d e
#
# In AST terms:
# expansion(b, expansions(c, d), e)
# -->
# expansions( expansion(b, c, e), expansion(b, d, e) )
self._flatten(tree)
for i, child in enumerate(tree.children):
if isinstance(child, Tree) and child.data == 'expansions':
tree.data = 'expansions'
tree.children = [self.visit(ST('expansion', [option if i==j else other
for j, other in enumerate(tree.children)]))
for option in set(child.children)]
self._flatten(tree)
break
def alias(self, tree):
rule, alias_name = tree.children
if rule.data == 'expansions':
aliases = []
for child in tree.children[0].children:
aliases.append(ST('alias', [child, alias_name]))
tree.data = 'expansions'
tree.children = aliases
def expansions(self, tree):
self._flatten(tree)
tree.children = list(set(tree.children))
class RuleTreeToText(Transformer):
def expansions(self, x):
return x
def expansion(self, symbols):
return symbols, None
def alias(self, x):
(expansion, _alias), alias = x
assert _alias is None, (alias, expansion, '-', _alias) # Double alias not allowed
return expansion, alias.value
@inline_args
class CanonizeTree(Transformer_InPlace):
def maybe(self, expr):
return ST('expr', [expr, Token('OP', '?', -1)])
def tokenmods(self, *args):
if len(args) == 1:
return list(args)
tokenmods, value = args
return tokenmods + [value]
class PrepareAnonTerminals(Transformer_InPlace):
"Create a unique list of anonymous terminals. Attempt to give meaningful names to them when we add them"
def __init__(self, terminals):
self.terminals = terminals
self.term_set = {td.name for td in self.terminals}
self.term_reverse = {td.pattern: td for td in terminals}
self.i = 0
@inline_args
def pattern(self, p):
value = p.value
if p in self.term_reverse and p.flags != self.term_reverse[p].pattern.flags:
raise GrammarError(u'Conflicting flags for the same terminal: %s' % p)
term_name = None
if isinstance(p, PatternStr):
try:
# If already defined, use the user-defined terminal name
term_name = self.term_reverse[p].name
except KeyError:
# Try to assign an indicative anon-terminal name
try:
term_name = _TERMINAL_NAMES[value]
except KeyError:
if value.isalnum() and value[0].isalpha() and value.upper() not in self.term_set:
with suppress(UnicodeEncodeError):
value.upper().encode('ascii') # Make sure we don't have unicode in our terminal names
term_name = value.upper()
if term_name in self.term_set:
term_name = None
elif isinstance(p, PatternRE):
if p in self.term_reverse: # Kind of a wierd placement.name
term_name = self.term_reverse[p].name
else:
assert False, p
if term_name is None:
term_name = '__ANON_%d' % self.i
self.i += 1
if term_name not in self.term_set:
assert p not in self.term_reverse
self.term_set.add(term_name)
termdef = TerminalDef(term_name, p)
self.term_reverse[p] = termdef
self.terminals.append(termdef)
return Terminal(term_name, filter_out=isinstance(p, PatternStr))
def _rfind(s, choices):
return max(s.rfind(c) for c in choices)
def _fix_escaping(s):
w = ''
i = iter(s)
for n in i:
w += n
if n == '\\':
n2 = next(i)
if n2 == '\\':
w += '\\\\'
elif n2 not in 'unftr':
w += '\\'
w += n2
w = w.replace('\\"', '"').replace("'", "\\'")
to_eval = "u'''%s'''" % w
try:
s = literal_eval(to_eval)
except SyntaxError as e:
raise ValueError(s, e)
return s
def _literal_to_pattern(literal):
v = literal.value
flag_start = _rfind(v, '/"')+1
assert flag_start > 0
flags = v[flag_start:]
assert all(f in _RE_FLAGS for f in flags), flags
v = v[:flag_start]
assert v[0] == v[-1] and v[0] in '"/'
x = v[1:-1]
s = _fix_escaping(x)
if literal.type == 'STRING':
s = s.replace('\\\\', '\\')
return { 'STRING': PatternStr,
'REGEXP': PatternRE }[literal.type](s, flags)
@inline_args
class PrepareLiterals(Transformer_InPlace):
def literal(self, literal):
return ST('pattern', [_literal_to_pattern(literal)])
def range(self, start, end):
assert start.type == end.type == 'STRING'
start = start.value[1:-1]
end = end.value[1:-1]
assert len(start) == len(end) == 1, (start, end, len(start), len(end))
regexp = '[%s-%s]' % (start, end)
return ST('pattern', [PatternRE(regexp)])
class TerminalTreeToPattern(Transformer):
def pattern(self, ps):
p ,= ps
return p
def expansion(self, items):
assert items
if len(items) == 1:
return items[0]
if len({i.flags for i in items}) > 1:
raise GrammarError("Lark doesn't support joining terminals with conflicting flags!")
return PatternRE(''.join(i.to_regexp() for i in items), items[0].flags if items else ())
def expansions(self, exps):
if len(exps) == 1:
return exps[0]
if len({i.flags for i in exps}) > 1:
raise GrammarError("Lark doesn't support joining terminals with conflicting flags!")
return PatternRE('(?:%s)' % ('|'.join(i.to_regexp() for i in exps)), exps[0].flags)
def expr(self, args):
inner, op = args[:2]
if op == '~':
if len(args) == 3:
op = "{%d}" % int(args[2])
else:
mn, mx = map(int, args[2:])
if mx < mn:
raise GrammarError("Bad Range for %s (%d..%d isn't allowed)" % (inner, mn, mx))
op = "{%d,%d}" % (mn, mx)
else:
assert len(args) == 2
return PatternRE('(?:%s)%s' % (inner.to_regexp(), op), inner.flags)
def alias(self, t):
raise GrammarError("Aliasing not allowed in terminals (You used -> in the wrong place)")
def value(self, v):
return v[0]
class PrepareSymbols(Transformer_InPlace):
def value(self, v):
v ,= v
if isinstance(v, Tree):
return v
elif v.type == 'RULE':
return NonTerminal(v.value)
elif v.type == 'TERMINAL':
return Terminal(v.value, filter_out=v.startswith('_'))
assert False
def _choice_of_rules(rules):
return ST('expansions', [ST('expansion', [Token('RULE', name)]) for name in rules])
class Grammar:
def __init__(self, rule_defs, term_defs, ignore):
self.term_defs = term_defs
self.rule_defs = rule_defs
self.ignore = ignore
def compile(self):
# We change the trees in-place (to support huge grammars)
# So deepcopy allows calling compile more than once.
term_defs = deepcopy(list(self.term_defs))
rule_defs = deepcopy(self.rule_defs)
# ===================
# Compile Terminals
# ===================
# Convert terminal-trees to strings/regexps
transformer = PrepareLiterals() * TerminalTreeToPattern()
for name, (term_tree, priority) in term_defs:
if term_tree is None: # Terminal added through %declare
continue
expansions = list(term_tree.find_data('expansion'))
if len(expansions) == 1 and not expansions[0].children:
raise GrammarError("Terminals cannot be empty (%s)" % name)
terminals = [TerminalDef(name, transformer.transform(term_tree), priority)
for name, (term_tree, priority) in term_defs if term_tree]
# =================
# Compile Rules
# =================
# 1. Pre-process terminals
transformer = PrepareLiterals() * PrepareSymbols() * PrepareAnonTerminals(terminals) # Adds to terminals
# 2. Convert EBNF to BNF (and apply step 1)
ebnf_to_bnf = EBNF_to_BNF()
rules = []
for name, rule_tree, options in rule_defs:
ebnf_to_bnf.rule_options = RuleOptions(keep_all_tokens=True) if options and options.keep_all_tokens else None
tree = transformer.transform(rule_tree)
rules.append((name, ebnf_to_bnf.transform(tree), options))
rules += ebnf_to_bnf.new_rules
assert len(rules) == len({name for name, _t, _o in rules}), "Whoops, name collision"
# 3. Compile tree to Rule objects
rule_tree_to_text = RuleTreeToText()
simplify_rule = SimplifyRule_Visitor()
compiled_rules = []
for name, tree, options in rules:
simplify_rule.visit(tree)
expansions = rule_tree_to_text.transform(tree)
for expansion, alias in expansions:
if alias and name.startswith('_'):
raise GrammarError("Rule %s is marked for expansion (it starts with an underscore) and isn't allowed to have aliases (alias=%s)" % (name, alias))
assert all(isinstance(x, Symbol) for x in expansion), expansion
rule = Rule(NonTerminal(name), expansion, alias, options)
compiled_rules.append(rule)
return terminals, compiled_rules, self.ignore
_imported_grammars = {}
def import_grammar(grammar_path, base_paths=[]):
if grammar_path not in _imported_grammars:
import_paths = base_paths + IMPORT_PATHS
for import_path in import_paths:
with suppress(IOError):
with open(os.path.join(import_path, grammar_path)) as f:
text = f.read()
grammar = load_grammar(text, grammar_path)
_imported_grammars[grammar_path] = grammar
break
else:
open(grammar_path)
assert False
return _imported_grammars[grammar_path]
def import_from_grammar_into_namespace(grammar, namespace, aliases):
imported_terms = dict(grammar.term_defs)
imported_rules = {n:(n,t,o) for n,t,o in grammar.rule_defs}
term_defs = []
rule_defs = []
def rule_dependencies(symbol):
if symbol.type != 'RULE':
return []
_, tree, _ = imported_rules[symbol]
return tree.scan_values(lambda x: x.type in ('RULE', 'TERMINAL'))
def get_namespace_name(name):
try:
return aliases[name].value
except KeyError:
return '%s.%s' % (namespace, name)
to_import = list(bfs(aliases, rule_dependencies))
for symbol in to_import:
if symbol.type == 'TERMINAL':
term_defs.append([get_namespace_name(symbol), imported_terms[symbol]])
else:
assert symbol.type == 'RULE'
rule = imported_rules[symbol]
for t in rule[1].iter_subtrees():
for i, c in enumerate(t.children):
if isinstance(c, Token) and c.type in ('RULE', 'TERMINAL'):
t.children[i] = Token(c.type, get_namespace_name(c))
rule_defs.append((get_namespace_name(symbol), rule[1], rule[2]))
return term_defs, rule_defs
def resolve_term_references(term_defs):
# TODO Cycles detection
# TODO Solve with transitive closure (maybe)
token_dict = {k:t for k, (t,_p) in term_defs}
assert len(token_dict) == len(term_defs), "Same name defined twice?"
while True:
changed = False
for name, (token_tree, _p) in term_defs:
if token_tree is None: # Terminal added through %declare
continue
for exp in token_tree.find_data('value'):
item ,= exp.children
if isinstance(item, Token):
if item.type == 'RULE':
raise GrammarError("Rules aren't allowed inside terminals (%s in %s)" % (item, name))
if item.type == 'TERMINAL':
exp.children[0] = token_dict[item]
changed = True
if not changed:
break
def options_from_rule(name, *x):
if len(x) > 1:
priority, expansions = x
priority = int(priority)
else:
expansions ,= x
priority = None
keep_all_tokens = name.startswith('!')
name = name.lstrip('!')
expand1 = name.startswith('?')
name = name.lstrip('?')
return name, expansions, RuleOptions(keep_all_tokens, expand1, priority=priority)
def symbols_from_strcase(expansion):
return [Terminal(x, filter_out=x.startswith('_')) if x.isupper() else NonTerminal(x) for x in expansion]
@inline_args
class PrepareGrammar(Transformer_InPlace):
def terminal(self, name):
return name
def nonterminal(self, name):
return name
class GrammarLoader:
def __init__(self):
terminals = [TerminalDef(name, PatternRE(value)) for name, value in TERMINALS.items()]
rules = [options_from_rule(name, x) for name, x in RULES.items()]
rules = [Rule(NonTerminal(r), symbols_from_strcase(x.split()), None, o) for r, xs, o in rules for x in xs]
callback = ParseTreeBuilder(rules, ST).create_callback()
lexer_conf = LexerConf(terminals, ['WS', 'COMMENT'])
parser_conf = ParserConf(rules, callback, 'start')
self.parser = LALR_TraditionalLexer(lexer_conf, parser_conf)
self.canonize_tree = CanonizeTree()
def load_grammar(self, grammar_text, grammar_name='<?>'):
"Parse grammar_text, verify, and create Grammar object. Display nice messages on error."
try:
tree = self.canonize_tree.transform( self.parser.parse(grammar_text+'\n') )
except UnexpectedCharacters as e:
context = e.get_context(grammar_text)
raise GrammarError("Unexpected input at line %d column %d in %s: \n\n%s" %
(e.line, e.column, grammar_name, context))
except UnexpectedToken as e:
context = e.get_context(grammar_text)
error = e.match_examples(self.parser.parse, {
'Unclosed parenthesis': ['a: (\n'],
'Umatched closing parenthesis': ['a: )\n', 'a: [)\n', 'a: (]\n'],
'Expecting rule or terminal definition (missing colon)': ['a\n', 'a->\n', 'A->\n', 'a A\n'],
'Alias expects lowercase name': ['a: -> "a"\n'],
'Unexpected colon': ['a::\n', 'a: b:\n', 'a: B:\n', 'a: "a":\n'],
'Misplaced operator': ['a: b??', 'a: b(?)', 'a:+\n', 'a:?\n', 'a:*\n', 'a:|*\n'],
'Expecting option ("|") or a new rule or terminal definition': ['a:a\n()\n'],
'%import expects a name': ['%import "a"\n'],
'%ignore expects a value': ['%ignore %import\n'],
})
if error:
raise GrammarError("%s at line %s column %s\n\n%s" % (error, e.line, e.column, context))
elif 'STRING' in e.expected:
raise GrammarError("Expecting a value at line %s column %s\n\n%s" % (e.line, e.column, context))
raise
tree = PrepareGrammar().transform(tree)
# Extract grammar items
defs = classify(tree.children, lambda c: c.data, lambda c: c.children)
term_defs = defs.pop('term', [])
rule_defs = defs.pop('rule', [])
statements = defs.pop('statement', [])
assert not defs
term_defs = [td if len(td)==3 else (td[0], 1, td[1]) for td in term_defs]
term_defs = [(name.value, (t, int(p))) for name, p, t in term_defs]
rule_defs = [options_from_rule(*x) for x in rule_defs]
# Execute statements
ignore = []
for (stmt,) in statements:
if stmt.data == 'ignore':
t ,= stmt.children
ignore.append(t)
elif stmt.data == 'import':
if len(stmt.children) > 1:
path_node, arg1 = stmt.children
else:
path_node ,= stmt.children
arg1 = None
if isinstance(arg1, Tree): # Multi import
dotted_path = path_node.children
names = arg1.children
aliases = names # Can't have aliased multi import, so all aliases will be the same as names
else: # Single import
dotted_path = path_node.children[:-1]
names = [path_node.children[-1]] # Get name from dotted path
aliases = [arg1] if arg1 else names # Aliases if exist
grammar_path = os.path.join(*dotted_path) + EXT
if path_node.data == 'import_lib': # Import from library
g = import_grammar(grammar_path)
else: # Relative import
if grammar_name == '<string>': # Import relative to script file path if grammar is coded in script
base_file = os.path.abspath(sys.modules['__main__'].__file__)
else:
base_file = grammar_name # Import relative to grammar file path if external grammar file
base_path = os.path.split(base_file)[0]
g = import_grammar(grammar_path, base_paths=[base_path])
aliases_dict = dict(zip(names, aliases))
new_td, new_rd = import_from_grammar_into_namespace(g, '.'.join(dotted_path), aliases_dict)
term_defs += new_td
rule_defs += new_rd
elif stmt.data == 'declare':
for t in stmt.children:
term_defs.append([t.value, (None, None)])
else:
assert False, stmt
# Verify correctness 1
for name, _ in term_defs:
if name.startswith('__'):
raise GrammarError('Names starting with double-underscore are reserved (Error at %s)' % name)
# Handle ignore tokens
# XXX A slightly hacky solution. Recognition of %ignore TERMINAL as separate comes from the lexer's
# inability to handle duplicate terminals (two names, one value)
ignore_names = []
for t in ignore:
if t.data=='expansions' and len(t.children) == 1:
t2 ,= t.children
if t2.data=='expansion' and len(t2.children) == 1:
item ,= t2.children
if item.data == 'value':
item ,= item.children
if isinstance(item, Token) and item.type == 'TERMINAL':
ignore_names.append(item.value)
continue
name = '__IGNORE_%d'% len(ignore_names)
ignore_names.append(name)
term_defs.append((name, (t, 0)))
# Verify correctness 2
terminal_names = set()
for name, _ in term_defs:
if name in terminal_names:
raise GrammarError("Terminal '%s' defined more than once" % name)
terminal_names.add(name)
if set(ignore_names) > terminal_names:
raise GrammarError("Terminals %s were marked to ignore but were not defined!" % (set(ignore_names) - terminal_names))
resolve_term_references(term_defs)
rules = rule_defs
rule_names = set()
for name, _x, _o in rules:
if name.startswith('__'):
raise GrammarError('Names starting with double-underscore are reserved (Error at %s)' % name)
if name in rule_names:
raise GrammarError("Rule '%s' defined more than once" % name)
rule_names.add(name)
for name, expansions, _o in rules:
used_symbols = {t for x in expansions.find_data('expansion')
for t in x.scan_values(lambda t: t.type in ('RULE', 'TERMINAL'))}
for sym in used_symbols:
if sym.type == 'TERMINAL':
if sym not in terminal_names:
raise GrammarError("Token '%s' used but not defined (in rule %s)" % (sym, name))
else:
if sym not in rule_names:
raise GrammarError("Rule '%s' used but not defined (in rule %s)" % (sym, name))
# TODO don't include unused terminals, they can only cause trouble!
return Grammar(rules, term_defs, ignore_names)
load_grammar = GrammarLoader().load_grammar

View File

@@ -0,0 +1,174 @@
from .exceptions import GrammarError
from .utils import suppress
from .lexer import Token
from .grammar import Rule
from .tree import Tree
from .visitors import InlineTransformer # XXX Deprecated
###{standalone
from functools import partial, wraps
class ExpandSingleChild:
def __init__(self, node_builder):
self.node_builder = node_builder
def __call__(self, children):
if len(children) == 1:
return children[0]
else:
return self.node_builder(children)
class PropagatePositions:
def __init__(self, node_builder):
self.node_builder = node_builder
def __call__(self, children):
res = self.node_builder(children)
if isinstance(res, Tree):
res.meta.empty = True
for c in children:
if isinstance(c, Tree) and c.children and not c.meta.empty:
res.meta.line = c.meta.line
res.meta.column = c.meta.column
res.meta.start_pos = c.meta.start_pos
res.meta.empty = False
break
elif isinstance(c, Token):
res.meta.line = c.line
res.meta.column = c.column
res.meta.start_pos = c.pos_in_stream
res.meta.empty = False
break
for c in reversed(children):
if isinstance(c, Tree) and c.children and not c.meta.empty:
res.meta.end_line = c.meta.end_line
res.meta.end_column = c.meta.end_column
res.meta.end_pos = c.meta.end_pos
res.meta.empty = False
break
elif isinstance(c, Token):
res.meta.end_line = c.end_line
res.meta.end_column = c.end_column
res.meta.end_pos = c.pos_in_stream + len(c.value)
res.meta.empty = False
break
return res
class ChildFilter:
def __init__(self, to_include, node_builder):
self.node_builder = node_builder
self.to_include = to_include
def __call__(self, children):
filtered = []
for i, to_expand in self.to_include:
if to_expand:
filtered += children[i].children
else:
filtered.append(children[i])
return self.node_builder(filtered)
class ChildFilterLALR(ChildFilter):
"Optimized childfilter for LALR (assumes no duplication in parse tree, so it's safe to change it)"
def __call__(self, children):
filtered = []
for i, to_expand in self.to_include:
if to_expand:
if filtered:
filtered += children[i].children
else: # Optimize for left-recursion
filtered = children[i].children
else:
filtered.append(children[i])
return self.node_builder(filtered)
def _should_expand(sym):
return not sym.is_term and sym.name.startswith('_')
def maybe_create_child_filter(expansion, keep_all_tokens, ambiguous):
to_include = [(i, _should_expand(sym)) for i, sym in enumerate(expansion)
if keep_all_tokens or not (sym.is_term and sym.filter_out)]
if len(to_include) < len(expansion) or any(to_expand for i, to_expand in to_include):
return partial(ChildFilter if ambiguous else ChildFilterLALR, to_include)
class Callback(object):
pass
def ptb_inline_args(func):
@wraps(func)
def f(children):
return func(*children)
return f
class ParseTreeBuilder:
def __init__(self, rules, tree_class, propagate_positions=False, keep_all_tokens=False, ambiguous=False):
self.tree_class = tree_class
self.propagate_positions = propagate_positions
self.always_keep_all_tokens = keep_all_tokens
self.ambiguous = ambiguous
self.rule_builders = list(self._init_builders(rules))
self.user_aliases = {}
def _init_builders(self, rules):
for rule in rules:
options = rule.options
keep_all_tokens = self.always_keep_all_tokens or (options.keep_all_tokens if options else False)
expand_single_child = options.expand1 if options else False
wrapper_chain = filter(None, [
(expand_single_child and not rule.alias) and ExpandSingleChild,
maybe_create_child_filter(rule.expansion, keep_all_tokens, self.ambiguous),
self.propagate_positions and PropagatePositions,
])
yield rule, wrapper_chain
def create_callback(self, transformer=None):
callback = Callback()
i = 0
for rule, wrapper_chain in self.rule_builders:
internal_callback_name = '_cb%d_%s' % (i, rule.origin)
i += 1
user_callback_name = rule.alias or rule.origin.name
try:
f = getattr(transformer, user_callback_name)
assert not getattr(f, 'meta', False), "Meta args not supported for internal transformer"
# XXX InlineTransformer is deprecated!
if getattr(f, 'inline', False) or isinstance(transformer, InlineTransformer):
f = ptb_inline_args(f)
except AttributeError:
f = partial(self.tree_class, user_callback_name)
self.user_aliases[rule] = rule.alias
rule.alias = internal_callback_name
for w in wrapper_chain:
f = w(f)
if hasattr(callback, internal_callback_name):
raise GrammarError("Rule '%s' already exists" % (rule,))
setattr(callback, internal_callback_name, f)
return callback
###}

View File

@@ -0,0 +1,192 @@
import re
from functools import partial
from .utils import get_regexp_width
from .parsers.grammar_analysis import GrammarAnalyzer
from .lexer import TraditionalLexer, ContextualLexer, Lexer, Token
from .parsers import lalr_parser, earley, xearley, resolve_ambig, cyk
from .tree import Tree
class WithLexer:
lexer = None
parser = None
lexer_conf = None
def init_traditional_lexer(self, lexer_conf):
self.lexer_conf = lexer_conf
self.lexer = TraditionalLexer(lexer_conf.tokens, ignore=lexer_conf.ignore, user_callbacks=lexer_conf.callbacks)
def init_contextual_lexer(self, lexer_conf):
self.lexer_conf = lexer_conf
states = {idx:list(t.keys()) for idx, t in self.parser._parse_table.states.items()}
always_accept = lexer_conf.postlex.always_accept if lexer_conf.postlex else ()
self.lexer = ContextualLexer(lexer_conf.tokens, states,
ignore=lexer_conf.ignore,
always_accept=always_accept,
user_callbacks=lexer_conf.callbacks)
def lex(self, text):
stream = self.lexer.lex(text)
if self.lexer_conf.postlex:
return self.lexer_conf.postlex.process(stream)
return stream
def parse(self, text):
token_stream = self.lex(text)
sps = self.lexer.set_parser_state
return self.parser.parse(token_stream, *[sps] if sps is not NotImplemented else [])
class LALR_TraditionalLexer(WithLexer):
def __init__(self, lexer_conf, parser_conf, options=None):
self.parser = lalr_parser.Parser(parser_conf)
self.init_traditional_lexer(lexer_conf)
class LALR_ContextualLexer(WithLexer):
def __init__(self, lexer_conf, parser_conf, options=None):
self.parser = lalr_parser.Parser(parser_conf)
self.init_contextual_lexer(lexer_conf)
class LALR_CustomLexer(WithLexer):
def __init__(self, lexer_cls, lexer_conf, parser_conf, options=None):
self.parser = lalr_parser.Parser(parser_conf)
self.lexer_conf = lexer_conf
self.lexer = lexer_cls(lexer_conf)
def get_ambiguity_resolver(options):
if not options or options.ambiguity == 'resolve':
return resolve_ambig.standard_resolve_ambig
elif options.ambiguity == 'resolve__antiscore_sum':
return resolve_ambig.antiscore_sum_resolve_ambig
elif options.ambiguity == 'explicit':
return None
raise ValueError(options)
def tokenize_text(text):
line = 1
col_start_pos = 0
for i, ch in enumerate(text):
if '\n' in ch:
line += ch.count('\n')
col_start_pos = i + ch.rindex('\n')
yield Token('CHAR', ch, line=line, column=i - col_start_pos)
class Earley(WithLexer):
def __init__(self, lexer_conf, parser_conf, options=None):
self.init_traditional_lexer(lexer_conf)
self.parser = earley.Parser(parser_conf, self.match,
resolve_ambiguity=get_ambiguity_resolver(options))
def match(self, term, token):
return term.name == token.type
class XEarley:
def __init__(self, lexer_conf, parser_conf, options=None, **kw):
self.token_by_name = {t.name:t for t in lexer_conf.tokens}
self._prepare_match(lexer_conf)
self.parser = xearley.Parser(parser_conf,
self.match,
resolve_ambiguity=get_ambiguity_resolver(options),
ignore=lexer_conf.ignore,
predict_all=options.earley__predict_all,
**kw
)
def match(self, term, text, index=0):
return self.regexps[term.name].match(text, index)
def _prepare_match(self, lexer_conf):
self.regexps = {}
for t in lexer_conf.tokens:
regexp = t.pattern.to_regexp()
try:
width = get_regexp_width(regexp)[0]
except ValueError:
raise ValueError("Bad regexp in token %s: %s" % (t.name, regexp))
else:
if width == 0:
raise ValueError("Dynamic Earley doesn't allow zero-width regexps", t)
self.regexps[t.name] = re.compile(regexp)
def parse(self, text):
return self.parser.parse(text)
class XEarley_CompleteLex(XEarley):
def __init__(self, *args, **kw):
super(self).__init__(*args, complete_lex=True, **kw)
class CYK(WithLexer):
def __init__(self, lexer_conf, parser_conf, options=None):
self.init_traditional_lexer(lexer_conf)
self._analysis = GrammarAnalyzer(parser_conf)
self._parser = cyk.Parser(parser_conf.rules, parser_conf.start)
self._postprocess = {}
for rule in parser_conf.rules:
a = rule.alias
self._postprocess[a] = a if callable(a) else (a and getattr(parser_conf.callback, a))
def parse(self, text):
tokens = list(self.lex(text))
parse = self._parser.parse(tokens)
parse = self._transform(parse)
return parse
def _transform(self, tree):
subtrees = list(tree.iter_subtrees())
for subtree in subtrees:
subtree.children = [self._apply_callback(c) if isinstance(c, Tree) else c for c in subtree.children]
return self._apply_callback(tree)
def _apply_callback(self, tree):
children = tree.children
callback = self._postprocess[tree.rule.alias]
assert callback, tree.rule.alias
r = callback(children)
return r
def get_frontend(parser, lexer):
if parser=='lalr':
if lexer is None:
raise ValueError('The LALR parser requires use of a lexer')
elif lexer == 'standard':
return LALR_TraditionalLexer
elif lexer == 'contextual':
return LALR_ContextualLexer
elif issubclass(lexer, Lexer):
return partial(LALR_CustomLexer, lexer)
else:
raise ValueError('Unknown lexer: %s' % lexer)
elif parser=='earley':
if lexer=='standard':
return Earley
elif lexer=='dynamic':
return XEarley
elif lexer=='dynamic_complete':
return XEarley_CompleteLex
elif lexer=='contextual':
raise ValueError('The Earley parser does not support the contextual parser')
else:
raise ValueError('Unknown lexer: %s' % lexer)
elif parser == 'cyk':
if lexer == 'standard':
return CYK
else:
raise ValueError('CYK parser requires using standard parser.')
else:
raise ValueError('Unknown parser: %s' % parser)

View File

@@ -0,0 +1,98 @@
"""This module implements an ABLALR(1) Parser
Ambiguous Backtracking Look-Ahead(1) LR(0)
"""
# Author: Erez Shinan (2017)
# Email : erezshin@gmail.com
from ..common import UnexpectedToken
from .lalr_analysis import LALR_Analyzer, Shift
class Parser:
def __init__(self, parser_conf):
assert all(r.options is None or r.options.priority is None
for r in parser_conf.rules), "LALR doesn't yet support prioritization"
self.analysis = analysis = LALR_Analyzer(parser_conf)
analysis.compute_lookahead()
callbacks = {rule: getattr(parser_conf.callback, rule.alias or rule.origin, None)
for rule in analysis.rules}
self.parser_conf = parser_conf
self.parser = _Parser(analysis.parse_table, callbacks)
self.parse = self.parser.parse
###{standalone
class _Parser:
def __init__(self, parse_table, callbacks):
self.states = parse_table.states
self.start_state = parse_table.start_state
self.end_state = parse_table.end_state
self.callbacks = callbacks
def parse(self, seq, set_state=None):
i = 0
token = None
stream = iter(seq)
states = self.states
state_stack = [self.start_state]
value_stack = []
if set_state: set_state(self.start_state)
def get_action(key):
state = state_stack[-1]
try:
return states[state][key]
except KeyError:
expected = states[state].keys()
raise UnexpectedToken(token, expected, seq, i)
def reduce(rule):
size = len(rule.expansion)
if size:
s = value_stack[-size:]
del state_stack[-size:]
del value_stack[-size:]
else:
s = []
value = self.callbacks[rule](s)
_action, new_state = get_action(rule.origin)
assert _action is Shift
state_stack.append(new_state)
value_stack.append(value)
# Main LALR-parser loop
try:
token = next(stream)
i += 1
while True:
action, arg = get_action(token.type)
assert arg != self.end_state
if action is Shift:
state_stack.append(arg)
value_stack.append(token)
if set_state: set_state(arg)
token = next(stream)
i += 1
else:
reduce(arg)
except StopIteration:
pass
while True:
_action, arg = get_action('$END')
if _action is Shift:
assert arg == self.end_state
val ,= value_stack
return val
else:
reduce(arg)
###}

View File

@@ -0,0 +1,343 @@
"""This module implements a CYK parser."""
# Author: https://github.com/ehudt (2018)
#
# Adapted by Erez
from collections import defaultdict
import itertools
from ..exceptions import ParseError
from ..lexer import Token
from ..tree import Tree
from ..grammar import Terminal as T, NonTerminal as NT, Symbol
try:
xrange
except NameError:
xrange = range
def match(t, s):
assert isinstance(t, T)
return t.name == s.type
class Rule(object):
"""Context-free grammar rule."""
def __init__(self, lhs, rhs, weight, alias):
super(Rule, self).__init__()
assert isinstance(lhs, NT), lhs
assert all(isinstance(x, NT) or isinstance(x, T) for x in rhs), rhs
self.lhs = lhs
self.rhs = rhs
self.weight = weight
self.alias = alias
def __str__(self):
return '%s -> %s' % (str(self.lhs), ' '.join(str(x) for x in self.rhs))
def __repr__(self):
return str(self)
def __hash__(self):
return hash((self.lhs, tuple(self.rhs)))
def __eq__(self, other):
return self.lhs == other.lhs and self.rhs == other.rhs
def __ne__(self, other):
return not (self == other)
class Grammar(object):
"""Context-free grammar."""
def __init__(self, rules):
self.rules = frozenset(rules)
def __eq__(self, other):
return self.rules == other.rules
def __str__(self):
return '\n' + '\n'.join(sorted(repr(x) for x in self.rules)) + '\n'
def __repr__(self):
return str(self)
# Parse tree data structures
class RuleNode(object):
"""A node in the parse tree, which also contains the full rhs rule."""
def __init__(self, rule, children, weight=0):
self.rule = rule
self.children = children
self.weight = weight
def __repr__(self):
return 'RuleNode(%s, [%s])' % (repr(self.rule.lhs), ', '.join(str(x) for x in self.children))
class Parser(object):
"""Parser wrapper."""
def __init__(self, rules, start):
super(Parser, self).__init__()
self.orig_rules = {rule.alias: rule for rule in rules}
rules = [self._to_rule(rule) for rule in rules]
self.grammar = to_cnf(Grammar(rules))
self.start = NT(start)
def _to_rule(self, lark_rule):
"""Converts a lark rule, (lhs, rhs, callback, options), to a Rule."""
assert isinstance(lark_rule.origin, NT)
assert all(isinstance(x, Symbol) for x in lark_rule.expansion)
return Rule(
lark_rule.origin, lark_rule.expansion,
weight=lark_rule.options.priority if lark_rule.options and lark_rule.options.priority else 0,
alias=lark_rule.alias)
def parse(self, tokenized): # pylint: disable=invalid-name
"""Parses input, which is a list of tokens."""
table, trees = _parse(tokenized, self.grammar)
# Check if the parse succeeded.
if all(r.lhs != self.start for r in table[(0, len(tokenized) - 1)]):
raise ParseError('Parsing failed.')
parse = trees[(0, len(tokenized) - 1)][self.start]
return self._to_tree(revert_cnf(parse))
def _to_tree(self, rule_node):
"""Converts a RuleNode parse tree to a lark Tree."""
orig_rule = self.orig_rules[rule_node.rule.alias]
children = []
for child in rule_node.children:
if isinstance(child, RuleNode):
children.append(self._to_tree(child))
else:
assert isinstance(child.name, Token)
children.append(child.name)
t = Tree(orig_rule.origin, children)
t.rule=orig_rule
return t
def print_parse(node, indent=0):
if isinstance(node, RuleNode):
print(' ' * (indent * 2) + str(node.rule.lhs))
for child in node.children:
print_parse(child, indent + 1)
else:
print(' ' * (indent * 2) + str(node.s))
def _parse(s, g):
"""Parses sentence 's' using CNF grammar 'g'."""
# The CYK table. Indexed with a 2-tuple: (start pos, end pos)
table = defaultdict(set)
# Top-level structure is similar to the CYK table. Each cell is a dict from
# rule name to the best (lightest) tree for that rule.
trees = defaultdict(dict)
# Populate base case with existing terminal production rules
for i, w in enumerate(s):
for terminal, rules in g.terminal_rules.items():
if match(terminal, w):
for rule in rules:
table[(i, i)].add(rule)
if (rule.lhs not in trees[(i, i)] or
rule.weight < trees[(i, i)][rule.lhs].weight):
trees[(i, i)][rule.lhs] = RuleNode(rule, [T(w)], weight=rule.weight)
# Iterate over lengths of sub-sentences
for l in xrange(2, len(s) + 1):
# Iterate over sub-sentences with the given length
for i in xrange(len(s) - l + 1):
# Choose partition of the sub-sentence in [1, l)
for p in xrange(i + 1, i + l):
span1 = (i, p - 1)
span2 = (p, i + l - 1)
for r1, r2 in itertools.product(table[span1], table[span2]):
for rule in g.nonterminal_rules.get((r1.lhs, r2.lhs), []):
table[(i, i + l - 1)].add(rule)
r1_tree = trees[span1][r1.lhs]
r2_tree = trees[span2][r2.lhs]
rule_total_weight = rule.weight + r1_tree.weight + r2_tree.weight
if (rule.lhs not in trees[(i, i + l - 1)]
or rule_total_weight < trees[(i, i + l - 1)][rule.lhs].weight):
trees[(i, i + l - 1)][rule.lhs] = RuleNode(rule, [r1_tree, r2_tree], weight=rule_total_weight)
return table, trees
# This section implements context-free grammar converter to Chomsky normal form.
# It also implements a conversion of parse trees from its CNF to the original
# grammar.
# Overview:
# Applies the following operations in this order:
# * TERM: Eliminates non-solitary terminals from all rules
# * BIN: Eliminates rules with more than 2 symbols on their right-hand-side.
# * UNIT: Eliminates non-terminal unit rules
#
# The following grammar characteristics aren't featured:
# * Start symbol appears on RHS
# * Empty rules (epsilon rules)
class CnfWrapper(object):
"""CNF wrapper for grammar.
Validates that the input grammar is CNF and provides helper data structures.
"""
def __init__(self, grammar):
super(CnfWrapper, self).__init__()
self.grammar = grammar
self.rules = grammar.rules
self.terminal_rules = defaultdict(list)
self.nonterminal_rules = defaultdict(list)
for r in self.rules:
# Validate that the grammar is CNF and populate auxiliary data structures.
assert isinstance(r.lhs, NT), r
if len(r.rhs) not in [1, 2]:
raise ParseError("CYK doesn't support empty rules")
if len(r.rhs) == 1 and isinstance(r.rhs[0], T):
self.terminal_rules[r.rhs[0]].append(r)
elif len(r.rhs) == 2 and all(isinstance(x, NT) for x in r.rhs):
self.nonterminal_rules[tuple(r.rhs)].append(r)
else:
assert False, r
def __eq__(self, other):
return self.grammar == other.grammar
def __repr__(self):
return repr(self.grammar)
class UnitSkipRule(Rule):
"""A rule that records NTs that were skipped during transformation."""
def __init__(self, lhs, rhs, skipped_rules, weight, alias):
super(UnitSkipRule, self).__init__(lhs, rhs, weight, alias)
self.skipped_rules = skipped_rules
def __eq__(self, other):
return isinstance(other, type(self)) and self.skipped_rules == other.skipped_rules
__hash__ = Rule.__hash__
def build_unit_skiprule(unit_rule, target_rule):
skipped_rules = []
if isinstance(unit_rule, UnitSkipRule):
skipped_rules += unit_rule.skipped_rules
skipped_rules.append(target_rule)
if isinstance(target_rule, UnitSkipRule):
skipped_rules += target_rule.skipped_rules
return UnitSkipRule(unit_rule.lhs, target_rule.rhs, skipped_rules,
weight=unit_rule.weight + target_rule.weight, alias=unit_rule.alias)
def get_any_nt_unit_rule(g):
"""Returns a non-terminal unit rule from 'g', or None if there is none."""
for rule in g.rules:
if len(rule.rhs) == 1 and isinstance(rule.rhs[0], NT):
return rule
return None
def _remove_unit_rule(g, rule):
"""Removes 'rule' from 'g' without changing the langugage produced by 'g'."""
new_rules = [x for x in g.rules if x != rule]
refs = [x for x in g.rules if x.lhs == rule.rhs[0]]
new_rules += [build_unit_skiprule(rule, ref) for ref in refs]
return Grammar(new_rules)
def _split(rule):
"""Splits a rule whose len(rhs) > 2 into shorter rules."""
rule_str = str(rule.lhs) + '__' + '_'.join(str(x) for x in rule.rhs)
rule_name = '__SP_%s' % (rule_str) + '_%d'
yield Rule(rule.lhs, [rule.rhs[0], NT(rule_name % 1)], weight=rule.weight, alias=rule.alias)
for i in xrange(1, len(rule.rhs) - 2):
yield Rule(NT(rule_name % i), [rule.rhs[i], NT(rule_name % (i + 1))], weight=0, alias='Split')
yield Rule(NT(rule_name % (len(rule.rhs) - 2)), rule.rhs[-2:], weight=0, alias='Split')
def _term(g):
"""Applies the TERM rule on 'g' (see top comment)."""
all_t = {x for rule in g.rules for x in rule.rhs if isinstance(x, T)}
t_rules = {t: Rule(NT('__T_%s' % str(t)), [t], weight=0, alias='Term') for t in all_t}
new_rules = []
for rule in g.rules:
if len(rule.rhs) > 1 and any(isinstance(x, T) for x in rule.rhs):
new_rhs = [t_rules[x].lhs if isinstance(x, T) else x for x in rule.rhs]
new_rules.append(Rule(rule.lhs, new_rhs, weight=rule.weight, alias=rule.alias))
new_rules.extend(v for k, v in t_rules.items() if k in rule.rhs)
else:
new_rules.append(rule)
return Grammar(new_rules)
def _bin(g):
"""Applies the BIN rule to 'g' (see top comment)."""
new_rules = []
for rule in g.rules:
if len(rule.rhs) > 2:
new_rules += _split(rule)
else:
new_rules.append(rule)
return Grammar(new_rules)
def _unit(g):
"""Applies the UNIT rule to 'g' (see top comment)."""
nt_unit_rule = get_any_nt_unit_rule(g)
while nt_unit_rule:
g = _remove_unit_rule(g, nt_unit_rule)
nt_unit_rule = get_any_nt_unit_rule(g)
return g
def to_cnf(g):
"""Creates a CNF grammar from a general context-free grammar 'g'."""
g = _unit(_bin(_term(g)))
return CnfWrapper(g)
def unroll_unit_skiprule(lhs, orig_rhs, skipped_rules, children, weight, alias):
if not skipped_rules:
return RuleNode(Rule(lhs, orig_rhs, weight=weight, alias=alias), children, weight=weight)
else:
weight = weight - skipped_rules[0].weight
return RuleNode(
Rule(lhs, [skipped_rules[0].lhs], weight=weight, alias=alias), [
unroll_unit_skiprule(skipped_rules[0].lhs, orig_rhs,
skipped_rules[1:], children,
skipped_rules[0].weight, skipped_rules[0].alias)
], weight=weight)
def revert_cnf(node):
"""Reverts a parse tree (RuleNode) to its original non-CNF form (Node)."""
if isinstance(node, T):
return node
# Reverts TERM rule.
if node.rule.lhs.name.startswith('__T_'):
return node.children[0]
else:
children = []
for child in map(revert_cnf, node.children):
# Reverts BIN rule.
if isinstance(child, RuleNode) and child.rule.lhs.name.startswith('__SP_'):
children += child.children
else:
children.append(child)
# Reverts UNIT rule.
if isinstance(node.rule, UnitSkipRule):
return unroll_unit_skiprule(node.rule.lhs, node.rule.rhs,
node.rule.skipped_rules, children,
node.rule.weight, node.rule.alias)
else:
return RuleNode(node.rule, children)

