Skip to content
Fetching contributors…
Cannot retrieve contributors at this time
3053 lines (2432 sloc) 101 KB
# coding: utf-8
Package resource API
A resource is a logical file contained within a package, or a logical
subdirectory thereof. The package resource API expects resource names
to have their path parts separated with ``/``, *not* whatever the local
path separator is. Do not use os.path operations to manipulate resource
names being passed into the API.
The package resource API is designed to work with normal filesystem packages,
.egg files, and unpacked .egg files. It can also work in a limited way with
.zip files and with custom PEP 302 loaders that support the ``get_data()``
from __future__ import absolute_import
import sys
import os
import io
import time
import re
import types
import zipfile
import zipimport
import warnings
import stat
import functools
import pkgutil
import operator
import platform
import collections
import plistlib
import email.parser
import tempfile
import textwrap
import itertools
from pkgutil import get_importer
import _imp
except ImportError:
# Python 3.2 compatibility
import imp as _imp
from pip._vendor import six
from pip._vendor.six.moves import urllib, map, filter
# capture these to bypass sandboxing
from os import utime
from os import mkdir, rename, unlink
except ImportError:
# no write support, probably under GAE
from os import open as os_open
from os.path import isdir, split
import importlib.machinery as importlib_machinery
# access attribute to force import under delayed import mechanisms.
except ImportError:
importlib_machinery = None
from pip._vendor import appdirs
from pip._vendor import packaging
if (3, 0) < sys.version_info < (3, 3):
msg = (
"Support for Python 3.0-3.2 has been dropped. Future versions "
"will fail here."
# declare some globals that will be defined later to
# satisfy the linters.
require = None
working_set = None
class PEP440Warning(RuntimeWarning):
Used when there is an issue with a version or specifier not complying with
PEP 440.
class _SetuptoolsVersionMixin(object):
def __hash__(self):
return super(_SetuptoolsVersionMixin, self).__hash__()
def __lt__(self, other):
if isinstance(other, tuple):
return tuple(self) < other
return super(_SetuptoolsVersionMixin, self).__lt__(other)
def __le__(self, other):
if isinstance(other, tuple):
return tuple(self) <= other
return super(_SetuptoolsVersionMixin, self).__le__(other)
def __eq__(self, other):
if isinstance(other, tuple):
return tuple(self) == other
return super(_SetuptoolsVersionMixin, self).__eq__(other)
def __ge__(self, other):
if isinstance(other, tuple):
return tuple(self) >= other
return super(_SetuptoolsVersionMixin, self).__ge__(other)
def __gt__(self, other):
if isinstance(other, tuple):
return tuple(self) > other
return super(_SetuptoolsVersionMixin, self).__gt__(other)
def __ne__(self, other):
if isinstance(other, tuple):
return tuple(self) != other
return super(_SetuptoolsVersionMixin, self).__ne__(other)
def __getitem__(self, key):
return tuple(self)[key]
def __iter__(self):
component_re = re.compile(r'(\d+ | [a-z]+ | \.| -)', re.VERBOSE)
replace = {
'pre': 'c',
'preview': 'c',
'-': 'final-',
'rc': 'c',
'dev': '@',
def _parse_version_parts(s):
for part in component_re.split(s):
part = replace(part, part)
if not part or part == '.':
if part[:1] in '0123456789':
# pad for numeric comparison
yield part.zfill(8)
yield '*' + part
# ensure that alpha/beta/candidate are before final
yield '*final'
def old_parse_version(s):
parts = []
for part in _parse_version_parts(s.lower()):
if part.startswith('*'):
# remove '-' before a prerelease tag
if part < '*final':
while parts and parts[-1] == '*final-':
# remove trailing zeros from each series of numeric parts
while parts and parts[-1] == '00000000':
return tuple(parts)
# Warn for use of this function
"You have iterated over the result of "
"pkg_resources.parse_version. This is a legacy behavior which is "
"inconsistent with the new version class introduced in setuptools "
"8.0. In most cases, conversion to a tuple is unnecessary. For "
"comparison of versions, sort the Version instances directly. If "
"you have another use case requiring the tuple, please file a "
"bug with the setuptools project describing that need.",
for part in old_parse_version(str(self)):
yield part
class SetuptoolsVersion(_SetuptoolsVersionMixin, packaging.version.Version):
class SetuptoolsLegacyVersion(_SetuptoolsVersionMixin,
def parse_version(v):
return SetuptoolsVersion(v)
except packaging.version.InvalidVersion:
return SetuptoolsLegacyVersion(v)
_state_vars = {}
def _declare_state(vartype, **kw):
_state_vars.update(dict.fromkeys(kw, vartype))
def __getstate__():
state = {}
g = globals()
for k, v in _state_vars.items():
state[k] = g['_sget_' + v](g[k])
return state
def __setstate__(state):
g = globals()
for k, v in state.items():
g['_sset_' + _state_vars[k]](k, g[k], v)
return state
def _sget_dict(val):
return val.copy()
def _sset_dict(key, ob, state):
def _sget_object(val):
return val.__getstate__()
def _sset_object(key, ob, state):
_sget_none = _sset_none = lambda *args: None
def get_supported_platform():
"""Return this platform's maximum compatible version.
distutils.util.get_platform() normally reports the minimum version
of Mac OS X that would be required to *use* extensions produced by
distutils. But what we want when checking compatibility is to know the
version of Mac OS X that we are *running*. To allow usage of packages that
explicitly require a newer version of Mac OS X, we must also know the
current version of the OS.
If this condition occurs for any other platform with a version in its
platform strings, this function should be extended accordingly.
plat = get_build_platform()
m = macosVersionString.match(plat)
if m is not None and sys.platform == "darwin":
plat = 'macosx-%s-%s' % ('.'.join(_macosx_vers()[:2]),
except ValueError:
# not Mac OS X
return plat
__all__ = [
# Basic resource access and distribution/entry point discovery
'require', 'run_script', 'get_provider', 'get_distribution',
'load_entry_point', 'get_entry_map', 'get_entry_info',
'resource_string', 'resource_stream', 'resource_filename',
'resource_listdir', 'resource_exists', 'resource_isdir',
# Environmental control
'declare_namespace', 'working_set', 'add_activation_listener',
'find_distributions', 'set_extraction_path', 'cleanup_resources',
# Primary implementation classes
'Environment', 'WorkingSet', 'ResourceManager',
'Distribution', 'Requirement', 'EntryPoint',
# Exceptions
'ResolutionError', 'VersionConflict', 'DistributionNotFound',
'UnknownExtra', 'ExtractionError',
# Warnings
# Parsing functions and string utilities
'parse_requirements', 'parse_version', 'safe_name', 'safe_version',
'get_platform', 'compatible_platforms', 'yield_lines', 'split_sections',
'safe_extra', 'to_filename', 'invalid_marker', 'evaluate_marker',
# filesystem utilities
'ensure_directory', 'normalize_path',
# Distribution "precedence" constants
# "Provider" interfaces, implementations, and registration/lookup APIs
'IMetadataProvider', 'IResourceProvider', 'FileMetadata',
'PathMetadata', 'EggMetadata', 'EmptyProvider', 'empty_provider',
'NullProvider', 'EggProvider', 'DefaultProvider', 'ZipProvider',
'register_finder', 'register_namespace_handler', 'register_loader_type',
'fixup_namespace_packages', 'get_importer',
# Deprecated/backward compatibility only
'run_main', 'AvailableDistributions',
class ResolutionError(Exception):
"""Abstract base for dependency resolution errors"""
def __repr__(self):
return self.__class__.__name__ + repr(self.args)
class VersionConflict(ResolutionError):
An already-installed version conflicts with the requested version.
Should be initialized with the installed Distribution and the requested
_template = "{self.dist} is installed but {self.req} is required"
def dist(self):
return self.args[0]
def req(self):
return self.args[1]
def report(self):
return self._template.format(**locals())
def with_context(self, required_by):
If required_by is non-empty, return a version of self that is a
if not required_by:
return self
args = self.args + (required_by,)
return ContextualVersionConflict(*args)
class ContextualVersionConflict(VersionConflict):
A VersionConflict that accepts a third parameter, the set of the
requirements that required the installed Distribution.
_template = VersionConflict._template + ' by {self.required_by}'
def required_by(self):
return self.args[2]
class DistributionNotFound(ResolutionError):
"""A requested distribution was not found"""
_template = ("The '{self.req}' distribution was not found "
"and is required by {self.requirers_str}")
def req(self):
return self.args[0]
def requirers(self):
return self.args[1]
def requirers_str(self):
if not self.requirers:
return 'the application'
return ', '.join(self.requirers)
def report(self):
return self._template.format(**locals())
def __str__(self):
class UnknownExtra(ResolutionError):
"""Distribution doesn't have an "extra feature" of the given name"""
_provider_factories = {}
PY_MAJOR = sys.version[:3]
def register_loader_type(loader_type, provider_factory):
"""Register `provider_factory` to make providers for `loader_type`
`loader_type` is the type or class of a PEP 302 ``module.__loader__``,
and `provider_factory` is a function that, passed a *module* object,
returns an ``IResourceProvider`` for that module.
