branching: merge stable into default

contrib/heptapod-ci.yml

@@ -1,158 +1,139 b''
 stages:
   - tests
-  - phabricator
 
 image: registry.heptapod.net/mercurial/ci-images/mercurial-core:$HG_CI_IMAGE_TAG
 
 variables:
     PYTHON: python
     TEST_HGMODULEPOLICY: "allow"
     HG_CI_IMAGE_TAG: "v1.0"
     TEST_HGTESTS_ALLOW_NETIO: "0"
 
 .all_template: &all
   when: on_success
 
 .runtests_template: &runtests
     <<: *all
     stage: tests
     # The runner made a clone as root.
     # We make a new clone owned by user used to run the step.
     before_script:
       - hg clone . /tmp/mercurial-ci/ --noupdate --config phases.publish=no
       - hg -R /tmp/mercurial-ci/ update `hg log --rev '.' --template '{node}'`
       - cd /tmp/mercurial-ci/
       - ls -1 tests/test-check-*.* > /tmp/check-tests.txt
       - black --version
       - clang-format --version
     script:
         - echo "python used, $PYTHON"
         - echo "$RUNTEST_ARGS"
         - HGTESTS_ALLOW_NETIO="$TEST_HGTESTS_ALLOW_NETIO" HGMODULEPOLICY="$TEST_HGMODULEPOLICY" "$PYTHON" tests/run-tests.py --color=always $RUNTEST_ARGS
 
 checks:
     <<: *runtests
     variables:
         RUNTEST_ARGS: "--time --test-list /tmp/check-tests.txt"
         PYTHON: python3
 
 rust-cargo-test:
     <<: *all
     stage: tests
     script:
         - echo "python used, $PYTHON"
         - make rust-tests
     variables:
         PYTHON: python3
 
-phabricator-refresh:
-    stage: phabricator
-    except:
-      refs:
-        - merge_requests
-      variables:
-        - $PHABRICATOR_TOKEN == "NO-PHAB"
-    variables:
-      DEFAULT_COMMENT: ":white_check_mark: refresh by Heptapod after a successful CI run (:octopus: :green_heart:)"
-      STABLE_COMMENT: ":white_check_mark: refresh by Heptapod after a successful CI run (:octopus: :green_heart:)\n⚠ This patch is intended for stable ⚠\n{image https://media.giphy.com/media/nYI8SmmChYXK0/source.gif}"
-    script:
-      - |
-        if [ `hg branch` == "stable" ]; then
-            ./contrib/phab-refresh-stack.sh --comment "$STABLE_COMMENT";
-        else
-            ./contrib/phab-refresh-stack.sh --comment "$DEFAULT_COMMENT";
-        fi
-
 test-c:
     <<: *runtests
     variables:
         RUNTEST_ARGS: " --no-rust --blacklist /tmp/check-tests.txt"
         PYTHON: python3
         TEST_HGMODULEPOLICY: "c"
         TEST_HGTESTS_ALLOW_NETIO: "1"
 
 test-pure:
     <<: *runtests
     variables:
         RUNTEST_ARGS: "--pure --blacklist /tmp/check-tests.txt"
         PYTHON: python3
         TEST_HGMODULEPOLICY: "py"
 
 test-rust:
     <<: *runtests
     variables:
         HGWITHRUSTEXT: cpython
         RUNTEST_ARGS: "--rust --blacklist /tmp/check-tests.txt"
         PYTHON: python3
         TEST_HGMODULEPOLICY: "rust+c"
 
 test-rhg:
     <<: *runtests
     variables:
         HGWITHRUSTEXT: cpython
         RUNTEST_ARGS: "--rust --rhg --blacklist /tmp/check-tests.txt"
         PYTHON: python3
         TEST_HGMODULEPOLICY: "rust+c"
 
 test-chg:
     <<: *runtests
     variables:
         PYTHON: python3
         RUNTEST_ARGS: "--blacklist /tmp/check-tests.txt --chg"
         TEST_HGMODULEPOLICY: "c"
 
 check-pytype:
     extends: .runtests_template
     before_script:
       - hg clone . /tmp/mercurial-ci/ --noupdate --config phases.publish=no
       - hg -R /tmp/mercurial-ci/ update `hg log --rev '.' --template '{node}'`
       - cd /tmp/mercurial-ci/
       - make local PYTHON=$PYTHON
       - $PYTHON -m pip install --user -U pytype==2021.04.15
     variables:
         RUNTEST_ARGS: " --allow-slow-tests tests/test-check-pytype.t"
         HGTEST_SLOWTIMEOUT: "3600"
         PYTHON: python3
         TEST_HGMODULEPOLICY: "c"
 