View File

@@ -0,0 +1,239 @@
"This module implements an Earley Parser"
# The parser uses a parse-forest to keep track of derivations and ambiguations.
# When the parse ends successfully, a disambiguation stage resolves all ambiguity
# (right now ambiguity resolution is not developed beyond the needs of lark)
# Afterwards the parse tree is reduced (transformed) according to user callbacks.
# I use the no-recursion version of Transformer, because the tree might be
# deeper than Python's recursion limit (a bit absurd, but that's life)
#
# The algorithm keeps track of each state set, using a corresponding Column instance.
# Column keeps track of new items using NewsList instances.
#
# Author: Erez Shinan (2017)
# Email : erezshin@gmail.com
from ..tree import Tree
from ..visitors import Transformer_InPlace, v_args
from ..exceptions import ParseError, UnexpectedToken
from .grammar_analysis import GrammarAnalyzer
from ..grammar import NonTerminal
class Derivation(Tree):
def __init__(self, rule, items=None):
Tree.__init__(self, 'drv', items or [])
self.meta.rule = rule
self._hash = None
def _pretty_label(self): # Nicer pretty for debugging the parser
return self.meta.rule.origin.name if self.meta.rule else self.data
def __hash__(self):
if self._hash is None:
self._hash = Tree.__hash__(self)
return self._hash
class Item(object):
"An Earley Item, the atom of the algorithm."
def __init__(self, rule, ptr, start, tree):
self.rule = rule
self.ptr = ptr
self.start = start
self.tree = tree if tree is not None else Derivation(self.rule)
@property
def expect(self):
return self.rule.expansion[self.ptr]
@property
def is_complete(self):
return self.ptr == len(self.rule.expansion)
def advance(self, tree):
assert self.tree.data == 'drv'
new_tree = Derivation(self.rule, self.tree.children + [tree])
return self.__class__(self.rule, self.ptr+1, self.start, new_tree)
def __eq__(self, other):
return self.start is other.start and self.ptr == other.ptr and self.rule == other.rule
def __hash__(self):
return hash((self.rule, self.ptr, id(self.start))) # Always runs Derivation.__hash__
def __repr__(self):
before = list(map(str, self.rule.expansion[:self.ptr]))
after = list(map(str, self.rule.expansion[self.ptr:]))
return '<(%d) %s : %s * %s>' % (id(self.start), self.rule.origin, ' '.join(before), ' '.join(after))
class NewsList(list):
"Keeps track of newly added items (append-only)"
def __init__(self, initial=None):
list.__init__(self, initial or [])
self.last_iter = 0
def get_news(self):
i = self.last_iter
self.last_iter = len(self)
return self[i:]
class Column:
"An entry in the table, aka Earley Chart. Contains lists of items."
def __init__(self, i, FIRST, predict_all=False):
self.i = i
self.to_reduce = NewsList()
self.to_predict = NewsList()
self.to_scan = []
self.item_count = 0
self.FIRST = FIRST
self.predicted = set()
self.completed = {}
self.predict_all = predict_all
def add(self, items):
"""Sort items into scan/predict/reduce newslists
Makes sure only unique items are added.
"""
for item in items:
item_key = item, item.tree # Elsewhere, tree is not part of the comparison
if item.is_complete:
# XXX Potential bug: What happens if there's ambiguity in an empty rule?
if item.rule.expansion and item_key in self.completed:
old_tree = self.completed[item_key].tree
if old_tree == item.tree:
is_empty = not self.FIRST[item.rule.origin]
if not is_empty:
continue
if old_tree.data != '_ambig':
new_tree = old_tree.copy()
new_tree.meta.rule = old_tree.meta.rule
old_tree.set('_ambig', [new_tree])
old_tree.meta.rule = None # No longer a 'drv' node
if item.tree.children[0] is old_tree: # XXX a little hacky!
raise ParseError("Infinite recursion in grammar! (Rule %s)" % item.rule)
if item.tree not in old_tree.children:
old_tree.children.append(item.tree)
# old_tree.children.append(item.tree)
else:
self.completed[item_key] = item
self.to_reduce.append(item)
else:
if item.expect.is_term:
self.to_scan.append(item)
else:
k = item_key if self.predict_all else item
if k in self.predicted:
continue
self.predicted.add(k)
self.to_predict.append(item)
self.item_count += 1 # Only count if actually added
def __bool__(self):
return bool(self.item_count)
__nonzero__ = __bool__ # Py2 backwards-compatibility
class Parser:
def __init__(self, parser_conf, term_matcher, resolve_ambiguity=None):
analysis = GrammarAnalyzer(parser_conf)
self.parser_conf = parser_conf
self.resolve_ambiguity = resolve_ambiguity
self.FIRST = analysis.FIRST
self.postprocess = {}
self.predictions = {}
for rule in parser_conf.rules:
self.postprocess[rule] = rule.alias if callable(rule.alias) else getattr(parser_conf.callback, rule.alias)
self.predictions[rule.origin] = [x.rule for x in analysis.expand_rule(rule.origin)]
self.term_matcher = term_matcher
def parse(self, stream, start_symbol=None):
# Define parser functions
start_symbol = NonTerminal(start_symbol or self.parser_conf.start)
_Item = Item
match = self.term_matcher
def predict(nonterm, column):
assert not nonterm.is_term, nonterm
return [_Item(rule, 0, column, None) for rule in self.predictions[nonterm]]
def complete(item):
name = item.rule.origin
return [i.advance(item.tree) for i in item.start.to_predict if i.expect == name]
def predict_and_complete(column):
while True:
to_predict = {x.expect for x in column.to_predict.get_news()
if x.ptr} # if not part of an already predicted batch
to_reduce = set(column.to_reduce.get_news())
if not (to_predict or to_reduce):
break
for nonterm in to_predict:
column.add( predict(nonterm, column) )
for item in to_reduce:
new_items = list(complete(item))
if item in new_items:
raise ParseError('Infinite recursion detected! (rule %s)' % item.rule)
column.add(new_items)
def scan(i, token, column):
next_set = Column(i, self.FIRST)
next_set.add(item.advance(token) for item in column.to_scan if match(item.expect, token))
if not next_set:
expect = {i.expect.name for i in column.to_scan}
raise UnexpectedToken(token, expect, considered_rules=set(column.to_scan))
return next_set
# Main loop starts
column0 = Column(0, self.FIRST)
column0.add(predict(start_symbol, column0))
column = column0
for i, token in enumerate(stream):
predict_and_complete(column)
column = scan(i, token, column)
predict_and_complete(column)
# Parse ended. Now build a parse tree
solutions = [n.tree for n in column.to_reduce
if n.rule.origin==start_symbol and n.start is column0]
if not solutions:
raise ParseError('Incomplete parse: Could not find a solution to input')
elif len(solutions) == 1:
tree = solutions[0]
else:
tree = Tree('_ambig', solutions)
if self.resolve_ambiguity:
tree = self.resolve_ambiguity(tree)
return ApplyCallbacks(self.postprocess).transform(tree)
class ApplyCallbacks(Transformer_InPlace):
def __init__(self, postprocess):
self.postprocess = postprocess
@v_args(meta=True)
def drv(self, children, meta):
return self.postprocess[meta.rule](children)

View File

@@ -0,0 +1,149 @@
from ..utils import bfs, fzset, classify
from ..exceptions import GrammarError
from ..grammar import Rule, Terminal, NonTerminal
class RulePtr(object):
__slots__ = ('rule', 'index')
def __init__(self, rule, index):
assert isinstance(rule, Rule)
assert index <= len(rule.expansion)
self.rule = rule
self.index = index
def __repr__(self):
before = self.rule.expansion[:self.index]
after = self.rule.expansion[self.index:]
return '<%s : %s * %s>' % (self.rule.origin, ' '.join(before), ' '.join(after))
@property
def next(self):
return self.rule.expansion[self.index]
def advance(self, sym):
assert self.next == sym
return RulePtr(self.rule, self.index+1)
@property
def is_satisfied(self):
return self.index == len(self.rule.expansion)
def __eq__(self, other):
return self.rule == other.rule and self.index == other.index
def __hash__(self):
return hash((self.rule, self.index))
def update_set(set1, set2):
if not set2:
return False
copy = set(set1)
set1 |= set2
return set1 != copy
def calculate_sets(rules):
"""Calculate FOLLOW sets.
Adapted from: http://lara.epfl.ch/w/cc09:algorithm_for_first_and_follow_sets"""
symbols = {sym for rule in rules for sym in rule.expansion} | {rule.origin for rule in rules}
# foreach grammar rule X ::= Y(1) ... Y(k)
# if k=0 or {Y(1),...,Y(k)} subset of NULLABLE then
# NULLABLE = NULLABLE union {X}
# for i = 1 to k
# if i=1 or {Y(1),...,Y(i-1)} subset of NULLABLE then
# FIRST(X) = FIRST(X) union FIRST(Y(i))
# for j = i+1 to k
# if i=k or {Y(i+1),...Y(k)} subset of NULLABLE then
# FOLLOW(Y(i)) = FOLLOW(Y(i)) union FOLLOW(X)
# if i+1=j or {Y(i+1),...,Y(j-1)} subset of NULLABLE then
# FOLLOW(Y(i)) = FOLLOW(Y(i)) union FIRST(Y(j))
# until none of NULLABLE,FIRST,FOLLOW changed in last iteration
NULLABLE = set()
FIRST = {}
FOLLOW = {}
for sym in symbols:
FIRST[sym]={sym} if sym.is_term else set()
FOLLOW[sym]=set()
# Calculate NULLABLE and FIRST
changed = True
while changed:
changed = False
for rule in rules:
if set(rule.expansion) <= NULLABLE:
if update_set(NULLABLE, {rule.origin}):
changed = True
for i, sym in enumerate(rule.expansion):
if set(rule.expansion[:i]) <= NULLABLE:
if update_set(FIRST[rule.origin], FIRST[sym]):
changed = True
# Calculate FOLLOW
changed = True
while changed:
changed = False
for rule in rules:
for i, sym in enumerate(rule.expansion):
if i==len(rule.expansion)-1 or set(rule.expansion[i+1:]) <= NULLABLE:
if update_set(FOLLOW[sym], FOLLOW[rule.origin]):
changed = True
for j in range(i+1, len(rule.expansion)):
if set(rule.expansion[i+1:j]) <= NULLABLE:
if update_set(FOLLOW[sym], FIRST[rule.expansion[j]]):
changed = True
return FIRST, FOLLOW, NULLABLE
class GrammarAnalyzer(object):
def __init__(self, parser_conf, debug=False):
self.debug = debug
rules = parser_conf.rules + [Rule(NonTerminal('$root'), [NonTerminal(parser_conf.start), Terminal('$END')])]
self.rules_by_origin = classify(rules, lambda r: r.origin)
assert len(rules) == len(set(rules))
for r in rules:
for sym in r.expansion:
if not (sym.is_term or sym in self.rules_by_origin):
raise GrammarError("Using an undefined rule: %s" % sym) # TODO test validation
self.start_state = self.expand_rule(NonTerminal('$root'))
self.FIRST, self.FOLLOW, self.NULLABLE = calculate_sets(rules)
def expand_rule(self, rule):
"Returns all init_ptrs accessible by rule (recursive)"
init_ptrs = set()
def _expand_rule(rule):
assert not rule.is_term, rule
for r in self.rules_by_origin[rule]:
init_ptr = RulePtr(r, 0)
init_ptrs.add(init_ptr)
if r.expansion: # if not empty rule
new_r = init_ptr.next
if not new_r.is_term:
yield new_r
for _ in bfs([rule], _expand_rule):
pass
return fzset(init_ptrs)
def _first(self, r):
if r.is_term:
return {r}
else:
return {rp.next for rp in self.expand_rule(r) if rp.next.is_term}

View File

@@ -0,0 +1,109 @@
"""This module builds a LALR(1) transition-table for lalr_parser.py
For now, shift/reduce conflicts are automatically resolved as shifts.
"""
# Author: Erez Shinan (2017)
# Email : erezshin@gmail.com
import logging
from collections import defaultdict
from ..utils import classify, classify_bool, bfs, fzset
from ..exceptions import GrammarError
from .grammar_analysis import GrammarAnalyzer, Terminal
class Action:
def __init__(self, name):
self.name = name
def __str__(self):
return self.name
def __repr__(self):
return str(self)
Shift = Action('Shift')
Reduce = Action('Reduce')
class ParseTable:
def __init__(self, states, start_state, end_state):
self.states = states
self.start_state = start_state
self.end_state = end_state
class IntParseTable(ParseTable):
@classmethod
def from_ParseTable(cls, parse_table):
enum = list(parse_table.states)
state_to_idx = {s:i for i,s in enumerate(enum)}
int_states = {}
for s, la in parse_table.states.items():
la = {k:(v[0], state_to_idx[v[1]]) if v[0] is Shift else v
for k,v in la.items()}
int_states[ state_to_idx[s] ] = la
start_state = state_to_idx[parse_table.start_state]
end_state = state_to_idx[parse_table.end_state]
return cls(int_states, start_state, end_state)
class LALR_Analyzer(GrammarAnalyzer):
def compute_lookahead(self):
self.end_states = []
self.states = {}
def step(state):
lookahead = defaultdict(list)
sat, unsat = classify_bool(state, lambda rp: rp.is_satisfied)
for rp in sat:
for term in self.FOLLOW.get(rp.rule.origin, ()):
lookahead[term].append((Reduce, rp.rule))
d = classify(unsat, lambda rp: rp.next)
for sym, rps in d.items():
rps = {rp.advance(sym) for rp in rps}
for rp in set(rps):
if not rp.is_satisfied and not rp.next.is_term:
rps |= self.expand_rule(rp.next)
new_state = fzset(rps)
lookahead[sym].append((Shift, new_state))
if sym == Terminal('$END'):
self.end_states.append( new_state )
yield new_state
for k, v in lookahead.items():
if len(v) > 1:
if self.debug:
logging.warn("Shift/reduce conflict for %s: %s. Resolving as shift.", k, v)
for x in v:
# XXX resolving shift/reduce into shift, like PLY
# Give a proper warning
if x[0] is Shift:
lookahead[k] = [x]
for k, v in lookahead.items():
if not len(v) == 1:
raise GrammarError("Collision in %s: %s" %(k, ', '.join(['\n * %s: %s' % x for x in v])))
self.states[state] = {k.name:v[0] for k, v in lookahead.items()}
for _ in bfs([self.start_state], step):
pass
self.end_state ,= self.end_states
self._parse_table = ParseTable(self.states, self.start_state, self.end_state)
if self.debug:
self.parse_table = self._parse_table
else:
self.parse_table = IntParseTable.from_ParseTable(self._parse_table)

View File

@@ -0,0 +1,91 @@
"""This module implements a LALR(1) Parser
"""
# Author: Erez Shinan (2017)
# Email : erezshin@gmail.com
from ..exceptions import UnexpectedToken
from ..lexer import Token
from .lalr_analysis import LALR_Analyzer, Shift
class Parser:
def __init__(self, parser_conf):
assert all(r.options is None or r.options.priority is None
for r in parser_conf.rules), "LALR doesn't yet support prioritization"
analysis = LALR_Analyzer(parser_conf)
analysis.compute_lookahead()
callbacks = {rule: getattr(parser_conf.callback, rule.alias or rule.origin, None)
for rule in parser_conf.rules}
self._parse_table = analysis.parse_table
self.parser_conf = parser_conf
self.parser = _Parser(analysis.parse_table, callbacks)
self.parse = self.parser.parse
###{standalone
class _Parser:
def __init__(self, parse_table, callbacks):
self.states = parse_table.states
self.start_state = parse_table.start_state
self.end_state = parse_table.end_state
self.callbacks = callbacks
def parse(self, seq, set_state=None):
token = None
stream = iter(seq)
states = self.states
state_stack = [self.start_state]
value_stack = []
if set_state: set_state(self.start_state)
def get_action(token):
state = state_stack[-1]
try:
return states[state][token.type]
except KeyError:
expected = [s for s in states[state].keys() if s.isupper()]
raise UnexpectedToken(token, expected, state=state)
def reduce(rule):
size = len(rule.expansion)
if size:
s = value_stack[-size:]
del state_stack[-size:]
del value_stack[-size:]
else:
s = []
value = self.callbacks[rule](s)
_action, new_state = states[state_stack[-1]][rule.origin.name]
assert _action is Shift
state_stack.append(new_state)
value_stack.append(value)
# Main LALR-parser loop
for token in stream:
while True:
action, arg = get_action(token)
assert arg != self.end_state
if action is Shift:
state_stack.append(arg)
value_stack.append(token)
if set_state: set_state(arg)
break # next token
else:
reduce(arg)
token = Token.new_borrow_pos('$END', '', token) if token else Token('$END', '', 0, 1, 1)
while True:
_action, arg = get_action(token)
if _action is Shift:
assert arg == self.end_state
val ,= value_stack
return val
else:
reduce(arg)
###}

View File

@@ -0,0 +1,109 @@
from ..utils import compare
from functools import cmp_to_key
from ..tree import Tree
# Standard ambiguity resolver (uses comparison)
#
# Author: Erez Sh
def _compare_rules(rule1, rule2):
return -compare( len(rule1.expansion), len(rule2.expansion))
def _sum_priority(tree):
p = 0
for n in tree.iter_subtrees():
try:
p += n.meta.rule.options.priority or 0
except AttributeError:
pass
return p
def _compare_priority(tree1, tree2):
tree1.iter_subtrees()
def _compare_drv(tree1, tree2):
try:
rule1 = tree1.meta.rule
except AttributeError:
rule1 = None
try:
rule2 = tree2.meta.rule
except AttributeError:
rule2 = None
if None == rule1 == rule2:
return compare(tree1, tree2)
elif rule1 is None:
return -1
elif rule2 is None:
return 1
assert tree1.data != '_ambig'
assert tree2.data != '_ambig'
p1 = _sum_priority(tree1)
p2 = _sum_priority(tree2)
c = (p1 or p2) and compare(p1, p2)
if c:
return c
c = _compare_rules(tree1.meta.rule, tree2.meta.rule)
if c:
return c
# rules are "equal", so compare trees
if len(tree1.children) == len(tree2.children):
for t1, t2 in zip(tree1.children, tree2.children):
c = _compare_drv(t1, t2)
if c:
return c
return compare(len(tree1.children), len(tree2.children))
def _standard_resolve_ambig(tree):
assert tree.data == '_ambig'
key_f = cmp_to_key(_compare_drv)
best = max(tree.children, key=key_f)
assert best.data == 'drv'
tree.set('drv', best.children)
tree.meta.rule = best.meta.rule # needed for applying callbacks
def standard_resolve_ambig(tree):
for ambig in tree.find_data('_ambig'):
_standard_resolve_ambig(ambig)
return tree
# Anti-score Sum
#
# Author: Uriva (https://github.com/uriva)
def _antiscore_sum_drv(tree):
if not isinstance(tree, Tree):
return 0
assert tree.data != '_ambig'
return _sum_priority(tree)
def _antiscore_sum_resolve_ambig(tree):
assert tree.data == '_ambig'
best = min(tree.children, key=_antiscore_sum_drv)
assert best.data == 'drv'
tree.set('drv', best.children)
tree.meta.rule = best.meta.rule # needed for applying callbacks
def antiscore_sum_resolve_ambig(tree):
for ambig in tree.find_data('_ambig'):
_antiscore_sum_resolve_ambig(ambig)
return tree

View File

@@ -0,0 +1,158 @@
"This module implements an experimental Earley Parser with a dynamic lexer"
# The parser uses a parse-forest to keep track of derivations and ambiguations.
# When the parse ends successfully, a disambiguation stage resolves all ambiguity
# (right now ambiguity resolution is not developed beyond the needs of lark)
# Afterwards the parse tree is reduced (transformed) according to user callbacks.
# I use the no-recursion version of Transformer and Visitor, because the tree might be
# deeper than Python's recursion limit (a bit absurd, but that's life)
#
# The algorithm keeps track of each state set, using a corresponding Column instance.
# Column keeps track of new items using NewsList instances.
#
# Instead of running a lexer beforehand, or using a costy char-by-char method, this parser
# uses regular expressions by necessity, achieving high-performance while maintaining all of
# Earley's power in parsing any CFG.
#
#
# Author: Erez Shinan (2017)
# Email : erezshin@gmail.com
from collections import defaultdict
from ..exceptions import ParseError, UnexpectedCharacters
from ..lexer import Token
from ..tree import Tree
from .grammar_analysis import GrammarAnalyzer
from ..grammar import NonTerminal, Terminal
from .earley import ApplyCallbacks, Item, Column
class Parser:
def __init__(self, parser_conf, term_matcher, resolve_ambiguity=None, ignore=(), predict_all=False, complete_lex=False):
self.analysis = GrammarAnalyzer(parser_conf)
self.parser_conf = parser_conf
self.resolve_ambiguity = resolve_ambiguity
self.ignore = [Terminal(t) for t in ignore]
self.predict_all = predict_all
self.complete_lex = complete_lex
self.FIRST = self.analysis.FIRST
self.postprocess = {}
self.predictions = {}
for rule in parser_conf.rules:
self.postprocess[rule] = getattr(parser_conf.callback, rule.alias)
self.predictions[rule.origin] = [x.rule for x in self.analysis.expand_rule(rule.origin)]
self.term_matcher = term_matcher
def parse(self, stream, start_symbol=None):
# Define parser functions
start_symbol = NonTerminal(start_symbol or self.parser_conf.start)
delayed_matches = defaultdict(list)
match = self.term_matcher
text_line = 1
text_column = 1
def predict(nonterm, column):
assert not nonterm.is_term, nonterm
return [Item(rule, 0, column, None) for rule in self.predictions[nonterm]]
def complete(item):
name = item.rule.origin
return [i.advance(item.tree) for i in item.start.to_predict if i.expect == name]
def predict_and_complete(column):
while True:
to_predict = {x.expect for x in column.to_predict.get_news()
if x.ptr} # if not part of an already predicted batch
to_reduce = column.to_reduce.get_news()
if not (to_predict or to_reduce):
break
for nonterm in to_predict:
column.add( predict(nonterm, column) )
for item in to_reduce:
new_items = list(complete(item))
if item in new_items:
raise ParseError('Infinite recursion detected! (rule %s)' % item.rule)
column.add(new_items)
def scan(i, column):
to_scan = column.to_scan
for x in self.ignore:
m = match(x, stream, i)
if m:
delayed_matches[m.end()] += set(to_scan)
delayed_matches[m.end()] += set(column.to_reduce)
# TODO add partial matches for ignore too?
# s = m.group(0)
# for j in range(1, len(s)):
# m = x.match(s[:-j])
# if m:
# delayed_matches[m.end()] += to_scan
for item in to_scan:
m = match(item.expect, stream, i)
if m:
t = Token(item.expect.name, m.group(0), i, text_line, text_column)
delayed_matches[m.end()].append(item.advance(t))
if self.complete_lex:
s = m.group(0)
for j in range(1, len(s)):
m = match(item.expect, s[:-j])
if m:
t = Token(item.expect.name, m.group(0), i, text_line, text_column)
delayed_matches[i+m.end()].append(item.advance(t))
next_set = Column(i+1, self.FIRST, predict_all=self.predict_all)
next_set.add(delayed_matches[i+1])
del delayed_matches[i+1] # No longer needed, so unburden memory
if not next_set and not delayed_matches:
raise UnexpectedCharacters(stream, i, text_line, text_column, {item.expect for item in to_scan}, set(to_scan))
return next_set
# Main loop starts
column0 = Column(0, self.FIRST, predict_all=self.predict_all)
column0.add(predict(start_symbol, column0))
column = column0
for i, token in enumerate(stream):
predict_and_complete(column)
column = scan(i, column)
if token == '\n':
text_line += 1
text_column = 1
else:
text_column += 1
predict_and_complete(column)
# Parse ended. Now build a parse tree
solutions = [n.tree for n in column.to_reduce
if n.rule.origin==start_symbol and n.start is column0]
if not solutions:
expected_tokens = [t.expect for t in column.to_scan]
raise ParseError('Unexpected end of input! Expecting a terminal of: %s' % expected_tokens)
elif len(solutions) == 1:
tree = solutions[0]
else:
tree = Tree('_ambig', solutions)
if self.resolve_ambiguity:
tree = self.resolve_ambiguity(tree)
return ApplyCallbacks(self.postprocess).transform(tree)

View File

@@ -0,0 +1,130 @@
from collections import defaultdict
from .tree import Tree
from .visitors import Transformer_InPlace
from .common import ParserConf
from .lexer import Token, PatternStr
from .parsers import earley, resolve_ambig
from .grammar import Rule, Terminal, NonTerminal
def is_discarded_terminal(t):
return t.is_term and t.filter_out
def is_iter_empty(i):
try:
_ = next(i)
return False
except StopIteration:
return True
class WriteTokensTransformer(Transformer_InPlace):
def __init__(self, tokens):
self.tokens = tokens
def __default__(self, data, children, meta):
# if not isinstance(t, MatchTree):
# return t
if not getattr(meta, 'match_tree', False):
return Tree(data, children)
iter_args = iter(children)
to_write = []
for sym in meta.orig_expansion:
if is_discarded_terminal(sym):
t = self.tokens[sym.name]
assert isinstance(t.pattern, PatternStr)
to_write.append(t.pattern.value)
else:
x = next(iter_args)
if isinstance(x, list):
to_write += x
else:
if isinstance(x, Token):
assert Terminal(x.type) == sym, x
else:
assert NonTerminal(x.data) == sym, (sym, x)
to_write.append(x)
assert is_iter_empty(iter_args)
return to_write
class MatchTree(Tree):
pass
class MakeMatchTree:
def __init__(self, name, expansion):
self.name = name
self.expansion = expansion
def __call__(self, args):
t = MatchTree(self.name, args)
t.meta.match_tree = True
t.meta.orig_expansion = self.expansion
return t
class Reconstructor:
def __init__(self, parser):
# XXX TODO calling compile twice returns different results!
tokens, rules, _grammar_extra = parser.grammar.compile()
self.write_tokens = WriteTokensTransformer({t.name:t for t in tokens})
self.rules = list(self._build_recons_rules(rules))
def _build_recons_rules(self, rules):
expand1s = {r.origin for r in rules if r.options and r.options.expand1}
aliases = defaultdict(list)
for r in rules:
if r.alias:
aliases[r.origin].append( r.alias )
rule_names = {r.origin for r in rules}
nonterminals = {sym for sym in rule_names
if sym.name.startswith('_') or sym in expand1s or sym in aliases }
for r in rules:
recons_exp = [sym if sym in nonterminals else Terminal(sym.name)
for sym in r.expansion if not is_discarded_terminal(sym)]
# Skip self-recursive constructs
if recons_exp == [r.origin]:
continue
sym = NonTerminal(r.alias) if r.alias else r.origin
yield Rule(sym, recons_exp, MakeMatchTree(sym.name, r.expansion))
for origin, rule_aliases in aliases.items():
for alias in rule_aliases:
yield Rule(origin, [Terminal(alias)], MakeMatchTree(origin.name, [NonTerminal(alias)]))
yield Rule(origin, [Terminal(origin.name)], MakeMatchTree(origin.name, [origin]))
def _match(self, term, token):
if isinstance(token, Tree):
return Terminal(token.data) == term
elif isinstance(token, Token):
return term == Terminal(token.type)
assert False
def _reconstruct(self, tree):
# TODO: ambiguity?
parser = earley.Parser(ParserConf(self.rules, None, tree.data), self._match, resolve_ambiguity=resolve_ambig.standard_resolve_ambig)
unreduced_tree = parser.parse(tree.children) # find a full derivation
assert unreduced_tree.data == tree.data
res = self.write_tokens.transform(unreduced_tree)
for item in res:
if isinstance(item, Tree):
for x in self._reconstruct(item):
yield x
else:
yield item
def reconstruct(self, tree):
return ''.join(self._reconstruct(tree))

View File

@@ -0,0 +1,107 @@
import re
from collections import defaultdict
from .tree import Tree
from .common import is_terminal, ParserConf, PatternStr, Terminal
from .lexer import Token
from .parsers import earley
def is_discarded_terminal(t):
return is_terminal(t) and t.startswith('_')
def is_iter_empty(i):
try:
_ = next(i)
return False
except StopIteration:
return True
class Reconstructor:
def __init__(self, parser):
# Recreate the rules to assume a standard lexer
_tokens, rules, _grammar_extra = parser.grammar.compile(lexer='standard', start='whatever')
tokens = {t.name:t for t in _tokens}
token_res = {t.name:re.compile(t.pattern.to_regexp()) for t in _tokens}
class MatchTerminal(Terminal):
def match(self, other):
if isinstance(other, Tree):
return False
return token_res[self.data].match(other) is not None
class MatchTree(Terminal):
def match(self, other):
try:
return self.data == other.data
except AttributeError:
return False
class WriteTokens:
def __init__(self, name, expansion):
self.name = name
self.expansion = expansion
def f(self, args):
args2 = iter(args)
to_write = []
for sym in self.expansion:
if is_discarded_terminal(sym):
t = tokens[sym]
assert isinstance(t.pattern, PatternStr)
to_write.append(t.pattern.value)
else:
x = next(args2)
if isinstance(x, list):
to_write += x
else:
if isinstance(x, Token):
assert x.type == sym, x
else:
assert x.data == sym, x
to_write.append(x)
assert is_iter_empty(args2)
return to_write
d = defaultdict(list)
for name, (expansions, _o) in rules.items():
for expansion, alias in expansions:
if alias:
d[alias].append(expansion)
d[name].append([alias])
else:
d[name].append(expansion)
rules = []
expand1s = {name for name, (_x, options) in parser.rules.items()
if options and options.expand1}
for name, expansions in d.items():
for expansion in expansions:
reduced = [sym if sym.startswith('_') or sym in expand1s else
MatchTerminal(sym) if is_terminal(sym) else MatchTree(sym)
for sym in expansion if not is_discarded_terminal(sym)]
rules.append((name, reduced, WriteTokens(name, expansion).f, None))
self.rules = rules
def _reconstruct(self, tree):
# TODO: ambiguity?
parser = earley.Parser(self.rules, tree.data, {})
res = parser.parse(tree.children)
for item in res:
if isinstance(item, Tree):
for x in self._reconstruct(item):
yield x
else:
yield item
def reconstruct(self, tree):
return ''.join(self._reconstruct(tree))

View File

@@ -0,0 +1,186 @@
"Converts between Lark and Nearley grammars. Work in progress!"
import os.path
import sys
import codecs
from lark import Lark, InlineTransformer
nearley_grammar = r"""
start: (ruledef|directive)+
directive: "@" NAME (STRING|NAME)
| "@" JS -> js_code
ruledef: NAME "->" expansions
| NAME REGEXP "->" expansions -> macro
expansions: expansion ("|" expansion)*
expansion: expr+ js
?expr: item [":" /[+*?]/]
?item: rule|string|regexp
| "(" expansions ")"
rule: NAME
string: STRING
regexp: REGEXP
JS: /{%.*?%}/s
js: JS?
NAME: /[a-zA-Z_$]\w*/
COMMENT: /#[^\n]*/
REGEXP: /\[.*?\]/
STRING: /".*?"/
%import common.WS
%ignore WS
%ignore COMMENT
"""
nearley_grammar_parser = Lark(nearley_grammar, parser='earley', lexer='standard')
def _get_rulename(name):
name = {'_': '_ws_maybe', '__':'_ws'}.get(name, name)
return 'n_' + name.replace('$', '__DOLLAR__').lower()
class NearleyToLark(InlineTransformer):
def __init__(self):
self._count = 0
self.extra_rules = {}
self.extra_rules_rev = {}
self.alias_js_code = {}
def _new_function(self, code):
name = 'alias_%d' % self._count
self._count += 1
self.alias_js_code[name] = code
return name
def _extra_rule(self, rule):
if rule in self.extra_rules_rev:
return self.extra_rules_rev[rule]
name = 'xrule_%d' % len(self.extra_rules)
assert name not in self.extra_rules
self.extra_rules[name] = rule
self.extra_rules_rev[rule] = name
return name
def rule(self, name):
return _get_rulename(name)
def ruledef(self, name, exps):
return '!%s: %s' % (_get_rulename(name), exps)
def expr(self, item, op):
rule = '(%s)%s' % (item, op)
return self._extra_rule(rule)
def regexp(self, r):
return '/%s/' % r
def string(self, s):
return self._extra_rule(s)
def expansion(self, *x):
x, js = x[:-1], x[-1]
if js.children:
js_code ,= js.children
js_code = js_code[2:-2]
alias = '-> ' + self._new_function(js_code)
else:
alias = ''
return ' '.join(x) + alias
def expansions(self, *x):
return '%s' % ('\n |'.join(x))
def start(self, *rules):
return '\n'.join(filter(None, rules))
def _nearley_to_lark(g, builtin_path, n2l, js_code, folder_path, includes):
rule_defs = []
tree = nearley_grammar_parser.parse(g)
for statement in tree.children:
if statement.data == 'directive':
directive, arg = statement.children
if directive in ('builtin', 'include'):
folder = builtin_path if directive == 'builtin' else folder_path
path = os.path.join(folder, arg[1:-1])
if path not in includes:
includes.add(path)
with codecs.open(path, encoding='utf8') as f:
text = f.read()
rule_defs += _nearley_to_lark(text, builtin_path, n2l, js_code, os.path.abspath(os.path.dirname(path)), includes)
else:
assert False, directive
elif statement.data == 'js_code':
code ,= statement.children
code = code[2:-2]
js_code.append(code)
elif statement.data == 'macro':
pass # TODO Add support for macros!
elif statement.data == 'ruledef':
rule_defs.append( n2l.transform(statement) )
else:
raise Exception("Unknown statement: %s" % statement)
return rule_defs
def create_code_for_nearley_grammar(g, start, builtin_path, folder_path):
import js2py
emit_code = []
def emit(x=None):
if x:
emit_code.append(x)
emit_code.append('\n')
js_code = ['function id(x) {return x[0];}']
n2l = NearleyToLark()
rule_defs = _nearley_to_lark(g, builtin_path, n2l, js_code, folder_path, set())
lark_g = '\n'.join(rule_defs)
lark_g += '\n'+'\n'.join('!%s: %s' % item for item in n2l.extra_rules.items())
emit('from lark import Lark, Transformer')
emit()
emit('grammar = ' + repr(lark_g))
emit()
for alias, code in n2l.alias_js_code.items():
js_code.append('%s = (%s);' % (alias, code))
emit(js2py.translate_js('\n'.join(js_code)))
emit('class TransformNearley(Transformer):')
for alias in n2l.alias_js_code:
emit(" %s = var.get('%s').to_python()" % (alias, alias))
emit(" __default__ = lambda self, n, c, m: c if c else None")
emit()
emit('parser = Lark(grammar, start="n_%s")' % start)
emit('def parse(text):')
emit(' return TransformNearley().transform(parser.parse(text))')
return ''.join(emit_code)
def main(fn, start, nearley_lib):
with codecs.open(fn, encoding='utf8') as f:
grammar = f.read()
return create_code_for_nearley_grammar(grammar, start, os.path.join(nearley_lib, 'builtin'), os.path.abspath(os.path.dirname(fn)))
if __name__ == '__main__':
if len(sys.argv) < 4:
print("Reads Nearley grammar (with js functions) outputs an equivalent lark parser.")
print("Usage: %s <nearley_grammar_path> <start_rule> <nearley_lib_path>" % sys.argv[0])
sys.exit(1)
fn, start, nearley_lib = sys.argv[1:]
print(main(fn, start, nearley_lib))

View File

@@ -0,0 +1,251 @@
###{standalone
#
#
# Lark Stand-alone Generator Tool
# ----------------------------------
# Generates a stand-alone LALR(1) parser with a standard lexer
#
# Git: https://github.com/erezsh/lark
# Author: Erez Shinan (erezshin@gmail.com)
#
#
# >>> LICENSE
#
# This tool and its generated code use a separate license from Lark.
#
# It is licensed under GPLv2 or above.
#
# If you wish to purchase a commercial license for this tool and its
# generated code, contact me via email.
#
# If GPL is incompatible with your free or open-source project,
# contact me and we'll work it out (for free).
#
# This program is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 2 of the License, or
# (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# See <http://www.gnu.org/licenses/>.
#
#
###}
import codecs
import sys
import os
from pprint import pprint
from os import path
from collections import defaultdict
import lark
from lark import Lark
from lark.parsers.lalr_analysis import Shift, Reduce
from ..grammar import Rule
_dir = path.dirname(__file__)
_larkdir = path.join(_dir, path.pardir)
EXTRACT_STANDALONE_FILES = [
'tools/standalone.py',
'exceptions.py',
'utils.py',
'tree.py',
'visitors.py',
'indenter.py',
'lexer.py',
'parse_tree_builder.py',
'parsers/lalr_parser.py',
]
def extract_sections(lines):
section = None
text = []
sections = defaultdict(list)
for l in lines:
if l.startswith('###'):
if l[3] == '{':
section = l[4:].strip()
elif l[3] == '}':
sections[section] += text
section = None
text = []
else:
raise ValueError(l)
elif section:
text.append(l)
return {name:''.join(text) for name, text in sections.items()}
def _prepare_mres(mres):
return [(p.pattern,{i: t for i, t in d.items()}) for p,d in mres]
class TraditionalLexerAtoms:
def __init__(self, lexer):
self.mres = _prepare_mres(lexer.mres)
self.newline_types = lexer.newline_types
self.ignore_types = lexer.ignore_types
self.callback = {name:_prepare_mres(c.mres)
for name, c in lexer.callback.items()}
def print_python(self):
print('import re')
print('class LexerRegexps: pass')
print('NEWLINE_TYPES = %s' % self.newline_types)
print('IGNORE_TYPES = %s' % self.ignore_types)
self._print_python('lexer')
def _print_python(self, var_name):
print('MRES = (')
pprint(self.mres)
print(')')
print('LEXER_CALLBACK = (')
pprint(self.callback)
print(')')
print('lexer_regexps = LexerRegexps()')
print('lexer_regexps.mres = [(re.compile(p), d) for p, d in MRES]')
print('lexer_regexps.callback = {n: UnlessCallback([(re.compile(p), d) for p, d in mres])')
print(' for n, mres in LEXER_CALLBACK.items()}')
print('%s = (lexer_regexps)' % var_name)
class ContextualLexerAtoms:
def __init__(self, lexer):
self.lexer_atoms = {state: TraditionalLexerAtoms(lexer) for state, lexer in lexer.lexers.items()}
self.root_lexer_atoms = TraditionalLexerAtoms(lexer.root_lexer)
def print_python(self):
print('import re')
print('class LexerRegexps: pass')
print('NEWLINE_TYPES = %s' % self.root_lexer_atoms.newline_types)
print('IGNORE_TYPES = %s' % self.root_lexer_atoms.ignore_types)
print('LEXERS = {}')
for state, lexer_atoms in self.lexer_atoms.items():
lexer_atoms._print_python('LEXERS[%d]' % state)
print('class ContextualLexer:')
print(' def __init__(self):')
print(' self.lexers = LEXERS')
print(' self.set_parser_state(None)')
print(' def set_parser_state(self, state):')
print(' self.parser_state = state')
print(' def lex(self, stream):')
print(' newline_types = NEWLINE_TYPES')
print(' ignore_types = IGNORE_TYPES')
print(' lexers = LEXERS')
print(' l = _Lex(lexers[self.parser_state], self.parser_state)')
print(' for x in l.lex(stream, newline_types, ignore_types):')
print(' yield x')
print(' l.lexer = lexers[self.parser_state]')
print(' l.state = self.parser_state')
print('CON_LEXER = ContextualLexer()')
print('def lex(stream):')
print(' return CON_LEXER.lex(stream)')
class GetRule:
def __init__(self, rule_id):
self.rule_id = rule_id
def __repr__(self):
return 'RULES[%d]' % self.rule_id
rule_ids = {}
token_types = {}
def _get_token_type(token_type):
if token_type not in token_types:
token_types[token_type] = len(token_types)
return token_types[token_type]
class ParserAtoms:
def __init__(self, parser):
self.parse_table = parser._parse_table
def print_python(self):
print('class ParseTable: pass')
print('parse_table = ParseTable()')
print('STATES = {')
for state, actions in self.parse_table.states.items():
print(' %r: %r,' % (state, {_get_token_type(token): ((1, rule_ids[arg]) if action is Reduce else (0, arg))
for token, (action, arg) in actions.items()}))
print('}')
print('TOKEN_TYPES = (')
pprint({v:k for k, v in token_types.items()})
print(')')
print('parse_table.states = {s: {TOKEN_TYPES[t]: (a, RULES[x] if a is Reduce else x) for t, (a, x) in acts.items()}')
print(' for s, acts in STATES.items()}')
print('parse_table.start_state = %s' % self.parse_table.start_state)
print('parse_table.end_state = %s' % self.parse_table.end_state)
print('class Lark_StandAlone:')
print(' def __init__(self, transformer=None, postlex=None):')
print(' callback = parse_tree_builder.create_callback(transformer=transformer)')
print(' callbacks = {rule: getattr(callback, rule.alias or rule.origin, None) for rule in RULES.values()}')
print(' self.parser = _Parser(parse_table, callbacks)')
print(' self.postlex = postlex')
print(' def parse(self, stream):')
print(' tokens = lex(stream)')
print(' sps = CON_LEXER.set_parser_state')
print(' if self.postlex: tokens = self.postlex.process(tokens)')
print(' return self.parser.parse(tokens, sps)')
class TreeBuilderAtoms:
def __init__(self, lark):
self.rules = lark.rules
self.ptb = lark._parse_tree_builder
def print_python(self):
# print('class InlineTransformer: pass')
print('RULES = {')
for i, r in enumerate(self.rules):
rule_ids[r] = i
print(' %d: Rule(%r, [%s], %r, %r),' % (i, r.origin, ', '.join(s.fullrepr for s in r.expansion), self.ptb.user_aliases[r], r.options ))
print('}')
print('parse_tree_builder = ParseTreeBuilder(RULES.values(), Tree)')
def main(fobj, start):
lark_inst = Lark(fobj, parser="lalr", lexer="contextual", start=start)
lexer_atoms = ContextualLexerAtoms(lark_inst.parser.lexer)
parser_atoms = ParserAtoms(lark_inst.parser.parser)
tree_builder_atoms = TreeBuilderAtoms(lark_inst)
print('# The file was automatically generated by Lark v%s' % lark.__version__)
for pyfile in EXTRACT_STANDALONE_FILES:
with open(os.path.join(_larkdir, pyfile)) as f:
print (extract_sections(f)['standalone'])
with open(os.path.join(_larkdir, 'grammar.py')) as grammar_py:
print(grammar_py.read())
print('Shift = 0')
print('Reduce = 1')
lexer_atoms.print_python()
tree_builder_atoms.print_python()
parser_atoms.print_python()
if __name__ == '__main__':
if len(sys.argv) < 2:
print("Lark Stand-alone Generator Tool")
print("Usage: python -m lark.tools.standalone <grammar-file> [<start>]")
sys.exit(1)
if len(sys.argv) == 3:
fn, start = sys.argv[1:]
elif len(sys.argv) == 2:
fn, start = sys.argv[1], 'start'
else:
assert False, sys.argv
with codecs.open(fn, encoding='utf8') as f:
main(f, start)

View File

@@ -0,0 +1,168 @@
try:
from future_builtins import filter
except ImportError:
pass
from copy import deepcopy
###{standalone
class Meta:
pass
class Tree(object):
def __init__(self, data, children, meta=None):
self.data = data
self.children = children
self._meta = meta
@property
def meta(self):
if self._meta is None:
self._meta = Meta()
return self._meta
def __repr__(self):
return 'Tree(%s, %s)' % (self.data, self.children)
def _pretty_label(self):
return self.data
def _pretty(self, level, indent_str):
if len(self.children) == 1 and not isinstance(self.children[0], Tree):
return [ indent_str*level, self._pretty_label(), '\t', '%s' % (self.children[0],), '\n']
l = [ indent_str*level, self._pretty_label(), '\n' ]
for n in self.children:
if isinstance(n, Tree):
l += n._pretty(level+1, indent_str)
else:
l += [ indent_str*(level+1), '%s' % (n,), '\n' ]
return l
def pretty(self, indent_str=' '):
return ''.join(self._pretty(0, indent_str))
def __eq__(self, other):
try:
return self.data == other.data and self.children == other.children
except AttributeError:
return False
def __ne__(self, other):
return not (self == other)
def __hash__(self):
return hash((self.data, tuple(self.children)))
###}
def expand_kids_by_index(self, *indices):
"Expand (inline) children at the given indices"
for i in sorted(indices, reverse=True): # reverse so that changing tail won't affect indices
kid = self.children[i]
self.children[i:i+1] = kid.children
def find_pred(self, pred):
"Find all nodes where pred(tree) == True"
return filter(pred, self.iter_subtrees())
def find_data(self, data):
"Find all nodes where tree.data == data"
return self.find_pred(lambda t: t.data == data)
def scan_values(self, pred):
for c in self.children:
if isinstance(c, Tree):
for t in c.scan_values(pred):
yield t
else:
if pred(c):
yield c
def iter_subtrees(self):
# TODO: Re-write as a more efficient version
visited = set()
q = [self]
l = []
while q:
subtree = q.pop()
l.append( subtree )
if id(subtree) in visited:
continue # already been here from another branch
visited.add(id(subtree))
q += [c for c in subtree.children if isinstance(c, Tree)]
seen = set()
for x in reversed(l):
if id(x) not in seen:
yield x
seen.add(id(x))
def __deepcopy__(self, memo):
return type(self)(self.data, deepcopy(self.children, memo))
def copy(self):
return type(self)(self.data, self.children)
def set(self, data, children):
self.data = data
self.children = children
# XXX Deprecated! Here for backwards compatibility <0.6.0
@property
def line(self):
return self.meta.line
@property
def column(self):
return self.meta.column
@property
def end_line(self):
return self.meta.end_line
@property
def end_column(self):
return self.meta.end_column
class SlottedTree(Tree):
__slots__ = 'data', 'children', 'rule', '_meta'
def pydot__tree_to_png(tree, filename, rankdir="LR"):
"""Creates a colorful image that represents the tree (data+children, without meta)
Possible values for `rankdir` are "TB", "LR", "BT", "RL", corresponding to
directed graphs drawn from top to bottom, from left to right, from bottom to
top, and from right to left, respectively. See:
https://www.graphviz.org/doc/info/attrs.html#k:rankdir
"""
import pydot
graph = pydot.Dot(graph_type='digraph', rankdir=rankdir)
i = [0]
def new_leaf(leaf):
node = pydot.Node(i[0], label=repr(leaf))
i[0] += 1
graph.add_node(node)
return node
def _to_pydot(subtree):
color = hash(subtree.data) & 0xffffff
color |= 0x808080
subnodes = [_to_pydot(child) if isinstance(child, Tree) else new_leaf(child)
for child in subtree.children]
node = pydot.Node(i[0], style="filled", fillcolor="#%x"%color, label=subtree.data)
i[0] += 1
graph.add_node(node)
for subnode in subnodes:
graph.add_edge(pydot.Edge(node, subnode))
return node
_to_pydot(tree)
graph.write_png(filename)