_provider_factories[loader_type] = provider_factory
def get_provider(moduleOrReq):
"""Return an IResourceProvider for the named module or requirement"""
if isinstance(moduleOrReq, Requirement):
return working_set.find(moduleOrReq) or require(str(moduleOrReq))[0]
module = sys.modules[moduleOrReq]
except KeyError:
module = sys.modules[moduleOrReq]
loader = getattr(module, '__loader__', None)
return _find_adapter(_provider_factories, loader)(module)
def _macosx_vers(_cache=[]):
if not _cache:
version = platform.mac_ver()[0]
# fallback for MacPorts
if version == '':
plist = '/System/Library/CoreServices/SystemVersion.plist'
if os.path.exists(plist):
if hasattr(plistlib, 'readPlist'):
plist_content = plistlib.readPlist(plist)
if 'ProductVersion' in plist_content:
version = plist_content['ProductVersion']
return _cache[0]
def _macosx_arch(machine):
return {'PowerPC': 'ppc', 'Power_Macintosh': 'ppc'}.get(machine, machine)
def get_build_platform():
"""Return this platform's string for platform-specific distributions
XXX Currently this is the same as ``distutils.util.get_platform()``, but it
needs some hacks for Linux and Mac OS X.
# Python 2.7 or >=3.2
from sysconfig import get_platform
except ImportError:
from distutils.util import get_platform
plat = get_platform()
if sys.platform == "darwin" and not plat.startswith('macosx-'):
version = _macosx_vers()
machine = os.uname()[4].replace(" ", "_")
return "macosx-%d.%d-%s" % (int(version[0]), int(version[1]),
except ValueError:
# if someone is running a non-Mac darwin system, this will fall
# through to the default implementation
return plat
macosVersionString = re.compile(r"macosx-(\d+)\.(\d+)-(.*)")
darwinVersionString = re.compile(r"darwin-(\d+)\.(\d+)\.(\d+)-(.*)")
# XXX backward compat
get_platform = get_build_platform
def compatible_platforms(provided, required):
"""Can code for the `provided` platform run on the `required` platform?
Returns true if either platform is ``None``, or the platforms are equal.
XXX Needs compatibility checks for Linux and other unixy OSes.
if provided is None or required is None or provided == required:
# easy case
return True
# Mac OS X special cases
reqMac = macosVersionString.match(required)
if reqMac:
provMac = macosVersionString.match(provided)
# is this a Mac package?
if not provMac:
# this is backwards compatibility for packages built before
# setuptools 0.6. All packages built after this point will
# use the new macosx designation.
provDarwin = darwinVersionString.match(provided)
if provDarwin:
dversion = int(
macosversion = "%s.%s" % (,
if dversion == 7 and macosversion >= "10.3" or \
dversion == 8 and macosversion >= "10.4":
return True
# egg isn't macosx or legacy darwin
return False
# are they the same major version and machine type?
if != or \ !=
return False
# is the required OS major update >= the provided one?
if int( > int(
return False
return True
# XXX Linux and other platforms' special cases should go here
return False
def run_script(dist_spec, script_name):
"""Locate distribution `dist_spec` and run its `script_name` script"""
ns = sys._getframe(1).f_globals
name = ns['__name__']
ns['__name__'] = name
require(dist_spec)[0].run_script(script_name, ns)
# backward compatibility
run_main = run_script
def get_distribution(dist):
"""Return a current distribution object for a Requirement or string"""
if isinstance(dist, six.string_types):
dist = Requirement.parse(dist)
if isinstance(dist, Requirement):
dist = get_provider(dist)
if not isinstance(dist, Distribution):
raise TypeError("Expected string, Requirement, or Distribution", dist)
return dist
def load_entry_point(dist, group, name):
"""Return `name` entry point of `group` for `dist` or raise ImportError"""
return get_distribution(dist).load_entry_point(group, name)
def get_entry_map(dist, group=None):
"""Return the entry point map for `group`, or the full entry map"""
return get_distribution(dist).get_entry_map(group)
def get_entry_info(dist, group, name):
"""Return the EntryPoint object for `group`+`name`, or ``None``"""
return get_distribution(dist).get_entry_info(group, name)
class IMetadataProvider:
def has_metadata(name):
"""Does the package's distribution contain the named metadata?"""
def get_metadata(name):
"""The named metadata resource as a string"""
def get_metadata_lines(name):
"""Yield named metadata resource as list of non-blank non-comment lines
Leading and trailing whitespace is stripped from each line, and lines
with ``#`` as the first non-blank character are omitted."""
def metadata_isdir(name):
"""Is the named metadata a directory? (like ``os.path.isdir()``)"""
def metadata_listdir(name):
"""List of metadata names in the directory (like ``os.listdir()``)"""
def run_script(script_name, namespace):
"""Execute the named script in the supplied namespace dictionary"""
class IResourceProvider(IMetadataProvider):
"""An object that provides access to package resources"""
def get_resource_filename(manager, resource_name):
"""Return a true filesystem path for `resource_name`
`manager` must be an ``IResourceManager``"""
def get_resource_stream(manager, resource_name):
"""Return a readable file-like object for `resource_name`
`manager` must be an ``IResourceManager``"""
def get_resource_string(manager, resource_name):
"""Return a string containing the contents of `resource_name`
`manager` must be an ``IResourceManager``"""
def has_resource(resource_name):
"""Does the package contain the named resource?"""
def resource_isdir(resource_name):
"""Is the named resource a directory? (like ``os.path.isdir()``)"""
def resource_listdir(resource_name):
"""List of resource names in the directory (like ``os.listdir()``)"""
class WorkingSet(object):
"""A collection of active distributions on sys.path (or a similar list)"""
def __init__(self, entries=None):
"""Create working set from list of path entries (default=sys.path)"""
self.entries = []
self.entry_keys = {}
self.by_key = {}
self.callbacks = []
if entries is None:
entries = sys.path
for entry in entries:
def _build_master(cls):
Prepare the master working set.
ws = cls()
from __main__ import __requires__
except ImportError:
# The main program does not list any requirements
return ws
# ensure the requirements are met
except VersionConflict:
return cls._build_from_requirements(__requires__)
return ws
def _build_from_requirements(cls, req_spec):
Build a working set from a requirement spec. Rewrites sys.path.
# try it without defaults already on sys.path
# by starting with an empty path
ws = cls([])
reqs = parse_requirements(req_spec)
dists = ws.resolve(reqs, Environment())
for dist in dists:
# add any missing entries from sys.path
for entry in sys.path:
if entry not in ws.entries:
# then copy back to sys.path
sys.path[:] = ws.entries
return ws
def add_entry(self, entry):
"""Add a path item to ``.entries``, finding any distributions on it
``find_distributions(entry, True)`` is used to find distributions
corresponding to the path entry, and they are added. `entry` is
always appended to ``.entries``, even if it is already present.
(This is because ``sys.path`` can contain the same value more than
once, and the ``.entries`` of the ``sys.path`` WorkingSet should always
equal ``sys.path``.)
self.entry_keys.setdefault(entry, [])
for dist in find_distributions(entry, True):
self.add(dist, entry, False)
def __contains__(self, dist):
"""True if `dist` is the active distribution for its project"""
return self.by_key.get(dist.key) == dist
def find(self, req):
"""Find a distribution matching requirement `req`
If there is an active distribution for the requested project, this
returns it as long as it meets the version requirement specified by
`req`. But, if there is an active distribution for the project and it
does *not* meet the `req` requirement, ``VersionConflict`` is raised.
If there is no active distribution for the requested project, ``None``
is returned.
dist = self.by_key.get(req.key)
if dist is not None and dist not in req:
# XXX add more info
raise VersionConflict(dist, req)
return dist
def iter_entry_points(self, group, name=None):
"""Yield entry point objects from `group` matching `name`
If `name` is None, yields all entry points in `group` from all
distributions in the working set, otherwise only ones matching
both `group` and `name` are yielded (in distribution order).
for dist in self:
entries = dist.get_entry_map(group)
if name is None:
for ep in entries.values():
yield ep
elif name in entries:
yield entries[name]
def run_script(self, requires, script_name):
"""Locate distribution for `requires` and run `script_name` script"""
ns = sys._getframe(1).f_globals
name = ns['__name__']
ns['__name__'] = name
self.require(requires)[0].run_script(script_name, ns)
def __iter__(self):
"""Yield distributions for non-duplicate projects in the working set
The yield order is the order in which the items' path entries were
added to the working set.
seen = {}
for item in self.entries:
if item not in self.entry_keys:
# workaround a cache issue
for key in self.entry_keys[item]:
if key not in seen:
seen[key] = 1
yield self.by_key[key]
def add(self, dist, entry=None, insert=True, replace=False):
"""Add `dist` to working set, associated with `entry`
If `entry` is unspecified, it defaults to the ``.location`` of `dist`.
On exit from this routine, `entry` is added to the end of the working
set's ``.entries`` (if it wasn't already present).
`dist` is only added to the working set if it's for a project that
doesn't already have a distribution in the set, unless `replace=True`.