 # `sh.exe --login` sets a couple of extra environment variables that are defined
 # in the MinGW shell, but switches CWD to /home/$username.  The previous value
 # is stored in OLDPWD.  Of the added variables, MSYSTEM is crucial to running
 # run-tests.py- it is needed to make run-tests.py generate a `python3` script
 # that satisfies the various shebang lines and delegates to `py -3`.
 .window_runtests_template: &windows_runtests
     <<: *all
     when: manual  # we don't have any Windows runners anymore at the moment
     stage: tests
     before_script:
       - C:/MinGW/msys/1.0/bin/sh.exe --login -c 'cd "$OLDPWD" && ls -1 tests/test-check-*.* > C:/Temp/check-tests.txt'
       # TODO: find/install cvs, bzr, perforce, gpg, sqlite3
 
     script:
         - echo "Entering script section"
         - echo "python used, $Env:PYTHON"
         - Invoke-Expression "$Env:PYTHON -V"
         - Invoke-Expression "$Env:PYTHON -m black --version"
         - echo "$Env:RUNTEST_ARGS"
         - echo "$Env:TMP"
         - echo "$Env:TEMP"
 
         - C:/MinGW/msys/1.0/bin/sh.exe --login -c 'cd "$OLDPWD" && HGTESTS_ALLOW_NETIO="$TEST_HGTESTS_ALLOW_NETIO" HGMODULEPOLICY="$TEST_HGMODULEPOLICY" $PYTHON tests/run-tests.py --color=always $RUNTEST_ARGS'
 
 windows:
     <<: *windows_runtests
     tags:
       - windows
     variables:
         TEST_HGMODULEPOLICY: "c"
         RUNTEST_ARGS: "--blacklist C:/Temp/check-tests.txt"
         PYTHON: py -3
 
 windows-pyox:
     <<: *windows_runtests
     tags:
       - windows
     variables:
         TEST_HGMODULEPOLICY: "c"
         RUNTEST_ARGS: "--blacklist C:/Temp/check-tests.txt --pyoxidized"
         PYTHON: py -3

contrib/packaging/docker/rhel8

@@ -1,18 +1,21 b''
 FROM rockylinux/rockylinux:8
 
 RUN groupadd -g %GID% build && \
     useradd -u %UID% -g %GID% -s /bin/bash -d /build -m build
 
 RUN yum install -y \
 	gcc \
 	gettext \
 	make \
 	python3-devel \
 	python3-docutils \
 	rpm-build
 
 # For creating repo meta data
 RUN yum install -y createrepo
 
 # For rust extensions
 RUN yum install -y cargo
+
+# avoid incorrect docker image permissions on /tmp preventing writes by non-root users
+RUN chmod 1777 /tmp

hgdemandimport/__init__.py

@@ -1,80 +1,83 b''
 # hgdemandimport - global demand-loading of modules for Mercurial
 #
 # Copyright 2017 Facebook Inc.
 #
 # This software may be used and distributed according to the terms of the
 # GNU General Public License version 2 or any later version.
 
 '''demandimport - automatic demand-loading of modules'''
 
 # This is in a separate package from mercurial because in Python 3,
 # demand loading is per-package. Keeping demandimport in the mercurial package
 # would disable demand loading for any modules in mercurial.
 
 
 import os
 import sys
 
 from . import demandimportpy3 as demandimport
 
 # Full module names which can't be lazy imported.
 # Extensions can add to this set.
 IGNORES = {
     '__future__',
     '_hashlib',
     # ImportError during pkg_resources/__init__.py:fixup_namespace_package
     '_imp',
     '_xmlplus',
     'fcntl',
     'nt',  # pathlib2 tests the existence of built-in 'nt' module
     'win32com.gen_py',
     'win32com.shell',  # 'appdirs' tries to import win32com.shell
     '_winreg',  # 2.7 mimetypes needs immediate ImportError
     'pythoncom',
     # imported by tarfile, not available under Windows
     'pwd',
     'grp',
     # imported by profile, itself imported by hotshot.stats,
     # not available under Windows
     'resource',
     # this trips up many extension authors
     'gtk',
     # setuptools' pkg_resources.py expects "from __main__ import x" to
     # raise ImportError if x not defined
     '__main__',
     '_ast',  # https://bugs.python.org/issue41631
     '_ssl',  # conditional imports in the stdlib, issue1964
     '_sre',  # issue4920
     'rfc822',
     'mimetools',
     'sqlalchemy.events',  # has import-time side effects (issue5085)
     'sqlalchemy.dialects',  # similar problems as above
     # setuptools 8 expects this module to explode early when not on windows
     'distutils.msvc9compiler',
     '__builtin__',
     'builtins',
     'urwid.command_map',  # for pudb
     'lzma',
 }
 