View File

@@ -0,0 +1,120 @@
import sys
from collections import deque
Py36 = (sys.version_info[:2] >= (3, 6))
class fzset(frozenset):
def __repr__(self):
return '{%s}' % ', '.join(map(repr, self))
def classify_bool(seq, pred):
true_elems = []
false_elems = []
for elem in seq:
if pred(elem):
true_elems.append(elem)
else:
false_elems.append(elem)
return true_elems, false_elems
def classify(seq, key=None, value=None):
d = {}
for item in seq:
k = key(item) if (key is not None) else item
v = value(item) if (value is not None) else item
if k in d:
d[k].append(v)
else:
d[k] = [v]
return d
def bfs(initial, expand):
open_q = deque(list(initial))
visited = set(open_q)
while open_q:
node = open_q.popleft()
yield node
for next_node in expand(node):
if next_node not in visited:
visited.add(next_node)
open_q.append(next_node)
###{standalone
try:
STRING_TYPE = basestring
except NameError: # Python 3
STRING_TYPE = str
import types
from functools import wraps, partial
from contextlib import contextmanager
Str = type(u'')
def smart_decorator(f, create_decorator):
if isinstance(f, types.FunctionType):
return wraps(f)(create_decorator(f, True))
elif isinstance(f, (type, types.BuiltinFunctionType)):
return wraps(f)(create_decorator(f, False))
elif isinstance(f, types.MethodType):
return wraps(f)(create_decorator(f.__func__, True))
elif isinstance(f, partial):
# wraps does not work for partials in 2.7: https://bugs.python.org/issue3445
return create_decorator(f.__func__, True)
else:
return create_decorator(f.__func__.__call__, True)
###}
try:
from contextlib import suppress # Python 3
except ImportError:
@contextmanager
def suppress(*excs):
'''Catch and dismiss the provided exception
>>> x = 'hello'
>>> with suppress(IndexError):
... x = x[10]
>>> x
'hello'
'''
try:
yield
except excs:
pass
try:
compare = cmp
except NameError:
def compare(a, b):
if a == b:
return 0
elif a > b:
return 1
return -1
import sre_parse
import sre_constants
def get_regexp_width(regexp):
try:
return sre_parse.parse(regexp).getwidth()
except sre_constants.error:
raise ValueError(regexp)

View File

@@ -0,0 +1,259 @@
from functools import wraps
from .utils import smart_decorator
from .tree import Tree
###{standalone
from inspect import getmembers, getmro
class Discard(Exception):
pass
# Transformers
class Transformer:
"""Visits the tree recursively, starting with the leaves and finally the root (bottom-up)
Calls its methods (provided by user via inheritance) according to tree.data
The returned value replaces the old one in the structure.
Can be used to implement map or reduce.
"""
def _call_userfunc(self, tree, new_children=None):
# Assumes tree is already transformed
children = new_children if new_children is not None else tree.children
try:
f = getattr(self, tree.data)
except AttributeError:
return self.__default__(tree.data, children, tree.meta)
else:
if getattr(f, 'meta', False):
return f(children, tree.meta)
elif getattr(f, 'inline', False):
return f(*children)
elif getattr(f, 'whole_tree', False):
if new_children is not None:
raise NotImplementedError("Doesn't work with the base Transformer class")
return f(tree)
else:
return f(children)
def _transform_children(self, children):
for c in children:
try:
yield self._transform_tree(c) if isinstance(c, Tree) else c
except Discard:
pass
def _transform_tree(self, tree):
children = list(self._transform_children(tree.children))
return self._call_userfunc(tree, children)
def transform(self, tree):
return self._transform_tree(tree)
def __mul__(self, other):
return TransformerChain(self, other)
def __default__(self, data, children, meta):
"Default operation on tree (for override)"
return Tree(data, children, meta)
@classmethod
def _apply_decorator(cls, decorator, **kwargs):
mro = getmro(cls)
assert mro[0] is cls
libmembers = {name for _cls in mro[1:] for name, _ in getmembers(_cls)}
for name, value in getmembers(cls):
if name.startswith('_') or name in libmembers:
continue
if isinstance(cls.__dict__[name], (staticmethod, classmethod)):
kwargs['static'] = True
setattr(cls, name, decorator(value, **kwargs))
return cls
class InlineTransformer(Transformer): # XXX Deprecated
def _call_userfunc(self, tree, new_children=None):
# Assumes tree is already transformed
children = new_children if new_children is not None else tree.children
try:
f = getattr(self, tree.data)
except AttributeError:
return self.__default__(tree.data, children, tree.meta)
else:
return f(*children)
class TransformerChain(object):
def __init__(self, *transformers):
self.transformers = transformers
def transform(self, tree):
for t in self.transformers:
tree = t.transform(tree)
return tree
def __mul__(self, other):
return TransformerChain(*self.transformers + (other,))
class Transformer_InPlace(Transformer):
"Non-recursive. Changes the tree in-place instead of returning new instances"
def _transform_tree(self, tree): # Cancel recursion
return self._call_userfunc(tree)
def transform(self, tree):
for subtree in tree.iter_subtrees():
subtree.children = list(self._transform_children(subtree.children))
return self._transform_tree(tree)
class Transformer_InPlaceRecursive(Transformer):
"Recursive. Changes the tree in-place instead of returning new instances"
def _transform_tree(self, tree):
tree.children = list(self._transform_children(tree.children))
return self._call_userfunc(tree)
# Visitors
class VisitorBase:
def _call_userfunc(self, tree):
return getattr(self, tree.data, self.__default__)(tree)
def __default__(self, tree):
"Default operation on tree (for override)"
return tree
class Visitor(VisitorBase):
"""Bottom-up visitor, non-recursive
Visits the tree, starting with the leaves and finally the root (bottom-up)
Calls its methods (provided by user via inheritance) according to tree.data
"""
def visit(self, tree):
for subtree in tree.iter_subtrees():
self._call_userfunc(subtree)
return tree
class Visitor_Recursive(VisitorBase):
"""Bottom-up visitor, recursive
Visits the tree, starting with the leaves and finally the root (bottom-up)
Calls its methods (provided by user via inheritance) according to tree.data
"""
def visit(self, tree):
for child in tree.children:
if isinstance(child, Tree):
self.visit(child)
f = getattr(self, tree.data, self.__default__)
f(tree)
return tree
def visit_children_decor(func):
"See Interpreter"
@wraps(func)
def inner(cls, tree):
values = cls.visit_children(tree)
return func(cls, values)
return inner
class Interpreter:
"""Top-down visitor, recursive
Visits the tree, starting with the root and finally the leaves (top-down)
Calls its methods (provided by user via inheritance) according to tree.data
Unlike Transformer and Visitor, the Interpreter doesn't automatically visit its sub-branches.
The user has to explicitly call visit_children, or use the @visit_children_decor
"""
def visit(self, tree):
return getattr(self, tree.data)(tree)
def visit_children(self, tree):
return [self.visit(child) if isinstance(child, Tree) else child
for child in tree.children]
def __getattr__(self, name):
return self.__default__
def __default__(self, tree):
return self.visit_children(tree)
# Decorators
def _apply_decorator(obj, decorator, **kwargs):
try:
_apply = obj._apply_decorator
except AttributeError:
return decorator(obj, **kwargs)
else:
return _apply(decorator, **kwargs)
def _inline_args__func(func):
@wraps(func)
def create_decorator(_f, with_self):
if with_self:
def f(self, children):
return _f(self, *children)
else:
def f(self, children):
return _f(*children)
return f
return smart_decorator(func, create_decorator)
def inline_args(obj): # XXX Deprecated
return _apply_decorator(obj, _inline_args__func)
def _visitor_args_func_dec(func, inline=False, meta=False, whole_tree=False, static=False):
assert [whole_tree, meta, inline].count(True) <= 1
def create_decorator(_f, with_self):
if with_self:
def f(self, *args, **kwargs):
return _f(self, *args, **kwargs)
else:
def f(self, *args, **kwargs):
return _f(*args, **kwargs)
return f
if static:
f = wraps(func)(create_decorator(func, False))
else:
f = smart_decorator(func, create_decorator)
f.inline = inline
f.meta = meta
f.whole_tree = whole_tree
return f
def v_args(inline=False, meta=False, tree=False):
"A convenience decorator factory, for modifying the behavior of user-supplied visitor methods"
if [tree, meta, inline].count(True) > 1:
raise ValueError("Visitor functions can either accept tree, or meta, or be inlined. These cannot be combined.")
def _visitor_args_dec(obj):
return _apply_decorator(obj, _visitor_args_func_dec, inline=inline, meta=meta, whole_tree=tree)
return _visitor_args_dec
###}

View File

@@ -0,0 +1,42 @@
Metadata-Version: 1.1
Name: lark-parser
Version: 0.6.5
Summary: a modern parsing library
Home-page: https://github.com/erezsh/lark
Author: Erez Shinan
Author-email: erezshin@gmail.com
License: MIT
Download-URL: https://github.com/erezsh/lark/tarball/master
Description:
Lark is a modern general-purpose parsing library for Python.
With Lark, you can parse any context-free grammar, efficiently, with very little code.
Main Features:
- Builds a parse-tree (AST) automagically, based on the structure of the grammar
- Earley parser
- Can parse all context-free grammars
- Full support for ambiguous grammars
- LALR(1) parser
- Fast and light, competitive with PLY
- Can generate a stand-alone parser
- CYK parser, for highly ambiguous grammars
- EBNF grammar
- Unicode fully supported
- Python 2 & 3 compatible
- Automatic line & column tracking
- Standard library of terminals (strings, numbers, names, etc.)
- Import grammars from Nearley.js
- Extensive test suite
- And much more!
Keywords: Earley LALR parser parsing ast
Platform: UNKNOWN
Classifier: Development Status :: 5 - Production/Stable
Classifier: Intended Audience :: Developers
Classifier: Programming Language :: Python :: 2.7
Classifier: Programming Language :: Python :: 3
Classifier: Topic :: Software Development :: Libraries :: Python Modules
Classifier: Topic :: Text Processing :: General
Classifier: Topic :: Text Processing :: Linguistic
Classifier: License :: OSI Approved :: MIT License

View File

@@ -0,0 +1,80 @@
LICENSE
MANIFEST.in
README.md
setup.cfg
setup.py
docs/classes.md
docs/comparison_memory.png
docs/comparison_runtime.png
docs/features.md
docs/grammar.md
docs/how_to_use.md
docs/index.md
docs/json_tutorial.md
docs/lark_cheatsheet.pdf
docs/philosophy.md
docs/recipes.md
docs/tree_construction.md
examples/__init__.py
examples/calc.py
examples/comments.py
examples/conf_earley.py
examples/conf_lalr.py
examples/custom_lexer.py
examples/error_handling.py
examples/error_reporting_lalr.py
examples/fruitflies.png
examples/fruitflies.py
examples/indented_tree.py
examples/json_parser.py
examples/lark.lark
examples/lark_grammar.py
examples/python2.lark
examples/python3.lark
examples/python_parser.py
examples/qscintilla_json.py
examples/reconstruct_json.py
examples/turtle_dsl.py
lark/__init__.py
lark/common.py
lark/exceptions.py
lark/grammar.py
lark/indenter.py
lark/lark.py
lark/lexer.py
lark/load_grammar.py
lark/parse_tree_builder.py
lark/parser_frontends.py
lark/reconstruct.py
lark/reconstruct2.py
lark/tree.py
lark/utils.py
lark/visitors.py
lark/grammars/common.lark
lark/parsers/__init__.py
lark/parsers/ablalr.py
lark/parsers/cyk.py
lark/parsers/earley.py
lark/parsers/grammar_analysis.py
lark/parsers/lalr_analysis.py
lark/parsers/lalr_parser.py
lark/parsers/resolve_ambig.py
lark/parsers/xearley.py
lark/tools/__init__.py
lark/tools/nearley.py
lark/tools/standalone.py
lark_parser.egg-info/PKG-INFO
lark_parser.egg-info/SOURCES.txt
lark_parser.egg-info/dependency_links.txt
lark_parser.egg-info/not-zip-safe
lark_parser.egg-info/top_level.txt
tests/__init__.py
tests/__main__.py
tests/test_parser.py
tests/test_reconstructor.py
tests/test_tools.py
tests/test_trees.py
tests/test_nearley/__init__.py
tests/test_nearley/test_nearley.py
tests/test_nearley/grammars/include_unicode.ne
tests/test_nearley/grammars/unicode.ne

View File

@@ -0,0 +1,60 @@
../lark/__init__.py
../lark/__pycache__/__init__.cpython-36.pyc
../lark/__pycache__/common.cpython-36.pyc
../lark/__pycache__/exceptions.cpython-36.pyc
../lark/__pycache__/grammar.cpython-36.pyc
../lark/__pycache__/indenter.cpython-36.pyc
../lark/__pycache__/lark.cpython-36.pyc
../lark/__pycache__/lexer.cpython-36.pyc
../lark/__pycache__/load_grammar.cpython-36.pyc
../lark/__pycache__/parse_tree_builder.cpython-36.pyc
../lark/__pycache__/parser_frontends.cpython-36.pyc
../lark/__pycache__/reconstruct.cpython-36.pyc
../lark/__pycache__/reconstruct2.cpython-36.pyc
../lark/__pycache__/tree.cpython-36.pyc
../lark/__pycache__/utils.cpython-36.pyc
../lark/__pycache__/visitors.cpython-36.pyc
../lark/common.py
../lark/exceptions.py
../lark/grammar.py
../lark/grammars/common.lark
../lark/indenter.py
../lark/lark.py
../lark/lexer.py
../lark/load_grammar.py
../lark/parse_tree_builder.py
../lark/parser_frontends.py
../lark/parsers/__init__.py
../lark/parsers/__pycache__/__init__.cpython-36.pyc
../lark/parsers/__pycache__/ablalr.cpython-36.pyc
../lark/parsers/__pycache__/cyk.cpython-36.pyc
../lark/parsers/__pycache__/earley.cpython-36.pyc
../lark/parsers/__pycache__/grammar_analysis.cpython-36.pyc
../lark/parsers/__pycache__/lalr_analysis.cpython-36.pyc
../lark/parsers/__pycache__/lalr_parser.cpython-36.pyc
../lark/parsers/__pycache__/resolve_ambig.cpython-36.pyc
../lark/parsers/__pycache__/xearley.cpython-36.pyc
../lark/parsers/ablalr.py
../lark/parsers/cyk.py
../lark/parsers/earley.py
../lark/parsers/grammar_analysis.py
../lark/parsers/lalr_analysis.py
../lark/parsers/lalr_parser.py
../lark/parsers/resolve_ambig.py
../lark/parsers/xearley.py
../lark/reconstruct.py
../lark/reconstruct2.py
../lark/tools/__init__.py
../lark/tools/__pycache__/__init__.cpython-36.pyc
../lark/tools/__pycache__/nearley.cpython-36.pyc
../lark/tools/__pycache__/standalone.cpython-36.pyc
../lark/tools/nearley.py
../lark/tools/standalone.py
../lark/tree.py
../lark/utils.py
../lark/visitors.py
PKG-INFO
SOURCES.txt
dependency_links.txt
not-zip-safe
top_level.txt

View File

@@ -0,0 +1,69 @@
Metadata-Version: 2.1
Name: pip
Version: 10.0.1
Summary: The PyPA recommended tool for installing Python packages.
Home-page: https://pip.pypa.io/
Author: The pip developers
Author-email: python-virtualenv@groups.google.com
License: MIT
Description: pip
===
The `PyPA recommended`_ tool for installing Python packages.
.. image:: https://img.shields.io/pypi/v/pip.svg
:target: https://pypi.org/project/pip/
.. image:: https://img.shields.io/travis/pypa/pip/master.svg
:target: http://travis-ci.org/pypa/pip
.. image:: https://img.shields.io/appveyor/ci/pypa/pip.svg
:target: https://ci.appveyor.com/project/pypa/pip/history
.. image:: https://readthedocs.org/projects/pip/badge/?version=latest
:target: https://pip.pypa.io/en/latest
* `Installation`_
* `Documentation`_
* `Changelog`_
* `GitHub Page`_
* `Issue Tracking`_
* `User mailing list`_
* `Dev mailing list`_
* User IRC: #pypa on Freenode.
* Dev IRC: #pypa-dev on Freenode.
Code of Conduct
---------------
Everyone interacting in the pip project's codebases, issue trackers, chat
rooms and mailing lists is expected to follow the `PyPA Code of Conduct`_.
.. _PyPA recommended: https://packaging.python.org/en/latest/current/
.. _Installation: https://pip.pypa.io/en/stable/installing.html
.. _Documentation: https://pip.pypa.io/en/stable/
.. _Changelog: https://pip.pypa.io/en/stable/news.html
.. _GitHub Page: https://github.com/pypa/pip
.. _Issue Tracking: https://github.com/pypa/pip/issues
.. _User mailing list: http://groups.google.com/group/python-virtualenv
.. _Dev mailing list: http://groups.google.com/group/pypa-dev
.. _PyPA Code of Conduct: https://www.pypa.io/en/latest/code-of-conduct/
Keywords: easy_install distutils setuptools egg virtualenv
Platform: UNKNOWN
Classifier: Development Status :: 5 - Production/Stable
Classifier: Intended Audience :: Developers
Classifier: License :: OSI Approved :: MIT License
Classifier: Topic :: Software Development :: Build Tools
Classifier: Programming Language :: Python
Classifier: Programming Language :: Python :: 2
Classifier: Programming Language :: Python :: 2.7
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.3
Classifier: Programming Language :: Python :: 3.4
Classifier: Programming Language :: Python :: 3.5
Classifier: Programming Language :: Python :: 3.6
Classifier: Programming Language :: Python :: Implementation :: CPython
Classifier: Programming Language :: Python :: Implementation :: PyPy
Requires-Python: >=2.7,!=3.0.*,!=3.1.*,!=3.2.*
Provides-Extra: testing

View File

@@ -0,0 +1,347 @@
AUTHORS.txt
LICENSE.txt
MANIFEST.in
NEWS.rst
README.rst
pyproject.toml
setup.cfg
setup.py
docs/Makefile
docs/__init__.py
docs/conf.py
docs/configuration.rst
docs/cookbook.rst
docs/development.rst
docs/docutils.conf
docs/index.rst
docs/installing.rst
docs/logic.rst
docs/make.bat
docs/news.rst
docs/pipext.py
docs/quickstart.rst
docs/usage.rst
docs/user_guide.rst
docs/man/pip.rst
docs/man/commands/check.rst
docs/man/commands/config.rst
docs/man/commands/download.rst
docs/man/commands/freeze.rst
docs/man/commands/hash.rst
docs/man/commands/help.rst
docs/man/commands/install.rst
docs/man/commands/list.rst
docs/man/commands/search.rst
docs/man/commands/show.rst
docs/man/commands/uninstall.rst
docs/man/commands/wheel.rst
docs/reference/index.rst
docs/reference/pip.rst
docs/reference/pip_check.rst
docs/reference/pip_config.rst
docs/reference/pip_download.rst
docs/reference/pip_freeze.rst
docs/reference/pip_hash.rst
docs/reference/pip_install.rst
docs/reference/pip_list.rst
docs/reference/pip_search.rst
docs/reference/pip_show.rst
docs/reference/pip_uninstall.rst
docs/reference/pip_wheel.rst
src/pip/__init__.py
src/pip/__main__.py
src/pip.egg-info/PKG-INFO
src/pip.egg-info/SOURCES.txt
src/pip.egg-info/dependency_links.txt
src/pip.egg-info/entry_points.txt
src/pip.egg-info/not-zip-safe
src/pip.egg-info/requires.txt
src/pip.egg-info/top_level.txt
src/pip/_internal/__init__.py
src/pip/_internal/basecommand.py
src/pip/_internal/baseparser.py
src/pip/_internal/build_env.py
src/pip/_internal/cache.py
src/pip/_internal/cmdoptions.py
src/pip/_internal/compat.py
src/pip/_internal/configuration.py
src/pip/_internal/download.py
src/pip/_internal/exceptions.py
src/pip/_internal/index.py
src/pip/_internal/locations.py
src/pip/_internal/pep425tags.py
src/pip/_internal/resolve.py
src/pip/_internal/status_codes.py
src/pip/_internal/wheel.py
src/pip/_internal/commands/__init__.py
src/pip/_internal/commands/check.py
src/pip/_internal/commands/completion.py
src/pip/_internal/commands/configuration.py
src/pip/_internal/commands/download.py
src/pip/_internal/commands/freeze.py
src/pip/_internal/commands/hash.py
src/pip/_internal/commands/help.py
src/pip/_internal/commands/install.py
src/pip/_internal/commands/list.py
src/pip/_internal/commands/search.py
src/pip/_internal/commands/show.py
src/pip/_internal/commands/uninstall.py
src/pip/_internal/commands/wheel.py
src/pip/_internal/models/__init__.py
src/pip/_internal/models/index.py
src/pip/_internal/operations/__init__.py
src/pip/_internal/operations/check.py
src/pip/_internal/operations/freeze.py
src/pip/_internal/operations/prepare.py
src/pip/_internal/req/__init__.py
src/pip/_internal/req/req_file.py
src/pip/_internal/req/req_install.py
src/pip/_internal/req/req_set.py
src/pip/_internal/req/req_uninstall.py
src/pip/_internal/utils/__init__.py
src/pip/_internal/utils/appdirs.py
src/pip/_internal/utils/deprecation.py
src/pip/_internal/utils/encoding.py
src/pip/_internal/utils/filesystem.py
src/pip/_internal/utils/glibc.py
src/pip/_internal/utils/hashes.py
src/pip/_internal/utils/logging.py
src/pip/_internal/utils/misc.py
src/pip/_internal/utils/outdated.py
src/pip/_internal/utils/packaging.py
src/pip/_internal/utils/setuptools_build.py
src/pip/_internal/utils/temp_dir.py
src/pip/_internal/utils/typing.py
src/pip/_internal/utils/ui.py
src/pip/_internal/vcs/__init__.py
src/pip/_internal/vcs/bazaar.py
src/pip/_internal/vcs/git.py
src/pip/_internal/vcs/mercurial.py
src/pip/_internal/vcs/subversion.py
src/pip/_vendor/README.rst
src/pip/_vendor/__init__.py
src/pip/_vendor/appdirs.py
src/pip/_vendor/distro.py
src/pip/_vendor/ipaddress.py
src/pip/_vendor/pyparsing.py
src/pip/_vendor/retrying.py
src/pip/_vendor/six.py
src/pip/_vendor/vendor.txt
src/pip/_vendor/cachecontrol/__init__.py
src/pip/_vendor/cachecontrol/_cmd.py
src/pip/_vendor/cachecontrol/adapter.py
src/pip/_vendor/cachecontrol/cache.py
src/pip/_vendor/cachecontrol/compat.py
src/pip/_vendor/cachecontrol/controller.py
src/pip/_vendor/cachecontrol/filewrapper.py
src/pip/_vendor/cachecontrol/heuristics.py
src/pip/_vendor/cachecontrol/serialize.py
src/pip/_vendor/cachecontrol/wrapper.py
src/pip/_vendor/cachecontrol/caches/__init__.py
src/pip/_vendor/cachecontrol/caches/file_cache.py
src/pip/_vendor/cachecontrol/caches/redis_cache.py
src/pip/_vendor/certifi/__init__.py
src/pip/_vendor/certifi/__main__.py
src/pip/_vendor/certifi/cacert.pem
src/pip/_vendor/certifi/core.py
src/pip/_vendor/chardet/__init__.py
src/pip/_vendor/chardet/big5freq.py
src/pip/_vendor/chardet/big5prober.py
src/pip/_vendor/chardet/chardistribution.py
src/pip/_vendor/chardet/charsetgroupprober.py
src/pip/_vendor/chardet/charsetprober.py
src/pip/_vendor/chardet/codingstatemachine.py
src/pip/_vendor/chardet/compat.py
src/pip/_vendor/chardet/cp949prober.py
src/pip/_vendor/chardet/enums.py
src/pip/_vendor/chardet/escprober.py
src/pip/_vendor/chardet/escsm.py
src/pip/_vendor/chardet/eucjpprober.py
src/pip/_vendor/chardet/euckrfreq.py
src/pip/_vendor/chardet/euckrprober.py
src/pip/_vendor/chardet/euctwfreq.py
src/pip/_vendor/chardet/euctwprober.py
src/pip/_vendor/chardet/gb2312freq.py
src/pip/_vendor/chardet/gb2312prober.py
src/pip/_vendor/chardet/hebrewprober.py
src/pip/_vendor/chardet/jisfreq.py
src/pip/_vendor/chardet/jpcntx.py
src/pip/_vendor/chardet/langbulgarianmodel.py
src/pip/_vendor/chardet/langcyrillicmodel.py
src/pip/_vendor/chardet/langgreekmodel.py
src/pip/_vendor/chardet/langhebrewmodel.py
src/pip/_vendor/chardet/langhungarianmodel.py
src/pip/_vendor/chardet/langthaimodel.py
src/pip/_vendor/chardet/langturkishmodel.py
src/pip/_vendor/chardet/latin1prober.py
src/pip/_vendor/chardet/mbcharsetprober.py
src/pip/_vendor/chardet/mbcsgroupprober.py
src/pip/_vendor/chardet/mbcssm.py
src/pip/_vendor/chardet/sbcharsetprober.py
src/pip/_vendor/chardet/sbcsgroupprober.py
src/pip/_vendor/chardet/sjisprober.py
src/pip/_vendor/chardet/universaldetector.py
src/pip/_vendor/chardet/utf8prober.py
src/pip/_vendor/chardet/version.py
src/pip/_vendor/chardet/cli/__init__.py
src/pip/_vendor/chardet/cli/chardetect.py
src/pip/_vendor/colorama/__init__.py
src/pip/_vendor/colorama/ansi.py
src/pip/_vendor/colorama/ansitowin32.py
src/pip/_vendor/colorama/initialise.py
src/pip/_vendor/colorama/win32.py
src/pip/_vendor/colorama/winterm.py
src/pip/_vendor/distlib/__init__.py
src/pip/_vendor/distlib/compat.py
src/pip/_vendor/distlib/database.py
src/pip/_vendor/distlib/index.py
src/pip/_vendor/distlib/locators.py
src/pip/_vendor/distlib/manifest.py
src/pip/_vendor/distlib/markers.py
src/pip/_vendor/distlib/metadata.py
src/pip/_vendor/distlib/resources.py
src/pip/_vendor/distlib/scripts.py
src/pip/_vendor/distlib/t32.exe
src/pip/_vendor/distlib/t64.exe
src/pip/_vendor/distlib/util.py
src/pip/_vendor/distlib/version.py
src/pip/_vendor/distlib/w32.exe
src/pip/_vendor/distlib/w64.exe
src/pip/_vendor/distlib/wheel.py
src/pip/_vendor/distlib/_backport/__init__.py
src/pip/_vendor/distlib/_backport/misc.py
src/pip/_vendor/distlib/_backport/shutil.py
src/pip/_vendor/distlib/_backport/sysconfig.cfg
src/pip/_vendor/distlib/_backport/sysconfig.py
src/pip/_vendor/distlib/_backport/tarfile.py
src/pip/_vendor/html5lib/__init__.py
src/pip/_vendor/html5lib/_ihatexml.py
src/pip/_vendor/html5lib/_inputstream.py
src/pip/_vendor/html5lib/_tokenizer.py
src/pip/_vendor/html5lib/_utils.py
src/pip/_vendor/html5lib/constants.py
src/pip/_vendor/html5lib/html5parser.py
src/pip/_vendor/html5lib/serializer.py
src/pip/_vendor/html5lib/_trie/__init__.py
src/pip/_vendor/html5lib/_trie/_base.py
src/pip/_vendor/html5lib/_trie/datrie.py
src/pip/_vendor/html5lib/_trie/py.py
src/pip/_vendor/html5lib/filters/__init__.py
src/pip/_vendor/html5lib/filters/alphabeticalattributes.py
src/pip/_vendor/html5lib/filters/base.py
src/pip/_vendor/html5lib/filters/inject_meta_charset.py
src/pip/_vendor/html5lib/filters/lint.py
src/pip/_vendor/html5lib/filters/optionaltags.py
src/pip/_vendor/html5lib/filters/sanitizer.py
src/pip/_vendor/html5lib/filters/whitespace.py
src/pip/_vendor/html5lib/treeadapters/__init__.py
src/pip/_vendor/html5lib/treeadapters/genshi.py
src/pip/_vendor/html5lib/treeadapters/sax.py
src/pip/_vendor/html5lib/treebuilders/__init__.py
src/pip/_vendor/html5lib/treebuilders/base.py
src/pip/_vendor/html5lib/treebuilders/dom.py
src/pip/_vendor/html5lib/treebuilders/etree.py
src/pip/_vendor/html5lib/treebuilders/etree_lxml.py
src/pip/_vendor/html5lib/treewalkers/__init__.py
src/pip/_vendor/html5lib/treewalkers/base.py
src/pip/_vendor/html5lib/treewalkers/dom.py
src/pip/_vendor/html5lib/treewalkers/etree.py
src/pip/_vendor/html5lib/treewalkers/etree_lxml.py
src/pip/_vendor/html5lib/treewalkers/genshi.py
src/pip/_vendor/idna/__init__.py
src/pip/_vendor/idna/codec.py
src/pip/_vendor/idna/compat.py
src/pip/_vendor/idna/core.py
src/pip/_vendor/idna/idnadata.py
src/pip/_vendor/idna/intranges.py
src/pip/_vendor/idna/package_data.py
src/pip/_vendor/idna/uts46data.py
src/pip/_vendor/lockfile/__init__.py
src/pip/_vendor/lockfile/linklockfile.py
src/pip/_vendor/lockfile/mkdirlockfile.py
src/pip/_vendor/lockfile/pidlockfile.py
src/pip/_vendor/lockfile/sqlitelockfile.py
src/pip/_vendor/lockfile/symlinklockfile.py
src/pip/_vendor/msgpack/__init__.py
src/pip/_vendor/msgpack/_version.py
src/pip/_vendor/msgpack/exceptions.py
src/pip/_vendor/msgpack/fallback.py
src/pip/_vendor/packaging/__about__.py
src/pip/_vendor/packaging/__init__.py
src/pip/_vendor/packaging/_compat.py
src/pip/_vendor/packaging/_structures.py
src/pip/_vendor/packaging/markers.py
src/pip/_vendor/packaging/requirements.py
src/pip/_vendor/packaging/specifiers.py
src/pip/_vendor/packaging/utils.py
src/pip/_vendor/packaging/version.py
src/pip/_vendor/pkg_resources/__init__.py
src/pip/_vendor/pkg_resources/py31compat.py
src/pip/_vendor/progress/__init__.py
src/pip/_vendor/progress/bar.py
src/pip/_vendor/progress/counter.py
src/pip/_vendor/progress/helpers.py
src/pip/_vendor/progress/spinner.py
src/pip/_vendor/pytoml/__init__.py
src/pip/_vendor/pytoml/core.py
src/pip/_vendor/pytoml/parser.py
src/pip/_vendor/pytoml/writer.py
src/pip/_vendor/requests/__init__.py
src/pip/_vendor/requests/__version__.py
src/pip/_vendor/requests/_internal_utils.py
src/pip/_vendor/requests/adapters.py
src/pip/_vendor/requests/api.py
src/pip/_vendor/requests/auth.py
src/pip/_vendor/requests/certs.py
src/pip/_vendor/requests/compat.py
src/pip/_vendor/requests/cookies.py
src/pip/_vendor/requests/exceptions.py
src/pip/_vendor/requests/help.py
src/pip/_vendor/requests/hooks.py
src/pip/_vendor/requests/models.py
src/pip/_vendor/requests/packages.py
src/pip/_vendor/requests/sessions.py
src/pip/_vendor/requests/status_codes.py
src/pip/_vendor/requests/structures.py
src/pip/_vendor/requests/utils.py
src/pip/_vendor/urllib3/__init__.py
src/pip/_vendor/urllib3/_collections.py
src/pip/_vendor/urllib3/connection.py
src/pip/_vendor/urllib3/connectionpool.py
src/pip/_vendor/urllib3/exceptions.py
src/pip/_vendor/urllib3/fields.py
src/pip/_vendor/urllib3/filepost.py
src/pip/_vendor/urllib3/poolmanager.py
src/pip/_vendor/urllib3/request.py
src/pip/_vendor/urllib3/response.py
src/pip/_vendor/urllib3/contrib/__init__.py
src/pip/_vendor/urllib3/contrib/appengine.py
src/pip/_vendor/urllib3/contrib/ntlmpool.py
src/pip/_vendor/urllib3/contrib/pyopenssl.py
src/pip/_vendor/urllib3/contrib/securetransport.py
src/pip/_vendor/urllib3/contrib/socks.py
src/pip/_vendor/urllib3/contrib/_securetransport/__init__.py
src/pip/_vendor/urllib3/contrib/_securetransport/bindings.py
src/pip/_vendor/urllib3/contrib/_securetransport/low_level.py
src/pip/_vendor/urllib3/packages/__init__.py
src/pip/_vendor/urllib3/packages/ordered_dict.py
src/pip/_vendor/urllib3/packages/six.py
src/pip/_vendor/urllib3/packages/backports/__init__.py
src/pip/_vendor/urllib3/packages/backports/makefile.py
src/pip/_vendor/urllib3/packages/ssl_match_hostname/__init__.py
src/pip/_vendor/urllib3/packages/ssl_match_hostname/_implementation.py
src/pip/_vendor/urllib3/util/__init__.py
src/pip/_vendor/urllib3/util/connection.py
src/pip/_vendor/urllib3/util/request.py
src/pip/_vendor/urllib3/util/response.py
src/pip/_vendor/urllib3/util/retry.py
src/pip/_vendor/urllib3/util/selectors.py
src/pip/_vendor/urllib3/util/ssl_.py
src/pip/_vendor/urllib3/util/timeout.py
src/pip/_vendor/urllib3/util/url.py
src/pip/_vendor/urllib3/util/wait.py
src/pip/_vendor/webencodings/__init__.py
src/pip/_vendor/webencodings/labels.py
src/pip/_vendor/webencodings/mklabels.py
src/pip/_vendor/webencodings/tests.py
src/pip/_vendor/webencodings/x_user_defined.py

View File

@@ -0,0 +1,5 @@
[console_scripts]
pip = pip._internal:main
pip3 = pip._internal:main
pip3.6 = pip._internal:main

View File

@@ -0,0 +1,8 @@
[testing]
pytest
mock
pretend
scripttest>=1.3
virtualenv>=1.10
freezegun

View File

@@ -0,0 +1 @@
__version__ = "10.0.1"

View File

@@ -0,0 +1,19 @@
from __future__ import absolute_import
import os
import sys
# If we are running from a wheel, add the wheel to sys.path
# This allows the usage python pip-*.whl/pip install pip-*.whl
if __package__ == '':
# __file__ is pip-*.whl/pip/__main__.py
# first dirname call strips of '/__main__.py', second strips off '/pip'
# Resulting path is the name of the wheel itself
# Add that to sys.path so we can import pip
path = os.path.dirname(os.path.dirname(__file__))
sys.path.insert(0, path)
from pip._internal import main as _main # noqa
if __name__ == '__main__':
sys.exit(_main())

View File

@@ -0,0 +1,246 @@
#!/usr/bin/env python
from __future__ import absolute_import
import locale
import logging
import os
import optparse
import warnings
import sys
# 2016-06-17 barry@debian.org: urllib3 1.14 added optional support for socks,
# but if invoked (i.e. imported), it will issue a warning to stderr if socks
# isn't available. requests unconditionally imports urllib3's socks contrib
# module, triggering this warning. The warning breaks DEP-8 tests (because of
# the stderr output) and is just plain annoying in normal usage. I don't want
# to add socks as yet another dependency for pip, nor do I want to allow-stder
# in the DEP-8 tests, so just suppress the warning. pdb tells me this has to
# be done before the import of pip.vcs.
from pip._vendor.urllib3.exceptions import DependencyWarning
warnings.filterwarnings("ignore", category=DependencyWarning) # noqa
# We want to inject the use of SecureTransport as early as possible so that any
# references or sessions or what have you are ensured to have it, however we
# only want to do this in the case that we're running on macOS and the linked
# OpenSSL is too old to handle TLSv1.2
try:
import ssl
except ImportError:
pass
else:
# Checks for OpenSSL 1.0.1 on MacOS
if sys.platform == "darwin" and ssl.OPENSSL_VERSION_NUMBER < 0x1000100f:
try:
from pip._vendor.urllib3.contrib import securetransport
except (ImportError, OSError):
pass
else:
securetransport.inject_into_urllib3()
from pip import __version__
from pip._internal import cmdoptions
from pip._internal.exceptions import CommandError, PipError
from pip._internal.utils.misc import get_installed_distributions, get_prog
from pip._internal.utils import deprecation
from pip._internal.vcs import git, mercurial, subversion, bazaar # noqa
from pip._internal.baseparser import (
ConfigOptionParser, UpdatingDefaultsHelpFormatter,
)
from pip._internal.commands import get_summaries, get_similar_commands
from pip._internal.commands import commands_dict
from pip._vendor.urllib3.exceptions import InsecureRequestWarning
logger = logging.getLogger(__name__)
# Hide the InsecureRequestWarning from urllib3
warnings.filterwarnings("ignore", category=InsecureRequestWarning)
def autocomplete():
"""Command and option completion for the main option parser (and options)
and its subcommands (and options).
Enable by sourcing one of the completion shell scripts (bash, zsh or fish).
"""
# Don't complete if user hasn't sourced bash_completion file.
if 'PIP_AUTO_COMPLETE' not in os.environ:
return
cwords = os.environ['COMP_WORDS'].split()[1:]
cword = int(os.environ['COMP_CWORD'])
try:
current = cwords[cword - 1]
except IndexError:
current = ''
subcommands = [cmd for cmd, summary in get_summaries()]
options = []
# subcommand
try:
subcommand_name = [w for w in cwords if w in subcommands][0]
except IndexError:
subcommand_name = None
parser = create_main_parser()
# subcommand options
if subcommand_name:
# special case: 'help' subcommand has no options
if subcommand_name == 'help':
sys.exit(1)
# special case: list locally installed dists for show and uninstall
should_list_installed = (
subcommand_name in ['show', 'uninstall'] and
not current.startswith('-')
)
if should_list_installed:
installed = []
lc = current.lower()
for dist in get_installed_distributions(local_only=True):
if dist.key.startswith(lc) and dist.key not in cwords[1:]:
installed.append(dist.key)
# if there are no dists installed, fall back to option completion
if installed:
for dist in installed:
print(dist)
sys.exit(1)
subcommand = commands_dict[subcommand_name]()
for opt in subcommand.parser.option_list_all:
if opt.help != optparse.SUPPRESS_HELP:
for opt_str in opt._long_opts + opt._short_opts:
options.append((opt_str, opt.nargs))
# filter out previously specified options from available options
prev_opts = [x.split('=')[0] for x in cwords[1:cword - 1]]
options = [(x, v) for (x, v) in options if x not in prev_opts]
# filter options by current input
options = [(k, v) for k, v in options if k.startswith(current)]
for option in options:
opt_label = option[0]
# append '=' to options which require args
if option[1] and option[0][:2] == "--":
opt_label += '='
print(opt_label)
else:
# show main parser options only when necessary
if current.startswith('-') or current.startswith('--'):
opts = [i.option_list for i in parser.option_groups]
opts.append(parser.option_list)
opts = (o for it in opts for o in it)
for opt in opts:
if opt.help != optparse.SUPPRESS_HELP:
subcommands += opt._long_opts + opt._short_opts
print(' '.join([x for x in subcommands if x.startswith(current)]))
sys.exit(1)
def create_main_parser():
parser_kw = {
'usage': '\n%prog <command> [options]',
'add_help_option': False,
'formatter': UpdatingDefaultsHelpFormatter(),
'name': 'global',
'prog': get_prog(),
}
parser = ConfigOptionParser(**parser_kw)
parser.disable_interspersed_args()
pip_pkg_dir = os.path.dirname(os.path.dirname(os.path.abspath(__file__)))
parser.version = 'pip %s from %s (python %s)' % (
__version__, pip_pkg_dir, sys.version[:3],
)
# add the general options
gen_opts = cmdoptions.make_option_group(cmdoptions.general_group, parser)
parser.add_option_group(gen_opts)
parser.main = True # so the help formatter knows
# create command listing for description
command_summaries = get_summaries()
description = [''] + ['%-27s %s' % (i, j) for i, j in command_summaries]
parser.description = '\n'.join(description)
return parser
def parseopts(args):
parser = create_main_parser()
# Note: parser calls disable_interspersed_args(), so the result of this
# call is to split the initial args into the general options before the
# subcommand and everything else.
# For example:
# args: ['--timeout=5', 'install', '--user', 'INITools']
# general_options: ['--timeout==5']
# args_else: ['install', '--user', 'INITools']
general_options, args_else = parser.parse_args(args)
# --version
if general_options.version:
sys.stdout.write(parser.version)
sys.stdout.write(os.linesep)
sys.exit()
# pip || pip help -> print_help()
if not args_else or (args_else[0] == 'help' and len(args_else) == 1):
parser.print_help()
sys.exit()
# the subcommand name
cmd_name = args_else[0]
if cmd_name not in commands_dict:
guess = get_similar_commands(cmd_name)
msg = ['unknown command "%s"' % cmd_name]
if guess:
msg.append('maybe you meant "%s"' % guess)
raise CommandError(' - '.join(msg))
# all the args without the subcommand
cmd_args = args[:]
cmd_args.remove(cmd_name)
return cmd_name, cmd_args
def check_isolated(args):
isolated = False
if "--isolated" in args:
isolated = True
return isolated
def main(args=None):
if args is None:
args = sys.argv[1:]
# Configure our deprecation warnings to be sent through loggers
deprecation.install_warning_logger()
autocomplete()
try:
cmd_name, cmd_args = parseopts(args)
except PipError as exc:
sys.stderr.write("ERROR: %s" % exc)
sys.stderr.write(os.linesep)
sys.exit(1)
# Needed for locale.getpreferredencoding(False) to work
# in pip._internal.utils.encoding.auto_decode
try:
locale.setlocale(locale.LC_ALL, '')
except locale.Error as e:
# setlocale can apparently crash if locale are uninitialized
logger.debug("Ignoring error %s when setting locale", e)
command = commands_dict[cmd_name](isolated=check_isolated(cmd_args))
return command.main(cmd_args)