If it's added, any callbacks registered with the ``subscribe()`` method
will be called.
if insert:
dist.insert_on(self.entries, entry, replace=replace)
if entry is None:
entry = dist.location
keys = self.entry_keys.setdefault(entry, [])
keys2 = self.entry_keys.setdefault(dist.location, [])
if not replace and dist.key in self.by_key:
# ignore hidden distros
self.by_key[dist.key] = dist
if dist.key not in keys:
if dist.key not in keys2:
def resolve(self, requirements, env=None, installer=None,
"""List all distributions needed to (recursively) meet `requirements`
`requirements` must be a sequence of ``Requirement`` objects. `env`,
if supplied, should be an ``Environment`` instance. If
not supplied, it defaults to all distributions available within any
entry or distribution in the working set. `installer`, if supplied,
will be invoked with each requirement that cannot be met by an
already-installed distribution; it should return a ``Distribution`` or
Unless `replace_conflicting=True`, raises a VersionConflict exception if
any requirements are found on the path that have the correct name but
the wrong version. Otherwise, if an `installer` is supplied it will be
invoked to obtain the correct version of the requirement and activate
# set up the stack
requirements = list(requirements)[::-1]
# set of processed requirements
processed = {}
# key -> dist
best = {}
to_activate = []
req_extras = _ReqExtras()
# Mapping of requirement to set of distributions that required it;
# useful for reporting info about conflicts.
required_by = collections.defaultdict(set)
while requirements:
# process dependencies breadth-first
req = requirements.pop(0)
if req in processed:
# Ignore cyclic or redundant dependencies
if not req_extras.markers_pass(req):
dist = best.get(req.key)
if dist is None:
# Find the best distribution and add it to the map
dist = self.by_key.get(req.key)
if dist is None or (dist not in req and replace_conflicting):
ws = self
if env is None:
if dist is None:
env = Environment(self.entries)
# Use an empty environment and workingset to avoid
# any further conflicts with the conflicting
# distribution
env = Environment([])
ws = WorkingSet([])
dist = best[req.key] = env.best_match(req, ws, installer)
if dist is None:
requirers = required_by.get(req, None)
raise DistributionNotFound(req, requirers)
if dist not in req:
# Oops, the "best" so far conflicts with a dependency
dependent_req = required_by[req]
raise VersionConflict(dist, req).with_context(dependent_req)
# push the new requirements onto the stack
new_requirements = dist.requires(req.extras)[::-1]
# Register the new requirements needed by req
for new_requirement in new_requirements:
req_extras[new_requirement] = req.extras
processed[req] = True
# return list of distros to activate
return to_activate
def find_plugins(self, plugin_env, full_env=None, installer=None,
"""Find all activatable distributions in `plugin_env`
Example usage::
distributions, errors = working_set.find_plugins(
# add plugins+libs to sys.path
map(working_set.add, distributions)
# display errors
print('Could not load', errors)
The `plugin_env` should be an ``Environment`` instance that contains
only distributions that are in the project's "plugin directory" or
directories. The `full_env`, if supplied, should be an ``Environment``
contains all currently-available distributions. If `full_env` is not
supplied, one is created automatically from the ``WorkingSet`` this
method is called on, which will typically mean that every directory on
``sys.path`` will be scanned for distributions.
`installer` is a standard installer callback as used by the
``resolve()`` method. The `fallback` flag indicates whether we should
attempt to resolve older versions of a plugin if the newest version
cannot be resolved.
This method returns a 2-tuple: (`distributions`, `error_info`), where
`distributions` is a list of the distributions found in `plugin_env`
that were loadable, along with any other distributions that are needed
to resolve their dependencies. `error_info` is a dictionary mapping
unloadable plugin distributions to an exception instance describing the
error that occurred. Usually this will be a ``DistributionNotFound`` or
``VersionConflict`` instance.
plugin_projects = list(plugin_env)
# scan project names in alphabetic order
error_info = {}
distributions = {}
if full_env is None:
env = Environment(self.entries)
env += plugin_env
env = full_env + plugin_env
shadow_set = self.__class__([])
# put all our entries in shadow_set
list(map(shadow_set.add, self))
for project_name in plugin_projects:
for dist in plugin_env[project_name]:
req = [dist.as_requirement()]
resolvees = shadow_set.resolve(req, env, installer)
except ResolutionError as v:
# save error info
error_info[dist] = v
if fallback:
# try the next older version of project
# give up on this project, keep going
list(map(shadow_set.add, resolvees))
# success, no need to try any more versions of this project
distributions = list(distributions)
return distributions, error_info
def require(self, *requirements):
"""Ensure that distributions matching `requirements` are activated
`requirements` must be a string or a (possibly-nested) sequence
thereof, specifying the distributions and versions required. The
return value is a sequence of the distributions that needed to be
activated to fulfill the requirements; all relevant distributions are
included, even if they were already activated in this working set.
needed = self.resolve(parse_requirements(requirements))
for dist in needed:
return needed
def subscribe(self, callback, existing=True):
"""Invoke `callback` for all distributions
If `existing=True` (default),
call on all existing ones, as well.
if callback in self.callbacks:
if not existing:
for dist in self:
def _added_new(self, dist):
for callback in self.callbacks:
def __getstate__(self):
return (
self.entries[:], self.entry_keys.copy(), self.by_key.copy(),
def __setstate__(self, e_k_b_c):
entries, keys, by_key, callbacks = e_k_b_c
self.entries = entries[:]
self.entry_keys = keys.copy()
self.by_key = by_key.copy()
self.callbacks = callbacks[:]
class _ReqExtras(dict):
Map each requirement to the extras that demanded it.
def markers_pass(self, req):
Evaluate markers for req against each extra that
demanded it.
Return False if the req has a marker and fails
evaluation. Otherwise, return True.
extra_evals = (
req.marker.evaluate({'extra': extra})
for extra in self.get(req, ()) + (None,)
return not req.marker or any(extra_evals)
class Environment(object):
"""Searchable snapshot of distributions on a search path"""
def __init__(self, search_path=None, platform=get_supported_platform(),
"""Snapshot distributions available on a search path
Any distributions found on `search_path` are added to the environment.
`search_path` should be a sequence of ``sys.path`` items. If not
supplied, ``sys.path`` is used.
`platform` is an optional string specifying the name of the platform
that platform-specific distributions must be compatible with. If
unspecified, it defaults to the current platform. `python` is an
optional string naming the desired version of Python (e.g. ``'3.3'``);
it defaults to the current version.
You may explicitly set `platform` (and/or `python`) to ``None`` if you
wish to map *all* distributions, not just those compatible with the
running platform or Python version.
self._distmap = {}
self.platform = platform
self.python = python
def can_add(self, dist):
"""Is distribution `dist` acceptable for this environment?
The distribution must match the platform and python version
requirements specified when this environment was created, or False
is returned.
return (self.python is None or dist.py_version is None
or dist.py_version == self.python) \
and compatible_platforms(dist.platform, self.platform)
def remove(self, dist):
"""Remove `dist` from the environment"""
def scan(self, search_path=None):
"""Scan `search_path` for distributions usable in this environment
Any distributions found are added to the environment.
`search_path` should be a sequence of ``sys.path`` items. If not
supplied, ``sys.path`` is used. Only distributions conforming to
the platform/python version defined at initialization are added.
if search_path is None:
search_path = sys.path
for item in search_path:
for dist in find_distributions(item):
def __getitem__(self, project_name):
"""Return a newest-to-oldest list of distributions for `project_name`
Uses case-insensitive `project_name` comparison, assuming all the
project's distributions use their project's name converted to all
lowercase as their key.
distribution_key = project_name.lower()
return self._distmap.get(distribution_key, [])
def add(self, dist):
"""Add `dist` if we ``can_add()`` it and it has not already been added
if self.can_add(dist) and dist.has_version():
dists = self._distmap.setdefault(dist.key, [])
if dist not in dists:
dists.sort(key=operator.attrgetter('hashcmp'), reverse=True)
def best_match(self, req, working_set, installer=None):
"""Find distribution best matching `req` and usable on `working_set`
This calls the ``find(req)`` method of the `working_set` to see if a
suitable distribution is already active. (This may raise
``VersionConflict`` if an unsuitable version of the project is already
active in the specified `working_set`.) If a suitable distribution
isn't active, this method returns the newest distribution in the
environment that meets the ``Requirement`` in `req`. If no suitable
distribution is found, and `installer` is supplied, then the result of
calling the environment's ``obtain(req, installer)`` method will be
dist = working_set.find(req)
if dist is not None:
return dist
for dist in self[req.key]:
if dist in req:
return dist
# try to download/install
return self.obtain(req, installer)
def obtain(self, requirement, installer=None):
"""Obtain a distribution matching `requirement` (e.g. via download)
Obtain a distro that matches requirement (e.g. via download). In the
base ``Environment`` class, this routine just returns
``installer(requirement)``, unless `installer` is None, in which case
None is returned instead. This method is a hook that allows subclasses
to attempt other ways of obtaining a distribution before falling back
to the `installer` argument."""
if installer is not None:
return installer(requirement)
def __iter__(self):
"""Yield the unique project names of the available distributions"""
for key in self._distmap.keys():
if self[key]:
yield key
def __iadd__(self, other):
"""In-place addition of a distribution or environment"""
if isinstance(other, Distribution):
elif isinstance(other, Environment):
for project in other:
for dist in other[project]:
raise TypeError("Can't add %r to environment" % (other,))
return self
def __add__(self, other):
"""Add an environment or distribution to an environment"""
new = self.__class__([], platform=None, python=None)
for env in self, other:
new += env
return new
# XXX backward compatibility
AvailableDistributions = Environment
class ExtractionError(RuntimeError):
"""An error occurred extracting a resource
The following attributes are available from instances of this exception:
The resource manager that raised this exception
The base directory for resource extraction
The exception instance that caused extraction to fail
class ResourceManager:
"""Manage resource extraction and packages"""
extraction_path = None
def __init__(self):
self.cached_files = {}
def resource_exists(self, package_or_requirement, resource_name):
"""Does the named resource exist?"""