View File

@@ -0,0 +1,373 @@
"""Base Command class, and related routines"""
from __future__ import absolute_import
import logging
import logging.config
import optparse
import os
import sys
import warnings
from pip._internal import cmdoptions
from pip._internal.baseparser import (
ConfigOptionParser, UpdatingDefaultsHelpFormatter,
)
from pip._internal.compat import WINDOWS
from pip._internal.download import PipSession
from pip._internal.exceptions import (
BadCommand, CommandError, InstallationError, PreviousBuildDirError,
UninstallationError,
)
from pip._internal.index import PackageFinder
from pip._internal.locations import running_under_virtualenv
from pip._internal.req.req_file import parse_requirements
from pip._internal.req.req_install import InstallRequirement
from pip._internal.status_codes import (
ERROR, PREVIOUS_BUILD_DIR_ERROR, SUCCESS, UNKNOWN_ERROR,
VIRTUALENV_NOT_FOUND,
)
from pip._internal.utils import deprecation
from pip._internal.utils.logging import IndentingFormatter
from pip._internal.utils.misc import get_prog, normalize_path
from pip._internal.utils.outdated import pip_version_check
from pip._internal.utils.typing import MYPY_CHECK_RUNNING
if MYPY_CHECK_RUNNING:
from typing import Optional
__all__ = ['Command']
logger = logging.getLogger(__name__)
class Command(object):
name = None # type: Optional[str]
usage = None # type: Optional[str]
hidden = False # type: bool
ignore_require_venv = False # type: bool
log_streams = ("ext://sys.stdout", "ext://sys.stderr")
def __init__(self, isolated=False):
parser_kw = {
'usage': self.usage,
'prog': '%s %s' % (get_prog(), self.name),
'formatter': UpdatingDefaultsHelpFormatter(),
'add_help_option': False,
'name': self.name,
'description': self.__doc__,
'isolated': isolated,
}
self.parser = ConfigOptionParser(**parser_kw)
# Commands should add options to this option group
optgroup_name = '%s Options' % self.name.capitalize()
self.cmd_opts = optparse.OptionGroup(self.parser, optgroup_name)
# Add the general options
gen_opts = cmdoptions.make_option_group(
cmdoptions.general_group,
self.parser,
)
self.parser.add_option_group(gen_opts)
def _build_session(self, options, retries=None, timeout=None):
session = PipSession(
cache=(
normalize_path(os.path.join(options.cache_dir, "http"))
if options.cache_dir else None
),
retries=retries if retries is not None else options.retries,
insecure_hosts=options.trusted_hosts,
)
# Handle custom ca-bundles from the user
if options.cert:
session.verify = options.cert
# Handle SSL client certificate
if options.client_cert:
session.cert = options.client_cert
# Handle timeouts
if options.timeout or timeout:
session.timeout = (
timeout if timeout is not None else options.timeout
)
# Handle configured proxies
if options.proxy:
session.proxies = {
"http": options.proxy,
"https": options.proxy,
}
# Determine if we can prompt the user for authentication or not
session.auth.prompting = not options.no_input
return session
def parse_args(self, args):
# factored out for testability
return self.parser.parse_args(args)
def main(self, args):
options, args = self.parse_args(args)
# Set verbosity so that it can be used elsewhere.
self.verbosity = options.verbose - options.quiet
if self.verbosity >= 1:
level = "DEBUG"
elif self.verbosity == -1:
level = "WARNING"
elif self.verbosity == -2:
level = "ERROR"
elif self.verbosity <= -3:
level = "CRITICAL"
else:
level = "INFO"
# The root logger should match the "console" level *unless* we
# specified "--log" to send debug logs to a file.
root_level = level
if options.log:
root_level = "DEBUG"
logger_class = "pip._internal.utils.logging.ColorizedStreamHandler"
handler_class = "pip._internal.utils.logging.BetterRotatingFileHandler"
logging.config.dictConfig({
"version": 1,
"disable_existing_loggers": False,
"filters": {
"exclude_warnings": {
"()": "pip._internal.utils.logging.MaxLevelFilter",
"level": logging.WARNING,
},
},
"formatters": {
"indent": {
"()": IndentingFormatter,
"format": "%(message)s",
},
},
"handlers": {
"console": {
"level": level,
"class": logger_class,
"no_color": options.no_color,
"stream": self.log_streams[0],
"filters": ["exclude_warnings"],
"formatter": "indent",
},
"console_errors": {
"level": "WARNING",
"class": logger_class,
"no_color": options.no_color,
"stream": self.log_streams[1],
"formatter": "indent",
},
"user_log": {
"level": "DEBUG",
"class": handler_class,
"filename": options.log or "/dev/null",
"delay": True,
"formatter": "indent",
},
},
"root": {
"level": root_level,
"handlers": list(filter(None, [
"console",
"console_errors",
"user_log" if options.log else None,
])),
},
# Disable any logging besides WARNING unless we have DEBUG level
# logging enabled. These use both pip._vendor and the bare names
# for the case where someone unbundles our libraries.
"loggers": {
name: {
"level": (
"WARNING" if level in ["INFO", "ERROR"] else "DEBUG"
)
} for name in [
"pip._vendor", "distlib", "requests", "urllib3"
]
},
})
if sys.version_info[:2] == (3, 3):
warnings.warn(
"Python 3.3 supported has been deprecated and support for it "
"will be dropped in the future. Please upgrade your Python.",
deprecation.RemovedInPip11Warning,
)
# TODO: try to get these passing down from the command?
# without resorting to os.environ to hold these.
if options.no_input:
os.environ['PIP_NO_INPUT'] = '1'
if options.exists_action:
os.environ['PIP_EXISTS_ACTION'] = ' '.join(options.exists_action)
if options.require_venv and not self.ignore_require_venv:
# If a venv is required check if it can really be found
if not running_under_virtualenv():
logger.critical(
'Could not find an activated virtualenv (required).'
)
sys.exit(VIRTUALENV_NOT_FOUND)
original_root_handlers = set(logging.root.handlers)
try:
status = self.run(options, args)
# FIXME: all commands should return an exit status
# and when it is done, isinstance is not needed anymore
if isinstance(status, int):
return status
except PreviousBuildDirError as exc:
logger.critical(str(exc))
logger.debug('Exception information:', exc_info=True)
return PREVIOUS_BUILD_DIR_ERROR
except (InstallationError, UninstallationError, BadCommand) as exc:
logger.critical(str(exc))
logger.debug('Exception information:', exc_info=True)
return ERROR
except CommandError as exc:
logger.critical('ERROR: %s', exc)
logger.debug('Exception information:', exc_info=True)
return ERROR
except KeyboardInterrupt:
logger.critical('Operation cancelled by user')
logger.debug('Exception information:', exc_info=True)
return ERROR
except:
logger.critical('Exception:', exc_info=True)
return UNKNOWN_ERROR
finally:
# Check if we're using the latest version of pip available
if (not options.disable_pip_version_check and not
getattr(options, "no_index", False)):
with self._build_session(
options,
retries=0,
timeout=min(5, options.timeout)) as session:
pip_version_check(session, options)
# Avoid leaking loggers
for handler in set(logging.root.handlers) - original_root_handlers:
# this method benefit from the Logger class internal lock
logging.root.removeHandler(handler)
return SUCCESS
class RequirementCommand(Command):
@staticmethod
def populate_requirement_set(requirement_set, args, options, finder,
session, name, wheel_cache):
"""
Marshal cmd line args into a requirement set.
"""
# NOTE: As a side-effect, options.require_hashes and
# requirement_set.require_hashes may be updated
for filename in options.constraints:
for req_to_add in parse_requirements(
filename,
constraint=True, finder=finder, options=options,
session=session, wheel_cache=wheel_cache):
req_to_add.is_direct = True
requirement_set.add_requirement(req_to_add)
for req in args:
req_to_add = InstallRequirement.from_line(
req, None, isolated=options.isolated_mode,
wheel_cache=wheel_cache
)
req_to_add.is_direct = True
requirement_set.add_requirement(req_to_add)
for req in options.editables:
req_to_add = InstallRequirement.from_editable(
req,
isolated=options.isolated_mode,
wheel_cache=wheel_cache
)
req_to_add.is_direct = True
requirement_set.add_requirement(req_to_add)
for filename in options.requirements:
for req_to_add in parse_requirements(
filename,
finder=finder, options=options, session=session,
wheel_cache=wheel_cache):
req_to_add.is_direct = True
requirement_set.add_requirement(req_to_add)
# If --require-hashes was a line in a requirements file, tell
# RequirementSet about it:
requirement_set.require_hashes = options.require_hashes
if not (args or options.editables or options.requirements):
opts = {'name': name}
if options.find_links:
raise CommandError(
'You must give at least one requirement to %(name)s '
'(maybe you meant "pip %(name)s %(links)s"?)' %
dict(opts, links=' '.join(options.find_links)))
else:
raise CommandError(
'You must give at least one requirement to %(name)s '
'(see "pip help %(name)s")' % opts)
# On Windows, any operation modifying pip should be run as:
# python -m pip ...
# See https://github.com/pypa/pip/issues/1299 for more discussion
should_show_use_python_msg = (
WINDOWS and
requirement_set.has_requirement("pip") and
os.path.basename(sys.argv[0]).startswith("pip")
)
if should_show_use_python_msg:
new_command = [
sys.executable, "-m", "pip"
] + sys.argv[1:]
raise CommandError(
'To modify pip, please run the following command:\n{}'
.format(" ".join(new_command))
)
def _build_package_finder(self, options, session,
platform=None, python_versions=None,
abi=None, implementation=None):
"""
Create a package finder appropriate to this requirement command.
"""
index_urls = [options.index_url] + options.extra_index_urls
if options.no_index:
logger.debug('Ignoring indexes: %s', ','.join(index_urls))
index_urls = []
return PackageFinder(
find_links=options.find_links,
format_control=options.format_control,
index_urls=index_urls,
trusted_hosts=options.trusted_hosts,
allow_all_prereleases=options.pre,
process_dependency_links=options.process_dependency_links,
session=session,
platform=platform,
versions=python_versions,
abi=abi,
implementation=implementation,
)

View File

@@ -0,0 +1,240 @@
"""Base option parser setup"""
from __future__ import absolute_import
import logging
import optparse
import sys
import textwrap
from distutils.util import strtobool
from pip._vendor.six import string_types
from pip._internal.compat import get_terminal_size
from pip._internal.configuration import Configuration, ConfigurationError
logger = logging.getLogger(__name__)
class PrettyHelpFormatter(optparse.IndentedHelpFormatter):
"""A prettier/less verbose help formatter for optparse."""
def __init__(self, *args, **kwargs):
# help position must be aligned with __init__.parseopts.description
kwargs['max_help_position'] = 30
kwargs['indent_increment'] = 1
kwargs['width'] = get_terminal_size()[0] - 2
optparse.IndentedHelpFormatter.__init__(self, *args, **kwargs)
def format_option_strings(self, option):
return self._format_option_strings(option, ' <%s>', ', ')
def _format_option_strings(self, option, mvarfmt=' <%s>', optsep=', '):
"""
Return a comma-separated list of option strings and metavars.
:param option: tuple of (short opt, long opt), e.g: ('-f', '--format')
:param mvarfmt: metavar format string - evaluated as mvarfmt % metavar
:param optsep: separator
"""
opts = []
if option._short_opts:
opts.append(option._short_opts[0])
if option._long_opts:
opts.append(option._long_opts[0])
if len(opts) > 1:
opts.insert(1, optsep)
if option.takes_value():
metavar = option.metavar or option.dest.lower()
opts.append(mvarfmt % metavar.lower())
return ''.join(opts)
def format_heading(self, heading):
if heading == 'Options':
return ''
return heading + ':\n'
def format_usage(self, usage):
"""
Ensure there is only one newline between usage and the first heading
if there is no description.
"""
msg = '\nUsage: %s\n' % self.indent_lines(textwrap.dedent(usage), " ")
return msg
def format_description(self, description):
# leave full control over description to us
if description:
if hasattr(self.parser, 'main'):
label = 'Commands'
else:
label = 'Description'
# some doc strings have initial newlines, some don't
description = description.lstrip('\n')
# some doc strings have final newlines and spaces, some don't
description = description.rstrip()
# dedent, then reindent
description = self.indent_lines(textwrap.dedent(description), " ")
description = '%s:\n%s\n' % (label, description)
return description
else:
return ''
def format_epilog(self, epilog):
# leave full control over epilog to us
if epilog:
return epilog
else:
return ''
def indent_lines(self, text, indent):
new_lines = [indent + line for line in text.split('\n')]
return "\n".join(new_lines)
class UpdatingDefaultsHelpFormatter(PrettyHelpFormatter):
"""Custom help formatter for use in ConfigOptionParser.
This is updates the defaults before expanding them, allowing
them to show up correctly in the help listing.
"""
def expand_default(self, option):
if self.parser is not None:
self.parser._update_defaults(self.parser.defaults)
return optparse.IndentedHelpFormatter.expand_default(self, option)
class CustomOptionParser(optparse.OptionParser):
def insert_option_group(self, idx, *args, **kwargs):
"""Insert an OptionGroup at a given position."""
group = self.add_option_group(*args, **kwargs)
self.option_groups.pop()
self.option_groups.insert(idx, group)
return group
@property
def option_list_all(self):
"""Get a list of all options, including those in option groups."""
res = self.option_list[:]
for i in self.option_groups:
res.extend(i.option_list)
return res
class ConfigOptionParser(CustomOptionParser):
"""Custom option parser which updates its defaults by checking the
configuration files and environmental variables"""
def __init__(self, *args, **kwargs):
self.name = kwargs.pop('name')
isolated = kwargs.pop("isolated", False)
self.config = Configuration(isolated)
assert self.name
optparse.OptionParser.__init__(self, *args, **kwargs)
def check_default(self, option, key, val):
try:
return option.check_value(key, val)
except optparse.OptionValueError as exc:
print("An error occurred during configuration: %s" % exc)
sys.exit(3)
def _get_ordered_configuration_items(self):
# Configuration gives keys in an unordered manner. Order them.
override_order = ["global", self.name, ":env:"]
# Pool the options into different groups
section_items = {name: [] for name in override_order}
for section_key, val in self.config.items():
# ignore empty values
if not val:
logger.debug(
"Ignoring configuration key '%s' as it's value is empty.",
section_key
)
continue
section, key = section_key.split(".", 1)
if section in override_order:
section_items[section].append((key, val))
# Yield each group in their override order
for section in override_order:
for key, val in section_items[section]:
yield key, val
def _update_defaults(self, defaults):
"""Updates the given defaults with values from the config files and
the environ. Does a little special handling for certain types of
options (lists)."""
# Accumulate complex default state.
self.values = optparse.Values(self.defaults)
late_eval = set()
# Then set the options with those values
for key, val in self._get_ordered_configuration_items():
# '--' because configuration supports only long names
option = self.get_option('--' + key)
# Ignore options not present in this parser. E.g. non-globals put
# in [global] by users that want them to apply to all applicable
# commands.
if option is None:
continue
if option.action in ('store_true', 'store_false', 'count'):
val = strtobool(val)
elif option.action == 'append':
val = val.split()
val = [self.check_default(option, key, v) for v in val]
elif option.action == 'callback':
late_eval.add(option.dest)
opt_str = option.get_opt_string()
val = option.convert_value(opt_str, val)
# From take_action
args = option.callback_args or ()
kwargs = option.callback_kwargs or {}
option.callback(option, opt_str, val, self, *args, **kwargs)
else:
val = self.check_default(option, key, val)
defaults[option.dest] = val
for key in late_eval:
defaults[key] = getattr(self.values, key)
self.values = None
return defaults
def get_default_values(self):
"""Overriding to make updating the defaults after instantiation of
the option parser possible, _update_defaults() does the dirty work."""
if not self.process_default_values:
# Old, pre-Optik 1.5 behaviour.
return optparse.Values(self.defaults)
# Load the configuration, or error out in case of an error
try:
self.config.load()
except ConfigurationError as err:
self.exit(2, err.args[0])
defaults = self._update_defaults(self.defaults.copy()) # ours
for option in self._get_all_options():
default = defaults.get(option.dest)
if isinstance(default, string_types):
opt_str = option.get_opt_string()
defaults[option.dest] = option.check_value(opt_str, default)
return optparse.Values(defaults)
def error(self, msg):
self.print_usage(sys.stderr)
self.exit(2, "%s\n" % msg)

View File

@@ -0,0 +1,92 @@
"""Build Environment used for isolation during sdist building
"""
import os
from distutils.sysconfig import get_python_lib
from sysconfig import get_paths
from pip._internal.utils.temp_dir import TempDirectory
class BuildEnvironment(object):
"""Creates and manages an isolated environment to install build deps
"""
def __init__(self, no_clean):
self._temp_dir = TempDirectory(kind="build-env")
self._no_clean = no_clean
@property
def path(self):
return self._temp_dir.path
def __enter__(self):
self._temp_dir.create()
self.save_path = os.environ.get('PATH', None)
self.save_pythonpath = os.environ.get('PYTHONPATH', None)
self.save_nousersite = os.environ.get('PYTHONNOUSERSITE', None)
install_scheme = 'nt' if (os.name == 'nt') else 'posix_prefix'
install_dirs = get_paths(install_scheme, vars={
'base': self.path,
'platbase': self.path,
})
scripts = install_dirs['scripts']
if self.save_path:
os.environ['PATH'] = scripts + os.pathsep + self.save_path
else:
os.environ['PATH'] = scripts + os.pathsep + os.defpath
# Note: prefer distutils' sysconfig to get the
# library paths so PyPy is correctly supported.
purelib = get_python_lib(plat_specific=0, prefix=self.path)
platlib = get_python_lib(plat_specific=1, prefix=self.path)
if purelib == platlib:
lib_dirs = purelib
else:
lib_dirs = purelib + os.pathsep + platlib
if self.save_pythonpath:
os.environ['PYTHONPATH'] = lib_dirs + os.pathsep + \
self.save_pythonpath
else:
os.environ['PYTHONPATH'] = lib_dirs
os.environ['PYTHONNOUSERSITE'] = '1'
return self.path
def __exit__(self, exc_type, exc_val, exc_tb):
if not self._no_clean:
self._temp_dir.cleanup()
def restore_var(varname, old_value):
if old_value is None:
os.environ.pop(varname, None)
else:
os.environ[varname] = old_value
restore_var('PATH', self.save_path)
restore_var('PYTHONPATH', self.save_pythonpath)
restore_var('PYTHONNOUSERSITE', self.save_nousersite)
def cleanup(self):
self._temp_dir.cleanup()
class NoOpBuildEnvironment(BuildEnvironment):
"""A no-op drop-in replacement for BuildEnvironment
"""
def __init__(self, no_clean):
pass
def __enter__(self):
pass
def __exit__(self, exc_type, exc_val, exc_tb):
pass
def cleanup(self):
pass

View File

@@ -0,0 +1,202 @@
"""Cache Management
"""
import errno
import hashlib
import logging
import os
from pip._vendor.packaging.utils import canonicalize_name
from pip._internal import index
from pip._internal.compat import expanduser
from pip._internal.download import path_to_url
from pip._internal.utils.temp_dir import TempDirectory
from pip._internal.wheel import InvalidWheelFilename, Wheel
logger = logging.getLogger(__name__)
class Cache(object):
"""An abstract class - provides cache directories for data from links
:param cache_dir: The root of the cache.
:param format_control: A pip.index.FormatControl object to limit
binaries being read from the cache.
:param allowed_formats: which formats of files the cache should store.
('binary' and 'source' are the only allowed values)
"""
def __init__(self, cache_dir, format_control, allowed_formats):
super(Cache, self).__init__()
self.cache_dir = expanduser(cache_dir) if cache_dir else None
self.format_control = format_control
self.allowed_formats = allowed_formats
_valid_formats = {"source", "binary"}
assert self.allowed_formats.union(_valid_formats) == _valid_formats
def _get_cache_path_parts(self, link):
"""Get parts of part that must be os.path.joined with cache_dir
"""
# We want to generate an url to use as our cache key, we don't want to
# just re-use the URL because it might have other items in the fragment
# and we don't care about those.
key_parts = [link.url_without_fragment]
if link.hash_name is not None and link.hash is not None:
key_parts.append("=".join([link.hash_name, link.hash]))
key_url = "#".join(key_parts)
# Encode our key url with sha224, we'll use this because it has similar
# security properties to sha256, but with a shorter total output (and
# thus less secure). However the differences don't make a lot of
# difference for our use case here.
hashed = hashlib.sha224(key_url.encode()).hexdigest()
# We want to nest the directories some to prevent having a ton of top
# level directories where we might run out of sub directories on some
# FS.
parts = [hashed[:2], hashed[2:4], hashed[4:6], hashed[6:]]
return parts
def _get_candidates(self, link, package_name):
can_not_cache = (
not self.cache_dir or
not package_name or
not link
)
if can_not_cache:
return []
canonical_name = canonicalize_name(package_name)
formats = index.fmt_ctl_formats(
self.format_control, canonical_name
)
if not self.allowed_formats.intersection(formats):
return []
root = self.get_path_for_link(link)
try:
return os.listdir(root)
except OSError as err:
if err.errno in {errno.ENOENT, errno.ENOTDIR}:
return []
raise
def get_path_for_link(self, link):
"""Return a directory to store cached items in for link.
"""
raise NotImplementedError()
def get(self, link, package_name):
"""Returns a link to a cached item if it exists, otherwise returns the
passed link.
"""
raise NotImplementedError()
def _link_for_candidate(self, link, candidate):
root = self.get_path_for_link(link)
path = os.path.join(root, candidate)
return index.Link(path_to_url(path))
def cleanup(self):
pass
class SimpleWheelCache(Cache):
"""A cache of wheels for future installs.
"""
def __init__(self, cache_dir, format_control):
super(SimpleWheelCache, self).__init__(
cache_dir, format_control, {"binary"}
)
def get_path_for_link(self, link):
"""Return a directory to store cached wheels for link
Because there are M wheels for any one sdist, we provide a directory
to cache them in, and then consult that directory when looking up
cache hits.
We only insert things into the cache if they have plausible version
numbers, so that we don't contaminate the cache with things that were
not unique. E.g. ./package might have dozens of installs done for it
and build a version of 0.0...and if we built and cached a wheel, we'd
end up using the same wheel even if the source has been edited.
:param link: The link of the sdist for which this will cache wheels.
"""
parts = self._get_cache_path_parts(link)
# Store wheels within the root cache_dir
return os.path.join(self.cache_dir, "wheels", *parts)
def get(self, link, package_name):
candidates = []
for wheel_name in self._get_candidates(link, package_name):
try:
wheel = Wheel(wheel_name)
except InvalidWheelFilename:
continue
if not wheel.supported():
# Built for a different python/arch/etc
continue
candidates.append((wheel.support_index_min(), wheel_name))
if not candidates:
return link
return self._link_for_candidate(link, min(candidates)[1])
class EphemWheelCache(SimpleWheelCache):
"""A SimpleWheelCache that creates it's own temporary cache directory
"""
def __init__(self, format_control):
self._temp_dir = TempDirectory(kind="ephem-wheel-cache")
self._temp_dir.create()
super(EphemWheelCache, self).__init__(
self._temp_dir.path, format_control
)
def cleanup(self):
self._temp_dir.cleanup()
class WheelCache(Cache):
"""Wraps EphemWheelCache and SimpleWheelCache into a single Cache
This Cache allows for gracefully degradation, using the ephem wheel cache
when a certain link is not found in the simple wheel cache first.
"""
def __init__(self, cache_dir, format_control):
super(WheelCache, self).__init__(
cache_dir, format_control, {'binary'}
)
self._wheel_cache = SimpleWheelCache(cache_dir, format_control)
self._ephem_cache = EphemWheelCache(format_control)
def get_path_for_link(self, link):
return self._wheel_cache.get_path_for_link(link)
def get_ephem_path_for_link(self, link):
return self._ephem_cache.get_path_for_link(link)
def get(self, link, package_name):
retval = self._wheel_cache.get(link, package_name)
if retval is link:
retval = self._ephem_cache.get(link, package_name)
return retval
def cleanup(self):
self._wheel_cache.cleanup()
self._ephem_cache.cleanup()

View File

@@ -0,0 +1,609 @@
"""
shared options and groups
The principle here is to define options once, but *not* instantiate them
globally. One reason being that options with action='append' can carry state
between parses. pip parses general options twice internally, and shouldn't
pass on state. To be consistent, all options will follow this design.
"""
from __future__ import absolute_import
import warnings
from functools import partial
from optparse import SUPPRESS_HELP, Option, OptionGroup
from pip._internal.index import (
FormatControl, fmt_ctl_handle_mutual_exclude, fmt_ctl_no_binary,
)
from pip._internal.locations import USER_CACHE_DIR, src_prefix
from pip._internal.models import PyPI
from pip._internal.utils.hashes import STRONG_HASHES
from pip._internal.utils.typing import MYPY_CHECK_RUNNING
from pip._internal.utils.ui import BAR_TYPES
if MYPY_CHECK_RUNNING:
from typing import Any
def make_option_group(group, parser):
"""
Return an OptionGroup object
group -- assumed to be dict with 'name' and 'options' keys
parser -- an optparse Parser
"""
option_group = OptionGroup(parser, group['name'])
for option in group['options']:
option_group.add_option(option())
return option_group
def check_install_build_global(options, check_options=None):
"""Disable wheels if per-setup.py call options are set.
:param options: The OptionParser options to update.
:param check_options: The options to check, if not supplied defaults to
options.
"""
if check_options is None:
check_options = options
def getname(n):
return getattr(check_options, n, None)
names = ["build_options", "global_options", "install_options"]
if any(map(getname, names)):
control = options.format_control
fmt_ctl_no_binary(control)
warnings.warn(
'Disabling all use of wheels due to the use of --build-options '
'/ --global-options / --install-options.', stacklevel=2,
)
###########
# options #
###########
help_ = partial(
Option,
'-h', '--help',
dest='help',
action='help',
help='Show help.',
) # type: Any
isolated_mode = partial(
Option,
"--isolated",
dest="isolated_mode",
action="store_true",
default=False,
help=(
"Run pip in an isolated mode, ignoring environment variables and user "
"configuration."
),
)
require_virtualenv = partial(
Option,
# Run only if inside a virtualenv, bail if not.
'--require-virtualenv', '--require-venv',
dest='require_venv',
action='store_true',
default=False,
help=SUPPRESS_HELP
) # type: Any
verbose = partial(
Option,
'-v', '--verbose',
dest='verbose',
action='count',
default=0,
help='Give more output. Option is additive, and can be used up to 3 times.'
)
no_color = partial(
Option,
'--no-color',
dest='no_color',
action='store_true',
default=False,
help="Suppress colored output",
)
version = partial(
Option,
'-V', '--version',
dest='version',
action='store_true',
help='Show version and exit.',
) # type: Any
quiet = partial(
Option,
'-q', '--quiet',
dest='quiet',
action='count',
default=0,
help=(
'Give less output. Option is additive, and can be used up to 3'
' times (corresponding to WARNING, ERROR, and CRITICAL logging'
' levels).'
),
) # type: Any
progress_bar = partial(
Option,
'--progress-bar',
dest='progress_bar',
type='choice',
choices=list(BAR_TYPES.keys()),
default='on',
help=(
'Specify type of progress to be displayed [' +
'|'.join(BAR_TYPES.keys()) + '] (default: %default)'
),
) # type: Any
log = partial(
Option,
"--log", "--log-file", "--local-log",
dest="log",
metavar="path",
help="Path to a verbose appending log."
) # type: Any
no_input = partial(
Option,
# Don't ask for input
'--no-input',
dest='no_input',
action='store_true',
default=False,
help=SUPPRESS_HELP
) # type: Any
proxy = partial(
Option,
'--proxy',
dest='proxy',
type='str',
default='',
help="Specify a proxy in the form [user:passwd@]proxy.server:port."
) # type: Any
retries = partial(
Option,
'--retries',
dest='retries',
type='int',
default=5,
help="Maximum number of retries each connection should attempt "
"(default %default times).",
) # type: Any
timeout = partial(
Option,
'--timeout', '--default-timeout',
metavar='sec',
dest='timeout',
type='float',
default=15,
help='Set the socket timeout (default %default seconds).',
) # type: Any
skip_requirements_regex = partial(
Option,
# A regex to be used to skip requirements
'--skip-requirements-regex',
dest='skip_requirements_regex',
type='str',
default='',
help=SUPPRESS_HELP,
) # type: Any
def exists_action():
return Option(
# Option when path already exist
'--exists-action',
dest='exists_action',
type='choice',
choices=['s', 'i', 'w', 'b', 'a'],
default=[],
action='append',
metavar='action',
help="Default action when a path already exists: "
"(s)witch, (i)gnore, (w)ipe, (b)ackup, (a)bort).",
)
cert = partial(
Option,
'--cert',
dest='cert',
type='str',
metavar='path',
help="Path to alternate CA bundle.",
) # type: Any
client_cert = partial(
Option,
'--client-cert',
dest='client_cert',
type='str',
default=None,
metavar='path',
help="Path to SSL client certificate, a single file containing the "
"private key and the certificate in PEM format.",
) # type: Any
index_url = partial(
Option,
'-i', '--index-url', '--pypi-url',
dest='index_url',
metavar='URL',
default=PyPI.simple_url,
help="Base URL of Python Package Index (default %default). "
"This should point to a repository compliant with PEP 503 "
"(the simple repository API) or a local directory laid out "
"in the same format.",
) # type: Any
def extra_index_url():
return Option(
'--extra-index-url',
dest='extra_index_urls',
metavar='URL',
action='append',
default=[],
help="Extra URLs of package indexes to use in addition to "
"--index-url. Should follow the same rules as "
"--index-url.",
)
no_index = partial(
Option,
'--no-index',
dest='no_index',
action='store_true',
default=False,
help='Ignore package index (only looking at --find-links URLs instead).',
) # type: Any
def find_links():
return Option(
'-f', '--find-links',
dest='find_links',
action='append',
default=[],
metavar='url',
help="If a url or path to an html file, then parse for links to "
"archives. If a local path or file:// url that's a directory, "
"then look for archives in the directory listing.",
)
def trusted_host():
return Option(
"--trusted-host",
dest="trusted_hosts",
action="append",
metavar="HOSTNAME",
default=[],
help="Mark this host as trusted, even though it does not have valid "
"or any HTTPS.",
)
# Remove after 1.5
process_dependency_links = partial(
Option,
"--process-dependency-links",
dest="process_dependency_links",
action="store_true",
default=False,
help="Enable the processing of dependency links.",
) # type: Any
def constraints():
return Option(
'-c', '--constraint',
dest='constraints',
action='append',
default=[],
metavar='file',
help='Constrain versions using the given constraints file. '
'This option can be used multiple times.'
)
def requirements():
return Option(
'-r', '--requirement',
dest='requirements',
action='append',
default=[],
metavar='file',
help='Install from the given requirements file. '
'This option can be used multiple times.'
)
def editable():
return Option(
'-e', '--editable',
dest='editables',
action='append',
default=[],
metavar='path/url',
help=('Install a project in editable mode (i.e. setuptools '
'"develop mode") from a local project path or a VCS url.'),
)
src = partial(
Option,
'--src', '--source', '--source-dir', '--source-directory',
dest='src_dir',
metavar='dir',
default=src_prefix,
help='Directory to check out editable projects into. '
'The default in a virtualenv is "<venv path>/src". '
'The default for global installs is "<current dir>/src".'
) # type: Any
def _get_format_control(values, option):
"""Get a format_control object."""
return getattr(values, option.dest)
def _handle_no_binary(option, opt_str, value, parser):
existing = getattr(parser.values, option.dest)
fmt_ctl_handle_mutual_exclude(
value, existing.no_binary, existing.only_binary,
)
def _handle_only_binary(option, opt_str, value, parser):
existing = getattr(parser.values, option.dest)
fmt_ctl_handle_mutual_exclude(
value, existing.only_binary, existing.no_binary,
)
def no_binary():
return Option(
"--no-binary", dest="format_control", action="callback",
callback=_handle_no_binary, type="str",
default=FormatControl(set(), set()),
help="Do not use binary packages. Can be supplied multiple times, and "
"each time adds to the existing value. Accepts either :all: to "
"disable all binary packages, :none: to empty the set, or one or "
"more package names with commas between them. Note that some "
"packages are tricky to compile and may fail to install when "
"this option is used on them.",
)
def only_binary():
return Option(
"--only-binary", dest="format_control", action="callback",
callback=_handle_only_binary, type="str",
default=FormatControl(set(), set()),
help="Do not use source packages. Can be supplied multiple times, and "
"each time adds to the existing value. Accepts either :all: to "
"disable all source packages, :none: to empty the set, or one or "
"more package names with commas between them. Packages without "
"binary distributions will fail to install when this option is "
"used on them.",
)
cache_dir = partial(
Option,
"--cache-dir",
dest="cache_dir",
default=USER_CACHE_DIR,
metavar="dir",
help="Store the cache data in <dir>."
)
no_cache = partial(
Option,
"--no-cache-dir",
dest="cache_dir",
action="store_false",
help="Disable the cache.",
)
no_deps = partial(
Option,
'--no-deps', '--no-dependencies',
dest='ignore_dependencies',
action='store_true',
default=False,
help="Don't install package dependencies.",
) # type: Any
build_dir = partial(
Option,
'-b', '--build', '--build-dir', '--build-directory',
dest='build_dir',
metavar='dir',
help='Directory to unpack packages into and build in. Note that '
'an initial build still takes place in a temporary directory. '
'The location of temporary directories can be controlled by setting '
'the TMPDIR environment variable (TEMP on Windows) appropriately. '
'When passed, build directories are not cleaned in case of failures.'
) # type: Any
ignore_requires_python = partial(
Option,
'--ignore-requires-python',
dest='ignore_requires_python',
action='store_true',
help='Ignore the Requires-Python information.'
) # type: Any
no_build_isolation = partial(
Option,
'--no-build-isolation',
dest='build_isolation',
action='store_false',
default=True,
help='Disable isolation when building a modern source distribution. '
'Build dependencies specified by PEP 518 must be already installed '
'if this option is used.'
) # type: Any
install_options = partial(
Option,
'--install-option',
dest='install_options',
action='append',
metavar='options',
help="Extra arguments to be supplied to the setup.py install "
"command (use like --install-option=\"--install-scripts=/usr/local/"
"bin\"). Use multiple --install-option options to pass multiple "
"options to setup.py install. If you are using an option with a "
"directory path, be sure to use absolute path.",
) # type: Any
global_options = partial(
Option,
'--global-option',
dest='global_options',
action='append',
metavar='options',
help="Extra global options to be supplied to the setup.py "
"call before the install command.",
) # type: Any
no_clean = partial(
Option,
'--no-clean',
action='store_true',
default=False,
help="Don't clean up build directories)."
) # type: Any
pre = partial(
Option,
'--pre',
action='store_true',
default=False,
help="Include pre-release and development versions. By default, "
"pip only finds stable versions.",
) # type: Any
disable_pip_version_check = partial(
Option,
"--disable-pip-version-check",
dest="disable_pip_version_check",
action="store_true",
default=False,
help="Don't periodically check PyPI to determine whether a new version "
"of pip is available for download. Implied with --no-index.",
) # type: Any
# Deprecated, Remove later
always_unzip = partial(
Option,
'-Z', '--always-unzip',
dest='always_unzip',
action='store_true',
help=SUPPRESS_HELP,
) # type: Any
def _merge_hash(option, opt_str, value, parser):
"""Given a value spelled "algo:digest", append the digest to a list
pointed to in a dict by the algo name."""
if not parser.values.hashes:
parser.values.hashes = {}
try:
algo, digest = value.split(':', 1)
except ValueError:
parser.error('Arguments to %s must be a hash name '
'followed by a value, like --hash=sha256:abcde...' %
opt_str)
if algo not in STRONG_HASHES:
parser.error('Allowed hash algorithms for %s are %s.' %
(opt_str, ', '.join(STRONG_HASHES)))
parser.values.hashes.setdefault(algo, []).append(digest)
hash = partial(
Option,
'--hash',
# Hash values eventually end up in InstallRequirement.hashes due to
# __dict__ copying in process_line().
dest='hashes',
action='callback',
callback=_merge_hash,
type='string',
help="Verify that the package's archive matches this "
'hash before installing. Example: --hash=sha256:abcdef...',
) # type: Any
require_hashes = partial(
Option,
'--require-hashes',
dest='require_hashes',
action='store_true',
default=False,
help='Require a hash to check each requirement against, for '
'repeatable installs. This option is implied when any package in a '
'requirements file has a --hash option.',
) # type: Any
##########
# groups #
##########
general_group = {
'name': 'General Options',
'options': [
help_,
isolated_mode,
require_virtualenv,
verbose,
version,
quiet,
log,
no_input,
proxy,
retries,
timeout,
skip_requirements_regex,
exists_action,
trusted_host,
cert,
client_cert,
cache_dir,
no_cache,
disable_pip_version_check,
no_color,
]
}
index_group = {
'name': 'Package Index Options',
'options': [
index_url,
extra_index_url,
no_index,
find_links,
process_dependency_links,
]
}

View File

@@ -0,0 +1,79 @@
"""
Package containing all pip commands
"""
from __future__ import absolute_import
from pip._internal.commands.completion import CompletionCommand
from pip._internal.commands.configuration import ConfigurationCommand
from pip._internal.commands.download import DownloadCommand
from pip._internal.commands.freeze import FreezeCommand
from pip._internal.commands.hash import HashCommand
from pip._internal.commands.help import HelpCommand
from pip._internal.commands.list import ListCommand
from pip._internal.commands.check import CheckCommand
from pip._internal.commands.search import SearchCommand
from pip._internal.commands.show import ShowCommand
from pip._internal.commands.install import InstallCommand
from pip._internal.commands.uninstall import UninstallCommand
from pip._internal.commands.wheel import WheelCommand
from pip._internal.utils.typing import MYPY_CHECK_RUNNING
if MYPY_CHECK_RUNNING:
from typing import List, Type
from pip._internal.basecommand import Command
commands_order = [
InstallCommand,
DownloadCommand,
UninstallCommand,
FreezeCommand,
ListCommand,
ShowCommand,
CheckCommand,
ConfigurationCommand,
SearchCommand,
WheelCommand,
HashCommand,
CompletionCommand,
HelpCommand,
] # type: List[Type[Command]]
commands_dict = {c.name: c for c in commands_order}
def get_summaries(ordered=True):
"""Yields sorted (command name, command summary) tuples."""
if ordered:
cmditems = _sort_commands(commands_dict, commands_order)
else:
cmditems = commands_dict.items()
for name, command_class in cmditems:
yield (name, command_class.summary)
def get_similar_commands(name):
"""Command name auto-correct."""
from difflib import get_close_matches
name = name.lower()
close_commands = get_close_matches(name, commands_dict.keys())
if close_commands:
return close_commands[0]
else:
return False
def _sort_commands(cmddict, order):
def keyfn(key):
try:
return order.index(key[1])
except ValueError:
# unordered items should come last
return 0xff
return sorted(cmddict.items(), key=keyfn)

View File

@@ -0,0 +1,42 @@
import logging
from pip._internal.basecommand import Command
from pip._internal.operations.check import (
check_package_set, create_package_set_from_installed,
)
from pip._internal.utils.misc import get_installed_distributions
logger = logging.getLogger(__name__)
class CheckCommand(Command):
"""Verify installed packages have compatible dependencies."""
name = 'check'
usage = """
%prog [options]"""
summary = 'Verify installed packages have compatible dependencies.'
def run(self, options, args):
package_set = create_package_set_from_installed()
missing, conflicting = check_package_set(package_set)
for project_name in missing:
version = package_set[project_name].version
for dependency in missing[project_name]:
logger.info(
"%s %s requires %s, which is not installed.",
project_name, version, dependency[0],
)
for project_name in conflicting:
version = package_set[project_name].version
for dep_name, dep_version, req in conflicting[project_name]:
logger.info(
"%s %s has requirement %s, but you have %s %s.",
project_name, version, req, dep_name, dep_version,
)
if missing or conflicting:
return 1
else:
logger.info("No broken requirements found.")

View File

@@ -0,0 +1,94 @@
from __future__ import absolute_import
import sys
import textwrap
from pip._internal.basecommand import Command
from pip._internal.utils.misc import get_prog
BASE_COMPLETION = """
# pip %(shell)s completion start%(script)s# pip %(shell)s completion end
"""
COMPLETION_SCRIPTS = {
'bash': """
_pip_completion()
{
COMPREPLY=( $( COMP_WORDS="${COMP_WORDS[*]}" \\
COMP_CWORD=$COMP_CWORD \\
PIP_AUTO_COMPLETE=1 $1 ) )
}
complete -o default -F _pip_completion %(prog)s
""",
'zsh': """
function _pip_completion {
local words cword
read -Ac words
read -cn cword
reply=( $( COMP_WORDS="$words[*]" \\
COMP_CWORD=$(( cword-1 )) \\
PIP_AUTO_COMPLETE=1 $words[1] ) )
}
compctl -K _pip_completion %(prog)s
""",
'fish': """
function __fish_complete_pip
set -lx COMP_WORDS (commandline -o) ""
set -lx COMP_CWORD ( \\
math (contains -i -- (commandline -t) $COMP_WORDS)-1 \\
)
set -lx PIP_AUTO_COMPLETE 1
string split \\ -- (eval $COMP_WORDS[1])
end
complete -fa "(__fish_complete_pip)" -c %(prog)s
""",
}
class CompletionCommand(Command):
"""A helper command to be used for command completion."""
name = 'completion'
summary = 'A helper command used for command completion.'
ignore_require_venv = True
def __init__(self, *args, **kw):
super(CompletionCommand, self).__init__(*args, **kw)
cmd_opts = self.cmd_opts
cmd_opts.add_option(
'--bash', '-b',
action='store_const',
const='bash',
dest='shell',
help='Emit completion code for bash')
cmd_opts.add_option(
'--zsh', '-z',
action='store_const',
const='zsh',
dest='shell',
help='Emit completion code for zsh')
cmd_opts.add_option(
'--fish', '-f',
action='store_const',
const='fish',
dest='shell',
help='Emit completion code for fish')
self.parser.insert_option_group(0, cmd_opts)
def run(self, options, args):
"""Prints the completion code of the given shell"""
shells = COMPLETION_SCRIPTS.keys()
shell_options = ['--' + shell for shell in sorted(shells)]
if options.shell in shells:
script = textwrap.dedent(
COMPLETION_SCRIPTS.get(options.shell, '') % {
'prog': get_prog(),
}
)
print(BASE_COMPLETION % {'script': script, 'shell': options.shell})
else:
sys.stderr.write(
'ERROR: You must pass %s\n' % ' or '.join(shell_options)
)

View File

@@ -0,0 +1,227 @@
import logging
import os
import subprocess
from pip._internal.basecommand import Command
from pip._internal.configuration import Configuration, kinds
from pip._internal.exceptions import PipError
from pip._internal.locations import venv_config_file
from pip._internal.status_codes import ERROR, SUCCESS
from pip._internal.utils.misc import get_prog
logger = logging.getLogger(__name__)
class ConfigurationCommand(Command):
"""Manage local and global configuration.
Subcommands:
list: List the active configuration (or from the file specified)
edit: Edit the configuration file in an editor
get: Get the value associated with name
set: Set the name=value
unset: Unset the value associated with name
If none of --user, --global and --venv are passed, a virtual
environment configuration file is used if one is active and the file
exists. Otherwise, all modifications happen on the to the user file by
default.
"""
name = 'config'
usage = """
%prog [<file-option>] list
%prog [<file-option>] [--editor <editor-path>] edit
%prog [<file-option>] get name
%prog [<file-option>] set name value
%prog [<file-option>] unset name
"""
summary = "Manage local and global configuration."
def __init__(self, *args, **kwargs):
super(ConfigurationCommand, self).__init__(*args, **kwargs)
self.configuration = None
self.cmd_opts.add_option(
'--editor',
dest='editor',
action='store',
default=None,
help=(
'Editor to use to edit the file. Uses VISUAL or EDITOR '
'environment variables if not provided.'
)
)
self.cmd_opts.add_option(
'--global',
dest='global_file',
action='store_true',
default=False,
help='Use the system-wide configuration file only'
)
self.cmd_opts.add_option(
'--user',
dest='user_file',
action='store_true',
default=False,
help='Use the user configuration file only'
)
self.cmd_opts.add_option(
'--venv',
dest='venv_file',
action='store_true',
default=False,
help='Use the virtualenv configuration file only'
)
self.parser.insert_option_group(0, self.cmd_opts)
def run(self, options, args):
handlers = {
"list": self.list_values,
"edit": self.open_in_editor,
"get": self.get_name,
"set": self.set_name_value,
"unset": self.unset_name
}
# Determine action
if not args or args[0] not in handlers:
logger.error("Need an action ({}) to perform.".format(
", ".join(sorted(handlers)))
)
return ERROR
action = args[0]
# Determine which configuration files are to be loaded
# Depends on whether the command is modifying.
try:
load_only = self._determine_file(
options, need_value=(action in ["get", "set", "unset", "edit"])
)
except PipError as e:
logger.error(e.args[0])
return ERROR
# Load a new configuration
self.configuration = Configuration(
isolated=options.isolated_mode, load_only=load_only
)
self.configuration.load()
# Error handling happens here, not in the action-handlers.
try:
handlers[action](options, args[1:])
except PipError as e:
logger.error(e.args[0])
return ERROR
return SUCCESS
def _determine_file(self, options, need_value):
file_options = {
kinds.USER: options.user_file,
kinds.GLOBAL: options.global_file,
kinds.VENV: options.venv_file
}
if sum(file_options.values()) == 0:
if not need_value:
return None
# Default to user, unless there's a virtualenv file.
elif os.path.exists(venv_config_file):
return kinds.VENV
else:
return kinds.USER
elif sum(file_options.values()) == 1:
# There's probably a better expression for this.
return [key for key in file_options if file_options[key]][0]
raise PipError(
"Need exactly one file to operate upon "
"(--user, --venv, --global) to perform."
)
def list_values(self, options, args):
self._get_n_args(args, "list", n=0)
for key, value in sorted(self.configuration.items()):
logger.info("%s=%r", key, value)
def get_name(self, options, args):
key = self._get_n_args(args, "get [name]", n=1)
value = self.configuration.get_value(key)
logger.info("%s", value)
def set_name_value(self, options, args):
key, value = self._get_n_args(args, "set [name] [value]", n=2)
self.configuration.set_value(key, value)
self._save_configuration()
def unset_name(self, options, args):
key = self._get_n_args(args, "unset [name]", n=1)
self.configuration.unset_value(key)
self._save_configuration()
def open_in_editor(self, options, args):
editor = self._determine_editor(options)
fname = self.configuration.get_file_to_edit()
if fname is None:
raise PipError("Could not determine appropriate file.")
try:
subprocess.check_call([editor, fname])
except subprocess.CalledProcessError as e:
raise PipError(
"Editor Subprocess exited with exit code {}"
.format(e.returncode)
)
def _get_n_args(self, args, example, n):
"""Helper to make sure the command got the right number of arguments
"""
if len(args) != n:
msg = (
'Got unexpected number of arguments, expected {}. '
'(example: "{} config {}")'
).format(n, get_prog(), example)
raise PipError(msg)
if n == 1:
return args[0]
else:
return args
def _save_configuration(self):
# We successfully ran a modifying command. Need to save the
# configuration.
try:
self.configuration.save()
except Exception:
logger.error(
"Unable to save configuration. Please report this as a bug.",
exc_info=1
)
raise PipError("Internal Error.")
def _determine_editor(self, options):
if options.editor is not None:
return options.editor
elif "VISUAL" in os.environ:
return os.environ["VISUAL"]
elif "EDITOR" in os.environ:
return os.environ["EDITOR"]
else:
raise PipError("Could not determine editor to use.")

View File

@@ -0,0 +1,233 @@
from __future__ import absolute_import
import logging
import os
from pip._internal import cmdoptions
from pip._internal.basecommand import RequirementCommand
from pip._internal.exceptions import CommandError
from pip._internal.index import FormatControl
from pip._internal.operations.prepare import RequirementPreparer
from pip._internal.req import RequirementSet
from pip._internal.resolve import Resolver
from pip._internal.utils.filesystem import check_path_owner
from pip._internal.utils.misc import ensure_dir, normalize_path
from pip._internal.utils.temp_dir import TempDirectory
logger = logging.getLogger(__name__)
class DownloadCommand(RequirementCommand):
"""
Download packages from:
- PyPI (and other indexes) using requirement specifiers.
- VCS project urls.
- Local project directories.
- Local or remote source archives.
pip also supports downloading from "requirements files", which provide
an easy way to specify a whole environment to be downloaded.
"""
name = 'download'
usage = """
%prog [options] <requirement specifier> [package-index-options] ...
%prog [options] -r <requirements file> [package-index-options] ...
%prog [options] <vcs project url> ...
%prog [options] <local project path> ...
%prog [options] <archive url/path> ..."""
summary = 'Download packages.'
def __init__(self, *args, **kw):
super(DownloadCommand, self).__init__(*args, **kw)
cmd_opts = self.cmd_opts
cmd_opts.add_option(cmdoptions.constraints())
cmd_opts.add_option(cmdoptions.requirements())
cmd_opts.add_option(cmdoptions.build_dir())
cmd_opts.add_option(cmdoptions.no_deps())
cmd_opts.add_option(cmdoptions.global_options())
cmd_opts.add_option(cmdoptions.no_binary())
cmd_opts.add_option(cmdoptions.only_binary())
cmd_opts.add_option(cmdoptions.src())
cmd_opts.add_option(cmdoptions.pre())
cmd_opts.add_option(cmdoptions.no_clean())
cmd_opts.add_option(cmdoptions.require_hashes())
cmd_opts.add_option(cmdoptions.progress_bar())
cmd_opts.add_option(cmdoptions.no_build_isolation())
cmd_opts.add_option(
'-d', '--dest', '--destination-dir', '--destination-directory',
dest='download_dir',
metavar='dir',
default=os.curdir,
help=("Download packages into <dir>."),
)
cmd_opts.add_option(
'--platform',
dest='platform',
metavar='platform',
default=None,
help=("Only download wheels compatible with <platform>. "
"Defaults to the platform of the running system."),
)
cmd_opts.add_option(
'--python-version',
dest='python_version',
metavar='python_version',
default=None,
help=("Only download wheels compatible with Python "
"interpreter version <version>. If not specified, then the "
"current system interpreter minor version is used. A major "
"version (e.g. '2') can be specified to match all "
"minor revs of that major version. A minor version "
"(e.g. '34') can also be specified."),
)
cmd_opts.add_option(
'--implementation',
dest='implementation',
metavar='implementation',
default=None,
help=("Only download wheels compatible with Python "
"implementation <implementation>, e.g. 'pp', 'jy', 'cp', "
" or 'ip'. If not specified, then the current "
"interpreter implementation is used. Use 'py' to force "
"implementation-agnostic wheels."),
)
cmd_opts.add_option(
'--abi',
dest='abi',
metavar='abi',
default=None,
help=("Only download wheels compatible with Python "
"abi <abi>, e.g. 'pypy_41'. If not specified, then the "
"current interpreter abi tag is used. Generally "
"you will need to specify --implementation, "
"--platform, and --python-version when using "
"this option."),
)
index_opts = cmdoptions.make_option_group(
cmdoptions.index_group,
self.parser,
)
self.parser.insert_option_group(0, index_opts)
self.parser.insert_option_group(0, cmd_opts)
def run(self, options, args):
options.ignore_installed = True
# editable doesn't really make sense for `pip download`, but the bowels
# of the RequirementSet code require that property.
options.editables = []
if options.python_version:
python_versions = [options.python_version]
else:
python_versions = None
dist_restriction_set = any([
options.python_version,
options.platform,
options.abi,
options.implementation,
])
binary_only = FormatControl(set(), {':all:'})
no_sdist_dependencies = (
options.format_control != binary_only and
not options.ignore_dependencies
)
if dist_restriction_set and no_sdist_dependencies:
raise CommandError(
"When restricting platform and interpreter constraints using "
"--python-version, --platform, --abi, or --implementation, "
"either --no-deps must be set, or --only-binary=:all: must be "
"set and --no-binary must not be set (or must be set to "
":none:)."
)
options.src_dir = os.path.abspath(options.src_dir)
options.download_dir = normalize_path(options.download_dir)
ensure_dir(options.download_dir)
with self._build_session(options) as session:
finder = self._build_package_finder(
options=options,
session=session,
platform=options.platform,
python_versions=python_versions,
abi=options.abi,
implementation=options.implementation,
)
build_delete = (not (options.no_clean or options.build_dir))
if options.cache_dir and not check_path_owner(options.cache_dir):
logger.warning(
"The directory '%s' or its parent directory is not owned "
"by the current user and caching wheels has been "
"disabled. check the permissions and owner of that "
"directory. If executing pip with sudo, you may want "
"sudo's -H flag.",
options.cache_dir,
)
options.cache_dir = None
with TempDirectory(
options.build_dir, delete=build_delete, kind="download"
) as directory:
requirement_set = RequirementSet(
require_hashes=options.require_hashes,
)
self.populate_requirement_set(
requirement_set,
args,
options,
finder,
session,
self.name,
None
)
preparer = RequirementPreparer(
build_dir=directory.path,
src_dir=options.src_dir,
download_dir=options.download_dir,
wheel_download_dir=None,
progress_bar=options.progress_bar,
build_isolation=options.build_isolation,
)
resolver = Resolver(
preparer=preparer,
finder=finder,
session=session,
wheel_cache=None,
use_user_site=False,
upgrade_strategy="to-satisfy-only",
force_reinstall=False,
ignore_dependencies=options.ignore_dependencies,
ignore_requires_python=False,
ignore_installed=True,
isolated=options.isolated_mode,
)
resolver.resolve(requirement_set)
downloaded = ' '.join([
req.name for req in requirement_set.successfully_downloaded
])
if downloaded:
logger.info('Successfully downloaded %s', downloaded)
# Clean up
if not options.no_clean:
requirement_set.cleanup_files()
return requirement_set

View File

@@ -0,0 +1,96 @@
from __future__ import absolute_import
import sys
from pip._internal import index
from pip._internal.basecommand import Command
from pip._internal.cache import WheelCache
from pip._internal.compat import stdlib_pkgs
from pip._internal.operations.freeze import freeze
DEV_PKGS = {'pip', 'setuptools', 'distribute', 'wheel'}
class FreezeCommand(Command):
"""
Output installed packages in requirements format.
packages are listed in a case-insensitive sorted order.
"""
name = 'freeze'
usage = """
%prog [options]"""
summary = 'Output installed packages in requirements format.'
log_streams = ("ext://sys.stderr", "ext://sys.stderr")
def __init__(self, *args, **kw):
super(FreezeCommand, self).__init__(*args, **kw)
self.cmd_opts.add_option(
'-r', '--requirement',
dest='requirements',
action='append',
default=[],
metavar='file',
help="Use the order in the given requirements file and its "
"comments when generating output. This option can be "
"used multiple times.")
self.cmd_opts.add_option(
'-f', '--find-links',
dest='find_links',
action='append',
default=[],
metavar='URL',
help='URL for finding packages, which will be added to the '
'output.')
self.cmd_opts.add_option(
'-l', '--local',
dest='local',
action='store_true',
default=False,
help='If in a virtualenv that has global access, do not output '
'globally-installed packages.')
self.cmd_opts.add_option(
'--user',
dest='user',
action='store_true',
default=False,
help='Only output packages installed in user-site.')
self.cmd_opts.add_option(
'--all',
dest='freeze_all',
action='store_true',
help='Do not skip these packages in the output:'
' %s' % ', '.join(DEV_PKGS))
self.cmd_opts.add_option(
'--exclude-editable',
dest='exclude_editable',
action='store_true',
help='Exclude editable package from output.')
self.parser.insert_option_group(0, self.cmd_opts)
def run(self, options, args):
format_control = index.FormatControl(set(), set())
wheel_cache = WheelCache(options.cache_dir, format_control)
skip = set(stdlib_pkgs)
if not options.freeze_all:
skip.update(DEV_PKGS)
freeze_kwargs = dict(
requirement=options.requirements,
find_links=options.find_links,
local_only=options.local,
user_only=options.user,
skip_regex=options.skip_requirements_regex,
isolated=options.isolated_mode,
wheel_cache=wheel_cache,
skip=skip,
exclude_editable=options.exclude_editable,
)
try:
for line in freeze(**freeze_kwargs):
sys.stdout.write(line + '\n')
finally:
wheel_cache.cleanup()

View File

@@ -0,0 +1,57 @@
from __future__ import absolute_import
import hashlib
import logging
import sys
from pip._internal.basecommand import Command
from pip._internal.status_codes import ERROR
from pip._internal.utils.hashes import FAVORITE_HASH, STRONG_HASHES
from pip._internal.utils.misc import read_chunks
logger = logging.getLogger(__name__)
class HashCommand(Command):
"""
Compute a hash of a local package archive.
These can be used with --hash in a requirements file to do repeatable
installs.
"""
name = 'hash'
usage = '%prog [options] <file> ...'
summary = 'Compute hashes of package archives.'
ignore_require_venv = True
def __init__(self, *args, **kw):
super(HashCommand, self).__init__(*args, **kw)
self.cmd_opts.add_option(
'-a', '--algorithm',
dest='algorithm',
choices=STRONG_HASHES,
action='store',
default=FAVORITE_HASH,
help='The hash algorithm to use: one of %s' %
', '.join(STRONG_HASHES))
self.parser.insert_option_group(0, self.cmd_opts)
def run(self, options, args):
if not args:
self.parser.print_usage(sys.stderr)
return ERROR
algorithm = options.algorithm
for path in args:
logger.info('%s:\n--hash=%s:%s',
path, algorithm, _hash_of_file(path, algorithm))
def _hash_of_file(path, algorithm):
"""Return the hash digest of a file."""
with open(path, 'rb') as archive:
hash = hashlib.new(algorithm)
for chunk in read_chunks(archive):
hash.update(chunk)
return hash.hexdigest()

View File

@@ -0,0 +1,36 @@
from __future__ import absolute_import
from pip._internal.basecommand import SUCCESS, Command
from pip._internal.exceptions import CommandError
class HelpCommand(Command):
"""Show help for commands"""
name = 'help'
usage = """
%prog <command>"""
summary = 'Show help for commands.'
ignore_require_venv = True
def run(self, options, args):
from pip._internal.commands import commands_dict, get_similar_commands
try:
# 'pip help' with no args is handled by pip.__init__.parseopt()
cmd_name = args[0] # the command we need help for
except IndexError:
return SUCCESS
if cmd_name not in commands_dict:
guess = get_similar_commands(cmd_name)
msg = ['unknown command "%s"' % cmd_name]
if guess:
msg.append('maybe you meant "%s"' % guess)
raise CommandError(' - '.join(msg))
command = commands_dict[cmd_name]()
command.parser.print_help()
return SUCCESS

View File

@@ -0,0 +1,502 @@
from __future__ import absolute_import
import errno
import logging
import operator
import os
import shutil
from optparse import SUPPRESS_HELP
from pip._internal import cmdoptions
from pip._internal.basecommand import RequirementCommand
from pip._internal.cache import WheelCache
from pip._internal.exceptions import (
CommandError, InstallationError, PreviousBuildDirError,
)
from pip._internal.locations import distutils_scheme, virtualenv_no_global
from pip._internal.operations.check import check_install_conflicts
from pip._internal.operations.prepare import RequirementPreparer
from pip._internal.req import RequirementSet, install_given_reqs
from pip._internal.resolve import Resolver
from pip._internal.status_codes import ERROR
from pip._internal.utils.filesystem import check_path_owner
from pip._internal.utils.misc import ensure_dir, get_installed_version
from pip._internal.utils.temp_dir import TempDirectory
from pip._internal.wheel import WheelBuilder
try:
import wheel
except ImportError:
wheel = None
logger = logging.getLogger(__name__)
class InstallCommand(RequirementCommand):
"""
Install packages from:
- PyPI (and other indexes) using requirement specifiers.
- VCS project urls.
- Local project directories.
- Local or remote source archives.
pip also supports installing from "requirements files", which provide
an easy way to specify a whole environment to be installed.
"""
name = 'install'
usage = """
%prog [options] <requirement specifier> [package-index-options] ...
%prog [options] -r <requirements file> [package-index-options] ...
%prog [options] [-e] <vcs project url> ...
%prog [options] [-e] <local project path> ...
%prog [options] <archive url/path> ..."""
summary = 'Install packages.'
def __init__(self, *args, **kw):
super(InstallCommand, self).__init__(*args, **kw)
cmd_opts = self.cmd_opts
cmd_opts.add_option(cmdoptions.requirements())
cmd_opts.add_option(cmdoptions.constraints())
cmd_opts.add_option(cmdoptions.no_deps())
cmd_opts.add_option(cmdoptions.pre())
cmd_opts.add_option(cmdoptions.editable())
cmd_opts.add_option(
'-t', '--target',
dest='target_dir',
metavar='dir',
default=None,
help='Install packages into <dir>. '
'By default this will not replace existing files/folders in '
'<dir>. Use --upgrade to replace existing packages in <dir> '
'with new versions.'
)
cmd_opts.add_option(
'--user',
dest='use_user_site',
action='store_true',
help="Install to the Python user install directory for your "
"platform. Typically ~/.local/, or %APPDATA%\\Python on "
"Windows. (See the Python documentation for site.USER_BASE "
"for full details.)")
cmd_opts.add_option(
'--no-user',
dest='use_user_site',
action='store_false',
help=SUPPRESS_HELP)
cmd_opts.add_option(
'--root',
dest='root_path',
metavar='dir',
default=None,
help="Install everything relative to this alternate root "
"directory.")
cmd_opts.add_option(
'--prefix',
dest='prefix_path',
metavar='dir',
default=None,
help="Installation prefix where lib, bin and other top-level "
"folders are placed")
cmd_opts.add_option(cmdoptions.build_dir())
cmd_opts.add_option(cmdoptions.src())
cmd_opts.add_option(
'-U', '--upgrade',
dest='upgrade',
action='store_true',
help='Upgrade all specified packages to the newest available '
'version. The handling of dependencies depends on the '
'upgrade-strategy used.'
)
cmd_opts.add_option(
'--upgrade-strategy',
dest='upgrade_strategy',
default='only-if-needed',
choices=['only-if-needed', 'eager'],
help='Determines how dependency upgrading should be handled '
'[default: %default]. '
'"eager" - dependencies are upgraded regardless of '
'whether the currently installed version satisfies the '
'requirements of the upgraded package(s). '
'"only-if-needed" - are upgraded only when they do not '
'satisfy the requirements of the upgraded package(s).'
)
cmd_opts.add_option(
'--force-reinstall',
dest='force_reinstall',
action='store_true',
help='Reinstall all packages even if they are already '
'up-to-date.')
cmd_opts.add_option(
'-I', '--ignore-installed',
dest='ignore_installed',
action='store_true',
help='Ignore the installed packages (reinstalling instead).')
cmd_opts.add_option(cmdoptions.ignore_requires_python())
cmd_opts.add_option(cmdoptions.no_build_isolation())
cmd_opts.add_option(cmdoptions.install_options())
cmd_opts.add_option(cmdoptions.global_options())
cmd_opts.add_option(
"--compile",
action="store_true",
dest="compile",
default=True,
help="Compile Python source files to bytecode",
)
cmd_opts.add_option(
"--no-compile",
action="store_false",
dest="compile",
help="Do not compile Python source files to bytecode",
)
cmd_opts.add_option(
"--no-warn-script-location",
action="store_false",
dest="warn_script_location",
default=True,
help="Do not warn when installing scripts outside PATH",
)
cmd_opts.add_option(
"--no-warn-conflicts",
action="store_false",
dest="warn_about_conflicts",
default=True,
help="Do not warn about broken dependencies",
)
cmd_opts.add_option(cmdoptions.no_binary())
cmd_opts.add_option(cmdoptions.only_binary())
cmd_opts.add_option(cmdoptions.no_clean())
cmd_opts.add_option(cmdoptions.require_hashes())
cmd_opts.add_option(cmdoptions.progress_bar())
index_opts = cmdoptions.make_option_group(
cmdoptions.index_group,
self.parser,
)
self.parser.insert_option_group(0, index_opts)
self.parser.insert_option_group(0, cmd_opts)
def run(self, options, args):
cmdoptions.check_install_build_global(options)
upgrade_strategy = "to-satisfy-only"
if options.upgrade:
upgrade_strategy = options.upgrade_strategy
if options.build_dir:
options.build_dir = os.path.abspath(options.build_dir)
options.src_dir = os.path.abspath(options.src_dir)
install_options = options.install_options or []
if options.use_user_site:
if options.prefix_path:
raise CommandError(
"Can not combine '--user' and '--prefix' as they imply "
"different installation locations"
)
if virtualenv_no_global():
raise InstallationError(
"Can not perform a '--user' install. User site-packages "
"are not visible in this virtualenv."
)
install_options.append('--user')
install_options.append('--prefix=')
target_temp_dir = TempDirectory(kind="target")
if options.target_dir:
options.ignore_installed = True
options.target_dir = os.path.abspath(options.target_dir)
if (os.path.exists(options.target_dir) and not
os.path.isdir(options.target_dir)):
raise CommandError(
"Target path exists but is not a directory, will not "
"continue."
)
# Create a target directory for using with the target option
target_temp_dir.create()
install_options.append('--home=' + target_temp_dir.path)
global_options = options.global_options or []
with self._build_session(options) as session:
finder = self._build_package_finder(options, session)
build_delete = (not (options.no_clean or options.build_dir))
wheel_cache = WheelCache(options.cache_dir, options.format_control)
if options.cache_dir and not check_path_owner(options.cache_dir):
logger.warning(
"The directory '%s' or its parent directory is not owned "
"by the current user and caching wheels has been "
"disabled. check the permissions and owner of that "
"directory. If executing pip with sudo, you may want "
"sudo's -H flag.",
options.cache_dir,
)
options.cache_dir = None
with TempDirectory(
options.build_dir, delete=build_delete, kind="install"
) as directory:
requirement_set = RequirementSet(
require_hashes=options.require_hashes,
)
try:
self.populate_requirement_set(
requirement_set, args, options, finder, session,
self.name, wheel_cache
)
preparer = RequirementPreparer(
build_dir=directory.path,
src_dir=options.src_dir,
download_dir=None,
wheel_download_dir=None,
progress_bar=options.progress_bar,
build_isolation=options.build_isolation,
)
resolver = Resolver(
preparer=preparer,
finder=finder,
session=session,
wheel_cache=wheel_cache,
use_user_site=options.use_user_site,
upgrade_strategy=upgrade_strategy,
force_reinstall=options.force_reinstall,
ignore_dependencies=options.ignore_dependencies,
ignore_requires_python=options.ignore_requires_python,
ignore_installed=options.ignore_installed,
isolated=options.isolated_mode,
)
resolver.resolve(requirement_set)
# If caching is disabled or wheel is not installed don't
# try to build wheels.
if wheel and options.cache_dir:
# build wheels before install.
wb = WheelBuilder(
finder, preparer, wheel_cache,
build_options=[], global_options=[],
)
# Ignore the result: a failed wheel will be
# installed from the sdist/vcs whatever.
wb.build(
requirement_set.requirements.values(),
session=session, autobuilding=True
)
to_install = resolver.get_installation_order(
requirement_set
)
# Consistency Checking of the package set we're installing.
should_warn_about_conflicts = (
not options.ignore_dependencies and
options.warn_about_conflicts
)
if should_warn_about_conflicts:
self._warn_about_conflicts(to_install)
# Don't warn about script install locations if
# --target has been specified
warn_script_location = options.warn_script_location
if options.target_dir:
warn_script_location = False
installed = install_given_reqs(
to_install,
install_options,
global_options,
root=options.root_path,
home=target_temp_dir.path,
prefix=options.prefix_path,
pycompile=options.compile,
warn_script_location=warn_script_location,
use_user_site=options.use_user_site,
)
possible_lib_locations = get_lib_location_guesses(
user=options.use_user_site,
home=target_temp_dir.path,
root=options.root_path,
prefix=options.prefix_path,
isolated=options.isolated_mode,
)
reqs = sorted(installed, key=operator.attrgetter('name'))
items = []
for req in reqs:
item = req.name
try:
installed_version = get_installed_version(
req.name, possible_lib_locations
)
if installed_version:
item += '-' + installed_version
except Exception:
pass
items.append(item)
installed = ' '.join(items)
if installed:
logger.info('Successfully installed %s', installed)
except EnvironmentError as error:
show_traceback = (self.verbosity >= 1)
message = create_env_error_message(
error, show_traceback, options.use_user_site,
)
logger.error(message, exc_info=show_traceback)
return ERROR
except PreviousBuildDirError:
options.no_clean = True
raise
finally:
# Clean up
if not options.no_clean:
requirement_set.cleanup_files()
wheel_cache.cleanup()
if options.target_dir:
self._handle_target_dir(
options.target_dir, target_temp_dir, options.upgrade
)
return requirement_set
def _handle_target_dir(self, target_dir, target_temp_dir, upgrade):
ensure_dir(target_dir)
# Checking both purelib and platlib directories for installed
# packages to be moved to target directory
lib_dir_list = []
with target_temp_dir:
# Checking both purelib and platlib directories for installed
# packages to be moved to target directory
scheme = distutils_scheme('', home=target_temp_dir.path)
purelib_dir = scheme['purelib']
platlib_dir = scheme['platlib']
data_dir = scheme['data']
if os.path.exists(purelib_dir):
lib_dir_list.append(purelib_dir)
if os.path.exists(platlib_dir) and platlib_dir != purelib_dir:
lib_dir_list.append(platlib_dir)
if os.path.exists(data_dir):
lib_dir_list.append(data_dir)
for lib_dir in lib_dir_list:
for item in os.listdir(lib_dir):
if lib_dir == data_dir:
ddir = os.path.join(data_dir, item)
if any(s.startswith(ddir) for s in lib_dir_list[:-1]):
continue
target_item_dir = os.path.join(target_dir, item)
if os.path.exists(target_item_dir):
if not upgrade:
logger.warning(
'Target directory %s already exists. Specify '
'--upgrade to force replacement.',
target_item_dir
)
continue
if os.path.islink(target_item_dir):
logger.warning(
'Target directory %s already exists and is '
'a link. Pip will not automatically replace '
'links, please remove if replacement is '
'desired.',
target_item_dir
)
continue
if os.path.isdir(target_item_dir):
shutil.rmtree(target_item_dir)
else:
os.remove(target_item_dir)
shutil.move(
os.path.join(lib_dir, item),
target_item_dir
)
def _warn_about_conflicts(self, to_install):
package_set, _dep_info = check_install_conflicts(to_install)
missing, conflicting = _dep_info
# NOTE: There is some duplication here from pip check
for project_name in missing:
version = package_set[project_name][0]
for dependency in missing[project_name]:
logger.critical(
"%s %s requires %s, which is not installed.",
project_name, version, dependency[1],
)
for project_name in conflicting:
version = package_set[project_name][0]
for dep_name, dep_version, req in conflicting[project_name]:
logger.critical(
"%s %s has requirement %s, but you'll have %s %s which is "
"incompatible.",
project_name, version, req, dep_name, dep_version,
)
def get_lib_location_guesses(*args, **kwargs):
scheme = distutils_scheme('', *args, **kwargs)
return [scheme['purelib'], scheme['platlib']]
def create_env_error_message(error, show_traceback, using_user_site):
"""Format an error message for an EnvironmentError
It may occur anytime during the execution of the install command.
"""
parts = []
# Mention the error if we are not going to show a traceback
parts.append("Could not install packages due to an EnvironmentError")
if not show_traceback:
parts.append(": ")
parts.append(str(error))
else:
parts.append(".")
# Spilt the error indication from a helper message (if any)
parts[-1] += "\n"
# Suggest useful actions to the user:
# (1) using user site-packages or (2) verifying the permissions
if error.errno == errno.EACCES:
user_option_part = "Consider using the `--user` option"
permissions_part = "Check the permissions"
if not using_user_site:
parts.extend([
user_option_part, " or ",
permissions_part.lower(),
])
else:
parts.append(permissions_part)
parts.append(".\n")
return "".join(parts).strip() + "\n"

View File

@@ -0,0 +1,343 @@
from __future__ import absolute_import
import json
import logging
import warnings
from pip._vendor import six
from pip._vendor.six.moves import zip_longest
from pip._internal.basecommand import Command
from pip._internal.cmdoptions import index_group, make_option_group
from pip._internal.exceptions import CommandError
from pip._internal.index import PackageFinder
from pip._internal.utils.deprecation import RemovedInPip11Warning
from pip._internal.utils.misc import (
dist_is_editable, get_installed_distributions,
)
from pip._internal.utils.packaging import get_installer
logger = logging.getLogger(__name__)
class ListCommand(Command):
"""
List installed packages, including editables.
Packages are listed in a case-insensitive sorted order.
"""
name = 'list'
usage = """
%prog [options]"""
summary = 'List installed packages.'
def __init__(self, *args, **kw):
super(ListCommand, self).__init__(*args, **kw)
cmd_opts = self.cmd_opts
cmd_opts.add_option(
'-o', '--outdated',
action='store_true',
default=False,
help='List outdated packages')
cmd_opts.add_option(
'-u', '--uptodate',
action='store_true',
default=False,
help='List uptodate packages')
cmd_opts.add_option(
'-e', '--editable',
action='store_true',
default=False,
help='List editable projects.')
cmd_opts.add_option(
'-l', '--local',
action='store_true',
default=False,
help=('If in a virtualenv that has global access, do not list '
'globally-installed packages.'),
)
self.cmd_opts.add_option(
'--user',
dest='user',
action='store_true',
default=False,
help='Only output packages installed in user-site.')
cmd_opts.add_option(
'--pre',
action='store_true',
default=False,
help=("Include pre-release and development versions. By default, "
"pip only finds stable versions."),
)
cmd_opts.add_option(
'--format',
action='store',
dest='list_format',
default="columns",
choices=('legacy', 'columns', 'freeze', 'json'),
help="Select the output format among: columns (default), freeze, "
"json, or legacy.",
)
cmd_opts.add_option(
'--not-required',
action='store_true',
dest='not_required',
help="List packages that are not dependencies of "
"installed packages.",
)
cmd_opts.add_option(
'--exclude-editable',
action='store_false',
dest='include_editable',
help='Exclude editable package from output.',
)
cmd_opts.add_option(
'--include-editable',
action='store_true',
dest='include_editable',
help='Include editable package from output.',
default=True,
)
index_opts = make_option_group(index_group, self.parser)
self.parser.insert_option_group(0, index_opts)
self.parser.insert_option_group(0, cmd_opts)
def _build_package_finder(self, options, index_urls, session):
"""
Create a package finder appropriate to this list command.
"""
return PackageFinder(
find_links=options.find_links,
index_urls=index_urls,
allow_all_prereleases=options.pre,
trusted_hosts=options.trusted_hosts,
process_dependency_links=options.process_dependency_links,
session=session,
)
def run(self, options, args):
if options.list_format == "legacy":
warnings.warn(
"The legacy format has been deprecated and will be removed "
"in the future.",
RemovedInPip11Warning,
)
if options.outdated and options.uptodate:
raise CommandError(
"Options --outdated and --uptodate cannot be combined.")
packages = get_installed_distributions(
local_only=options.local,
user_only=options.user,
editables_only=options.editable,
include_editables=options.include_editable,
)
if options.outdated:
packages = self.get_outdated(packages, options)
elif options.uptodate:
packages = self.get_uptodate(packages, options)
if options.not_required:
packages = self.get_not_required(packages, options)
self.output_package_listing(packages, options)
def get_outdated(self, packages, options):
return [
dist for dist in self.iter_packages_latest_infos(packages, options)
if dist.latest_version > dist.parsed_version
]
def get_uptodate(self, packages, options):
return [
dist for dist in self.iter_packages_latest_infos(packages, options)
if dist.latest_version == dist.parsed_version
]
def get_not_required(self, packages, options):
dep_keys = set()
for dist in packages:
dep_keys.update(requirement.key for requirement in dist.requires())
return {pkg for pkg in packages if pkg.key not in dep_keys}
def iter_packages_latest_infos(self, packages, options):
index_urls = [options.index_url] + options.extra_index_urls
if options.no_index:
logger.debug('Ignoring indexes: %s', ','.join(index_urls))
index_urls = []
dependency_links = []
for dist in packages:
if dist.has_metadata('dependency_links.txt'):
dependency_links.extend(
dist.get_metadata_lines('dependency_links.txt'),
)
with self._build_session(options) as session:
finder = self._build_package_finder(options, index_urls, session)
finder.add_dependency_links(dependency_links)
for dist in packages:
typ = 'unknown'
all_candidates = finder.find_all_candidates(dist.key)
if not options.pre:
# Remove prereleases
all_candidates = [candidate for candidate in all_candidates
if not candidate.version.is_prerelease]
if not all_candidates:
continue
best_candidate = max(all_candidates,
key=finder._candidate_sort_key)
remote_version = best_candidate.version
if best_candidate.location.is_wheel:
typ = 'wheel'
else:
typ = 'sdist'
# This is dirty but makes the rest of the code much cleaner
dist.latest_version = remote_version
dist.latest_filetype = typ
yield dist
def output_legacy(self, dist, options):
if options.verbose >= 1:
return '%s (%s, %s, %s)' % (
dist.project_name,
dist.version,
dist.location,
get_installer(dist),
)
elif dist_is_editable(dist):
return '%s (%s, %s)' % (
dist.project_name,
dist.version,
dist.location,
)
else:
return '%s (%s)' % (dist.project_name, dist.version)
def output_legacy_latest(self, dist, options):
return '%s - Latest: %s [%s]' % (
self.output_legacy(dist, options),
dist.latest_version,
dist.latest_filetype,
)
def output_package_listing(self, packages, options):
packages = sorted(
packages,
key=lambda dist: dist.project_name.lower(),
)
if options.list_format == 'columns' and packages:
data, header = format_for_columns(packages, options)
self.output_package_listing_columns(data, header)
elif options.list_format == 'freeze':
for dist in packages:
if options.verbose >= 1:
logger.info("%s==%s (%s)", dist.project_name,
dist.version, dist.location)
else:
logger.info("%s==%s", dist.project_name, dist.version)
elif options.list_format == 'json':
logger.info(format_for_json(packages, options))
elif options.list_format == "legacy":
for dist in packages:
if options.outdated:
logger.info(self.output_legacy_latest(dist, options))
else:
logger.info(self.output_legacy(dist, options))
def output_package_listing_columns(self, data, header):
# insert the header first: we need to know the size of column names
if len(data) > 0:
data.insert(0, header)
pkg_strings, sizes = tabulate(data)
# Create and add a separator.
if len(data) > 0:
pkg_strings.insert(1, " ".join(map(lambda x: '-' * x, sizes)))
for val in pkg_strings:
logger.info(val)
def tabulate(vals):
# From pfmoore on GitHub:
# https://github.com/pypa/pip/issues/3651#issuecomment-216932564
assert len(vals) > 0
sizes = [0] * max(len(x) for x in vals)
for row in vals:
sizes = [max(s, len(str(c))) for s, c in zip_longest(sizes, row)]
result = []
for row in vals:
display = " ".join([str(c).ljust(s) if c is not None else ''
for s, c in zip_longest(sizes, row)])
result.append(display)
return result, sizes
def format_for_columns(pkgs, options):
"""
Convert the package data into something usable
by output_package_listing_columns.
"""
running_outdated = options.outdated
# Adjust the header for the `pip list --outdated` case.
if running_outdated:
header = ["Package", "Version", "Latest", "Type"]
else:
header = ["Package", "Version"]
data = []
if options.verbose >= 1 or any(dist_is_editable(x) for x in pkgs):
header.append("Location")
if options.verbose >= 1:
header.append("Installer")
for proj in pkgs:
# if we're working on the 'outdated' list, separate out the
# latest_version and type
row = [proj.project_name, proj.version]
if running_outdated:
row.append(proj.latest_version)
row.append(proj.latest_filetype)
if options.verbose >= 1 or dist_is_editable(proj):
row.append(proj.location)
if options.verbose >= 1:
row.append(get_installer(proj))
data.append(row)
return data, header
def format_for_json(packages, options):
data = []
for dist in packages:
info = {
'name': dist.project_name,
'version': six.text_type(dist.version),
}
if options.verbose >= 1:
info['location'] = dist.location
info['installer'] = get_installer(dist)
if options.outdated:
info['latest_version'] = six.text_type(dist.latest_version)
info['latest_filetype'] = dist.latest_filetype
data.append(info)
return json.dumps(data)

View File

@@ -0,0 +1,135 @@
from __future__ import absolute_import
import logging
import sys
import textwrap
from collections import OrderedDict
from pip._vendor import pkg_resources
from pip._vendor.packaging.version import parse as parse_version
# NOTE: XMLRPC Client is not annotated in typeshed as on 2017-07-17, which is
# why we ignore the type on this import
from pip._vendor.six.moves import xmlrpc_client # type: ignore
from pip._internal.basecommand import SUCCESS, Command
from pip._internal.compat import get_terminal_size
from pip._internal.download import PipXmlrpcTransport
from pip._internal.exceptions import CommandError
from pip._internal.models import PyPI
from pip._internal.status_codes import NO_MATCHES_FOUND
from pip._internal.utils.logging import indent_log
logger = logging.getLogger(__name__)
class SearchCommand(Command):
"""Search for PyPI packages whose name or summary contains <query>."""
name = 'search'
usage = """
%prog [options] <query>"""
summary = 'Search PyPI for packages.'
ignore_require_venv = True
def __init__(self, *args, **kw):
super(SearchCommand, self).__init__(*args, **kw)
self.cmd_opts.add_option(
'-i', '--index',
dest='index',
metavar='URL',
default=PyPI.pypi_url,
help='Base URL of Python Package Index (default %default)')
self.parser.insert_option_group(0, self.cmd_opts)
def run(self, options, args):
if not args:
raise CommandError('Missing required argument (search query).')
query = args
pypi_hits = self.search(query, options)
hits = transform_hits(pypi_hits)
terminal_width = None
if sys.stdout.isatty():
terminal_width = get_terminal_size()[0]
print_results(hits, terminal_width=terminal_width)
if pypi_hits:
return SUCCESS
return NO_MATCHES_FOUND
def search(self, query, options):
index_url = options.index
with self._build_session(options) as session:
transport = PipXmlrpcTransport(index_url, session)
pypi = xmlrpc_client.ServerProxy(index_url, transport)
hits = pypi.search({'name': query, 'summary': query}, 'or')
return hits
def transform_hits(hits):
"""
The list from pypi is really a list of versions. We want a list of
packages with the list of versions stored inline. This converts the
list from pypi into one we can use.
"""
packages = OrderedDict()
for hit in hits:
name = hit['name']
summary = hit['summary']
version = hit['version']
if name not in packages.keys():
packages[name] = {
'name': name,
'summary': summary,
'versions': [version],
}
else:
packages[name]['versions'].append(version)
# if this is the highest version, replace summary and score
if version == highest_version(packages[name]['versions']):
packages[name]['summary'] = summary
return list(packages.values())
def print_results(hits, name_column_width=None, terminal_width=None):
if not hits:
return
if name_column_width is None:
name_column_width = max([
len(hit['name']) + len(highest_version(hit.get('versions', ['-'])))
for hit in hits
]) + 4
installed_packages = [p.project_name for p in pkg_resources.working_set]
for hit in hits:
name = hit['name']
summary = hit['summary'] or ''
latest = highest_version(hit.get('versions', ['-']))
if terminal_width is not None:
target_width = terminal_width - name_column_width - 5
if target_width > 10:
# wrap and indent summary to fit terminal
summary = textwrap.wrap(summary, target_width)
summary = ('\n' + ' ' * (name_column_width + 3)).join(summary)
line = '%-*s - %s' % (name_column_width,
'%s (%s)' % (name, latest), summary)
try:
logger.info(line)
if name in installed_packages:
dist = pkg_resources.get_distribution(name)
with indent_log():
if dist.version == latest:
logger.info('INSTALLED: %s (latest)', dist.version)
else:
logger.info('INSTALLED: %s', dist.version)
logger.info('LATEST: %s', latest)
except UnicodeEncodeError:
pass
def highest_version(versions):
return max(versions, key=parse_version)

View File

@@ -0,0 +1,164 @@
from __future__ import absolute_import
import logging
import os
from email.parser import FeedParser # type: ignore
from pip._vendor import pkg_resources
from pip._vendor.packaging.utils import canonicalize_name
from pip._internal.basecommand import Command
from pip._internal.status_codes import ERROR, SUCCESS
logger = logging.getLogger(__name__)
class ShowCommand(Command):
"""Show information about one or more installed packages."""
name = 'show'
usage = """
%prog [options] <package> ..."""
summary = 'Show information about installed packages.'
ignore_require_venv = True
def __init__(self, *args, **kw):
super(ShowCommand, self).__init__(*args, **kw)
self.cmd_opts.add_option(
'-f', '--files',
dest='files',
action='store_true',
default=False,
help='Show the full list of installed files for each package.')
self.parser.insert_option_group(0, self.cmd_opts)
def run(self, options, args):
if not args:
logger.warning('ERROR: Please provide a package name or names.')
return ERROR
query = args
results = search_packages_info(query)
if not print_results(
results, list_files=options.files, verbose=options.verbose):
return ERROR
return SUCCESS
def search_packages_info(query):
"""
Gather details from installed distributions. Print distribution name,
version, location, and installed files. Installed files requires a
pip generated 'installed-files.txt' in the distributions '.egg-info'
directory.
"""
installed = {}
for p in pkg_resources.working_set:
installed[canonicalize_name(p.project_name)] = p
query_names = [canonicalize_name(name) for name in query]
for dist in [installed[pkg] for pkg in query_names if pkg in installed]:
package = {
'name': dist.project_name,
'version': dist.version,
'location': dist.location,
'requires': [dep.project_name for dep in dist.requires()],
}
file_list = None
metadata = None
if isinstance(dist, pkg_resources.DistInfoDistribution):
# RECORDs should be part of .dist-info metadatas
if dist.has_metadata('RECORD'):
lines = dist.get_metadata_lines('RECORD')
paths = [l.split(',')[0] for l in lines]
paths = [os.path.join(dist.location, p) for p in paths]
file_list = [os.path.relpath(p, dist.location) for p in paths]
if dist.has_metadata('METADATA'):
metadata = dist.get_metadata('METADATA')
else:
# Otherwise use pip's log for .egg-info's
if dist.has_metadata('installed-files.txt'):
paths = dist.get_metadata_lines('installed-files.txt')
paths = [os.path.join(dist.egg_info, p) for p in paths]
file_list = [os.path.relpath(p, dist.location) for p in paths]
if dist.has_metadata('PKG-INFO'):
metadata = dist.get_metadata('PKG-INFO')
if dist.has_metadata('entry_points.txt'):
entry_points = dist.get_metadata_lines('entry_points.txt')
package['entry_points'] = entry_points
if dist.has_metadata('INSTALLER'):
for line in dist.get_metadata_lines('INSTALLER'):
if line.strip():
package['installer'] = line.strip()
break
# @todo: Should pkg_resources.Distribution have a
# `get_pkg_info` method?
feed_parser = FeedParser()
feed_parser.feed(metadata)
pkg_info_dict = feed_parser.close()
for key in ('metadata-version', 'summary',
'home-page', 'author', 'author-email', 'license'):
package[key] = pkg_info_dict.get(key)
# It looks like FeedParser cannot deal with repeated headers
classifiers = []
for line in metadata.splitlines():
if line.startswith('Classifier: '):
classifiers.append(line[len('Classifier: '):])
package['classifiers'] = classifiers
if file_list:
package['files'] = sorted(file_list)
yield package
def print_results(distributions, list_files=False, verbose=False):
"""
Print the informations from installed distributions found.
"""
results_printed = False
for i, dist in enumerate(distributions):
results_printed = True
if i > 0:
logger.info("---")
name = dist.get('name', '')
required_by = [
pkg.project_name for pkg in pkg_resources.working_set
if name in [required.name for required in pkg.requires()]
]
logger.info("Name: %s", name)
logger.info("Version: %s", dist.get('version', ''))
logger.info("Summary: %s", dist.get('summary', ''))
logger.info("Home-page: %s", dist.get('home-page', ''))
logger.info("Author: %s", dist.get('author', ''))
logger.info("Author-email: %s", dist.get('author-email', ''))
logger.info("License: %s", dist.get('license', ''))
logger.info("Location: %s", dist.get('location', ''))
logger.info("Requires: %s", ', '.join(dist.get('requires', [])))
logger.info("Required-by: %s", ', '.join(required_by))
if verbose:
logger.info("Metadata-Version: %s",
dist.get('metadata-version', ''))
logger.info("Installer: %s", dist.get('installer', ''))
logger.info("Classifiers:")
for classifier in dist.get('classifiers', []):
logger.info(" %s", classifier)
logger.info("Entry-points:")
for entry in dist.get('entry_points', []):
logger.info(" %s", entry.strip())
if list_files:
logger.info("Files:")
for line in dist.get('files', []):
logger.info(" %s", line.strip())
if "files" not in dist:
logger.info("Cannot locate installed-files.txt")
return results_printed