return get_provider(package_or_requirement).has_resource(resource_name)
def resource_isdir(self, package_or_requirement, resource_name):
"""Is the named resource an existing directory?"""
return get_provider(package_or_requirement).resource_isdir(
def resource_filename(self, package_or_requirement, resource_name):
"""Return a true filesystem path for specified resource"""
return get_provider(package_or_requirement).get_resource_filename(
self, resource_name
def resource_stream(self, package_or_requirement, resource_name):
"""Return a readable file-like object for specified resource"""
return get_provider(package_or_requirement).get_resource_stream(
self, resource_name
def resource_string(self, package_or_requirement, resource_name):
"""Return specified resource as a string"""
return get_provider(package_or_requirement).get_resource_string(
self, resource_name
def resource_listdir(self, package_or_requirement, resource_name):
"""List the contents of the named resource directory"""
return get_provider(package_or_requirement).resource_listdir(
def extraction_error(self):
"""Give an error message for problems extracting file(s)"""
old_exc = sys.exc_info()[1]
cache_path = self.extraction_path or get_default_cache()
tmpl = textwrap.dedent("""
Can't extract file(s) to egg cache
The following error occurred while trying to extract file(s) to the Python egg
The Python egg cache directory is currently set to:
Perhaps your account does not have write access to this directory? You can
change the cache directory by setting the PYTHON_EGG_CACHE environment
variable to point to an accessible directory.
err = ExtractionError(tmpl.format(**locals()))
err.manager = self
err.cache_path = cache_path
err.original_error = old_exc
raise err
def get_cache_path(self, archive_name, names=()):
"""Return absolute location in cache for `archive_name` and `names`
The parent directory of the resulting path will be created if it does
not already exist. `archive_name` should be the base filename of the
enclosing egg (which may not be the name of the enclosing zipfile!),
including its ".egg" extension. `names`, if provided, should be a
sequence of path name parts "under" the egg's extraction location.
This method should only be called by resource providers that need to
obtain an extraction location, and only for names they intend to
extract, as it tracks the generated names for possible cleanup later.
extract_path = self.extraction_path or get_default_cache()
target_path = os.path.join(extract_path, archive_name + '-tmp', *names)
self.cached_files[target_path] = 1
return target_path
def _warn_unsafe_extraction_path(path):
If the default extraction path is overridden and set to an insecure
location, such as /tmp, it opens up an opportunity for an attacker to
replace an extracted file with an unauthorized payload. Warn the user
if a known insecure location is used.
See Distribute #375 for more details.
if == 'nt' and not path.startswith(os.environ['windir']):
# On Windows, permissions are generally restrictive by default
# and temp directories are not writable by other users, so
# bypass the warning.
mode = os.stat(path).st_mode
if mode & stat.S_IWOTH or mode & stat.S_IWGRP:
msg = ("%s is writable by group/others and vulnerable to attack "
"when "
"used with get_resource_filename. Consider a more secure "
"location (set with .set_extraction_path or the "
"PYTHON_EGG_CACHE environment variable)." % path)
warnings.warn(msg, UserWarning)
def postprocess(self, tempname, filename):
"""Perform any platform-specific postprocessing of `tempname`
This is where Mac header rewrites should be done; other platforms don't
have anything special they should do.
Resource providers should call this method ONLY after successfully
extracting a compressed resource. They must NOT call it on resources
that are already in the filesystem.
`tempname` is the current (temporary) name of the file, and `filename`
is the name it will be renamed to by the caller after this routine
if == 'posix':
# Make the resource executable
mode = ((os.stat(tempname).st_mode) | 0o555) & 0o7777
os.chmod(tempname, mode)
def set_extraction_path(self, path):
"""Set the base path where resources will be extracted to, if needed.
If you do not call this routine before any extractions take place, the
path defaults to the return value of ``get_default_cache()``. (Which
is based on the ``PYTHON_EGG_CACHE`` environment variable, with various
platform-specific fallbacks. See that routine's documentation for more
Resources are extracted to subdirectories of this path based upon
information given by the ``IResourceProvider``. You may set this to a
temporary directory, but then you must call ``cleanup_resources()`` to
delete the extracted files when done. There is no guarantee that
``cleanup_resources()`` will be able to remove all extracted files.
(Note: you may not change the extraction path for a given resource
manager once resources have been extracted, unless you first call
if self.cached_files:
raise ValueError(
"Can't change extraction path, files already extracted"
self.extraction_path = path
def cleanup_resources(self, force=False):
Delete all extracted resource files and directories, returning a list
of the file and directory names that could not be successfully removed.
This function does not have any concurrency protection, so it should
generally only be called when the extraction path is a temporary
directory exclusive to a single process. This method is not
automatically called; you must call it explicitly or register it as an
``atexit`` function if you wish to ensure cleanup of a temporary
directory used for extractions.
def get_default_cache():
Return the ``PYTHON_EGG_CACHE`` environment variable
or a platform-relevant user cache dir for an app
named "Python-Eggs".
return (
or appdirs.user_cache_dir(appname='Python-Eggs')
def safe_name(name):
"""Convert an arbitrary string to a standard distribution name
Any runs of non-alphanumeric/. characters are replaced with a single '-'.
return re.sub('[^A-Za-z0-9.]+', '-', name)
def safe_version(version):
Convert an arbitrary string to a standard version string
# normalize the version
return str(packaging.version.Version(version))
except packaging.version.InvalidVersion:
version = version.replace(' ', '.')
return re.sub('[^A-Za-z0-9.]+', '-', version)
def safe_extra(extra):
"""Convert an arbitrary string to a standard 'extra' name
Any runs of non-alphanumeric characters are replaced with a single '_',
and the result is always lowercased.
return re.sub('[^A-Za-z0-9.-]+', '_', extra).lower()
def to_filename(name):
"""Convert a project or version name to its filename-escaped form
Any '-' characters are currently replaced with '_'.
return name.replace('-', '_')
def invalid_marker(text):
Validate text as a PEP 508 environment marker; return an exception
if invalid or False otherwise.
except SyntaxError as e:
e.filename = None
e.lineno = None
return e
return False
def evaluate_marker(text, extra=None):
Evaluate a PEP 508 environment marker.
Return a boolean indicating the marker result in this environment.
Raise SyntaxError if marker is invalid.
This implementation uses the 'pyparsing' module.
marker = packaging.markers.Marker(text)
return marker.evaluate()
except packaging.markers.InvalidMarker as e:
raise SyntaxError(e)
class NullProvider:
"""Try to implement resources and metadata for arbitrary PEP 302 loaders"""
egg_name = None
egg_info = None
loader = None
def __init__(self, module):
self.loader = getattr(module, '__loader__', None)
self.module_path = os.path.dirname(getattr(module, '__file__', ''))
def get_resource_filename(self, manager, resource_name):
return self._fn(self.module_path, resource_name)
def get_resource_stream(self, manager, resource_name):
return io.BytesIO(self.get_resource_string(manager, resource_name))
def get_resource_string(self, manager, resource_name):
return self._get(self._fn(self.module_path, resource_name))
def has_resource(self, resource_name):
return self._has(self._fn(self.module_path, resource_name))
def has_metadata(self, name):
return self.egg_info and self._has(self._fn(self.egg_info, name))
def get_metadata(self, name):
if not self.egg_info:
return ""
value = self._get(self._fn(self.egg_info, name))
return value.decode('utf-8') if six.PY3 else value
def get_metadata_lines(self, name):
return yield_lines(self.get_metadata(name))
def resource_isdir(self, resource_name):
return self._isdir(self._fn(self.module_path, resource_name))
def metadata_isdir(self, name):
return self.egg_info and self._isdir(self._fn(self.egg_info, name))
def resource_listdir(self, resource_name):
return self._listdir(self._fn(self.module_path, resource_name))
def metadata_listdir(self, name):
if self.egg_info:
return self._listdir(self._fn(self.egg_info, name))
return []
def run_script(self, script_name, namespace):
script = 'scripts/' + script_name
if not self.has_metadata(script):
raise ResolutionError("No script named %r" % script_name)
script_text = self.get_metadata(script).replace('\r\n', '\n')
script_text = script_text.replace('\r', '\n')
script_filename = self._fn(self.egg_info, script)
namespace['__file__'] = script_filename
if os.path.exists(script_filename):
source = open(script_filename).read()
code = compile(source, script_filename, 'exec')
exec(code, namespace, namespace)
from linecache import cache
cache[script_filename] = (
len(script_text), 0, script_text.split('\n'), script_filename
script_code = compile(script_text, script_filename, 'exec')
exec(script_code, namespace, namespace)
def _has(self, path):
raise NotImplementedError(
"Can't perform this operation for unregistered loader type"
def _isdir(self, path):
raise NotImplementedError(
"Can't perform this operation for unregistered loader type"
def _listdir(self, path):
raise NotImplementedError(
"Can't perform this operation for unregistered loader type"
def _fn(self, base, resource_name):
if resource_name:
return os.path.join(base, *resource_name.split('/'))
return base
def _get(self, path):
if hasattr(self.loader, 'get_data'):
return self.loader.get_data(path)
raise NotImplementedError(
"Can't perform this operation for loaders without 'get_data()'"
register_loader_type(object, NullProvider)
class EggProvider(NullProvider):
"""Provider based on a virtual filesystem"""
def __init__(self, module):
NullProvider.__init__(self, module)
def _setup_prefix(self):
# we assume here that our metadata may be nested inside a "basket"
# of multiple eggs; that's why we use module_path instead of .archive
path = self.module_path
old = None
while path != old:
if _is_unpacked_egg(path):
self.egg_name = os.path.basename(path)
self.egg_info = os.path.join(path, 'EGG-INFO')
self.egg_root = path
old = path
path, base = os.path.split(path)
class DefaultProvider(EggProvider):
"""Provides access to package resources in the filesystem"""
def _has(self, path):
return os.path.exists(path)
def _isdir(self, path):
return os.path.isdir(path)
def _listdir(self, path):
return os.listdir(path)
def get_resource_stream(self, manager, resource_name):
return open(self._fn(self.module_path, resource_name), 'rb')
def _get(self, path):
with open(path, 'rb') as stream:
def _register(cls):
loader_cls = getattr(importlib_machinery, 'SourceFileLoader',
register_loader_type(loader_cls, cls)
class EmptyProvider(NullProvider):
"""Provider that returns nothing for all requests"""
_isdir = _has = lambda self, path: False
_get = lambda self, path: ''
_listdir = lambda self, path: []
module_path = None
def __init__(self):
empty_provider = EmptyProvider()
class ZipManifests(dict):
zip manifest builder
def build(cls, path):
Build a dictionary similar to the zipimport directory
caches, except instead of tuples, store ZipInfo objects.