View File

@@ -0,0 +1,71 @@
from __future__ import absolute_import
from pip._vendor.packaging.utils import canonicalize_name
from pip._internal.basecommand import Command
from pip._internal.exceptions import InstallationError
from pip._internal.req import InstallRequirement, parse_requirements
class UninstallCommand(Command):
"""
Uninstall packages.
pip is able to uninstall most installed packages. Known exceptions are:
- Pure distutils packages installed with ``python setup.py install``, which
leave behind no metadata to determine what files were installed.
- Script wrappers installed by ``python setup.py develop``.
"""
name = 'uninstall'
usage = """
%prog [options] <package> ...
%prog [options] -r <requirements file> ..."""
summary = 'Uninstall packages.'
def __init__(self, *args, **kw):
super(UninstallCommand, self).__init__(*args, **kw)
self.cmd_opts.add_option(
'-r', '--requirement',
dest='requirements',
action='append',
default=[],
metavar='file',
help='Uninstall all the packages listed in the given requirements '
'file. This option can be used multiple times.',
)
self.cmd_opts.add_option(
'-y', '--yes',
dest='yes',
action='store_true',
help="Don't ask for confirmation of uninstall deletions.")
self.parser.insert_option_group(0, self.cmd_opts)
def run(self, options, args):
with self._build_session(options) as session:
reqs_to_uninstall = {}
for name in args:
req = InstallRequirement.from_line(
name, isolated=options.isolated_mode,
)
if req.name:
reqs_to_uninstall[canonicalize_name(req.name)] = req
for filename in options.requirements:
for req in parse_requirements(
filename,
options=options,
session=session):
if req.name:
reqs_to_uninstall[canonicalize_name(req.name)] = req
if not reqs_to_uninstall:
raise InstallationError(
'You must give at least one requirement to %(name)s (see '
'"pip help %(name)s")' % dict(name=self.name)
)
for req in reqs_to_uninstall.values():
uninstall_pathset = req.uninstall(
auto_confirm=options.yes, verbose=self.verbosity > 0,
)
if uninstall_pathset:
uninstall_pathset.commit()

View File

@@ -0,0 +1,179 @@
# -*- coding: utf-8 -*-
from __future__ import absolute_import
import logging
import os
from pip._internal import cmdoptions
from pip._internal.basecommand import RequirementCommand
from pip._internal.cache import WheelCache
from pip._internal.exceptions import CommandError, PreviousBuildDirError
from pip._internal.operations.prepare import RequirementPreparer
from pip._internal.req import RequirementSet
from pip._internal.resolve import Resolver
from pip._internal.utils.temp_dir import TempDirectory
from pip._internal.wheel import WheelBuilder
logger = logging.getLogger(__name__)
class WheelCommand(RequirementCommand):
"""
Build Wheel archives for your requirements and dependencies.
Wheel is a built-package format, and offers the advantage of not
recompiling your software during every install. For more details, see the
wheel docs: https://wheel.readthedocs.io/en/latest/
Requirements: setuptools>=0.8, and wheel.
'pip wheel' uses the bdist_wheel setuptools extension from the wheel
package to build individual wheels.
"""
name = 'wheel'
usage = """
%prog [options] <requirement specifier> ...
%prog [options] -r <requirements file> ...
%prog [options] [-e] <vcs project url> ...
%prog [options] [-e] <local project path> ...
%prog [options] <archive url/path> ..."""
summary = 'Build wheels from your requirements.'
def __init__(self, *args, **kw):
super(WheelCommand, self).__init__(*args, **kw)
cmd_opts = self.cmd_opts
cmd_opts.add_option(
'-w', '--wheel-dir',
dest='wheel_dir',
metavar='dir',
default=os.curdir,
help=("Build wheels into <dir>, where the default is the "
"current working directory."),
)
cmd_opts.add_option(cmdoptions.no_binary())
cmd_opts.add_option(cmdoptions.only_binary())
cmd_opts.add_option(
'--build-option',
dest='build_options',
metavar='options',
action='append',
help="Extra arguments to be supplied to 'setup.py bdist_wheel'.",
)
cmd_opts.add_option(cmdoptions.no_build_isolation())
cmd_opts.add_option(cmdoptions.constraints())
cmd_opts.add_option(cmdoptions.editable())
cmd_opts.add_option(cmdoptions.requirements())
cmd_opts.add_option(cmdoptions.src())
cmd_opts.add_option(cmdoptions.ignore_requires_python())
cmd_opts.add_option(cmdoptions.no_deps())
cmd_opts.add_option(cmdoptions.build_dir())
cmd_opts.add_option(cmdoptions.progress_bar())
cmd_opts.add_option(
'--global-option',
dest='global_options',
action='append',
metavar='options',
help="Extra global options to be supplied to the setup.py "
"call before the 'bdist_wheel' command.")
cmd_opts.add_option(
'--pre',
action='store_true',
default=False,
help=("Include pre-release and development versions. By default, "
"pip only finds stable versions."),
)
cmd_opts.add_option(cmdoptions.no_clean())
cmd_opts.add_option(cmdoptions.require_hashes())
index_opts = cmdoptions.make_option_group(
cmdoptions.index_group,
self.parser,
)
self.parser.insert_option_group(0, index_opts)
self.parser.insert_option_group(0, cmd_opts)
def run(self, options, args):
cmdoptions.check_install_build_global(options)
index_urls = [options.index_url] + options.extra_index_urls
if options.no_index:
logger.debug('Ignoring indexes: %s', ','.join(index_urls))
index_urls = []
if options.build_dir:
options.build_dir = os.path.abspath(options.build_dir)
options.src_dir = os.path.abspath(options.src_dir)
with self._build_session(options) as session:
finder = self._build_package_finder(options, session)
build_delete = (not (options.no_clean or options.build_dir))
wheel_cache = WheelCache(options.cache_dir, options.format_control)
with TempDirectory(
options.build_dir, delete=build_delete, kind="wheel"
) as directory:
requirement_set = RequirementSet(
require_hashes=options.require_hashes,
)
try:
self.populate_requirement_set(
requirement_set, args, options, finder, session,
self.name, wheel_cache
)
preparer = RequirementPreparer(
build_dir=directory.path,
src_dir=options.src_dir,
download_dir=None,
wheel_download_dir=options.wheel_dir,
progress_bar=options.progress_bar,
build_isolation=options.build_isolation,
)
resolver = Resolver(
preparer=preparer,
finder=finder,
session=session,
wheel_cache=wheel_cache,
use_user_site=False,
upgrade_strategy="to-satisfy-only",
force_reinstall=False,
ignore_dependencies=options.ignore_dependencies,
ignore_requires_python=options.ignore_requires_python,
ignore_installed=True,
isolated=options.isolated_mode,
)
resolver.resolve(requirement_set)
# build wheels
wb = WheelBuilder(
finder, preparer, wheel_cache,
build_options=options.build_options or [],
global_options=options.global_options or [],
no_clean=options.no_clean,
)
wheels_built_successfully = wb.build(
requirement_set.requirements.values(), session=session,
)
if not wheels_built_successfully:
raise CommandError(
"Failed to build one or more wheels"
)
except PreviousBuildDirError:
options.no_clean = True
raise
finally:
if not options.no_clean:
requirement_set.cleanup_files()
wheel_cache.cleanup()

View File

@@ -0,0 +1,235 @@
"""Stuff that differs in different Python versions and platform
distributions."""
from __future__ import absolute_import, division
import codecs
import locale
import logging
import os
import shutil
import sys
from pip._vendor.six import text_type
try:
import ipaddress
except ImportError:
try:
from pip._vendor import ipaddress # type: ignore
except ImportError:
import ipaddr as ipaddress # type: ignore
ipaddress.ip_address = ipaddress.IPAddress
ipaddress.ip_network = ipaddress.IPNetwork
__all__ = [
"ipaddress", "uses_pycache", "console_to_str", "native_str",
"get_path_uid", "stdlib_pkgs", "WINDOWS", "samefile", "get_terminal_size",
]
logger = logging.getLogger(__name__)
if sys.version_info >= (3, 4):
uses_pycache = True
from importlib.util import cache_from_source
else:
import imp
try:
cache_from_source = imp.cache_from_source # type: ignore
except AttributeError:
# does not use __pycache__
cache_from_source = None
uses_pycache = cache_from_source is not None
if sys.version_info >= (3, 5):
backslashreplace_decode = "backslashreplace"
else:
# In version 3.4 and older, backslashreplace exists
# but does not support use for decoding.
# We implement our own replace handler for this
# situation, so that we can consistently use
# backslash replacement for all versions.
def backslashreplace_decode_fn(err):
raw_bytes = (err.object[i] for i in range(err.start, err.end))
if sys.version_info[0] == 2:
# Python 2 gave us characters - convert to numeric bytes
raw_bytes = (ord(b) for b in raw_bytes)
return u"".join(u"\\x%x" % c for c in raw_bytes), err.end
codecs.register_error(
"backslashreplace_decode",
backslashreplace_decode_fn,
)
backslashreplace_decode = "backslashreplace_decode"
def console_to_str(data):
"""Return a string, safe for output, of subprocess output.
We assume the data is in the locale preferred encoding.
If it won't decode properly, we warn the user but decode as
best we can.
We also ensure that the output can be safely written to
standard output without encoding errors.
"""
# First, get the encoding we assume. This is the preferred
# encoding for the locale, unless that is not found, or
# it is ASCII, in which case assume UTF-8
encoding = locale.getpreferredencoding()
if (not encoding) or codecs.lookup(encoding).name == "ascii":
encoding = "utf-8"
# Now try to decode the data - if we fail, warn the user and
# decode with replacement.
try:
s = data.decode(encoding)
except UnicodeDecodeError:
logger.warning(
"Subprocess output does not appear to be encoded as %s",
encoding,
)
s = data.decode(encoding, errors=backslashreplace_decode)
# Make sure we can print the output, by encoding it to the output
# encoding with replacement of unencodable characters, and then
# decoding again.
# We use stderr's encoding because it's less likely to be
# redirected and if we don't find an encoding we skip this
# step (on the assumption that output is wrapped by something
# that won't fail).
# The double getattr is to deal with the possibility that we're
# being called in a situation where sys.__stderr__ doesn't exist,
# or doesn't have an encoding attribute. Neither of these cases
# should occur in normal pip use, but there's no harm in checking
# in case people use pip in (unsupported) unusual situations.
output_encoding = getattr(getattr(sys, "__stderr__", None),
"encoding", None)
if output_encoding:
s = s.encode(output_encoding, errors="backslashreplace")
s = s.decode(output_encoding)
return s
if sys.version_info >= (3,):
def native_str(s, replace=False):
if isinstance(s, bytes):
return s.decode('utf-8', 'replace' if replace else 'strict')
return s
else:
def native_str(s, replace=False):
# Replace is ignored -- unicode to UTF-8 can't fail
if isinstance(s, text_type):
return s.encode('utf-8')
return s
def get_path_uid(path):
"""
Return path's uid.
Does not follow symlinks:
https://github.com/pypa/pip/pull/935#discussion_r5307003
Placed this function in compat due to differences on AIX and
Jython, that should eventually go away.
:raises OSError: When path is a symlink or can't be read.
"""
if hasattr(os, 'O_NOFOLLOW'):
fd = os.open(path, os.O_RDONLY | os.O_NOFOLLOW)
file_uid = os.fstat(fd).st_uid
os.close(fd)
else: # AIX and Jython
# WARNING: time of check vulnerability, but best we can do w/o NOFOLLOW
if not os.path.islink(path):
# older versions of Jython don't have `os.fstat`
file_uid = os.stat(path).st_uid
else:
# raise OSError for parity with os.O_NOFOLLOW above
raise OSError(
"%s is a symlink; Will not return uid for symlinks" % path
)
return file_uid
def expanduser(path):
"""
Expand ~ and ~user constructions.
Includes a workaround for http://bugs.python.org/issue14768
"""
expanded = os.path.expanduser(path)
if path.startswith('~/') and expanded.startswith('//'):
expanded = expanded[1:]
return expanded
# packages in the stdlib that may have installation metadata, but should not be
# considered 'installed'. this theoretically could be determined based on
# dist.location (py27:`sysconfig.get_paths()['stdlib']`,
# py26:sysconfig.get_config_vars('LIBDEST')), but fear platform variation may
# make this ineffective, so hard-coding
stdlib_pkgs = {"python", "wsgiref", "argparse"}
# windows detection, covers cpython and ironpython
WINDOWS = (sys.platform.startswith("win") or
(sys.platform == 'cli' and os.name == 'nt'))
def samefile(file1, file2):
"""Provide an alternative for os.path.samefile on Windows/Python2"""
if hasattr(os.path, 'samefile'):
return os.path.samefile(file1, file2)
else:
path1 = os.path.normcase(os.path.abspath(file1))
path2 = os.path.normcase(os.path.abspath(file2))
return path1 == path2
if hasattr(shutil, 'get_terminal_size'):
def get_terminal_size():
"""
Returns a tuple (x, y) representing the width(x) and the height(y)
in characters of the terminal window.
"""
return tuple(shutil.get_terminal_size())
else:
def get_terminal_size():
"""
Returns a tuple (x, y) representing the width(x) and the height(y)
in characters of the terminal window.
"""
def ioctl_GWINSZ(fd):
try:
import fcntl
import termios
import struct
cr = struct.unpack_from(
'hh',
fcntl.ioctl(fd, termios.TIOCGWINSZ, '12345678')
)
except:
return None
if cr == (0, 0):
return None
return cr
cr = ioctl_GWINSZ(0) or ioctl_GWINSZ(1) or ioctl_GWINSZ(2)
if not cr:
try:
fd = os.open(os.ctermid(), os.O_RDONLY)
cr = ioctl_GWINSZ(fd)
os.close(fd)
except:
pass
if not cr:
cr = (os.environ.get('LINES', 25), os.environ.get('COLUMNS', 80))
return int(cr[1]), int(cr[0])

View File

@@ -0,0 +1,378 @@
"""Configuration management setup
Some terminology:
- name
As written in config files.
- value
Value associated with a name
- key
Name combined with it's section (section.name)
- variant
A single word describing where the configuration key-value pair came from
"""
import locale
import logging
import os
from pip._vendor import six
from pip._vendor.six.moves import configparser
from pip._internal.exceptions import ConfigurationError
from pip._internal.locations import (
legacy_config_file, new_config_file, running_under_virtualenv,
site_config_files, venv_config_file,
)
from pip._internal.utils.misc import ensure_dir, enum
from pip._internal.utils.typing import MYPY_CHECK_RUNNING
if MYPY_CHECK_RUNNING:
from typing import Any, Dict, Iterable, List, NewType, Optional, Tuple
RawConfigParser = configparser.RawConfigParser # Shorthand
Kind = NewType("Kind", str)
logger = logging.getLogger(__name__)
# NOTE: Maybe use the optionx attribute to normalize keynames.
def _normalize_name(name):
# type: (str) -> str
"""Make a name consistent regardless of source (environment or file)
"""
name = name.lower().replace('_', '-')
if name.startswith('--'):
name = name[2:] # only prefer long opts
return name
def _disassemble_key(name):
# type: (str) -> List[str]
return name.split(".", 1)
# The kinds of configurations there are.
kinds = enum(
USER="user", # User Specific
GLOBAL="global", # System Wide
VENV="venv", # Virtual Environment Specific
ENV="env", # from PIP_CONFIG_FILE
ENV_VAR="env-var", # from Environment Variables
)
class Configuration(object):
"""Handles management of configuration.
Provides an interface to accessing and managing configuration files.
This class converts provides an API that takes "section.key-name" style
keys and stores the value associated with it as "key-name" under the
section "section".
This allows for a clean interface wherein the both the section and the
key-name are preserved in an easy to manage form in the configuration files
and the data stored is also nice.
"""
def __init__(self, isolated, load_only=None):
# type: (bool, Kind) -> None
super(Configuration, self).__init__()
_valid_load_only = [kinds.USER, kinds.GLOBAL, kinds.VENV, None]
if load_only not in _valid_load_only:
raise ConfigurationError(
"Got invalid value for load_only - should be one of {}".format(
", ".join(map(repr, _valid_load_only[:-1]))
)
)
self.isolated = isolated # type: bool
self.load_only = load_only # type: Optional[Kind]
# The order here determines the override order.
self._override_order = [
kinds.GLOBAL, kinds.USER, kinds.VENV, kinds.ENV, kinds.ENV_VAR
]
self._ignore_env_names = ["version", "help"]
# Because we keep track of where we got the data from
self._parsers = {
variant: [] for variant in self._override_order
} # type: Dict[Kind, List[Tuple[str, RawConfigParser]]]
self._config = {
variant: {} for variant in self._override_order
} # type: Dict[Kind, Dict[str, Any]]
self._modified_parsers = [] # type: List[Tuple[str, RawConfigParser]]
def load(self):
# type: () -> None
"""Loads configuration from configuration files and environment
"""
self._load_config_files()
if not self.isolated:
self._load_environment_vars()
def get_file_to_edit(self):
# type: () -> Optional[str]
"""Returns the file with highest priority in configuration
"""
assert self.load_only is not None, \
"Need to be specified a file to be editing"
try:
return self._get_parser_to_modify()[0]
except IndexError:
return None
def items(self):
# type: () -> Iterable[Tuple[str, Any]]
"""Returns key-value pairs like dict.items() representing the loaded
configuration
"""
return self._dictionary.items()
def get_value(self, key):
# type: (str) -> Any
"""Get a value from the configuration.
"""
try:
return self._dictionary[key]
except KeyError:
raise ConfigurationError("No such key - {}".format(key))
def set_value(self, key, value):
# type: (str, Any) -> None
"""Modify a value in the configuration.
"""
self._ensure_have_load_only()
fname, parser = self._get_parser_to_modify()
if parser is not None:
section, name = _disassemble_key(key)
# Modify the parser and the configuration
if not parser.has_section(section):
parser.add_section(section)
parser.set(section, name, value)
self._config[self.load_only][key] = value
self._mark_as_modified(fname, parser)
def unset_value(self, key):
# type: (str) -> None
"""Unset a value in the configuration.
"""
self._ensure_have_load_only()
if key not in self._config[self.load_only]:
raise ConfigurationError("No such key - {}".format(key))
fname, parser = self._get_parser_to_modify()
if parser is not None:
section, name = _disassemble_key(key)
# Remove the key in the parser
modified_something = False
if parser.has_section(section):
# Returns whether the option was removed or not
modified_something = parser.remove_option(section, name)
if modified_something:
# name removed from parser, section may now be empty
section_iter = iter(parser.items(section))
try:
val = six.next(section_iter)
except StopIteration:
val = None
if val is None:
parser.remove_section(section)
self._mark_as_modified(fname, parser)
else:
raise ConfigurationError(
"Fatal Internal error [id=1]. Please report as a bug."
)
del self._config[self.load_only][key]
def save(self):
# type: () -> None
"""Save the currentin-memory state.
"""
self._ensure_have_load_only()
for fname, parser in self._modified_parsers:
logger.info("Writing to %s", fname)
# Ensure directory exists.
ensure_dir(os.path.dirname(fname))
with open(fname, "w") as f:
parser.write(f) # type: ignore
#
# Private routines
#
def _ensure_have_load_only(self):
# type: () -> None
if self.load_only is None:
raise ConfigurationError("Needed a specific file to be modifying.")
logger.debug("Will be working with %s variant only", self.load_only)
@property
def _dictionary(self):
# type: () -> Dict[str, Any]
"""A dictionary representing the loaded configuration.
"""
# NOTE: Dictionaries are not populated if not loaded. So, conditionals
# are not needed here.
retval = {}
for variant in self._override_order:
retval.update(self._config[variant])
return retval
def _load_config_files(self):
# type: () -> None
"""Loads configuration from configuration files
"""
config_files = dict(self._iter_config_files())
if config_files[kinds.ENV][0:1] == [os.devnull]:
logger.debug(
"Skipping loading configuration files due to "
"environment's PIP_CONFIG_FILE being os.devnull"
)
return
for variant, files in config_files.items():
for fname in files:
# If there's specific variant set in `load_only`, load only
# that variant, not the others.
if self.load_only is not None and variant != self.load_only:
logger.debug(
"Skipping file '%s' (variant: %s)", fname, variant
)
continue
parser = self._load_file(variant, fname)
# Keeping track of the parsers used
self._parsers[variant].append((fname, parser))
def _load_file(self, variant, fname):
# type: (Kind, str) -> RawConfigParser
logger.debug("For variant '%s', will try loading '%s'", variant, fname)
parser = self._construct_parser(fname)
for section in parser.sections():
items = parser.items(section)
self._config[variant].update(self._normalized_keys(section, items))
return parser
def _construct_parser(self, fname):
# type: (str) -> RawConfigParser
parser = configparser.RawConfigParser()
# If there is no such file, don't bother reading it but create the
# parser anyway, to hold the data.
# Doing this is useful when modifying and saving files, where we don't
# need to construct a parser.
if os.path.exists(fname):
try:
parser.read(fname)
except UnicodeDecodeError:
raise ConfigurationError((
"ERROR: "
"Configuration file contains invalid %s characters.\n"
"Please fix your configuration, located at %s\n"
) % (locale.getpreferredencoding(False), fname))
return parser
def _load_environment_vars(self):
# type: () -> None
"""Loads configuration from environment variables
"""
self._config[kinds.ENV_VAR].update(
self._normalized_keys(":env:", self._get_environ_vars())
)
def _normalized_keys(self, section, items):
# type: (str, Iterable[Tuple[str, Any]]) -> Dict[str, Any]
"""Normalizes items to construct a dictionary with normalized keys.
This routine is where the names become keys and are made the same
regardless of source - configuration files or environment.
"""
normalized = {}
for name, val in items:
key = section + "." + _normalize_name(name)
normalized[key] = val
return normalized
def _get_environ_vars(self):
# type: () -> Iterable[Tuple[str, str]]
"""Returns a generator with all environmental vars with prefix PIP_"""
for key, val in os.environ.items():
should_be_yielded = (
key.startswith("PIP_") and
key[4:].lower() not in self._ignore_env_names
)
if should_be_yielded:
yield key[4:].lower(), val
# XXX: This is patched in the tests.
def _iter_config_files(self):
# type: () -> Iterable[Tuple[Kind, List[str]]]
"""Yields variant and configuration files associated with it.
This should be treated like items of a dictionary.
"""
# SMELL: Move the conditions out of this function
# environment variables have the lowest priority
config_file = os.environ.get('PIP_CONFIG_FILE', None)
if config_file is not None:
yield kinds.ENV, [config_file]
else:
yield kinds.ENV, []
# at the base we have any global configuration
yield kinds.GLOBAL, list(site_config_files)
# per-user configuration next
should_load_user_config = not self.isolated and not (
config_file and os.path.exists(config_file)
)
if should_load_user_config:
# The legacy config file is overridden by the new config file
yield kinds.USER, [legacy_config_file, new_config_file]
# finally virtualenv configuration first trumping others
if running_under_virtualenv():
yield kinds.VENV, [venv_config_file]
def _get_parser_to_modify(self):
# type: () -> Tuple[str, RawConfigParser]
# Determine which parser to modify
parsers = self._parsers[self.load_only]
if not parsers:
# This should not happen if everything works correctly.
raise ConfigurationError(
"Fatal Internal error [id=2]. Please report as a bug."
)
# Use the highest priority parser.
return parsers[-1]
# XXX: This is patched in the tests.
def _mark_as_modified(self, fname, parser):
# type: (str, RawConfigParser) -> None
file_parser_tuple = (fname, parser)
if file_parser_tuple not in self._modified_parsers:
self._modified_parsers.append(file_parser_tuple)

View File

@@ -0,0 +1,922 @@
from __future__ import absolute_import
import cgi
import email.utils
import getpass
import json
import logging
import mimetypes
import os
import platform
import re
import shutil
import sys
from pip._vendor import requests, six, urllib3
from pip._vendor.cachecontrol import CacheControlAdapter
from pip._vendor.cachecontrol.caches import FileCache
from pip._vendor.lockfile import LockError
from pip._vendor.requests.adapters import BaseAdapter, HTTPAdapter
from pip._vendor.requests.auth import AuthBase, HTTPBasicAuth
from pip._vendor.requests.models import CONTENT_CHUNK_SIZE, Response
from pip._vendor.requests.structures import CaseInsensitiveDict
from pip._vendor.requests.utils import get_netrc_auth
# NOTE: XMLRPC Client is not annotated in typeshed as on 2017-07-17, which is
# why we ignore the type on this import
from pip._vendor.six.moves import xmlrpc_client # type: ignore
from pip._vendor.six.moves.urllib import parse as urllib_parse
from pip._vendor.six.moves.urllib import request as urllib_request
from pip._vendor.six.moves.urllib.parse import unquote as urllib_unquote
from pip._vendor.urllib3.util import IS_PYOPENSSL
import pip
from pip._internal.compat import WINDOWS
from pip._internal.exceptions import HashMismatch, InstallationError
from pip._internal.locations import write_delete_marker_file
from pip._internal.models import PyPI
from pip._internal.utils.encoding import auto_decode
from pip._internal.utils.filesystem import check_path_owner
from pip._internal.utils.glibc import libc_ver
from pip._internal.utils.logging import indent_log
from pip._internal.utils.misc import (
ARCHIVE_EXTENSIONS, ask_path_exists, backup_dir, call_subprocess, consume,
display_path, format_size, get_installed_version, rmtree, splitext,
unpack_file,
)
from pip._internal.utils.setuptools_build import SETUPTOOLS_SHIM
from pip._internal.utils.temp_dir import TempDirectory
from pip._internal.utils.ui import DownloadProgressProvider
from pip._internal.vcs import vcs
try:
import ssl # noqa
except ImportError:
ssl = None
HAS_TLS = (ssl is not None) or IS_PYOPENSSL
__all__ = ['get_file_content',
'is_url', 'url_to_path', 'path_to_url',
'is_archive_file', 'unpack_vcs_link',
'unpack_file_url', 'is_vcs_url', 'is_file_url',
'unpack_http_url', 'unpack_url']
logger = logging.getLogger(__name__)
def user_agent():
"""
Return a string representing the user agent.
"""
data = {
"installer": {"name": "pip", "version": pip.__version__},
"python": platform.python_version(),
"implementation": {
"name": platform.python_implementation(),
},
}
if data["implementation"]["name"] == 'CPython':
data["implementation"]["version"] = platform.python_version()
elif data["implementation"]["name"] == 'PyPy':
if sys.pypy_version_info.releaselevel == 'final':
pypy_version_info = sys.pypy_version_info[:3]
else:
pypy_version_info = sys.pypy_version_info
data["implementation"]["version"] = ".".join(
[str(x) for x in pypy_version_info]
)
elif data["implementation"]["name"] == 'Jython':
# Complete Guess
data["implementation"]["version"] = platform.python_version()
elif data["implementation"]["name"] == 'IronPython':
# Complete Guess
data["implementation"]["version"] = platform.python_version()
if sys.platform.startswith("linux"):
from pip._vendor import distro
distro_infos = dict(filter(
lambda x: x[1],
zip(["name", "version", "id"], distro.linux_distribution()),
))
libc = dict(filter(
lambda x: x[1],
zip(["lib", "version"], libc_ver()),
))
if libc:
distro_infos["libc"] = libc
if distro_infos:
data["distro"] = distro_infos
if sys.platform.startswith("darwin") and platform.mac_ver()[0]:
data["distro"] = {"name": "macOS", "version": platform.mac_ver()[0]}
if platform.system():
data.setdefault("system", {})["name"] = platform.system()
if platform.release():
data.setdefault("system", {})["release"] = platform.release()
if platform.machine():
data["cpu"] = platform.machine()
if HAS_TLS:
data["openssl_version"] = ssl.OPENSSL_VERSION
setuptools_version = get_installed_version("setuptools")
if setuptools_version is not None:
data["setuptools_version"] = setuptools_version
return "{data[installer][name]}/{data[installer][version]} {json}".format(
data=data,
json=json.dumps(data, separators=(",", ":"), sort_keys=True),
)
class MultiDomainBasicAuth(AuthBase):
def __init__(self, prompting=True):
self.prompting = prompting
self.passwords = {}
def __call__(self, req):
parsed = urllib_parse.urlparse(req.url)
# Get the netloc without any embedded credentials
netloc = parsed.netloc.rsplit("@", 1)[-1]
# Set the url of the request to the url without any credentials
req.url = urllib_parse.urlunparse(parsed[:1] + (netloc,) + parsed[2:])
# Use any stored credentials that we have for this netloc
username, password = self.passwords.get(netloc, (None, None))
# Extract credentials embedded in the url if we have none stored
if username is None:
username, password = self.parse_credentials(parsed.netloc)
# Get creds from netrc if we still don't have them
if username is None and password is None:
netrc_auth = get_netrc_auth(req.url)
username, password = netrc_auth if netrc_auth else (None, None)
if username or password:
# Store the username and password
self.passwords[netloc] = (username, password)
# Send the basic auth with this request
req = HTTPBasicAuth(username or "", password or "")(req)
# Attach a hook to handle 401 responses
req.register_hook("response", self.handle_401)
return req
def handle_401(self, resp, **kwargs):
# We only care about 401 responses, anything else we want to just
# pass through the actual response
if resp.status_code != 401:
return resp
# We are not able to prompt the user so simply return the response
if not self.prompting:
return resp
parsed = urllib_parse.urlparse(resp.url)
# Prompt the user for a new username and password
username = six.moves.input("User for %s: " % parsed.netloc)
password = getpass.getpass("Password: ")
# Store the new username and password to use for future requests
if username or password:
self.passwords[parsed.netloc] = (username, password)
# Consume content and release the original connection to allow our new
# request to reuse the same one.
resp.content
resp.raw.release_conn()
# Add our new username and password to the request
req = HTTPBasicAuth(username or "", password or "")(resp.request)
# Send our new request
new_resp = resp.connection.send(req, **kwargs)
new_resp.history.append(resp)
return new_resp
def parse_credentials(self, netloc):
if "@" in netloc:
userinfo = netloc.rsplit("@", 1)[0]
if ":" in userinfo:
user, pwd = userinfo.split(":", 1)
return (urllib_unquote(user), urllib_unquote(pwd))
return urllib_unquote(userinfo), None
return None, None
class LocalFSAdapter(BaseAdapter):
def send(self, request, stream=None, timeout=None, verify=None, cert=None,
proxies=None):
pathname = url_to_path(request.url)
resp = Response()
resp.status_code = 200
resp.url = request.url
try:
stats = os.stat(pathname)
except OSError as exc:
resp.status_code = 404
resp.raw = exc
else:
modified = email.utils.formatdate(stats.st_mtime, usegmt=True)
content_type = mimetypes.guess_type(pathname)[0] or "text/plain"
resp.headers = CaseInsensitiveDict({
"Content-Type": content_type,
"Content-Length": stats.st_size,
"Last-Modified": modified,
})
resp.raw = open(pathname, "rb")
resp.close = resp.raw.close
return resp
def close(self):
pass
class SafeFileCache(FileCache):
"""
A file based cache which is safe to use even when the target directory may
not be accessible or writable.
"""
def __init__(self, *args, **kwargs):
super(SafeFileCache, self).__init__(*args, **kwargs)
# Check to ensure that the directory containing our cache directory
# is owned by the user current executing pip. If it does not exist
# we will check the parent directory until we find one that does exist.
# If it is not owned by the user executing pip then we will disable
# the cache and log a warning.
if not check_path_owner(self.directory):
logger.warning(
"The directory '%s' or its parent directory is not owned by "
"the current user and the cache has been disabled. Please "
"check the permissions and owner of that directory. If "
"executing pip with sudo, you may want sudo's -H flag.",
self.directory,
)
# Set our directory to None to disable the Cache
self.directory = None
def get(self, *args, **kwargs):
# If we don't have a directory, then the cache should be a no-op.
if self.directory is None:
return
try:
return super(SafeFileCache, self).get(*args, **kwargs)
except (LockError, OSError, IOError):
# We intentionally silence this error, if we can't access the cache
# then we can just skip caching and process the request as if
# caching wasn't enabled.
pass
def set(self, *args, **kwargs):
# If we don't have a directory, then the cache should be a no-op.
if self.directory is None:
return
try:
return super(SafeFileCache, self).set(*args, **kwargs)
except (LockError, OSError, IOError):
# We intentionally silence this error, if we can't access the cache
# then we can just skip caching and process the request as if
# caching wasn't enabled.
pass
def delete(self, *args, **kwargs):
# If we don't have a directory, then the cache should be a no-op.
if self.directory is None:
return
try:
return super(SafeFileCache, self).delete(*args, **kwargs)
except (LockError, OSError, IOError):
# We intentionally silence this error, if we can't access the cache
# then we can just skip caching and process the request as if
# caching wasn't enabled.
pass
class InsecureHTTPAdapter(HTTPAdapter):
def cert_verify(self, conn, url, verify, cert):
conn.cert_reqs = 'CERT_NONE'
conn.ca_certs = None
class PipSession(requests.Session):
timeout = None
def __init__(self, *args, **kwargs):
retries = kwargs.pop("retries", 0)
cache = kwargs.pop("cache", None)
insecure_hosts = kwargs.pop("insecure_hosts", [])
super(PipSession, self).__init__(*args, **kwargs)
# Attach our User Agent to the request
self.headers["User-Agent"] = user_agent()
# Attach our Authentication handler to the session
self.auth = MultiDomainBasicAuth()
# Create our urllib3.Retry instance which will allow us to customize
# how we handle retries.
retries = urllib3.Retry(
# Set the total number of retries that a particular request can
# have.
total=retries,
# A 503 error from PyPI typically means that the Fastly -> Origin
# connection got interrupted in some way. A 503 error in general
# is typically considered a transient error so we'll go ahead and
# retry it.
# A 500 may indicate transient error in Amazon S3
# A 520 or 527 - may indicate transient error in CloudFlare
status_forcelist=[500, 503, 520, 527],
# Add a small amount of back off between failed requests in
# order to prevent hammering the service.
backoff_factor=0.25,
)
# We want to _only_ cache responses on securely fetched origins. We do
# this because we can't validate the response of an insecurely fetched
# origin, and we don't want someone to be able to poison the cache and
# require manual eviction from the cache to fix it.
if cache:
secure_adapter = CacheControlAdapter(
cache=SafeFileCache(cache, use_dir_lock=True),
max_retries=retries,
)
else:
secure_adapter = HTTPAdapter(max_retries=retries)
# Our Insecure HTTPAdapter disables HTTPS validation. It does not
# support caching (see above) so we'll use it for all http:// URLs as
# well as any https:// host that we've marked as ignoring TLS errors
# for.
insecure_adapter = InsecureHTTPAdapter(max_retries=retries)
self.mount("https://", secure_adapter)
self.mount("http://", insecure_adapter)
# Enable file:// urls
self.mount("file://", LocalFSAdapter())
# We want to use a non-validating adapter for any requests which are
# deemed insecure.
for host in insecure_hosts:
self.mount("https://{}/".format(host), insecure_adapter)
def request(self, method, url, *args, **kwargs):
# Allow setting a default timeout on a session
kwargs.setdefault("timeout", self.timeout)
# Dispatch the actual request
return super(PipSession, self).request(method, url, *args, **kwargs)
def get_file_content(url, comes_from=None, session=None):
"""Gets the content of a file; it may be a filename, file: URL, or
http: URL. Returns (location, content). Content is unicode.
:param url: File path or url.
:param comes_from: Origin description of requirements.
:param session: Instance of pip.download.PipSession.
"""
if session is None:
raise TypeError(
"get_file_content() missing 1 required keyword argument: 'session'"
)
match = _scheme_re.search(url)
if match:
scheme = match.group(1).lower()
if (scheme == 'file' and comes_from and
comes_from.startswith('http')):
raise InstallationError(
'Requirements file %s references URL %s, which is local'
% (comes_from, url))
if scheme == 'file':
path = url.split(':', 1)[1]
path = path.replace('\\', '/')
match = _url_slash_drive_re.match(path)
if match:
path = match.group(1) + ':' + path.split('|', 1)[1]
path = urllib_parse.unquote(path)
if path.startswith('/'):
path = '/' + path.lstrip('/')
url = path
else:
# FIXME: catch some errors
resp = session.get(url)
resp.raise_for_status()
return resp.url, resp.text
try:
with open(url, 'rb') as f:
content = auto_decode(f.read())
except IOError as exc:
raise InstallationError(
'Could not open requirements file: %s' % str(exc)
)
return url, content
_scheme_re = re.compile(r'^(http|https|file):', re.I)
_url_slash_drive_re = re.compile(r'/*([a-z])\|', re.I)
def is_url(name):
"""Returns true if the name looks like a URL"""
if ':' not in name:
return False
scheme = name.split(':', 1)[0].lower()
return scheme in ['http', 'https', 'file', 'ftp'] + vcs.all_schemes
def url_to_path(url):
"""
Convert a file: URL to a path.
"""
assert url.startswith('file:'), (
"You can only turn file: urls into filenames (not %r)" % url)
_, netloc, path, _, _ = urllib_parse.urlsplit(url)
# if we have a UNC path, prepend UNC share notation
if netloc:
netloc = '\\\\' + netloc
path = urllib_request.url2pathname(netloc + path)
return path
def path_to_url(path):
"""
Convert a path to a file: URL. The path will be made absolute and have
quoted path parts.
"""
path = os.path.normpath(os.path.abspath(path))
url = urllib_parse.urljoin('file:', urllib_request.pathname2url(path))
return url
def is_archive_file(name):
"""Return True if `name` is a considered as an archive file."""
ext = splitext(name)[1].lower()
if ext in ARCHIVE_EXTENSIONS:
return True
return False
def unpack_vcs_link(link, location):
vcs_backend = _get_used_vcs_backend(link)
vcs_backend.unpack(location)
def _get_used_vcs_backend(link):
for backend in vcs.backends:
if link.scheme in backend.schemes:
vcs_backend = backend(link.url)
return vcs_backend
def is_vcs_url(link):
return bool(_get_used_vcs_backend(link))
def is_file_url(link):
return link.url.lower().startswith('file:')
def is_dir_url(link):
"""Return whether a file:// Link points to a directory.
``link`` must not have any other scheme but file://. Call is_file_url()
first.
"""
link_path = url_to_path(link.url_without_fragment)
return os.path.isdir(link_path)
def _progress_indicator(iterable, *args, **kwargs):
return iterable
def _download_url(resp, link, content_file, hashes, progress_bar):
try:
total_length = int(resp.headers['content-length'])
except (ValueError, KeyError, TypeError):
total_length = 0
cached_resp = getattr(resp, "from_cache", False)
if logger.getEffectiveLevel() > logging.INFO:
show_progress = False
elif cached_resp:
show_progress = False
elif total_length > (40 * 1000):
show_progress = True
elif not total_length:
show_progress = True
else:
show_progress = False
show_url = link.show_url
def resp_read(chunk_size):
try:
# Special case for urllib3.
for chunk in resp.raw.stream(
chunk_size,
# We use decode_content=False here because we don't
# want urllib3 to mess with the raw bytes we get
# from the server. If we decompress inside of
# urllib3 then we cannot verify the checksum
# because the checksum will be of the compressed
# file. This breakage will only occur if the
# server adds a Content-Encoding header, which
# depends on how the server was configured:
# - Some servers will notice that the file isn't a
# compressible file and will leave the file alone
# and with an empty Content-Encoding
# - Some servers will notice that the file is
# already compressed and will leave the file
# alone and will add a Content-Encoding: gzip
# header
# - Some servers won't notice anything at all and
# will take a file that's already been compressed
# and compress it again and set the
# Content-Encoding: gzip header
#
# By setting this not to decode automatically we
# hope to eliminate problems with the second case.
decode_content=False):
yield chunk
except AttributeError:
# Standard file-like object.
while True:
chunk = resp.raw.read(chunk_size)
if not chunk:
break
yield chunk
def written_chunks(chunks):
for chunk in chunks:
content_file.write(chunk)
yield chunk
progress_indicator = _progress_indicator
if link.netloc == PyPI.netloc:
url = show_url
else:
url = link.url_without_fragment
if show_progress: # We don't show progress on cached responses
progress_indicator = DownloadProgressProvider(progress_bar,
max=total_length)
if total_length:
logger.info("Downloading %s (%s)", url, format_size(total_length))
else:
logger.info("Downloading %s", url)
elif cached_resp:
logger.info("Using cached %s", url)
else:
logger.info("Downloading %s", url)
logger.debug('Downloading from URL %s', link)
downloaded_chunks = written_chunks(
progress_indicator(
resp_read(CONTENT_CHUNK_SIZE),
CONTENT_CHUNK_SIZE
)
)
if hashes:
hashes.check_against_chunks(downloaded_chunks)
else:
consume(downloaded_chunks)
def _copy_file(filename, location, link):
copy = True
download_location = os.path.join(location, link.filename)
if os.path.exists(download_location):
response = ask_path_exists(
'The file %s exists. (i)gnore, (w)ipe, (b)ackup, (a)abort' %
display_path(download_location), ('i', 'w', 'b', 'a'))
if response == 'i':
copy = False
elif response == 'w':
logger.warning('Deleting %s', display_path(download_location))
os.remove(download_location)
elif response == 'b':
dest_file = backup_dir(download_location)
logger.warning(
'Backing up %s to %s',
display_path(download_location),
display_path(dest_file),
)
shutil.move(download_location, dest_file)
elif response == 'a':
sys.exit(-1)
if copy:
shutil.copy(filename, download_location)
logger.info('Saved %s', display_path(download_location))
def unpack_http_url(link, location, download_dir=None,
session=None, hashes=None, progress_bar="on"):
if session is None:
raise TypeError(
"unpack_http_url() missing 1 required keyword argument: 'session'"
)
with TempDirectory(kind="unpack") as temp_dir:
# If a download dir is specified, is the file already downloaded there?
already_downloaded_path = None
if download_dir:
already_downloaded_path = _check_download_dir(link,
download_dir,
hashes)
if already_downloaded_path:
from_path = already_downloaded_path
content_type = mimetypes.guess_type(from_path)[0]
else:
# let's download to a tmp dir
from_path, content_type = _download_http_url(link,
session,
temp_dir.path,
hashes,
progress_bar)
# unpack the archive to the build dir location. even when only
# downloading archives, they have to be unpacked to parse dependencies
unpack_file(from_path, location, content_type, link)
# a download dir is specified; let's copy the archive there
if download_dir and not already_downloaded_path:
_copy_file(from_path, download_dir, link)
if not already_downloaded_path:
os.unlink(from_path)
def unpack_file_url(link, location, download_dir=None, hashes=None):
"""Unpack link into location.
If download_dir is provided and link points to a file, make a copy
of the link file inside download_dir.
"""
link_path = url_to_path(link.url_without_fragment)
# If it's a url to a local directory
if is_dir_url(link):
if os.path.isdir(location):
rmtree(location)
shutil.copytree(link_path, location, symlinks=True)
if download_dir:
logger.info('Link is a directory, ignoring download_dir')
return
# If --require-hashes is off, `hashes` is either empty, the
# link's embedded hash, or MissingHashes; it is required to
# match. If --require-hashes is on, we are satisfied by any
# hash in `hashes` matching: a URL-based or an option-based
# one; no internet-sourced hash will be in `hashes`.
if hashes:
hashes.check_against_path(link_path)
# If a download dir is specified, is the file already there and valid?
already_downloaded_path = None
if download_dir:
already_downloaded_path = _check_download_dir(link,
download_dir,
hashes)
if already_downloaded_path:
from_path = already_downloaded_path
else:
from_path = link_path
content_type = mimetypes.guess_type(from_path)[0]
# unpack the archive to the build dir location. even when only downloading
# archives, they have to be unpacked to parse dependencies
unpack_file(from_path, location, content_type, link)
# a download dir is specified and not already downloaded
if download_dir and not already_downloaded_path:
_copy_file(from_path, download_dir, link)
def _copy_dist_from_dir(link_path, location):
"""Copy distribution files in `link_path` to `location`.
Invoked when user requests to install a local directory. E.g.:
pip install .
pip install ~/dev/git-repos/python-prompt-toolkit
"""
# Note: This is currently VERY SLOW if you have a lot of data in the
# directory, because it copies everything with `shutil.copytree`.
# What it should really do is build an sdist and install that.
# See https://github.com/pypa/pip/issues/2195
if os.path.isdir(location):
rmtree(location)
# build an sdist
setup_py = 'setup.py'
sdist_args = [sys.executable]
sdist_args.append('-c')
sdist_args.append(SETUPTOOLS_SHIM % setup_py)
sdist_args.append('sdist')
sdist_args += ['--dist-dir', location]
logger.info('Running setup.py sdist for %s', link_path)
with indent_log():
call_subprocess(sdist_args, cwd=link_path, show_stdout=False)
# unpack sdist into `location`
sdist = os.path.join(location, os.listdir(location)[0])
logger.info('Unpacking sdist %s into %s', sdist, location)
unpack_file(sdist, location, content_type=None, link=None)
class PipXmlrpcTransport(xmlrpc_client.Transport):
"""Provide a `xmlrpclib.Transport` implementation via a `PipSession`
object.
"""
def __init__(self, index_url, session, use_datetime=False):
xmlrpc_client.Transport.__init__(self, use_datetime)
index_parts = urllib_parse.urlparse(index_url)
self._scheme = index_parts.scheme
self._session = session
def request(self, host, handler, request_body, verbose=False):
parts = (self._scheme, host, handler, None, None, None)
url = urllib_parse.urlunparse(parts)
try:
headers = {'Content-Type': 'text/xml'}
response = self._session.post(url, data=request_body,
headers=headers, stream=True)
response.raise_for_status()
self.verbose = verbose
return self.parse_response(response.raw)
except requests.HTTPError as exc:
logger.critical(
"HTTP error %s while getting %s",
exc.response.status_code, url,
)
raise
def unpack_url(link, location, download_dir=None,
only_download=False, session=None, hashes=None,
progress_bar="on"):
"""Unpack link.
If link is a VCS link:
if only_download, export into download_dir and ignore location
else unpack into location
for other types of link:
- unpack into location
- if download_dir, copy the file into download_dir
- if only_download, mark location for deletion
:param hashes: A Hashes object, one of whose embedded hashes must match,
or HashMismatch will be raised. If the Hashes is empty, no matches are
required, and unhashable types of requirements (like VCS ones, which
would ordinarily raise HashUnsupported) are allowed.
"""
# non-editable vcs urls
if is_vcs_url(link):
unpack_vcs_link(link, location)
# file urls
elif is_file_url(link):
unpack_file_url(link, location, download_dir, hashes=hashes)
# http urls
else:
if session is None:
session = PipSession()
unpack_http_url(
link,
location,
download_dir,
session,
hashes=hashes,
progress_bar=progress_bar
)
if only_download:
write_delete_marker_file(location)
def _download_http_url(link, session, temp_dir, hashes, progress_bar):
"""Download link url into temp_dir using provided session"""
target_url = link.url.split('#', 1)[0]
try:
resp = session.get(
target_url,
# We use Accept-Encoding: identity here because requests
# defaults to accepting compressed responses. This breaks in
# a variety of ways depending on how the server is configured.
# - Some servers will notice that the file isn't a compressible
# file and will leave the file alone and with an empty
# Content-Encoding
# - Some servers will notice that the file is already
# compressed and will leave the file alone and will add a
# Content-Encoding: gzip header
# - Some servers won't notice anything at all and will take
# a file that's already been compressed and compress it again
# and set the Content-Encoding: gzip header
# By setting this to request only the identity encoding We're
# hoping to eliminate the third case. Hopefully there does not
# exist a server which when given a file will notice it is
# already compressed and that you're not asking for a
# compressed file and will then decompress it before sending
# because if that's the case I don't think it'll ever be
# possible to make this work.
headers={"Accept-Encoding": "identity"},
stream=True,
)
resp.raise_for_status()
except requests.HTTPError as exc:
logger.critical(
"HTTP error %s while getting %s", exc.response.status_code, link,
)
raise
content_type = resp.headers.get('content-type', '')
filename = link.filename # fallback
# Have a look at the Content-Disposition header for a better guess
content_disposition = resp.headers.get('content-disposition')
if content_disposition:
type, params = cgi.parse_header(content_disposition)
# We use ``or`` here because we don't want to use an "empty" value
# from the filename param.
filename = params.get('filename') or filename
ext = splitext(filename)[1]
if not ext:
ext = mimetypes.guess_extension(content_type)
if ext:
filename += ext
if not ext and link.url != resp.url:
ext = os.path.splitext(resp.url)[1]
if ext:
filename += ext
file_path = os.path.join(temp_dir, filename)
with open(file_path, 'wb') as content_file:
_download_url(resp, link, content_file, hashes, progress_bar)
return file_path, content_type
def _check_download_dir(link, download_dir, hashes):
""" Check download_dir for previously downloaded file with correct hash
If a correct file is found return its path else None
"""
download_path = os.path.join(download_dir, link.filename)
if os.path.exists(download_path):
# If already downloaded, does its hash match?
logger.info('File was already downloaded %s', download_path)
if hashes:
try:
hashes.check_against_path(download_path)
except HashMismatch:
logger.warning(
'Previously-downloaded file %s has bad hash. '
'Re-downloading.',
download_path
)
os.unlink(download_path)
return None
return download_path
return None

View File

@@ -0,0 +1,249 @@
"""Exceptions used throughout package"""
from __future__ import absolute_import
from itertools import chain, groupby, repeat
from pip._vendor.six import iteritems
class PipError(Exception):
"""Base pip exception"""
class ConfigurationError(PipError):
"""General exception in configuration"""
class InstallationError(PipError):
"""General exception during installation"""
class UninstallationError(PipError):
"""General exception during uninstallation"""
class DistributionNotFound(InstallationError):
"""Raised when a distribution cannot be found to satisfy a requirement"""
class RequirementsFileParseError(InstallationError):
"""Raised when a general error occurs parsing a requirements file line."""
class BestVersionAlreadyInstalled(PipError):
"""Raised when the most up-to-date version of a package is already
installed."""
class BadCommand(PipError):
"""Raised when virtualenv or a command is not found"""
class CommandError(PipError):
"""Raised when there is an error in command-line arguments"""
class PreviousBuildDirError(PipError):
"""Raised when there's a previous conflicting build directory"""
class InvalidWheelFilename(InstallationError):
"""Invalid wheel filename."""
class UnsupportedWheel(InstallationError):
"""Unsupported wheel."""
class HashErrors(InstallationError):
"""Multiple HashError instances rolled into one for reporting"""
def __init__(self):
self.errors = []
def append(self, error):
self.errors.append(error)
def __str__(self):
lines = []
self.errors.sort(key=lambda e: e.order)
for cls, errors_of_cls in groupby(self.errors, lambda e: e.__class__):
lines.append(cls.head)
lines.extend(e.body() for e in errors_of_cls)
if lines:
return '\n'.join(lines)
def __nonzero__(self):
return bool(self.errors)
def __bool__(self):
return self.__nonzero__()
class HashError(InstallationError):
"""
A failure to verify a package against known-good hashes
:cvar order: An int sorting hash exception classes by difficulty of
recovery (lower being harder), so the user doesn't bother fretting
about unpinned packages when he has deeper issues, like VCS
dependencies, to deal with. Also keeps error reports in a
deterministic order.
:cvar head: A section heading for display above potentially many
exceptions of this kind
:ivar req: The InstallRequirement that triggered this error. This is
pasted on after the exception is instantiated, because it's not
typically available earlier.
"""
req = None
head = ''
def body(self):
"""Return a summary of me for display under the heading.
This default implementation simply prints a description of the
triggering requirement.
:param req: The InstallRequirement that provoked this error, with
populate_link() having already been called
"""
return ' %s' % self._requirement_name()
def __str__(self):
return '%s\n%s' % (self.head, self.body())
def _requirement_name(self):
"""Return a description of the requirement that triggered me.
This default implementation returns long description of the req, with
line numbers
"""
return str(self.req) if self.req else 'unknown package'
class VcsHashUnsupported(HashError):
"""A hash was provided for a version-control-system-based requirement, but
we don't have a method for hashing those."""
order = 0
head = ("Can't verify hashes for these requirements because we don't "
"have a way to hash version control repositories:")
class DirectoryUrlHashUnsupported(HashError):
"""A hash was provided for a version-control-system-based requirement, but
we don't have a method for hashing those."""
order = 1
head = ("Can't verify hashes for these file:// requirements because they "
"point to directories:")
class HashMissing(HashError):
"""A hash was needed for a requirement but is absent."""
order = 2
head = ('Hashes are required in --require-hashes mode, but they are '
'missing from some requirements. Here is a list of those '
'requirements along with the hashes their downloaded archives '
'actually had. Add lines like these to your requirements files to '
'prevent tampering. (If you did not enable --require-hashes '
'manually, note that it turns on automatically when any package '
'has a hash.)')
def __init__(self, gotten_hash):
"""
:param gotten_hash: The hash of the (possibly malicious) archive we
just downloaded
"""
self.gotten_hash = gotten_hash
def body(self):
# Dodge circular import.
from pip._internal.utils.hashes import FAVORITE_HASH
package = None
if self.req:
# In the case of URL-based requirements, display the original URL
# seen in the requirements file rather than the package name,
# so the output can be directly copied into the requirements file.
package = (self.req.original_link if self.req.original_link
# In case someone feeds something downright stupid
# to InstallRequirement's constructor.
else getattr(self.req, 'req', None))
return ' %s --hash=%s:%s' % (package or 'unknown package',
FAVORITE_HASH,
self.gotten_hash)
class HashUnpinned(HashError):
"""A requirement had a hash specified but was not pinned to a specific
version."""
order = 3
head = ('In --require-hashes mode, all requirements must have their '
'versions pinned with ==. These do not:')
class HashMismatch(HashError):
"""
Distribution file hash values don't match.
:ivar package_name: The name of the package that triggered the hash
mismatch. Feel free to write to this after the exception is raise to
improve its error message.
"""
order = 4
head = ('THESE PACKAGES DO NOT MATCH THE HASHES FROM THE REQUIREMENTS '
'FILE. If you have updated the package versions, please update '
'the hashes. Otherwise, examine the package contents carefully; '
'someone may have tampered with them.')
def __init__(self, allowed, gots):
"""
:param allowed: A dict of algorithm names pointing to lists of allowed
hex digests
:param gots: A dict of algorithm names pointing to hashes we
actually got from the files under suspicion
"""
self.allowed = allowed
self.gots = gots
def body(self):
return ' %s:\n%s' % (self._requirement_name(),
self._hash_comparison())
def _hash_comparison(self):
"""
Return a comparison of actual and expected hash values.
Example::
Expected sha256 abcdeabcdeabcdeabcdeabcdeabcdeabcdeabcdeabcde
or 123451234512345123451234512345123451234512345
Got bcdefbcdefbcdefbcdefbcdefbcdefbcdefbcdefbcdef
"""
def hash_then_or(hash_name):
# For now, all the decent hashes have 6-char names, so we can get
# away with hard-coding space literals.
return chain([hash_name], repeat(' or'))
lines = []
for hash_name, expecteds in iteritems(self.allowed):
prefix = hash_then_or(hash_name)
lines.extend((' Expected %s %s' % (next(prefix), e))
for e in expecteds)
lines.append(' Got %s\n' %
self.gots[hash_name].hexdigest())
prefix = ' or'
return '\n'.join(lines)
class UnsupportedPythonVersion(InstallationError):
"""Unsupported python version according to Requires-Python package
metadata."""

File diff suppressed because it is too large Load Diff

View File

@@ -0,0 +1,194 @@
"""Locations where we look for configs, install stuff, etc"""
from __future__ import absolute_import
import os
import os.path
import platform
import site
import sys
import sysconfig
from distutils import sysconfig as distutils_sysconfig
from distutils.command.install import SCHEME_KEYS, install # type: ignore
from pip._internal.compat import WINDOWS, expanduser
from pip._internal.utils import appdirs
# Application Directories
USER_CACHE_DIR = appdirs.user_cache_dir("pip")
DELETE_MARKER_MESSAGE = '''\
This file is placed here by pip to indicate the source was put
here by pip.
Once this package is successfully installed this source code will be
deleted (unless you remove this file).
'''
PIP_DELETE_MARKER_FILENAME = 'pip-delete-this-directory.txt'
def write_delete_marker_file(directory):
"""
Write the pip delete marker file into this directory.
"""
filepath = os.path.join(directory, PIP_DELETE_MARKER_FILENAME)
with open(filepath, 'w') as marker_fp:
marker_fp.write(DELETE_MARKER_MESSAGE)
def running_under_virtualenv():
"""
Return True if we're running inside a virtualenv, False otherwise.
"""
if hasattr(sys, 'real_prefix'):
return True
elif sys.prefix != getattr(sys, "base_prefix", sys.prefix):
return True
return False
def virtualenv_no_global():
"""
Return True if in a venv and no system site packages.
"""
# this mirrors the logic in virtualenv.py for locating the
# no-global-site-packages.txt file
site_mod_dir = os.path.dirname(os.path.abspath(site.__file__))
no_global_file = os.path.join(site_mod_dir, 'no-global-site-packages.txt')
if running_under_virtualenv() and os.path.isfile(no_global_file):
return True
if running_under_virtualenv():
src_prefix = os.path.join(sys.prefix, 'src')
else:
# FIXME: keep src in cwd for now (it is not a temporary folder)
try:
src_prefix = os.path.join(os.getcwd(), 'src')
except OSError:
# In case the current working directory has been renamed or deleted
sys.exit(
"The folder you are executing pip from can no longer be found."
)
# under macOS + virtualenv sys.prefix is not properly resolved
# it is something like /path/to/python/bin/..
# Note: using realpath due to tmp dirs on OSX being symlinks
src_prefix = os.path.abspath(src_prefix)
# FIXME doesn't account for venv linked to global site-packages
site_packages = sysconfig.get_path("purelib")
# This is because of a bug in PyPy's sysconfig module, see
# https://bitbucket.org/pypy/pypy/issues/2506/sysconfig-returns-incorrect-paths
# for more information.
if platform.python_implementation().lower() == "pypy":
site_packages = distutils_sysconfig.get_python_lib()
try:
# Use getusersitepackages if this is present, as it ensures that the
# value is initialised properly.
user_site = site.getusersitepackages()
except AttributeError:
user_site = site.USER_SITE
user_dir = expanduser('~')
if WINDOWS:
bin_py = os.path.join(sys.prefix, 'Scripts')
bin_user = os.path.join(user_site, 'Scripts')
# buildout uses 'bin' on Windows too?
if not os.path.exists(bin_py):
bin_py = os.path.join(sys.prefix, 'bin')
bin_user = os.path.join(user_site, 'bin')
config_basename = 'pip.ini'
legacy_storage_dir = os.path.join(user_dir, 'pip')
legacy_config_file = os.path.join(
legacy_storage_dir,
config_basename,
)
else:
bin_py = os.path.join(sys.prefix, 'bin')
bin_user = os.path.join(user_site, 'bin')
config_basename = 'pip.conf'
legacy_storage_dir = os.path.join(user_dir, '.pip')
legacy_config_file = os.path.join(
legacy_storage_dir,
config_basename,
)
# Forcing to use /usr/local/bin for standard macOS framework installs
# Also log to ~/Library/Logs/ for use with the Console.app log viewer
if sys.platform[:6] == 'darwin' and sys.prefix[:16] == '/System/Library/':
bin_py = '/usr/local/bin'
site_config_files = [
os.path.join(path, config_basename)
for path in appdirs.site_config_dirs('pip')
]
venv_config_file = os.path.join(sys.prefix, config_basename)
new_config_file = os.path.join(appdirs.user_config_dir("pip"), config_basename)
def distutils_scheme(dist_name, user=False, home=None, root=None,
isolated=False, prefix=None):
"""
Return a distutils install scheme
"""
from distutils.dist import Distribution
scheme = {}
if isolated:
extra_dist_args = {"script_args": ["--no-user-cfg"]}
else:
extra_dist_args = {}
dist_args = {'name': dist_name}
dist_args.update(extra_dist_args)
d = Distribution(dist_args)
d.parse_config_files()
i = d.get_command_obj('install', create=True)
# NOTE: setting user or home has the side-effect of creating the home dir
# or user base for installations during finalize_options()
# ideally, we'd prefer a scheme class that has no side-effects.
assert not (user and prefix), "user={} prefix={}".format(user, prefix)
i.user = user or i.user
if user:
i.prefix = ""
i.prefix = prefix or i.prefix
i.home = home or i.home
i.root = root or i.root
i.finalize_options()
for key in SCHEME_KEYS:
scheme[key] = getattr(i, 'install_' + key)
# install_lib specified in setup.cfg should install *everything*
# into there (i.e. it takes precedence over both purelib and
# platlib). Note, i.install_lib is *always* set after
# finalize_options(); we only want to override here if the user
# has explicitly requested it hence going back to the config
if 'install_lib' in d.get_option_dict('install'):
scheme.update(dict(purelib=i.install_lib, platlib=i.install_lib))
if running_under_virtualenv():
scheme['headers'] = os.path.join(
sys.prefix,
'include',
'site',
'python' + sys.version[:3],
dist_name,
)
if root is not None:
path_no_drive = os.path.splitdrive(
os.path.abspath(scheme["headers"]))[1]
scheme["headers"] = os.path.join(
root,
path_no_drive[1:],
)
return scheme

View File

@@ -0,0 +1,4 @@
from pip._internal.models.index import Index, PyPI
__all__ = ["Index", "PyPI"]

View File

@@ -0,0 +1,15 @@
from pip._vendor.six.moves.urllib import parse as urllib_parse
class Index(object):
def __init__(self, url):
self.url = url
self.netloc = urllib_parse.urlsplit(url).netloc
self.simple_url = self.url_to_path('simple')
self.pypi_url = self.url_to_path('pypi')
def url_to_path(self, path):
return urllib_parse.urljoin(self.url, path)
PyPI = Index('https://pypi.org/')

View File

@@ -0,0 +1,106 @@
"""Validation of dependencies of packages
"""
from collections import namedtuple
from pip._vendor.packaging.utils import canonicalize_name
from pip._internal.operations.prepare import make_abstract_dist
from pip._internal.utils.misc import get_installed_distributions
from pip._internal.utils.typing import MYPY_CHECK_RUNNING
if MYPY_CHECK_RUNNING:
from pip._internal.req.req_install import InstallRequirement
from typing import Any, Dict, Iterator, Set, Tuple, List
# Shorthands
PackageSet = Dict[str, 'PackageDetails']
Missing = Tuple[str, Any]
Conflicting = Tuple[str, str, Any]
MissingDict = Dict[str, List[Missing]]
ConflictingDict = Dict[str, List[Conflicting]]
CheckResult = Tuple[MissingDict, ConflictingDict]
PackageDetails = namedtuple('PackageDetails', ['version', 'requires'])
def create_package_set_from_installed(**kwargs):
# type: (**Any) -> PackageSet
"""Converts a list of distributions into a PackageSet.
"""
# Default to using all packages installed on the system
if kwargs == {}:
kwargs = {"local_only": False, "skip": ()}
retval = {}
for dist in get_installed_distributions(**kwargs):
name = canonicalize_name(dist.project_name)
retval[name] = PackageDetails(dist.version, dist.requires())
return retval
def check_package_set(package_set):
# type: (PackageSet) -> CheckResult
"""Check if a package set is consistent
"""
missing = dict()
conflicting = dict()
for package_name in package_set:
# Info about dependencies of package_name
missing_deps = set() # type: Set[Missing]
conflicting_deps = set() # type: Set[Conflicting]
for req in package_set[package_name].requires:
name = canonicalize_name(req.project_name) # type: str
# Check if it's missing
if name not in package_set:
missed = True
if req.marker is not None:
missed = req.marker.evaluate()
if missed:
missing_deps.add((name, req))
continue
# Check if there's a conflict
version = package_set[name].version # type: str
if not req.specifier.contains(version, prereleases=True):
conflicting_deps.add((name, version, req))
def str_key(x):
return str(x)
if missing_deps:
missing[package_name] = sorted(missing_deps, key=str_key)
if conflicting_deps:
conflicting[package_name] = sorted(conflicting_deps, key=str_key)
return missing, conflicting
def check_install_conflicts(to_install):
# type: (List[InstallRequirement]) -> Tuple[PackageSet, CheckResult]
"""For checking if the dependency graph would be consistent after \
installing given requirements
"""
# Start from the current state
state = create_package_set_from_installed()
_simulate_installation_of(to_install, state)
return state, check_package_set(state)
# NOTE from @pradyunsg
# This required a minor update in dependency link handling logic over at
# operations.prepare.IsSDist.dist() to get it working
def _simulate_installation_of(to_install, state):
# type: (List[InstallRequirement], PackageSet) -> None
"""Computes the version of packages after installing to_install.
"""
# Modify it as installing requirement_set would (assuming no errors)
for inst_req in to_install:
dist = make_abstract_dist(inst_req).dist(finder=None)
name = canonicalize_name(dist.key)
state[name] = PackageDetails(dist.version, dist.requires())

View File

@@ -0,0 +1,252 @@
from __future__ import absolute_import
import collections
import logging
import os
import re
import warnings
from pip._vendor import pkg_resources, six
from pip._vendor.packaging.utils import canonicalize_name
from pip._vendor.pkg_resources import RequirementParseError
from pip._internal.exceptions import InstallationError
from pip._internal.req import InstallRequirement
from pip._internal.req.req_file import COMMENT_RE
from pip._internal.utils.deprecation import RemovedInPip11Warning
from pip._internal.utils.misc import (
dist_is_editable, get_installed_distributions,
)
logger = logging.getLogger(__name__)
def freeze(
requirement=None,
find_links=None, local_only=None, user_only=None, skip_regex=None,
isolated=False,
wheel_cache=None,
exclude_editable=False,
skip=()):
find_links = find_links or []
skip_match = None
if skip_regex:
skip_match = re.compile(skip_regex).search
dependency_links = []
for dist in pkg_resources.working_set:
if dist.has_metadata('dependency_links.txt'):
dependency_links.extend(
dist.get_metadata_lines('dependency_links.txt')
)
for link in find_links:
if '#egg=' in link:
dependency_links.append(link)
for link in find_links:
yield '-f %s' % link
installations = {}
for dist in get_installed_distributions(local_only=local_only,
skip=(),
user_only=user_only):
try:
req = FrozenRequirement.from_dist(
dist,
dependency_links
)
except RequirementParseError:
logger.warning(
"Could not parse requirement: %s",
dist.project_name
)
continue
if exclude_editable and req.editable:
continue
installations[req.name] = req
if requirement:
# the options that don't get turned into an InstallRequirement
# should only be emitted once, even if the same option is in multiple
# requirements files, so we need to keep track of what has been emitted
# so that we don't emit it again if it's seen again
emitted_options = set()
# keep track of which files a requirement is in so that we can
# give an accurate warning if a requirement appears multiple times.
req_files = collections.defaultdict(list)
for req_file_path in requirement:
with open(req_file_path) as req_file:
for line in req_file:
if (not line.strip() or
line.strip().startswith('#') or
(skip_match and skip_match(line)) or
line.startswith((
'-r', '--requirement',
'-Z', '--always-unzip',
'-f', '--find-links',
'-i', '--index-url',
'--pre',
'--trusted-host',
'--process-dependency-links',
'--extra-index-url'))):
line = line.rstrip()
if line not in emitted_options:
emitted_options.add(line)
yield line
continue
if line.startswith('-e') or line.startswith('--editable'):
if line.startswith('-e'):
line = line[2:].strip()
else:
line = line[len('--editable'):].strip().lstrip('=')
line_req = InstallRequirement.from_editable(
line,
isolated=isolated,
wheel_cache=wheel_cache,
)
else:
line_req = InstallRequirement.from_line(
COMMENT_RE.sub('', line).strip(),
isolated=isolated,
wheel_cache=wheel_cache,
)
if not line_req.name:
logger.info(
"Skipping line in requirement file [%s] because "
"it's not clear what it would install: %s",
req_file_path, line.strip(),
)
logger.info(
" (add #egg=PackageName to the URL to avoid"
" this warning)"
)
elif line_req.name not in installations:
# either it's not installed, or it is installed
# but has been processed already
if not req_files[line_req.name]:
logger.warning(
"Requirement file [%s] contains %s, but that "
"package is not installed",
req_file_path,
COMMENT_RE.sub('', line).strip(),
)
else:
req_files[line_req.name].append(req_file_path)
else:
yield str(installations[line_req.name]).rstrip()
del installations[line_req.name]
req_files[line_req.name].append(req_file_path)
# Warn about requirements that were included multiple times (in a
# single requirements file or in different requirements files).
for name, files in six.iteritems(req_files):
if len(files) > 1:
logger.warning("Requirement %s included multiple times [%s]",
name, ', '.join(sorted(set(files))))
yield(
'## The following requirements were added by '
'pip freeze:'
)
for installation in sorted(
installations.values(), key=lambda x: x.name.lower()):
if canonicalize_name(installation.name) not in skip:
yield str(installation).rstrip()
class FrozenRequirement(object):
def __init__(self, name, req, editable, comments=()):
self.name = name
self.req = req
self.editable = editable
self.comments = comments
_rev_re = re.compile(r'-r(\d+)$')
_date_re = re.compile(r'-(20\d\d\d\d\d\d)$')
@classmethod
def from_dist(cls, dist, dependency_links):
location = os.path.normcase(os.path.abspath(dist.location))
comments = []
from pip._internal.vcs import vcs, get_src_requirement
if dist_is_editable(dist) and vcs.get_backend_name(location):
editable = True
try:
req = get_src_requirement(dist, location)
except InstallationError as exc:
logger.warning(
"Error when trying to get requirement for VCS system %s, "
"falling back to uneditable format", exc
)
req = None
if req is None:
logger.warning(
'Could not determine repository location of %s', location
)
comments.append(
'## !! Could not determine repository location'
)
req = dist.as_requirement()
editable = False
else:
editable = False
req = dist.as_requirement()
specs = req.specs
assert len(specs) == 1 and specs[0][0] in ["==", "==="], \
'Expected 1 spec with == or ===; specs = %r; dist = %r' % \
(specs, dist)
version = specs[0][1]
ver_match = cls._rev_re.search(version)
date_match = cls._date_re.search(version)
if ver_match or date_match:
svn_backend = vcs.get_backend('svn')
if svn_backend:
svn_location = svn_backend().get_location(
dist,
dependency_links,
)
if not svn_location:
logger.warning(
'Warning: cannot find svn location for %s', req,
)
comments.append(
'## FIXME: could not find svn URL in dependency_links '
'for this package:'
)
else:
warnings.warn(
"SVN editable detection based on dependency links "
"will be dropped in the future.",
RemovedInPip11Warning,
)
comments.append(
'# Installing as editable to satisfy requirement %s:' %
req
)
if ver_match:
rev = ver_match.group(1)
else:
rev = '{%s}' % date_match.group(1)
editable = True
req = '%s@%s#egg=%s' % (
svn_location,
rev,
cls.egg_name(dist)
)
return cls(dist.project_name, req, editable, comments)
@staticmethod
def egg_name(dist):
name = dist.egg_name()
match = re.search(r'-py\d\.\d$', name)
if match:
name = name[:match.start()]
return name
def __str__(self):
req = self.req
if self.editable:
req = '-e %s' % req
return '\n'.join(list(self.comments) + [str(req)]) + '\n'

View File

@@ -0,0 +1,380 @@
"""Prepares a distribution for installation
"""
import itertools
import logging
import os
import sys
from copy import copy
from pip._vendor import pkg_resources, requests
from pip._internal.build_env import NoOpBuildEnvironment
from pip._internal.compat import expanduser
from pip._internal.download import (
is_dir_url, is_file_url, is_vcs_url, unpack_url, url_to_path,
)
from pip._internal.exceptions import (
DirectoryUrlHashUnsupported, HashUnpinned, InstallationError,
PreviousBuildDirError, VcsHashUnsupported,
)
from pip._internal.index import FormatControl
from pip._internal.req.req_install import InstallRequirement
from pip._internal.utils.hashes import MissingHashes
from pip._internal.utils.logging import indent_log
from pip._internal.utils.misc import (
call_subprocess, display_path, normalize_path,
)
from pip._internal.utils.ui import open_spinner
from pip._internal.vcs import vcs
logger = logging.getLogger(__name__)
def make_abstract_dist(req):
"""Factory to make an abstract dist object.
Preconditions: Either an editable req with a source_dir, or satisfied_by or
a wheel link, or a non-editable req with a source_dir.
:return: A concrete DistAbstraction.
"""
if req.editable:
return IsSDist(req)
elif req.link and req.link.is_wheel:
return IsWheel(req)
else:
return IsSDist(req)
def _install_build_reqs(finder, prefix, build_requirements):
# NOTE: What follows is not a very good thing.
# Eventually, this should move into the BuildEnvironment class and
# that should handle all the isolation and sub-process invocation.
finder = copy(finder)
finder.format_control = FormatControl(set(), set([":all:"]))
urls = [
finder.find_requirement(
InstallRequirement.from_line(r), upgrade=False).url
for r in build_requirements
]
args = [
sys.executable, '-m', 'pip', 'install', '--ignore-installed',
'--no-user', '--prefix', prefix,
] + list(urls)
with open_spinner("Installing build dependencies") as spinner:
call_subprocess(args, show_stdout=False, spinner=spinner)
class DistAbstraction(object):
"""Abstracts out the wheel vs non-wheel Resolver.resolve() logic.
The requirements for anything installable are as follows:
- we must be able to determine the requirement name
(or we can't correctly handle the non-upgrade case).
- we must be able to generate a list of run-time dependencies
without installing any additional packages (or we would
have to either burn time by doing temporary isolated installs
or alternatively violate pips 'don't start installing unless
all requirements are available' rule - neither of which are
desirable).
- for packages with setup requirements, we must also be able
to determine their requirements without installing additional
packages (for the same reason as run-time dependencies)
- we must be able to create a Distribution object exposing the
above metadata.
"""
def __init__(self, req):
self.req = req
def dist(self, finder):
"""Return a setuptools Dist object."""
raise NotImplementedError(self.dist)
def prep_for_dist(self, finder):
"""Ensure that we can get a Dist for this requirement."""
raise NotImplementedError(self.dist)
class IsWheel(DistAbstraction):
def dist(self, finder):
return list(pkg_resources.find_distributions(
self.req.source_dir))[0]
def prep_for_dist(self, finder, build_isolation):
# FIXME:https://github.com/pypa/pip/issues/1112
pass
class IsSDist(DistAbstraction):
def dist(self, finder):
dist = self.req.get_dist()
# FIXME: shouldn't be globally added.
if finder and dist.has_metadata('dependency_links.txt'):
finder.add_dependency_links(
dist.get_metadata_lines('dependency_links.txt')
)
return dist
def prep_for_dist(self, finder, build_isolation):
# Before calling "setup.py egg_info", we need to set-up the build
# environment.
build_requirements, isolate = self.req.get_pep_518_info()
should_isolate = build_isolation and isolate
minimum_requirements = ('setuptools', 'wheel')
missing_requirements = set(minimum_requirements) - set(
pkg_resources.Requirement(r).key
for r in build_requirements
)
if missing_requirements:
def format_reqs(rs):
return ' and '.join(map(repr, sorted(rs)))
logger.warning(
"Missing build time requirements in pyproject.toml for %s: "
"%s.", self.req, format_reqs(missing_requirements)
)
logger.warning(
"This version of pip does not implement PEP 517 so it cannot "
"build a wheel without %s.", format_reqs(minimum_requirements)
)
if should_isolate:
with self.req.build_env:
pass
_install_build_reqs(finder, self.req.build_env.path,
build_requirements)
else:
self.req.build_env = NoOpBuildEnvironment(no_clean=False)
self.req.run_egg_info()
self.req.assert_source_matches_version()
class Installed(DistAbstraction):
def dist(self, finder):
return self.req.satisfied_by
def prep_for_dist(self, finder):
pass
class RequirementPreparer(object):
"""Prepares a Requirement
"""
def __init__(self, build_dir, download_dir, src_dir, wheel_download_dir,
progress_bar, build_isolation):
super(RequirementPreparer, self).__init__()
self.src_dir = src_dir
self.build_dir = build_dir
# Where still packed archives should be written to. If None, they are
# not saved, and are deleted immediately after unpacking.
self.download_dir = download_dir
# Where still-packed .whl files should be written to. If None, they are
# written to the download_dir parameter. Separate to download_dir to
# permit only keeping wheel archives for pip wheel.
if wheel_download_dir:
wheel_download_dir = normalize_path(wheel_download_dir)
self.wheel_download_dir = wheel_download_dir
# NOTE
# download_dir and wheel_download_dir overlap semantically and may
# be combined if we're willing to have non-wheel archives present in
# the wheelhouse output by 'pip wheel'.
self.progress_bar = progress_bar
# Is build isolation allowed?
self.build_isolation = build_isolation
@property
def _download_should_save(self):
# TODO: Modify to reduce indentation needed
if self.download_dir:
self.download_dir = expanduser(self.download_dir)
if os.path.exists(self.download_dir):
return True
else:
logger.critical('Could not find download directory')
raise InstallationError(
"Could not find or access download directory '%s'"
% display_path(self.download_dir))
return False
def prepare_linked_requirement(self, req, session, finder,
upgrade_allowed, require_hashes):
"""Prepare a requirement that would be obtained from req.link
"""
# TODO: Breakup into smaller functions
if req.link and req.link.scheme == 'file':
path = url_to_path(req.link.url)
logger.info('Processing %s', display_path(path))
else:
logger.info('Collecting %s', req)
with indent_log():
# @@ if filesystem packages are not marked
# editable in a req, a non deterministic error
# occurs when the script attempts to unpack the
# build directory
req.ensure_has_source_dir(self.build_dir)
# If a checkout exists, it's unwise to keep going. version
# inconsistencies are logged later, but do not fail the
# installation.
# FIXME: this won't upgrade when there's an existing
# package unpacked in `req.source_dir`
# package unpacked in `req.source_dir`
if os.path.exists(os.path.join(req.source_dir, 'setup.py')):
raise PreviousBuildDirError(
"pip can't proceed with requirements '%s' due to a"
" pre-existing build directory (%s). This is "
"likely due to a previous installation that failed"
". pip is being responsible and not assuming it "
"can delete this. Please delete it and try again."
% (req, req.source_dir)
)
req.populate_link(finder, upgrade_allowed, require_hashes)
# We can't hit this spot and have populate_link return None.
# req.satisfied_by is None here (because we're
# guarded) and upgrade has no impact except when satisfied_by
# is not None.
# Then inside find_requirement existing_applicable -> False
# If no new versions are found, DistributionNotFound is raised,
# otherwise a result is guaranteed.
assert req.link
link = req.link
# Now that we have the real link, we can tell what kind of
# requirements we have and raise some more informative errors
# than otherwise. (For example, we can raise VcsHashUnsupported
# for a VCS URL rather than HashMissing.)
if require_hashes:
# We could check these first 2 conditions inside
# unpack_url and save repetition of conditions, but then
# we would report less-useful error messages for
# unhashable requirements, complaining that there's no
# hash provided.
if is_vcs_url(link):
raise VcsHashUnsupported()
elif is_file_url(link) and is_dir_url(link):
raise DirectoryUrlHashUnsupported()
if not req.original_link and not req.is_pinned:
# Unpinned packages are asking for trouble when a new
# version is uploaded. This isn't a security check, but
# it saves users a surprising hash mismatch in the
# future.
#
# file:/// URLs aren't pinnable, so don't complain
# about them not being pinned.
raise HashUnpinned()
hashes = req.hashes(trust_internet=not require_hashes)
if require_hashes and not hashes:
# Known-good hashes are missing for this requirement, so
# shim it with a facade object that will provoke hash
# computation and then raise a HashMissing exception
# showing the user what the hash should be.
hashes = MissingHashes()
try:
download_dir = self.download_dir
# We always delete unpacked sdists after pip ran.
autodelete_unpacked = True
if req.link.is_wheel and self.wheel_download_dir:
# when doing 'pip wheel` we download wheels to a
# dedicated dir.
download_dir = self.wheel_download_dir
if req.link.is_wheel:
if download_dir:
# When downloading, we only unpack wheels to get
# metadata.
autodelete_unpacked = True
else:
# When installing a wheel, we use the unpacked
# wheel.
autodelete_unpacked = False
unpack_url(
req.link, req.source_dir,
download_dir, autodelete_unpacked,
session=session, hashes=hashes,
progress_bar=self.progress_bar
)
except requests.HTTPError as exc:
logger.critical(
'Could not install requirement %s because of error %s',
req,
exc,
)
raise InstallationError(
'Could not install requirement %s because of HTTP '
'error %s for URL %s' %
(req, exc, req.link)
)
abstract_dist = make_abstract_dist(req)
abstract_dist.prep_for_dist(finder, self.build_isolation)
if self._download_should_save:
# Make a .zip of the source_dir we already created.
if req.link.scheme in vcs.all_schemes:
req.archive(self.download_dir)
return abstract_dist
def prepare_editable_requirement(self, req, require_hashes, use_user_site,
finder):
"""Prepare an editable requirement
"""
assert req.editable, "cannot prepare a non-editable req as editable"
logger.info('Obtaining %s', req)
with indent_log():
if require_hashes:
raise InstallationError(
'The editable requirement %s cannot be installed when '
'requiring hashes, because there is no single file to '
'hash.' % req
)
req.ensure_has_source_dir(self.src_dir)
req.update_editable(not self._download_should_save)
abstract_dist = make_abstract_dist(req)
abstract_dist.prep_for_dist(finder, self.build_isolation)
if self._download_should_save:
req.archive(self.download_dir)
req.check_if_exists(use_user_site)
return abstract_dist
def prepare_installed_requirement(self, req, require_hashes, skip_reason):
"""Prepare an already-installed requirement
"""
assert req.satisfied_by, "req should have been satisfied but isn't"
assert skip_reason is not None, (
"did not get skip reason skipped but req.satisfied_by "
"is set to %r" % (req.satisfied_by,)
)
logger.info(
'Requirement %s: %s (%s)',
skip_reason, req, req.satisfied_by.version
)
with indent_log():
if require_hashes:
logger.debug(
'Since it is already installed, we are trusting this '
'package without checking its hash. To ensure a '
'completely repeatable environment, install into an '
'empty virtualenv.'
)
abstract_dist = Installed(req)
return abstract_dist