Use a platform-specific path separator (os.sep) for the path keys
for compatibility with pypy on Windows.
with ContextualZipFile(path) as zfile:
items = (
name.replace('/', os.sep),
for name in zfile.namelist()
return dict(items)
load = build
class MemoizedZipManifests(ZipManifests):
Memoized zipfile manifests.
manifest_mod = collections.namedtuple('manifest_mod', 'manifest mtime')
def load(self, path):
Load a manifest at path or return a suitable manifest already loaded.
path = os.path.normpath(path)
mtime = os.stat(path).st_mtime
if path not in self or self[path].mtime != mtime:
manifest =
self[path] = self.manifest_mod(manifest, mtime)
return self[path].manifest
class ContextualZipFile(zipfile.ZipFile):
Supplement ZipFile class to support context manager for Python 2.6
def __enter__(self):
return self
def __exit__(self, type, value, traceback):
def __new__(cls, *args, **kwargs):
Construct a ZipFile or ContextualZipFile as appropriate
if hasattr(zipfile.ZipFile, '__exit__'):
return zipfile.ZipFile(*args, **kwargs)
return super(ContextualZipFile, cls).__new__(cls)
class ZipProvider(EggProvider):
"""Resource support for zips and eggs"""
eagers = None
_zip_manifests = MemoizedZipManifests()
def __init__(self, module):
EggProvider.__init__(self, module)
self.zip_pre = self.loader.archive + os.sep
def _zipinfo_name(self, fspath):
# Convert a virtual filename (full path to file) into a zipfile subpath
# usable with the zipimport directory cache for our target archive
if fspath.startswith(self.zip_pre):
return fspath[len(self.zip_pre):]
raise AssertionError(
"%s is not a subpath of %s" % (fspath, self.zip_pre)
def _parts(self, zip_path):
# Convert a zipfile subpath into an egg-relative path part list.
# pseudo-fs path
fspath = self.zip_pre + zip_path
if fspath.startswith(self.egg_root + os.sep):
return fspath[len(self.egg_root) + 1:].split(os.sep)
raise AssertionError(
"%s is not a subpath of %s" % (fspath, self.egg_root)
def zipinfo(self):
return self._zip_manifests.load(self.loader.archive)
def get_resource_filename(self, manager, resource_name):
if not self.egg_name:
raise NotImplementedError(
"resource_filename() only supported for .egg, not .zip"
# no need to lock for extraction, since we use temp names
zip_path = self._resource_to_zip(resource_name)
eagers = self._get_eager_resources()
if '/'.join(self._parts(zip_path)) in eagers:
for name in eagers:
self._extract_resource(manager, self._eager_to_zip(name))
return self._extract_resource(manager, zip_path)
def _get_date_and_size(zip_stat):
size = zip_stat.file_size
# ymdhms+wday, yday, dst
date_time = zip_stat.date_time + (0, 0, -1)
# 1980 offset already done
timestamp = time.mktime(date_time)
return timestamp, size
def _extract_resource(self, manager, zip_path):
if zip_path in self._index():
for name in self._index()[zip_path]:
last = self._extract_resource(
manager, os.path.join(zip_path, name)
# return the extracted directory name
return os.path.dirname(last)
timestamp, size = self._get_date_and_size(self.zipinfo[zip_path])
raise IOError('"os.rename" and "os.unlink" are not supported '
'on this platform')
real_path = manager.get_cache_path(
self.egg_name, self._parts(zip_path)
if self._is_current(real_path, zip_path):
return real_path
outf, tmpnam = _mkstemp(".$extract", dir=os.path.dirname(real_path))
os.write(outf, self.loader.get_data(zip_path))
utime(tmpnam, (timestamp, timestamp))
manager.postprocess(tmpnam, real_path)
rename(tmpnam, real_path)
except os.error:
if os.path.isfile(real_path):
if self._is_current(real_path, zip_path):
# the file became current since it was checked above,
# so proceed.
return real_path
# Windows, del old file and retry
elif == 'nt':
rename(tmpnam, real_path)
return real_path
except os.error:
# report a user-friendly error
return real_path
def _is_current(self, file_path, zip_path):
Return True if the file_path is current for this zip_path
timestamp, size = self._get_date_and_size(self.zipinfo[zip_path])
if not os.path.isfile(file_path):
return False
stat = os.stat(file_path)
if stat.st_size != size or stat.st_mtime != timestamp:
return False
# check that the contents match
zip_contents = self.loader.get_data(zip_path)
with open(file_path, 'rb') as f:
file_contents =
return zip_contents == file_contents
def _get_eager_resources(self):
if self.eagers is None:
eagers = []
for name in ('native_libs.txt', 'eager_resources.txt'):
if self.has_metadata(name):
self.eagers = eagers
return self.eagers
def _index(self):
return self._dirindex
except AttributeError:
ind = {}
for path in self.zipinfo:
parts = path.split(os.sep)
while parts:
parent = os.sep.join(parts[:-1])
if parent in ind:
ind[parent] = [parts.pop()]
self._dirindex = ind
return ind
def _has(self, fspath):
zip_path = self._zipinfo_name(fspath)
return zip_path in self.zipinfo or zip_path in self._index()
def _isdir(self, fspath):
return self._zipinfo_name(fspath) in self._index()
def _listdir(self, fspath):
return list(self._index().get(self._zipinfo_name(fspath), ()))
def _eager_to_zip(self, resource_name):
return self._zipinfo_name(self._fn(self.egg_root, resource_name))
def _resource_to_zip(self, resource_name):
return self._zipinfo_name(self._fn(self.module_path, resource_name))
register_loader_type(zipimport.zipimporter, ZipProvider)
class FileMetadata(EmptyProvider):
"""Metadata handler for standalone PKG-INFO files
metadata = FileMetadata("/path/to/PKG-INFO")
This provider rejects all data and metadata requests except for PKG-INFO,
which is treated as existing, and will be the contents of the file at
the provided location.
def __init__(self, path):
self.path = path
def has_metadata(self, name):
return name == 'PKG-INFO' and os.path.isfile(self.path)
def get_metadata(self, name):
if name != 'PKG-INFO':
raise KeyError("No metadata except PKG-INFO is available")
with, encoding='utf-8', errors="replace") as f:
metadata =
return metadata
def _warn_on_replacement(self, metadata):
# Python 2.6 and 3.2 compat for: replacement_char = '�'
replacement_char = b'\xef\xbf\xbd'.decode('utf-8')
if replacement_char in metadata:
tmpl = "{self.path} could not be properly decoded in UTF-8"
msg = tmpl.format(**locals())
def get_metadata_lines(self, name):
return yield_lines(self.get_metadata(name))
class PathMetadata(DefaultProvider):
"""Metadata provider for egg directories
# Development eggs:
egg_info = "/path/to/PackageName.egg-info"
base_dir = os.path.dirname(egg_info)
metadata = PathMetadata(base_dir, egg_info)
dist_name = os.path.splitext(os.path.basename(egg_info))[0]
dist = Distribution(basedir, project_name=dist_name, metadata=metadata)
# Unpacked egg directories:
egg_path = "/path/to/PackageName-ver-pyver-etc.egg"
metadata = PathMetadata(egg_path, os.path.join(egg_path,'EGG-INFO'))
dist = Distribution.from_filename(egg_path, metadata=metadata)
def __init__(self, path, egg_info):
self.module_path = path
self.egg_info = egg_info
class EggMetadata(ZipProvider):
"""Metadata provider for .egg files"""
def __init__(self, importer):
"""Create a metadata provider from a zipimporter"""
self.zip_pre = importer.archive + os.sep
self.loader = importer
if importer.prefix:
self.module_path = os.path.join(importer.archive, importer.prefix)
self.module_path = importer.archive
_declare_state('dict', _distribution_finders={})
def register_finder(importer_type, distribution_finder):
"""Register `distribution_finder` to find distributions in sys.path items
`importer_type` is the type or class of a PEP 302 "Importer" (sys.path item
handler), and `distribution_finder` is a callable that, passed a path
item and the importer instance, yields ``Distribution`` instances found on
that path item. See ``pkg_resources.find_on_path`` for an example."""