View File

@@ -0,0 +1,317 @@
"""Generate and work with PEP 425 Compatibility Tags."""
from __future__ import absolute_import
import distutils.util
import logging
import platform
import re
import sys
import sysconfig
import warnings
from collections import OrderedDict
import pip._internal.utils.glibc
logger = logging.getLogger(__name__)
_osx_arch_pat = re.compile(r'(.+)_(\d+)_(\d+)_(.+)')
def get_config_var(var):
try:
return sysconfig.get_config_var(var)
except IOError as e: # Issue #1074
warnings.warn("{}".format(e), RuntimeWarning)
return None
def get_abbr_impl():
"""Return abbreviated implementation name."""
if hasattr(sys, 'pypy_version_info'):
pyimpl = 'pp'
elif sys.platform.startswith('java'):
pyimpl = 'jy'
elif sys.platform == 'cli':
pyimpl = 'ip'
else:
pyimpl = 'cp'
return pyimpl
def get_impl_ver():
"""Return implementation version."""
impl_ver = get_config_var("py_version_nodot")
if not impl_ver or get_abbr_impl() == 'pp':
impl_ver = ''.join(map(str, get_impl_version_info()))
return impl_ver
def get_impl_version_info():
"""Return sys.version_info-like tuple for use in decrementing the minor
version."""
if get_abbr_impl() == 'pp':
# as per https://github.com/pypa/pip/issues/2882
return (sys.version_info[0], sys.pypy_version_info.major,
sys.pypy_version_info.minor)
else:
return sys.version_info[0], sys.version_info[1]
def get_impl_tag():
"""
Returns the Tag for this specific implementation.
"""
return "{}{}".format(get_abbr_impl(), get_impl_ver())
def get_flag(var, fallback, expected=True, warn=True):
"""Use a fallback method for determining SOABI flags if the needed config
var is unset or unavailable."""
val = get_config_var(var)
if val is None:
if warn:
logger.debug("Config variable '%s' is unset, Python ABI tag may "
"be incorrect", var)
return fallback()
return val == expected
def get_abi_tag():
"""Return the ABI tag based on SOABI (if available) or emulate SOABI
(CPython 2, PyPy)."""
soabi = get_config_var('SOABI')
impl = get_abbr_impl()
if not soabi and impl in {'cp', 'pp'} and hasattr(sys, 'maxunicode'):
d = ''
m = ''
u = ''
if get_flag('Py_DEBUG',
lambda: hasattr(sys, 'gettotalrefcount'),
warn=(impl == 'cp')):
d = 'd'
if get_flag('WITH_PYMALLOC',
lambda: impl == 'cp',
warn=(impl == 'cp')):
m = 'm'
if get_flag('Py_UNICODE_SIZE',
lambda: sys.maxunicode == 0x10ffff,
expected=4,
warn=(impl == 'cp' and
sys.version_info < (3, 3))) \
and sys.version_info < (3, 3):
u = 'u'
abi = '%s%s%s%s%s' % (impl, get_impl_ver(), d, m, u)
elif soabi and soabi.startswith('cpython-'):
abi = 'cp' + soabi.split('-')[1]
elif soabi:
abi = soabi.replace('.', '_').replace('-', '_')
else:
abi = None
return abi
def _is_running_32bit():
return sys.maxsize == 2147483647
def get_platform():
"""Return our platform name 'win32', 'linux_x86_64'"""
if sys.platform == 'darwin':
# distutils.util.get_platform() returns the release based on the value
# of MACOSX_DEPLOYMENT_TARGET on which Python was built, which may
# be significantly older than the user's current machine.
release, _, machine = platform.mac_ver()
split_ver = release.split('.')
if machine == "x86_64" and _is_running_32bit():
machine = "i386"
elif machine == "ppc64" and _is_running_32bit():
machine = "ppc"
return 'macosx_{}_{}_{}'.format(split_ver[0], split_ver[1], machine)
# XXX remove distutils dependency
result = distutils.util.get_platform().replace('.', '_').replace('-', '_')
if result == "linux_x86_64" and _is_running_32bit():
# 32 bit Python program (running on a 64 bit Linux): pip should only
# install and run 32 bit compiled extensions in that case.
result = "linux_i686"
return result
def is_manylinux1_compatible():
# Only Linux, and only x86-64 / i686
if get_platform() not in {"linux_x86_64", "linux_i686"}:
return False
# Check for presence of _manylinux module
try:
import _manylinux
return bool(_manylinux.manylinux1_compatible)
except (ImportError, AttributeError):
# Fall through to heuristic check below
pass
# Check glibc version. CentOS 5 uses glibc 2.5.
return pip._internal.utils.glibc.have_compatible_glibc(2, 5)
def get_darwin_arches(major, minor, machine):
"""Return a list of supported arches (including group arches) for
the given major, minor and machine architecture of an macOS machine.
"""
arches = []
def _supports_arch(major, minor, arch):
# Looking at the application support for macOS versions in the chart
# provided by https://en.wikipedia.org/wiki/OS_X#Versions it appears
# our timeline looks roughly like:
#
# 10.0 - Introduces ppc support.
# 10.4 - Introduces ppc64, i386, and x86_64 support, however the ppc64
# and x86_64 support is CLI only, and cannot be used for GUI
# applications.
# 10.5 - Extends ppc64 and x86_64 support to cover GUI applications.
# 10.6 - Drops support for ppc64
# 10.7 - Drops support for ppc
#
# Given that we do not know if we're installing a CLI or a GUI
# application, we must be conservative and assume it might be a GUI
# application and behave as if ppc64 and x86_64 support did not occur
# until 10.5.
#
# Note: The above information is taken from the "Application support"
# column in the chart not the "Processor support" since I believe
# that we care about what instruction sets an application can use
# not which processors the OS supports.
if arch == 'ppc':
return (major, minor) <= (10, 5)
if arch == 'ppc64':
return (major, minor) == (10, 5)
if arch == 'i386':
return (major, minor) >= (10, 4)
if arch == 'x86_64':
return (major, minor) >= (10, 5)
if arch in groups:
for garch in groups[arch]:
if _supports_arch(major, minor, garch):
return True
return False
groups = OrderedDict([
("fat", ("i386", "ppc")),
("intel", ("x86_64", "i386")),
("fat64", ("x86_64", "ppc64")),
("fat32", ("x86_64", "i386", "ppc")),
])
if _supports_arch(major, minor, machine):
arches.append(machine)
for garch in groups:
if machine in groups[garch] and _supports_arch(major, minor, garch):
arches.append(garch)
arches.append('universal')
return arches
def get_supported(versions=None, noarch=False, platform=None,
impl=None, abi=None):
"""Return a list of supported tags for each version specified in
`versions`.
:param versions: a list of string versions, of the form ["33", "32"],
or None. The first version will be assumed to support our ABI.
:param platform: specify the exact platform you want valid
tags for, or None. If None, use the local system platform.
:param impl: specify the exact implementation you want valid
tags for, or None. If None, use the local interpreter impl.
:param abi: specify the exact abi you want valid
tags for, or None. If None, use the local interpreter abi.
"""
supported = []
# Versions must be given with respect to the preference
if versions is None:
versions = []
version_info = get_impl_version_info()
major = version_info[:-1]
# Support all previous minor Python versions.
for minor in range(version_info[-1], -1, -1):
versions.append(''.join(map(str, major + (minor,))))
impl = impl or get_abbr_impl()
abis = []
abi = abi or get_abi_tag()
if abi:
abis[0:0] = [abi]
abi3s = set()
import imp
for suffix in imp.get_suffixes():
if suffix[0].startswith('.abi'):
abi3s.add(suffix[0].split('.', 2)[1])
abis.extend(sorted(list(abi3s)))
abis.append('none')
if not noarch:
arch = platform or get_platform()
if arch.startswith('macosx'):
# support macosx-10.6-intel on macosx-10.9-x86_64
match = _osx_arch_pat.match(arch)
if match:
name, major, minor, actual_arch = match.groups()
tpl = '{}_{}_%i_%s'.format(name, major)
arches = []
for m in reversed(range(int(minor) + 1)):
for a in get_darwin_arches(int(major), m, actual_arch):
arches.append(tpl % (m, a))
else:
# arch pattern didn't match (?!)
arches = [arch]
elif platform is None and is_manylinux1_compatible():
arches = [arch.replace('linux', 'manylinux1'), arch]
else:
arches = [arch]
# Current version, current API (built specifically for our Python):
for abi in abis:
for arch in arches:
supported.append(('%s%s' % (impl, versions[0]), abi, arch))
# abi3 modules compatible with older version of Python
for version in versions[1:]:
# abi3 was introduced in Python 3.2
if version in {'31', '30'}:
break
for abi in abi3s: # empty set if not Python 3
for arch in arches:
supported.append(("%s%s" % (impl, version), abi, arch))
# Has binaries, does not use the Python API:
for arch in arches:
supported.append(('py%s' % (versions[0][0]), 'none', arch))
# No abi / arch, but requires our implementation:
supported.append(('%s%s' % (impl, versions[0]), 'none', 'any'))
# Tagged specifically as being cross-version compatible
# (with just the major version specified)
supported.append(('%s%s' % (impl, versions[0][0]), 'none', 'any'))
# No abi / arch, generic Python
for i, version in enumerate(versions):
supported.append(('py%s' % (version,), 'none', 'any'))
if i == 0:
supported.append(('py%s' % (version[0]), 'none', 'any'))
return supported
implementation_tag = get_impl_tag()

View File

@@ -0,0 +1,69 @@
from __future__ import absolute_import
import logging
from .req_install import InstallRequirement
from .req_set import RequirementSet
from .req_file import parse_requirements
from pip._internal.utils.logging import indent_log
__all__ = [
"RequirementSet", "InstallRequirement",
"parse_requirements", "install_given_reqs",
]
logger = logging.getLogger(__name__)
def install_given_reqs(to_install, install_options, global_options=(),
*args, **kwargs):
"""
Install everything in the given list.
(to be called after having downloaded and unpacked the packages)
"""
if to_install:
logger.info(
'Installing collected packages: %s',
', '.join([req.name for req in to_install]),
)
with indent_log():
for requirement in to_install:
if requirement.conflicts_with:
logger.info(
'Found existing installation: %s',
requirement.conflicts_with,
)
with indent_log():
uninstalled_pathset = requirement.uninstall(
auto_confirm=True
)
try:
requirement.install(
install_options,
global_options,
*args,
**kwargs
)
except:
should_rollback = (
requirement.conflicts_with and
not requirement.install_succeeded
)
# if install did not succeed, rollback previous uninstall
if should_rollback:
uninstalled_pathset.rollback()
raise
else:
should_commit = (
requirement.conflicts_with and
requirement.install_succeeded
)
if should_commit:
uninstalled_pathset.commit()
requirement.remove_temporary_source()
return to_install

View File

@@ -0,0 +1,338 @@
"""
Requirements file parsing
"""
from __future__ import absolute_import
import optparse
import os
import re
import shlex
import sys
from pip._vendor.six.moves import filterfalse
from pip._vendor.six.moves.urllib import parse as urllib_parse
from pip._internal import cmdoptions
from pip._internal.download import get_file_content
from pip._internal.exceptions import RequirementsFileParseError
from pip._internal.req.req_install import InstallRequirement
__all__ = ['parse_requirements']
SCHEME_RE = re.compile(r'^(http|https|file):', re.I)
COMMENT_RE = re.compile(r'(^|\s)+#.*$')
# Matches environment variable-style values in '${MY_VARIABLE_1}' with the
# variable name consisting of only uppercase letters, digits or the '_'
# (underscore). This follows the POSIX standard defined in IEEE Std 1003.1,
# 2013 Edition.
ENV_VAR_RE = re.compile(r'(?P<var>\$\{(?P<name>[A-Z0-9_]+)\})')
SUPPORTED_OPTIONS = [
cmdoptions.constraints,
cmdoptions.editable,
cmdoptions.requirements,
cmdoptions.no_index,
cmdoptions.index_url,
cmdoptions.find_links,
cmdoptions.extra_index_url,
cmdoptions.always_unzip,
cmdoptions.no_binary,
cmdoptions.only_binary,
cmdoptions.pre,
cmdoptions.process_dependency_links,
cmdoptions.trusted_host,
cmdoptions.require_hashes,
]
# options to be passed to requirements
SUPPORTED_OPTIONS_REQ = [
cmdoptions.install_options,
cmdoptions.global_options,
cmdoptions.hash,
]
# the 'dest' string values
SUPPORTED_OPTIONS_REQ_DEST = [o().dest for o in SUPPORTED_OPTIONS_REQ]
def parse_requirements(filename, finder=None, comes_from=None, options=None,
session=None, constraint=False, wheel_cache=None):
"""Parse a requirements file and yield InstallRequirement instances.
:param filename: Path or url of requirements file.
:param finder: Instance of pip.index.PackageFinder.
:param comes_from: Origin description of requirements.
:param options: cli options.
:param session: Instance of pip.download.PipSession.
:param constraint: If true, parsing a constraint file rather than
requirements file.
:param wheel_cache: Instance of pip.wheel.WheelCache
"""
if session is None:
raise TypeError(
"parse_requirements() missing 1 required keyword argument: "
"'session'"
)
_, content = get_file_content(
filename, comes_from=comes_from, session=session
)
lines_enum = preprocess(content, options)
for line_number, line in lines_enum:
req_iter = process_line(line, filename, line_number, finder,
comes_from, options, session, wheel_cache,
constraint=constraint)
for req in req_iter:
yield req
def preprocess(content, options):
"""Split, filter, and join lines, and return a line iterator
:param content: the content of the requirements file
:param options: cli options
"""
lines_enum = enumerate(content.splitlines(), start=1)
lines_enum = join_lines(lines_enum)
lines_enum = ignore_comments(lines_enum)
lines_enum = skip_regex(lines_enum, options)
lines_enum = expand_env_variables(lines_enum)
return lines_enum
def process_line(line, filename, line_number, finder=None, comes_from=None,
options=None, session=None, wheel_cache=None,
constraint=False):
"""Process a single requirements line; This can result in creating/yielding
requirements, or updating the finder.
For lines that contain requirements, the only options that have an effect
are from SUPPORTED_OPTIONS_REQ, and they are scoped to the
requirement. Other options from SUPPORTED_OPTIONS may be present, but are
ignored.
For lines that do not contain requirements, the only options that have an
effect are from SUPPORTED_OPTIONS. Options from SUPPORTED_OPTIONS_REQ may
be present, but are ignored. These lines may contain multiple options
(although our docs imply only one is supported), and all our parsed and
affect the finder.
:param constraint: If True, parsing a constraints file.
:param options: OptionParser options that we may update
"""
parser = build_parser(line)
defaults = parser.get_default_values()
defaults.index_url = None
if finder:
# `finder.format_control` will be updated during parsing
defaults.format_control = finder.format_control
args_str, options_str = break_args_options(line)
if sys.version_info < (2, 7, 3):
# Prior to 2.7.3, shlex cannot deal with unicode entries
options_str = options_str.encode('utf8')
opts, _ = parser.parse_args(shlex.split(options_str), defaults)
# preserve for the nested code path
line_comes_from = '%s %s (line %s)' % (
'-c' if constraint else '-r', filename, line_number,
)
# yield a line requirement
if args_str:
isolated = options.isolated_mode if options else False
if options:
cmdoptions.check_install_build_global(options, opts)
# get the options that apply to requirements
req_options = {}
for dest in SUPPORTED_OPTIONS_REQ_DEST:
if dest in opts.__dict__ and opts.__dict__[dest]:
req_options[dest] = opts.__dict__[dest]
yield InstallRequirement.from_line(
args_str, line_comes_from, constraint=constraint,
isolated=isolated, options=req_options, wheel_cache=wheel_cache
)
# yield an editable requirement
elif opts.editables:
isolated = options.isolated_mode if options else False
yield InstallRequirement.from_editable(
opts.editables[0], comes_from=line_comes_from,
constraint=constraint, isolated=isolated, wheel_cache=wheel_cache
)
# parse a nested requirements file
elif opts.requirements or opts.constraints:
if opts.requirements:
req_path = opts.requirements[0]
nested_constraint = False
else:
req_path = opts.constraints[0]
nested_constraint = True
# original file is over http
if SCHEME_RE.search(filename):
# do a url join so relative paths work
req_path = urllib_parse.urljoin(filename, req_path)
# original file and nested file are paths
elif not SCHEME_RE.search(req_path):
# do a join so relative paths work
req_path = os.path.join(os.path.dirname(filename), req_path)
# TODO: Why not use `comes_from='-r {} (line {})'` here as well?
parser = parse_requirements(
req_path, finder, comes_from, options, session,
constraint=nested_constraint, wheel_cache=wheel_cache
)
for req in parser:
yield req
# percolate hash-checking option upward
elif opts.require_hashes:
options.require_hashes = opts.require_hashes
# set finder options
elif finder:
if opts.index_url:
finder.index_urls = [opts.index_url]
if opts.no_index is True:
finder.index_urls = []
if opts.extra_index_urls:
finder.index_urls.extend(opts.extra_index_urls)
if opts.find_links:
# FIXME: it would be nice to keep track of the source
# of the find_links: support a find-links local path
# relative to a requirements file.
value = opts.find_links[0]
req_dir = os.path.dirname(os.path.abspath(filename))
relative_to_reqs_file = os.path.join(req_dir, value)
if os.path.exists(relative_to_reqs_file):
value = relative_to_reqs_file
finder.find_links.append(value)
if opts.pre:
finder.allow_all_prereleases = True
if opts.process_dependency_links:
finder.process_dependency_links = True
if opts.trusted_hosts:
finder.secure_origins.extend(
("*", host, "*") for host in opts.trusted_hosts)
def break_args_options(line):
"""Break up the line into an args and options string. We only want to shlex
(and then optparse) the options, not the args. args can contain markers
which are corrupted by shlex.
"""
tokens = line.split(' ')
args = []
options = tokens[:]
for token in tokens:
if token.startswith('-') or token.startswith('--'):
break
else:
args.append(token)
options.pop(0)
return ' '.join(args), ' '.join(options)
def build_parser(line):
"""
Return a parser for parsing requirement lines
"""
parser = optparse.OptionParser(add_help_option=False)
option_factories = SUPPORTED_OPTIONS + SUPPORTED_OPTIONS_REQ
for option_factory in option_factories:
option = option_factory()
parser.add_option(option)
# By default optparse sys.exits on parsing errors. We want to wrap
# that in our own exception.
def parser_exit(self, msg):
# add offending line
msg = 'Invalid requirement: %s\n%s' % (line, msg)
raise RequirementsFileParseError(msg)
parser.exit = parser_exit
return parser
def join_lines(lines_enum):
"""Joins a line ending in '\' with the previous line (except when following
comments). The joined line takes on the index of the first line.
"""
primary_line_number = None
new_line = []
for line_number, line in lines_enum:
if not line.endswith('\\') or COMMENT_RE.match(line):
if COMMENT_RE.match(line):
# this ensures comments are always matched later
line = ' ' + line
if new_line:
new_line.append(line)
yield primary_line_number, ''.join(new_line)
new_line = []
else:
yield line_number, line
else:
if not new_line:
primary_line_number = line_number
new_line.append(line.strip('\\'))
# last line contains \
if new_line:
yield primary_line_number, ''.join(new_line)
# TODO: handle space after '\'.
def ignore_comments(lines_enum):
"""
Strips comments and filter empty lines.
"""
for line_number, line in lines_enum:
line = COMMENT_RE.sub('', line)
line = line.strip()
if line:
yield line_number, line
def skip_regex(lines_enum, options):
"""
Skip lines that match '--skip-requirements-regex' pattern
Note: the regex pattern is only built once
"""
skip_regex = options.skip_requirements_regex if options else None
if skip_regex:
pattern = re.compile(skip_regex)
lines_enum = filterfalse(lambda e: pattern.search(e[1]), lines_enum)
return lines_enum
def expand_env_variables(lines_enum):
"""Replace all environment variables that can be retrieved via `os.getenv`.
The only allowed format for environment variables defined in the
requirement file is `${MY_VARIABLE_1}` to ensure two things:
1. Strings that contain a `$` aren't accidentally (partially) expanded.
2. Ensure consistency across platforms for requirement files.
These points are the result of a discusssion on the `github pull
request #3514 <https://github.com/pypa/pip/pull/3514>`_.
Valid characters in variable names follow the `POSIX standard
<http://pubs.opengroup.org/onlinepubs/9699919799/>`_ and are limited
to uppercase letter, digits and the `_` (underscore).
"""
for line_number, line in lines_enum:
for env_var, var_name in ENV_VAR_RE.findall(line):
value = os.getenv(var_name)
if not value:
continue
line = line.replace(env_var, value)
yield line_number, line

View File

@@ -0,0 +1,164 @@
from __future__ import absolute_import
import logging
from collections import OrderedDict
from pip._internal.exceptions import InstallationError
from pip._internal.utils.logging import indent_log
from pip._internal.wheel import Wheel
logger = logging.getLogger(__name__)
class RequirementSet(object):
def __init__(self, require_hashes=False):
"""Create a RequirementSet.
:param wheel_cache: The pip wheel cache, for passing to
InstallRequirement.
"""
self.requirements = OrderedDict()
self.require_hashes = require_hashes
# Mapping of alias: real_name
self.requirement_aliases = {}
self.unnamed_requirements = []
self.successfully_downloaded = []
self.reqs_to_cleanup = []
def __str__(self):
reqs = [req for req in self.requirements.values()
if not req.comes_from]
reqs.sort(key=lambda req: req.name.lower())
return ' '.join([str(req.req) for req in reqs])
def __repr__(self):
reqs = [req for req in self.requirements.values()]
reqs.sort(key=lambda req: req.name.lower())
reqs_str = ', '.join([str(req.req) for req in reqs])
return ('<%s object; %d requirement(s): %s>'
% (self.__class__.__name__, len(reqs), reqs_str))
def add_requirement(self, install_req, parent_req_name=None,
extras_requested=None):
"""Add install_req as a requirement to install.
:param parent_req_name: The name of the requirement that needed this
added. The name is used because when multiple unnamed requirements
resolve to the same name, we could otherwise end up with dependency
links that point outside the Requirements set. parent_req must
already be added. Note that None implies that this is a user
supplied requirement, vs an inferred one.
:param extras_requested: an iterable of extras used to evaluate the
environment markers.
:return: Additional requirements to scan. That is either [] if
the requirement is not applicable, or [install_req] if the
requirement is applicable and has just been added.
"""
name = install_req.name
if not install_req.match_markers(extras_requested):
logger.info("Ignoring %s: markers '%s' don't match your "
"environment", install_req.name,
install_req.markers)
return [], None
# This check has to come after we filter requirements with the
# environment markers.
if install_req.link and install_req.link.is_wheel:
wheel = Wheel(install_req.link.filename)
if not wheel.supported():
raise InstallationError(
"%s is not a supported wheel on this platform." %
wheel.filename
)
# This next bit is really a sanity check.
assert install_req.is_direct == (parent_req_name is None), (
"a direct req shouldn't have a parent and also, "
"a non direct req should have a parent"
)
if not name:
# url or path requirement w/o an egg fragment
self.unnamed_requirements.append(install_req)
return [install_req], None
else:
try:
existing_req = self.get_requirement(name)
except KeyError:
existing_req = None
if (parent_req_name is None and existing_req and not
existing_req.constraint and
existing_req.extras == install_req.extras and not
existing_req.req.specifier == install_req.req.specifier):
raise InstallationError(
'Double requirement given: %s (already in %s, name=%r)'
% (install_req, existing_req, name))
if not existing_req:
# Add requirement
self.requirements[name] = install_req
# FIXME: what about other normalizations? E.g., _ vs. -?
if name.lower() != name:
self.requirement_aliases[name.lower()] = name
result = [install_req]
else:
# Assume there's no need to scan, and that we've already
# encountered this for scanning.
result = []
if not install_req.constraint and existing_req.constraint:
if (install_req.link and not (existing_req.link and
install_req.link.path == existing_req.link.path)):
self.reqs_to_cleanup.append(install_req)
raise InstallationError(
"Could not satisfy constraints for '%s': "
"installation from path or url cannot be "
"constrained to a version" % name,
)
# If we're now installing a constraint, mark the existing
# object for real installation.
existing_req.constraint = False
existing_req.extras = tuple(
sorted(set(existing_req.extras).union(
set(install_req.extras))))
logger.debug("Setting %s extras to: %s",
existing_req, existing_req.extras)
# And now we need to scan this.
result = [existing_req]
# Canonicalise to the already-added object for the backref
# check below.
install_req = existing_req
# We return install_req here to allow for the caller to add it to
# the dependency information for the parent package.
return result, install_req
def has_requirement(self, project_name):
name = project_name.lower()
if (name in self.requirements and
not self.requirements[name].constraint or
name in self.requirement_aliases and
not self.requirements[self.requirement_aliases[name]].constraint):
return True
return False
@property
def has_requirements(self):
return list(req for req in self.requirements.values() if not
req.constraint) or self.unnamed_requirements
def get_requirement(self, project_name):
for name in project_name, project_name.lower():
if name in self.requirements:
return self.requirements[name]
if name in self.requirement_aliases:
return self.requirements[self.requirement_aliases[name]]
raise KeyError("No project with the name %r" % project_name)
def cleanup_files(self):
"""Clean up files, remove builds."""
logger.debug('Cleaning up...')
with indent_log():
for req in self.reqs_to_cleanup:
req.remove_temporary_source()

View File

@@ -0,0 +1,455 @@
from __future__ import absolute_import
import csv
import functools
import logging
import os
import sys
import sysconfig
from pip._vendor import pkg_resources
from pip._internal.compat import WINDOWS, cache_from_source, uses_pycache
from pip._internal.exceptions import UninstallationError
from pip._internal.locations import bin_py, bin_user
from pip._internal.utils.logging import indent_log
from pip._internal.utils.misc import (
FakeFile, ask, dist_in_usersite, dist_is_local, egg_link_path, is_local,
normalize_path, renames,
)
from pip._internal.utils.temp_dir import TempDirectory
logger = logging.getLogger(__name__)
def _script_names(dist, script_name, is_gui):
"""Create the fully qualified name of the files created by
{console,gui}_scripts for the given ``dist``.
Returns the list of file names
"""
if dist_in_usersite(dist):
bin_dir = bin_user
else:
bin_dir = bin_py
exe_name = os.path.join(bin_dir, script_name)
paths_to_remove = [exe_name]
if WINDOWS:
paths_to_remove.append(exe_name + '.exe')
paths_to_remove.append(exe_name + '.exe.manifest')
if is_gui:
paths_to_remove.append(exe_name + '-script.pyw')
else:
paths_to_remove.append(exe_name + '-script.py')
return paths_to_remove
def _unique(fn):
@functools.wraps(fn)
def unique(*args, **kw):
seen = set()
for item in fn(*args, **kw):
if item not in seen:
seen.add(item)
yield item
return unique
@_unique
def uninstallation_paths(dist):
"""
Yield all the uninstallation paths for dist based on RECORD-without-.pyc
Yield paths to all the files in RECORD. For each .py file in RECORD, add
the .pyc in the same directory.
UninstallPathSet.add() takes care of the __pycache__ .pyc.
"""
r = csv.reader(FakeFile(dist.get_metadata_lines('RECORD')))
for row in r:
path = os.path.join(dist.location, row[0])
yield path
if path.endswith('.py'):
dn, fn = os.path.split(path)
base = fn[:-3]
path = os.path.join(dn, base + '.pyc')
yield path
def compact(paths):
"""Compact a path set to contain the minimal number of paths
necessary to contain all paths in the set. If /a/path/ and
/a/path/to/a/file.txt are both in the set, leave only the
shorter path."""
sep = os.path.sep
short_paths = set()
for path in sorted(paths, key=len):
should_add = any(
path.startswith(shortpath.rstrip("*")) and
path[len(shortpath.rstrip("*").rstrip(sep))] == sep
for shortpath in short_paths
)
if not should_add:
short_paths.add(path)
return short_paths
def compress_for_output_listing(paths):
"""Returns a tuple of 2 sets of which paths to display to user
The first set contains paths that would be deleted. Files of a package
are not added and the top-level directory of the package has a '*' added
at the end - to signify that all it's contents are removed.
The second set contains files that would have been skipped in the above
folders.
"""
will_remove = list(paths)
will_skip = set()
# Determine folders and files
folders = set()
files = set()
for path in will_remove:
if path.endswith(".pyc"):
continue
if path.endswith("__init__.py") or ".dist-info" in path:
folders.add(os.path.dirname(path))
files.add(path)
folders = compact(folders)
# This walks the tree using os.walk to not miss extra folders
# that might get added.
for folder in folders:
for dirpath, _, dirfiles in os.walk(folder):
for fname in dirfiles:
if fname.endswith(".pyc"):
continue
file_ = os.path.normcase(os.path.join(dirpath, fname))
if os.path.isfile(file_) and file_ not in files:
# We are skipping this file. Add it to the set.
will_skip.add(file_)
will_remove = files | {
os.path.join(folder, "*") for folder in folders
}
return will_remove, will_skip
class UninstallPathSet(object):
"""A set of file paths to be removed in the uninstallation of a
requirement."""
def __init__(self, dist):
self.paths = set()
self._refuse = set()
self.pth = {}
self.dist = dist
self.save_dir = TempDirectory(kind="uninstall")
self._moved_paths = []
def _permitted(self, path):
"""
Return True if the given path is one we are permitted to
remove/modify, False otherwise.
"""
return is_local(path)
def add(self, path):
head, tail = os.path.split(path)
# we normalize the head to resolve parent directory symlinks, but not
# the tail, since we only want to uninstall symlinks, not their targets
path = os.path.join(normalize_path(head), os.path.normcase(tail))
if not os.path.exists(path):
return
if self._permitted(path):
self.paths.add(path)
else:
self._refuse.add(path)
# __pycache__ files can show up after 'installed-files.txt' is created,
# due to imports
if os.path.splitext(path)[1] == '.py' and uses_pycache:
self.add(cache_from_source(path))
def add_pth(self, pth_file, entry):
pth_file = normalize_path(pth_file)
if self._permitted(pth_file):
if pth_file not in self.pth:
self.pth[pth_file] = UninstallPthEntries(pth_file)
self.pth[pth_file].add(entry)
else:
self._refuse.add(pth_file)
def _stash(self, path):
return os.path.join(
self.save_dir.path, os.path.splitdrive(path)[1].lstrip(os.path.sep)
)
def remove(self, auto_confirm=False, verbose=False):
"""Remove paths in ``self.paths`` with confirmation (unless
``auto_confirm`` is True)."""
if not self.paths:
logger.info(
"Can't uninstall '%s'. No files were found to uninstall.",
self.dist.project_name,
)
return
dist_name_version = (
self.dist.project_name + "-" + self.dist.version
)
logger.info('Uninstalling %s:', dist_name_version)
with indent_log():
if auto_confirm or self._allowed_to_proceed(verbose):
self.save_dir.create()
for path in sorted(compact(self.paths)):
new_path = self._stash(path)
logger.debug('Removing file or directory %s', path)
self._moved_paths.append(path)
renames(path, new_path)
for pth in self.pth.values():
pth.remove()
logger.info('Successfully uninstalled %s', dist_name_version)
def _allowed_to_proceed(self, verbose):
"""Display which files would be deleted and prompt for confirmation
"""
def _display(msg, paths):
if not paths:
return
logger.info(msg)
with indent_log():
for path in sorted(compact(paths)):
logger.info(path)
if not verbose:
will_remove, will_skip = compress_for_output_listing(self.paths)
else:
# In verbose mode, display all the files that are going to be
# deleted.
will_remove = list(self.paths)
will_skip = set()
_display('Would remove:', will_remove)
_display('Would not remove (might be manually added):', will_skip)
_display('Would not remove (outside of prefix):', self._refuse)
return ask('Proceed (y/n)? ', ('y', 'n')) == 'y'
def rollback(self):
"""Rollback the changes previously made by remove()."""
if self.save_dir.path is None:
logger.error(
"Can't roll back %s; was not uninstalled",
self.dist.project_name,
)
return False
logger.info('Rolling back uninstall of %s', self.dist.project_name)
for path in self._moved_paths:
tmp_path = self._stash(path)
logger.debug('Replacing %s', path)
renames(tmp_path, path)
for pth in self.pth.values():
pth.rollback()
def commit(self):
"""Remove temporary save dir: rollback will no longer be possible."""
self.save_dir.cleanup()
self._moved_paths = []
@classmethod
def from_dist(cls, dist):
dist_path = normalize_path(dist.location)
if not dist_is_local(dist):
logger.info(
"Not uninstalling %s at %s, outside environment %s",
dist.key,
dist_path,
sys.prefix,
)
return cls(dist)
if dist_path in {p for p in {sysconfig.get_path("stdlib"),
sysconfig.get_path("platstdlib")}
if p}:
logger.info(
"Not uninstalling %s at %s, as it is in the standard library.",
dist.key,
dist_path,
)
return cls(dist)
paths_to_remove = cls(dist)
develop_egg_link = egg_link_path(dist)
develop_egg_link_egg_info = '{}.egg-info'.format(
pkg_resources.to_filename(dist.project_name))
egg_info_exists = dist.egg_info and os.path.exists(dist.egg_info)
# Special case for distutils installed package
distutils_egg_info = getattr(dist._provider, 'path', None)
# Uninstall cases order do matter as in the case of 2 installs of the
# same package, pip needs to uninstall the currently detected version
if (egg_info_exists and dist.egg_info.endswith('.egg-info') and
not dist.egg_info.endswith(develop_egg_link_egg_info)):
# if dist.egg_info.endswith(develop_egg_link_egg_info), we
# are in fact in the develop_egg_link case
paths_to_remove.add(dist.egg_info)
if dist.has_metadata('installed-files.txt'):
for installed_file in dist.get_metadata(
'installed-files.txt').splitlines():
path = os.path.normpath(
os.path.join(dist.egg_info, installed_file)
)
paths_to_remove.add(path)
# FIXME: need a test for this elif block
# occurs with --single-version-externally-managed/--record outside
# of pip
elif dist.has_metadata('top_level.txt'):
if dist.has_metadata('namespace_packages.txt'):
namespaces = dist.get_metadata('namespace_packages.txt')
else:
namespaces = []
for top_level_pkg in [
p for p
in dist.get_metadata('top_level.txt').splitlines()
if p and p not in namespaces]:
path = os.path.join(dist.location, top_level_pkg)
paths_to_remove.add(path)
paths_to_remove.add(path + '.py')
paths_to_remove.add(path + '.pyc')
paths_to_remove.add(path + '.pyo')
elif distutils_egg_info:
raise UninstallationError(
"Cannot uninstall {!r}. It is a distutils installed project "
"and thus we cannot accurately determine which files belong "
"to it which would lead to only a partial uninstall.".format(
dist.project_name,
)
)
elif dist.location.endswith('.egg'):
# package installed by easy_install
# We cannot match on dist.egg_name because it can slightly vary
# i.e. setuptools-0.6c11-py2.6.egg vs setuptools-0.6rc11-py2.6.egg
paths_to_remove.add(dist.location)
easy_install_egg = os.path.split(dist.location)[1]
easy_install_pth = os.path.join(os.path.dirname(dist.location),
'easy-install.pth')
paths_to_remove.add_pth(easy_install_pth, './' + easy_install_egg)
elif egg_info_exists and dist.egg_info.endswith('.dist-info'):
for path in uninstallation_paths(dist):
paths_to_remove.add(path)
elif develop_egg_link:
# develop egg
with open(develop_egg_link, 'r') as fh:
link_pointer = os.path.normcase(fh.readline().strip())
assert (link_pointer == dist.location), (
'Egg-link %s does not match installed location of %s '
'(at %s)' % (link_pointer, dist.project_name, dist.location)
)
paths_to_remove.add(develop_egg_link)
easy_install_pth = os.path.join(os.path.dirname(develop_egg_link),
'easy-install.pth')
paths_to_remove.add_pth(easy_install_pth, dist.location)
else:
logger.debug(
'Not sure how to uninstall: %s - Check: %s',
dist, dist.location,
)
# find distutils scripts= scripts
if dist.has_metadata('scripts') and dist.metadata_isdir('scripts'):
for script in dist.metadata_listdir('scripts'):
if dist_in_usersite(dist):
bin_dir = bin_user
else:
bin_dir = bin_py
paths_to_remove.add(os.path.join(bin_dir, script))
if WINDOWS:
paths_to_remove.add(os.path.join(bin_dir, script) + '.bat')
# find console_scripts
_scripts_to_remove = []
console_scripts = dist.get_entry_map(group='console_scripts')
for name in console_scripts.keys():
_scripts_to_remove.extend(_script_names(dist, name, False))
# find gui_scripts
gui_scripts = dist.get_entry_map(group='gui_scripts')
for name in gui_scripts.keys():
_scripts_to_remove.extend(_script_names(dist, name, True))
for s in _scripts_to_remove:
paths_to_remove.add(s)
return paths_to_remove
class UninstallPthEntries(object):
def __init__(self, pth_file):
if not os.path.isfile(pth_file):
raise UninstallationError(
"Cannot remove entries from nonexistent file %s" % pth_file
)
self.file = pth_file
self.entries = set()
self._saved_lines = None
def add(self, entry):
entry = os.path.normcase(entry)
# On Windows, os.path.normcase converts the entry to use
# backslashes. This is correct for entries that describe absolute
# paths outside of site-packages, but all the others use forward
# slashes.
if WINDOWS and not os.path.splitdrive(entry)[0]:
entry = entry.replace('\\', '/')
self.entries.add(entry)
def remove(self):
logger.debug('Removing pth entries from %s:', self.file)
with open(self.file, 'rb') as fh:
# windows uses '\r\n' with py3k, but uses '\n' with py2.x
lines = fh.readlines()
self._saved_lines = lines
if any(b'\r\n' in line for line in lines):
endline = '\r\n'
else:
endline = '\n'
# handle missing trailing newline
if lines and not lines[-1].endswith(endline.encode("utf-8")):
lines[-1] = lines[-1] + endline.encode("utf-8")
for entry in self.entries:
try:
logger.debug('Removing entry: %s', entry)
lines.remove((entry + endline).encode("utf-8"))
except ValueError:
pass
with open(self.file, 'wb') as fh:
fh.writelines(lines)
def rollback(self):
if self._saved_lines is None:
logger.error(
'Cannot roll back changes to %s, none were made', self.file
)
return False
logger.debug('Rolling %s back to previous state', self.file)
with open(self.file, 'wb') as fh:
fh.writelines(self._saved_lines)
return True

View File

@@ -0,0 +1,354 @@
"""Dependency Resolution
The dependency resolution in pip is performed as follows:
for top-level requirements:
a. only one spec allowed per project, regardless of conflicts or not.
otherwise a "double requirement" exception is raised
b. they override sub-dependency requirements.
for sub-dependencies
a. "first found, wins" (where the order is breadth first)
"""
import logging
from collections import defaultdict
from itertools import chain
from pip._internal.exceptions import (
BestVersionAlreadyInstalled, DistributionNotFound, HashError, HashErrors,
UnsupportedPythonVersion,
)
from pip._internal.req.req_install import InstallRequirement
from pip._internal.utils.logging import indent_log
from pip._internal.utils.misc import dist_in_usersite, ensure_dir
from pip._internal.utils.packaging import check_dist_requires_python
logger = logging.getLogger(__name__)
class Resolver(object):
"""Resolves which packages need to be installed/uninstalled to perform \
the requested operation without breaking the requirements of any package.
"""
_allowed_strategies = {"eager", "only-if-needed", "to-satisfy-only"}
def __init__(self, preparer, session, finder, wheel_cache, use_user_site,
ignore_dependencies, ignore_installed, ignore_requires_python,
force_reinstall, isolated, upgrade_strategy):
super(Resolver, self).__init__()
assert upgrade_strategy in self._allowed_strategies
self.preparer = preparer
self.finder = finder
self.session = session
# NOTE: This would eventually be replaced with a cache that can give
# information about both sdist and wheels transparently.
self.wheel_cache = wheel_cache
self.require_hashes = None # This is set in resolve
self.upgrade_strategy = upgrade_strategy
self.force_reinstall = force_reinstall
self.isolated = isolated
self.ignore_dependencies = ignore_dependencies
self.ignore_installed = ignore_installed
self.ignore_requires_python = ignore_requires_python
self.use_user_site = use_user_site
self._discovered_dependencies = defaultdict(list)
def resolve(self, requirement_set):
"""Resolve what operations need to be done
As a side-effect of this method, the packages (and their dependencies)
are downloaded, unpacked and prepared for installation. This
preparation is done by ``pip.operations.prepare``.
Once PyPI has static dependency metadata available, it would be
possible to move the preparation to become a step separated from
dependency resolution.
"""
# make the wheelhouse
if self.preparer.wheel_download_dir:
ensure_dir(self.preparer.wheel_download_dir)
# If any top-level requirement has a hash specified, enter
# hash-checking mode, which requires hashes from all.
root_reqs = (
requirement_set.unnamed_requirements +
list(requirement_set.requirements.values())
)
self.require_hashes = (
requirement_set.require_hashes or
any(req.has_hash_options for req in root_reqs)
)
# Display where finder is looking for packages
locations = self.finder.get_formatted_locations()
if locations:
logger.info(locations)
# Actually prepare the files, and collect any exceptions. Most hash
# exceptions cannot be checked ahead of time, because
# req.populate_link() needs to be called before we can make decisions
# based on link type.
discovered_reqs = []
hash_errors = HashErrors()
for req in chain(root_reqs, discovered_reqs):
try:
discovered_reqs.extend(
self._resolve_one(requirement_set, req)
)
except HashError as exc:
exc.req = req
hash_errors.append(exc)
if hash_errors:
raise hash_errors
def _is_upgrade_allowed(self, req):
if self.upgrade_strategy == "to-satisfy-only":
return False
elif self.upgrade_strategy == "eager":
return True
else:
assert self.upgrade_strategy == "only-if-needed"
return req.is_direct
def _set_req_to_reinstall(self, req):
"""
Set a requirement to be installed.
"""
# Don't uninstall the conflict if doing a user install and the
# conflict is not a user install.
if not self.use_user_site or dist_in_usersite(req.satisfied_by):
req.conflicts_with = req.satisfied_by
req.satisfied_by = None
# XXX: Stop passing requirement_set for options
def _check_skip_installed(self, req_to_install):
"""Check if req_to_install should be skipped.
This will check if the req is installed, and whether we should upgrade
or reinstall it, taking into account all the relevant user options.
After calling this req_to_install will only have satisfied_by set to
None if the req_to_install is to be upgraded/reinstalled etc. Any
other value will be a dist recording the current thing installed that
satisfies the requirement.
Note that for vcs urls and the like we can't assess skipping in this
routine - we simply identify that we need to pull the thing down,
then later on it is pulled down and introspected to assess upgrade/
reinstalls etc.
:return: A text reason for why it was skipped, or None.
"""
if self.ignore_installed:
return None
req_to_install.check_if_exists(self.use_user_site)
if not req_to_install.satisfied_by:
return None
if self.force_reinstall:
self._set_req_to_reinstall(req_to_install)
return None
if not self._is_upgrade_allowed(req_to_install):
if self.upgrade_strategy == "only-if-needed":
return 'not upgraded as not directly required'
return 'already satisfied'
# Check for the possibility of an upgrade. For link-based
# requirements we have to pull the tree down and inspect to assess
# the version #, so it's handled way down.
if not req_to_install.link:
try:
self.finder.find_requirement(req_to_install, upgrade=True)
except BestVersionAlreadyInstalled:
# Then the best version is installed.
return 'already up-to-date'
except DistributionNotFound:
# No distribution found, so we squash the error. It will
# be raised later when we re-try later to do the install.
# Why don't we just raise here?
pass
self._set_req_to_reinstall(req_to_install)
return None
def _get_abstract_dist_for(self, req):
"""Takes a InstallRequirement and returns a single AbstractDist \
representing a prepared variant of the same.
"""
assert self.require_hashes is not None, (
"require_hashes should have been set in Resolver.resolve()"
)
if req.editable:
return self.preparer.prepare_editable_requirement(
req, self.require_hashes, self.use_user_site, self.finder,
)
# satisfied_by is only evaluated by calling _check_skip_installed,
# so it must be None here.
assert req.satisfied_by is None
skip_reason = self._check_skip_installed(req)
if req.satisfied_by:
return self.preparer.prepare_installed_requirement(
req, self.require_hashes, skip_reason
)
upgrade_allowed = self._is_upgrade_allowed(req)
abstract_dist = self.preparer.prepare_linked_requirement(
req, self.session, self.finder, upgrade_allowed,
self.require_hashes
)
# NOTE
# The following portion is for determining if a certain package is
# going to be re-installed/upgraded or not and reporting to the user.
# This should probably get cleaned up in a future refactor.
# req.req is only avail after unpack for URL
# pkgs repeat check_if_exists to uninstall-on-upgrade
# (#14)
if not self.ignore_installed:
req.check_if_exists(self.use_user_site)
if req.satisfied_by:
should_modify = (
self.upgrade_strategy != "to-satisfy-only" or
self.force_reinstall or
self.ignore_installed or
req.link.scheme == 'file'
)
if should_modify:
self._set_req_to_reinstall(req)
else:
logger.info(
'Requirement already satisfied (use --upgrade to upgrade):'
' %s', req,
)
return abstract_dist
def _resolve_one(self, requirement_set, req_to_install):
"""Prepare a single requirements file.
:return: A list of additional InstallRequirements to also install.
"""
# Tell user what we are doing for this requirement:
# obtain (editable), skipping, processing (local url), collecting
# (remote url or package name)
if req_to_install.constraint or req_to_install.prepared:
return []
req_to_install.prepared = True
# register tmp src for cleanup in case something goes wrong
requirement_set.reqs_to_cleanup.append(req_to_install)
abstract_dist = self._get_abstract_dist_for(req_to_install)
# Parse and return dependencies
dist = abstract_dist.dist(self.finder)
try:
check_dist_requires_python(dist)
except UnsupportedPythonVersion as err:
if self.ignore_requires_python:
logger.warning(err.args[0])
else:
raise
more_reqs = []
def add_req(subreq, extras_requested):
sub_install_req = InstallRequirement.from_req(
str(subreq),
req_to_install,
isolated=self.isolated,
wheel_cache=self.wheel_cache,
)
parent_req_name = req_to_install.name
to_scan_again, add_to_parent = requirement_set.add_requirement(
sub_install_req,
parent_req_name=parent_req_name,
extras_requested=extras_requested,
)
if parent_req_name and add_to_parent:
self._discovered_dependencies[parent_req_name].append(
add_to_parent
)
more_reqs.extend(to_scan_again)
with indent_log():
# We add req_to_install before its dependencies, so that we
# can refer to it when adding dependencies.
if not requirement_set.has_requirement(req_to_install.name):
# 'unnamed' requirements will get added here
req_to_install.is_direct = True
requirement_set.add_requirement(
req_to_install, parent_req_name=None,
)
if not self.ignore_dependencies:
if req_to_install.extras:
logger.debug(
"Installing extra requirements: %r",
','.join(req_to_install.extras),
)
missing_requested = sorted(
set(req_to_install.extras) - set(dist.extras)
)
for missing in missing_requested:
logger.warning(
'%s does not provide the extra \'%s\'',
dist, missing
)
available_requested = sorted(
set(dist.extras) & set(req_to_install.extras)
)
for subreq in dist.requires(available_requested):
add_req(subreq, extras_requested=available_requested)
if not req_to_install.editable and not req_to_install.satisfied_by:
# XXX: --no-install leads this to report 'Successfully
# downloaded' for only non-editable reqs, even though we took
# action on them.
requirement_set.successfully_downloaded.append(req_to_install)
return more_reqs
def get_installation_order(self, req_set):
"""Create the installation order.
The installation order is topological - requirements are installed
before the requiring thing. We break cycles at an arbitrary point,
and make no other guarantees.
"""
# The current implementation, which we may change at any point
# installs the user specified things in the order given, except when
# dependencies must come earlier to achieve topological order.
order = []
ordered_reqs = set()
def schedule(req):
if req.satisfied_by or req in ordered_reqs:
return
if req.constraint:
return
ordered_reqs.add(req)
for dep in self._discovered_dependencies[req.name]:
schedule(dep)
order.append(req)
for install_req in req_set.requirements.values():
schedule(install_req)
return order

Some files were not shown because too many files have changed in this diff Show More