_distribution_finders[importer_type] = distribution_finder
def find_distributions(path_item, only=False):
"""Yield distributions accessible via `path_item`"""
importer = get_importer(path_item)
finder = _find_adapter(_distribution_finders, importer)
return finder(importer, path_item, only)
def find_eggs_in_zip(importer, path_item, only=False):
Find eggs in zip files; possibly multiple nested eggs.
if importer.archive.endswith('.whl'):
# wheels are not supported with this finder
# they don't have PKG-INFO metadata, and won't ever contain eggs
metadata = EggMetadata(importer)
if metadata.has_metadata('PKG-INFO'):
yield Distribution.from_filename(path_item, metadata=metadata)
if only:
# don't yield nested distros
for subitem in metadata.resource_listdir('/'):
if _is_unpacked_egg(subitem):
subpath = os.path.join(path_item, subitem)
for dist in find_eggs_in_zip(zipimport.zipimporter(subpath), subpath):
yield dist
register_finder(zipimport.zipimporter, find_eggs_in_zip)
def find_nothing(importer, path_item, only=False):
return ()
register_finder(object, find_nothing)
def _by_version_descending(names):
Given a list of filenames, return them in descending order
by version number.
>>> names = 'bar', 'foo', 'Python-2.7.10.egg', 'Python-2.7.2.egg'
>>> _by_version_descending(names)
['Python-2.7.10.egg', 'Python-2.7.2.egg', 'foo', 'bar']
>>> names = 'Setuptools-1.2.3b1.egg', 'Setuptools-1.2.3.egg'
>>> _by_version_descending(names)
['Setuptools-1.2.3.egg', 'Setuptools-1.2.3b1.egg']
>>> names = 'Setuptools-1.2.3b1.egg', 'Setuptools-1.2.3.post1.egg'
>>> _by_version_descending(names)
['Setuptools-1.2.3.post1.egg', 'Setuptools-1.2.3b1.egg']
def _by_version(name):
Parse each component of the filename
name, ext = os.path.splitext(name)
parts = itertools.chain(name.split('-'), [ext])
return [packaging.version.parse(part) for part in parts]
return sorted(names, key=_by_version, reverse=True)
def find_on_path(importer, path_item, only=False):
"""Yield distributions accessible on a sys.path directory"""
path_item = _normalize_cached(path_item)
if os.path.isdir(path_item) and os.access(path_item, os.R_OK):
if _is_unpacked_egg(path_item):
yield Distribution.from_filename(
path_item, metadata=PathMetadata(
path_item, os.path.join(path_item, 'EGG-INFO')
# scan for .egg and .egg-info in directory
path_item_entries = _by_version_descending(os.listdir(path_item))
for entry in path_item_entries:
lower = entry.lower()
if lower.endswith('.egg-info') or lower.endswith('.dist-info'):
fullpath = os.path.join(path_item, entry)
if os.path.isdir(fullpath):
# egg-info directory, allow getting metadata
if len(os.listdir(fullpath)) == 0:
# Empty egg directory, skip.
metadata = PathMetadata(path_item, fullpath)
metadata = FileMetadata(fullpath)
yield Distribution.from_location(
path_item, entry, metadata, precedence=DEVELOP_DIST
elif not only and _is_unpacked_egg(entry):
dists = find_distributions(os.path.join(path_item, entry))
for dist in dists:
yield dist
elif not only and lower.endswith('.egg-link'):
with open(os.path.join(path_item, entry)) as entry_file:
entry_lines = entry_file.readlines()
for line in entry_lines:
if not line.strip():
path = os.path.join(path_item, line.rstrip())
dists = find_distributions(path)
for item in dists:
yield item
register_finder(pkgutil.ImpImporter, find_on_path)
if hasattr(importlib_machinery, 'FileFinder'):
register_finder(importlib_machinery.FileFinder, find_on_path)
_declare_state('dict', _namespace_handlers={})
_declare_state('dict', _namespace_packages={})
def register_namespace_handler(importer_type, namespace_handler):
"""Register `namespace_handler` to declare namespace packages
`importer_type` is the type or class of a PEP 302 "Importer" (sys.path item
handler), and `namespace_handler` is a callable like this::
def namespace_handler(importer, path_entry, moduleName, module):
# return a path_entry to use for child packages
Namespace handlers are only called if the importer object has already
agreed that it can handle the relevant path item, and they should only
return a subpath if the module __path__ does not already contain an
equivalent subpath. For an example namespace handler, see
_namespace_handlers[importer_type] = namespace_handler
def _handle_ns(packageName, path_item):
"""Ensure that named package includes a subpath of path_item (if needed)"""
importer = get_importer(path_item)
if importer is None:
return None
loader = importer.find_module(packageName)
if loader is None:
return None
module = sys.modules.get(packageName)
if module is None:
module = sys.modules[packageName] = types.ModuleType(packageName)
module.__path__ = []
elif not hasattr(module, '__path__'):
raise TypeError("Not a package:", packageName)
handler = _find_adapter(_namespace_handlers, importer)
subpath = handler(importer, path_item, packageName, module)
if subpath is not None:
path = module.__path__
_rebuild_mod_path(path, packageName, module)
return subpath
def _rebuild_mod_path(orig_path, package_name, module):
Rebuild module.__path__ ensuring that all entries are ordered
corresponding to their sys.path order
sys_path = [_normalize_cached(p) for p in sys.path]
def safe_sys_path_index(entry):
Workaround for #520 and #513.
return sys_path.index(entry)
except ValueError:
return float('inf')
def position_in_sys_path(path):
Return the ordinal of the path based on its position in sys.path
path_parts = path.split(os.sep)
module_parts = package_name.count('.') + 1
parts = path_parts[:-module_parts]
return safe_sys_path_index(_normalize_cached(os.sep.join(parts)))
module.__path__[:] = [_normalize_cached(p) for p in orig_path]
def declare_namespace(packageName):
"""Declare that package 'packageName' is a namespace package"""
if packageName in _namespace_packages:
path, parent = sys.path, None
if '.' in packageName:
parent = '.'.join(packageName.split('.')[:-1])
if parent not in _namespace_packages:
path = sys.modules[parent].__path__
except AttributeError:
raise TypeError("Not a package:", parent)
# Track what packages are namespaces, so when new path items are added,
# they can be updated
_namespace_packages.setdefault(parent, []).append(packageName)
_namespace_packages.setdefault(packageName, [])
for path_item in path:
# Ensure all the parent's path items are reflected in the child,
# if they apply
_handle_ns(packageName, path_item)
def fixup_namespace_packages(path_item, parent=None):
"""Ensure that previously-declared namespace packages include path_item"""
for package in _namespace_packages.get(parent, ()):
subpath = _handle_ns(package, path_item)
if subpath:
fixup_namespace_packages(subpath, package)
def file_ns_handler(importer, path_item, packageName, module):
"""Compute an ns-package subpath for a filesystem or zipfile importer"""
subpath = os.path.join(path_item, packageName.split('.')[-1])
normalized = _normalize_cached(subpath)
for item in module.__path__:
if _normalize_cached(item) == normalized:
# Only return the path if it's not already there
return subpath
register_namespace_handler(pkgutil.ImpImporter, file_ns_handler)
register_namespace_handler(zipimport.zipimporter, file_ns_handler)
if hasattr(importlib_machinery, 'FileFinder'):
register_namespace_handler(importlib_machinery.FileFinder, file_ns_handler)
def null_ns_handler(importer, path_item, packageName, module):
return None
register_namespace_handler(object, null_ns_handler)
def normalize_path(filename):
"""Normalize a file/dir name for comparison purposes"""
return os.path.normcase(os.path.realpath(filename))
def _normalize_cached(filename, _cache={}):
return _cache[filename]
except KeyError:
_cache[filename] = result = normalize_path(filename)
return result
def _is_unpacked_egg(path):
Determine if given path appears to be an unpacked egg.
return (
def _set_parent_ns(packageName):
parts = packageName.split('.')
name = parts.pop()
if parts:
parent = '.'.join(parts)
setattr(sys.modules[parent], name, sys.modules[packageName])
def yield_lines(strs):
"""Yield non-empty/non-comment lines of a string or sequence"""
if isinstance(strs, six.string_types):
for s in strs.splitlines():
s = s.strip()
# skip blank lines/comments
if s and not s.startswith('#'):
yield s
for ss in strs:
for s in yield_lines(ss):
yield s
MODULE = re.compile(r"\w+(\.\w+)*$").match
EGG_NAME = re.compile(
(?P<name>[^-]+) (
-(?P<ver>[^-]+) (
-py(?P<pyver>[^-]+) (
class EntryPoint(object):
"""Object representing an advertised importable object"""
def __init__(self, name, module_name, attrs=(), extras=(), dist=None):
if not MODULE(module_name):
raise ValueError("Invalid module name", module_name) = name
self.module_name = module_name
self.attrs = tuple(attrs)
self.extras = Requirement.parse(("x[%s]" % ','.join(extras))).extras
self.dist = dist
def __str__(self):
s = "%s = %s" % (, self.module_name)
if self.attrs:
s += ':' + '.'.join(self.attrs)
if self.extras:
s += ' [%s]' % ','.join(self.extras)
return s
def __repr__(self):
return "EntryPoint.parse(%r)" % str(self)
def load(self, require=True, *args, **kwargs):
Require packages for this EntryPoint, then resolve it.
if not require or args or kwargs:
"Parameters to load are deprecated. Call .resolve and "
".require separately.",
if require:
self.require(*args, **kwargs)
return self.resolve()
def resolve(self):
Resolve the entry point from its module and attrs.
module = __import__(self.module_name, fromlist=['__name__'], level=0)
return functools.reduce(getattr, self.attrs, module)
except AttributeError as exc:
raise ImportError(str(exc))
def require(self, env=None, installer=None):
if self.extras and not self.dist:
raise UnknownExtra("Can't require() without a distribution", self)
reqs = self.dist.requires(self.extras)
items = working_set.resolve(reqs, env, installer)
list(map(working_set.add, items))
pattern = re.compile(
def parse(cls, src, dist=None):
"""Parse a single entry point from string `src`
Entry point syntax follows the form::
name = some.module:some.attr [extra1, extra2]
The entry name and module name are required, but the ``:attrs`` and
``[extras]`` parts are optional
m = cls.pattern.match(src)
if not m:
msg = "EntryPoint must be in 'name=module:attrs [extras]' format"
raise ValueError(msg, src)
res = m.groupdict()
extras = cls._parse_extras(res['extras'])
attrs = res['attr'].split('.') if res['attr'] else ()
return cls(res['name'], res['module'], attrs, extras, dist)
def _parse_extras(cls, extras_spec):
if not extras_spec:
return ()
req = Requirement.parse('x' + extras_spec)
if req.specs:
raise ValueError()
return req.extras
def parse_group(cls, group, lines, dist=None):
"""Parse an entry point group"""
if not MODULE(group):
raise ValueError("Invalid group name", group)
this = {}
for line in yield_lines(lines):
ep = cls.parse(line, dist)
if in this:
raise ValueError("Duplicate entry point", group,
this[] = ep
return this
def parse_map(cls, data, dist=None):
"""Parse a map of entry point groups"""
if isinstance(data, dict):
data = data.items()
data = split_sections(data)
maps = {}
for group, lines in data:
if group is None:
if not lines:
raise ValueError("Entry points must be listed in groups")
group = group.strip()
if group in maps:
raise ValueError("Duplicate group name", group)
maps[group] = cls.parse_group(group, lines, dist)
return maps
def _remove_md5_fragment(location):
if not location:
return ''
parsed = urllib.parse.urlparse(location)
if parsed[-1].startswith('md5='):
return urllib.parse.urlunparse(parsed[:-1] + ('',))
return location
def _version_from_file(lines):
Given an iterable of lines from a Metadata file, return
the value of the Version field, if present, or None otherwise.
is_version_line = lambda line: line.lower().startswith('version:')
version_lines = filter(is_version_line, lines)
line = next(iter(version_lines), '')
_, _, value = line.partition(':')
return safe_version(value.strip()) or None
class Distribution(object):
"""Wrap an actual or potential sys.path entry w/metadata"""
def __init__(self, location=None, metadata=None, project_name=None,
version=None, py_version=PY_MAJOR, platform=None,
self.project_name = safe_name(project_name or 'Unknown')
if version is not None:
self._version = safe_version(version)
self.py_version = py_version
self.platform = platform
self.location = location
self.precedence = precedence
self._provider = metadata or empty_provider
def from_location(cls, location, basename, metadata=None, **kw):
project_name, version, py_version, platform = [None] * 4
basename, ext = os.path.splitext(basename)
if ext.lower() in _distributionImpl:
cls = _distributionImpl[ext.lower()]
match = EGG_NAME(basename)
if match:
project_name, version, py_version, platform =
'name', 'ver', 'pyver', 'plat'
return cls(
location, metadata, project_name=project_name, version=version,
py_version=py_version, platform=platform, **kw
def _reload_version(self):
return self
def hashcmp(self):
return (
self.py_version or '',
self.platform or '',
def __hash__(self):
return hash(self.hashcmp)
def __lt__(self, other):
return self.hashcmp < other.hashcmp
def __le__(self, other):
return self.hashcmp <= other.hashcmp
def __gt__(self, other):
return self.hashcmp > other.hashcmp
def __ge__(self, other):
return self.hashcmp >= other.hashcmp
def __eq__(self, other):
if not isinstance(other, self.__class__):
# It's not a Distribution, so they are not equal
return False
return self.hashcmp == other.hashcmp
def __ne__(self, other):
return not self == other
# These properties have to be lazy so that we don't have to load any
# metadata until/unless it's actually needed. (i.e., some distributions
# may not know their name or version without loading PKG-INFO)
def key(self):
return self._key
except AttributeError:
self._key = key = self.project_name.lower()
return key
def parsed_version(self):
if not hasattr(self, "_parsed_version"):
self._parsed_version = parse_version(self.version)
return self._parsed_version
def _warn_legacy_version(self):
LV = packaging.version.LegacyVersion
is_legacy = isinstance(self._parsed_version, LV)
if not is_legacy:
# While an empty version is technically a legacy version and
# is not a valid PEP 440 version, it's also unlikely to
# actually come from someone and instead it is more likely that
# it comes from setuptools attempting to parse a filename and
# including it in the list. So for that we'll gate this warning
# on if the version is anything at all or not.
if not self.version:
tmpl = textwrap.dedent("""
'{project_name} ({version})' is being parsed as a legacy,
non PEP 440,
version. You may find odd behavior and sort order.
In particular it will be sorted as less than 0.0. It
is recommended to migrate to PEP 440 compatible
""").strip().replace('\n', ' ')
warnings.warn(tmpl.format(**vars(self)), PEP440Warning)
def version(self):
return self._version
except AttributeError:
version = _version_from_file(self._get_metadata(self.PKG_INFO))
if version is None:
tmpl = "Missing 'Version:' header and/or %s file"
raise ValueError(tmpl % self.PKG_INFO, self)
return version
def _dep_map(self):
return self.__dep_map
except AttributeError:
dm = self.__dep_map = {None: []}
for name in 'requires.txt', 'depends.txt':
for extra, reqs in split_sections(self._get_metadata(name)):
if extra:
if ':' in extra:
extra, marker = extra.split(':', 1)
if invalid_marker(marker):
# XXX warn
reqs = []
elif not evaluate_marker(marker):
reqs = []
extra = safe_extra(extra) or None
dm.setdefault(extra, []).extend(parse_requirements(reqs))
return dm
def requires(self, extras=()):
"""List of Requirements needed for this distro if `extras` are used"""
dm = self._dep_map
deps = []
deps.extend(dm.get(None, ()))
for ext in extras:
except KeyError:
raise UnknownExtra(
"%s has no such extra feature %r" % (self, ext)
return deps
def _get_metadata(self, name):
if self.has_metadata(name):
for line in self.get_metadata_lines(name):
yield line
def activate(self, path=None, replace=False):
"""Ensure distribution is importable on `path` (default=sys.path)"""
if path is None:
path = sys.path
self.insert_on(path, replace=replace)
if path is sys.path:
for pkg in self._get_metadata('namespace_packages.txt'):
if pkg in sys.modules:
def egg_name(self):
"""Return what this distribution's standard .egg filename should be"""
filename = "%s-%s-py%s" % (
to_filename(self.project_name), to_filename(self.version),
self.py_version or PY_MAJOR
if self.platform:
filename += '-' + self.platform
return filename
def __repr__(self):
if self.location:
return "%s (%s)" % (self, self.location)
return str(self)
def __str__(self):
version = getattr(self, 'version', None)
except ValueError:
version = None
version = version or "[unknown version]"
return "%s %s" % (self.project_name, version)
def __getattr__(self, attr):
"""Delegate all unrecognized public attributes to .metadata provider"""
if attr.startswith('_'):
raise AttributeError(attr)
return getattr(self._provider, attr)
def from_filename(cls, filename, metadata=None, **kw):
return cls.from_location(
_normalize_cached(filename), os.path.basename(filename), metadata,
def as_requirement(self):
"""Return a ``Requirement`` that matches this distribution exactly"""
if isinstance(self.parsed_version, packaging.version.Version):
spec = "%s==%s" % (self.project_name, self.parsed_version)
spec = "%s===%s" % (self.project_name, self.parsed_version)
return Requirement.parse(spec)
def load_entry_point(self, group, name):
"""Return the `name` entry point of `group` or raise ImportError"""
ep = self.get_entry_info(group, name)
if ep is None:
raise ImportError("Entry point %r not found" % ((group, name),))
return ep.load()
def get_entry_map(self, group=None):
"""Return the entry point map for `group`, or the full entry map"""
ep_map = self._ep_map
except AttributeError:
ep_map = self._ep_map = EntryPoint.parse_map(
self._get_metadata('entry_points.txt'), self
if group is not None:
return ep_map.get(group, {})
return ep_map
def get_entry_info(self, group, name):
"""Return the EntryPoint object for `group`+`name`, or ``None``"""
return self.get_entry_map(group).get(name)
def insert_on(self, path, loc=None, replace=False):
"""Ensure self.location is on path
If replace=False (default):
- If location is already in path anywhere, do nothing.
- Else:
- If it's an egg and its parent directory is on path,
insert just ahead of the parent.
- Else: add to the end of path.
If replace=True:
- If location is already on path anywhere (not eggs)
or higher priority than its parent (eggs)
do nothing.
- Else:
- If it's an egg and its parent directory is on path,
insert just ahead of the parent,
removing any lower-priority entries.
- Else: add it to the front of path.
loc = loc or self.location
if not loc:
nloc = _normalize_cached(loc)
bdir = os.path.dirname(nloc)
npath = [(p and _normalize_cached(p) or p) for p in path]
for p, item in enumerate(npath):
if item == nloc:
if replace:
# don't modify path (even removing duplicates) if found and not replace
elif item == bdir and self.precedence == EGG_DIST:
# if it's an .egg, give it precedence over its directory
# UNLESS it's already been added to sys.path and replace=False
if (not replace) and nloc in npath[p:]:
if path is sys.path:
path.insert(p, loc)
npath.insert(p, nloc)
if path is sys.path:
if replace:
path.insert(0, loc)
# p is the spot where we found or inserted loc; now remove duplicates
while True:
np = npath.index(nloc, p + 1)
except ValueError:
del npath[np], path[np]
# ha!
p = np
def check_version_conflict(self):
if self.key == 'setuptools':
# ignore the inevitable setuptools self-conflicts :(
nsp = dict.fromkeys(self._get_metadata('namespace_packages.txt'))
loc = normalize_path(self.location)
for modname in self._get_metadata('top_level.txt'):
if (modname not in sys.modules or modname in nsp
or modname in _namespace_packages):
if modname in ('pkg_resources', 'setuptools', 'site'):
fn = getattr(sys.modules[modname], '__file__', None)
if fn and (normalize_path(fn).startswith(loc) or
"Module %s was already imported from %s, but %s is being added"
" to sys.path" % (modname, fn, self.location),
def has_version(self):
except ValueError:
issue_warning("Unbuilt egg for " + repr(self))
return False
return True
def clone(self, **kw):
"""Copy this distribution, substituting in any changed keyword args"""
names = 'project_name version py_version platform location precedence'
for attr in names.split():
kw.setdefault(attr, getattr(self, attr, None))
kw.setdefault('metadata', self._provider)
return self.__class__(**kw)
def extras(self):
return [dep for dep in self._dep_map if dep]
class EggInfoDistribution(Distribution):
def _reload_version(self):
Packages installed by distutils (e.g. numpy or scipy),
which uses an old safe_version, and so
their version numbers can get mangled when
converted to filenames (e.g., 1.11.0.dev0+2329eae to
1.11.0.dev0_2329eae). These distributions will not be
parsed properly
downstream by Distribution and safe_version, so
take an extra step and try to get the version number from
the metadata file itself instead of the filename.
md_version = _version_from_file(self._get_metadata(self.PKG_INFO))
if md_version:
self._version = md_version
return self
class DistInfoDistribution(Distribution):
"""Wrap an actual or potential sys.path entry w/metadata, .dist-info style"""
EQEQ = re.compile(r"([\(,])\s*(\d.*?)\s*([,\)])")
def _parsed_pkg_info(self):
"""Parse and cache metadata"""
return self._pkg_info
except AttributeError:
metadata = self.get_metadata(self.PKG_INFO)
self._pkg_info = email.parser.Parser().parsestr(metadata)
return self._pkg_info
def _dep_map(self):
return self.__dep_map
except AttributeError:
self.__dep_map = self._compute_dependencies()
return self.__dep_map
def _compute_dependencies(self):
"""Recompute this distribution's dependencies."""
dm = self.__dep_map = {None: []}
reqs = []
# Including any condition expressions
for req in self._parsed_pkg_info.get_all('Requires-Dist') or []:
def reqs_for_extra(extra):
for req in reqs:
if not req.marker or req.marker.evaluate({'extra': extra}):
yield req
common = frozenset(reqs_for_extra(None))
for extra in self._parsed_pkg_info.get_all('Provides-Extra') or []:
s_extra = safe_extra(extra.strip())
dm[s_extra] = list(frozenset(reqs_for_extra(extra)) - common)
return dm
_distributionImpl = {
'.egg': Distribution,
'.egg-info': EggInfoDistribution,
'.dist-info': DistInfoDistribution,
def issue_warning(*args, **kw):
level = 1
g = globals()
# find the first stack frame that is *not* code in
# the pkg_resources module, to use for the warning
while sys._getframe(level).f_globals is g:
level += 1
except ValueError:
warnings.warn(stacklevel=level + 1, *args, **kw)
class RequirementParseError(ValueError):
def __str__(self):
return ' '.join(self.args)
def parse_requirements(strs):
"""Yield ``Requirement`` objects for each specification in `strs`
`strs` must be a string, or a (possibly-nested) iterable thereof.
# create a steppable iterator, so we can handle \-continuations
lines = iter(yield_lines(strs))
for line in lines:
# Drop comments -- a hash without a space may be in a URL.
if ' #' in line:
line = line[:line.find(' #')]
# If there is a line continuation, drop it, and append the next line.
if line.endswith('\\'):
line = line[:-2].strip()
line += next(lines)
yield Requirement(line)
class Requirement(packaging.requirements.Requirement):
def __init__(self, requirement_string):
"""DO NOT CALL THIS UNDOCUMENTED METHOD; use Requirement.parse()!"""
super(Requirement, self).__init__(requirement_string)
except packaging.requirements.InvalidRequirement as e:
raise RequirementParseError(str(e))
self.unsafe_name =
project_name = safe_name(
self.project_name, self.key = project_name, project_name.lower()
self.specs = [
(spec.operator, spec.version) for spec in self.specifier]
self.extras = tuple(map(safe_extra, self.extras))
self.hashCmp = (
str(self.marker) if self.marker else None,
self.__hash = hash(self.hashCmp)
def __eq__(self, other):
return (
isinstance(other, Requirement) and
self.hashCmp == other.hashCmp
def __ne__(self, other):
return not self == other
def __contains__(self, item):
if isinstance(item, Distribution):
if item.key != self.key:
return False
item = item.version
# Allow prereleases always in order to match the previous behavior of
# this method. In the future this should be smarter and follow PEP 440
# more accurately.
return self.specifier.contains(item, prereleases=True)
def __hash__(self):
return self.__hash
def __repr__(self): return "Requirement.parse(%r)" % str(self)
def parse(s):
req, = parse_requirements(s)
return req
def _get_mro(cls):
"""Get an mro for a type or classic class"""
if not isinstance(cls, type):
class cls(cls, object):
return cls.__mro__[1:]
return cls.__mro__
def _find_adapter(registry, ob):
"""Return an adapter factory for `ob` from `registry`"""
for t in _get_mro(getattr(ob, '__class__', type(ob))):
if t in registry:
return registry[t]
def ensure_directory(path):
"""Ensure that the parent directory of `path` exists"""
dirname = os.path.dirname(path)
if not os.path.isdir(dirname):
def _bypass_ensure_directory(path):
"""Sandbox-bypassing version of ensure_directory()"""
raise IOError('"os.mkdir" not supported on this platform.')
dirname, filename = split(path)
if dirname and filename and not isdir(dirname):
mkdir(dirname, 0o755)
def split_sections(s):
"""Split a string or iterable thereof into (section, content) pairs
Each ``section`` is a stripped version of the section header ("[section]")
and each ``content`` is a list of stripped lines excluding blank lines and
comment-only lines. If there are any such lines before the first section
header, they're returned in a first ``section`` of ``None``.
section = None
content = []
for line in yield_lines(s):
if line.startswith("["):
if line.endswith("]"):
if section or content:
yield section, content
section = line[1:-1].strip()
content = []
raise ValueError("Invalid section heading", line)
# wrap up last segment
yield section, content
def _mkstemp(*args, **kw):
old_open =
# temporarily bypass sandboxing = os_open
return tempfile.mkstemp(*args, **kw)
# and then put it back = old_open
# Silence the PEP440Warning by default, so that end users don't get hit by it
# randomly just because they use pkg_resources. We want to append the rule
# because we want earlier uses of filterwarnings to take precedence over this
# one.
warnings.filterwarnings("ignore", category=PEP440Warning, append=True)
# from jaraco.functools 1.3
def _call_aside(f, *args, **kwargs):
f(*args, **kwargs)
return f
def _initialize(g=globals()):
"Set up global resource manager (deliberately not state-saved)"
manager = ResourceManager()
g['_manager'] = manager
for name in dir(manager):
if not name.startswith('_'):
g[name] = getattr(manager, name)
def _initialize_master_working_set():
Prepare the master working set and make the ``require()``
API available.
This function has explicit effects on the global state
of pkg_resources. It is intended to be invoked once at
the initialization of this module.
Invocation by other packages is unsupported and done
at their own risk.
working_set = WorkingSet._build_master()
_declare_state('object', working_set=working_set)
require = working_set.require
iter_entry_points = working_set.iter_entry_points
add_activation_listener = working_set.subscribe
run_script = working_set.run_script
# backward compatibility
run_main = run_script
# Activate all distributions already on sys.path with replace=False and
# ensure that all distributions added to the working set in the future
# (e.g. by calling ``require()``) will get activated as well,
# with higher priority (replace=True).
dist = None # ensure dist is defined for del dist below
for dist in working_set:
del dist
add_activation_listener(lambda dist: dist.activate(replace=True), existing=False)
working_set.entries = []
# match order
list(map(working_set.add_entry, sys.path))
You can’t perform that action at this time.