 _pypy = '__pypy__' in sys.builtin_module_names
 
 if _pypy:
     # _ctypes.pointer is shadowed by "from ... import pointer" (PyPy 5)
     IGNORES.add('_ctypes.pointer')
+    # pure Python module on PyPy, must be loaded to raise ModuleNotFoundError
+    # on non-Windows platforms
+    IGNORES.add('msvcrt')
 
 demandimport.init(IGNORES)
 
 # Re-export.
 isenabled = demandimport.isenabled
 disable = demandimport.disable
 deactivated = demandimport.deactivated
 
 
 def enable():
     # chg pre-imports modules so do not enable demandimport for it
     if (
         'CHGINTERNALMARK' not in os.environ
         and os.environ.get('HGDEMANDIMPORT') != 'disable'
     ):
         demandimport.enable()

mercurial/revset.py

@@ -1,2817 +1,2819 b''
 # revset.py - revision set queries for mercurial
 #
 # Copyright 2010 Olivia Mackall <[e-mail]>
 #
 # This software may be used and distributed according to the terms of the
 # GNU General Public License version 2 or any later version.
 
 
 import binascii
 import re
 
 from .i18n import _
 from .pycompat import getattr
 from .node import (
     bin,
     nullrev,
     wdirrev,
 )
 from . import (
     dagop,
     destutil,
     diffutil,
     encoding,
     error,
     grep as grepmod,
     hbisect,
     match as matchmod,
     obsolete as obsmod,
     obsutil,
     pathutil,
     phases,
     pycompat,
     registrar,
     repoview,
     revsetlang,
     scmutil,
     smartset,
     stack as stackmod,
     util,
 )
 from .utils import (
     dateutil,
     stringutil,
     urlutil,
 )
 
 # helpers for processing parsed tree
 getsymbol = revsetlang.getsymbol
 getstring = revsetlang.getstring
 getinteger = revsetlang.getinteger
 getboolean = revsetlang.getboolean
 getlist = revsetlang.getlist
 getintrange = revsetlang.getintrange
 getargs = revsetlang.getargs
 getargsdict = revsetlang.getargsdict
 
 baseset = smartset.baseset
 generatorset = smartset.generatorset
 spanset = smartset.spanset
 fullreposet = smartset.fullreposet
 
 # revisions not included in all(), but populated if specified
 _virtualrevs = (nullrev, wdirrev)
 
 # Constants for ordering requirement, used in getset():
 #
 # If 'define', any nested functions and operations MAY change the ordering of
 # the entries in the set (but if changes the ordering, it MUST ALWAYS change
 # it). If 'follow', any nested functions and operations MUST take the ordering
 # specified by the first operand to the '&' operator.
 #
 # For instance,
 #
 #   X & (Y | Z)
 #   ^   ^^^^^^^
 #   |   follow
 #   define
 #
 # will be evaluated as 'or(y(x()), z(x()))', where 'x()' can change the order
 # of the entries in the set, but 'y()', 'z()' and 'or()' shouldn't.
 #
 # 'any' means the order doesn't matter. For instance,
 #
 #   (X & !Y) | ancestors(Z)
 #         ^              ^
 #         any            any
 #
 # For 'X & !Y', 'X' decides the order and 'Y' is subtracted from 'X', so the
 # order of 'Y' does not matter. For 'ancestors(Z)', Z's order does not matter
 # since 'ancestors' does not care about the order of its argument.
 #
 # Currently, most revsets do not care about the order, so 'define' is
 # equivalent to 'follow' for them, and the resulting order is based on the
 # 'subset' parameter passed down to them:
 #
 #   m = revset.match(...)
 #   m(repo, subset, order=defineorder)
 #           ^^^^^^
 #      For most revsets, 'define' means using the order this subset provides
 #
 # There are a few revsets that always redefine the order if 'define' is
 # specified: 'sort(X)', 'reverse(X)', 'x:y'.
 anyorder = b'any'  # don't care the order, could be even random-shuffled
 defineorder = b'define'  # ALWAYS redefine, or ALWAYS follow the current order
 followorder = b'follow'  # MUST follow the current order
 
 # helpers
 
 
 def getset(repo, subset, x, order=defineorder):
     if not x:
         raise error.ParseError(_(b"missing argument"))
     return methods[x[0]](repo, subset, *x[1:], order=order)
 
 
 def _getrevsource(repo, r):
     extra = repo[r].extra()
     for label in (b'source', b'transplant_source', b'rebase_source'):
         if label in extra:
             try:
                 return repo[extra[label]].rev()
             except error.RepoLookupError:
                 pass
     return None
 
 
 def _sortedb(xs):
     return sorted(pycompat.rapply(pycompat.maybebytestr, xs))
 
 
 # operator methods
 
 
 def stringset(repo, subset, x, order):
     if not x:
         raise error.ParseError(_(b"empty string is not a valid revision"))
     x = scmutil.intrev(scmutil.revsymbol(repo, x))
     if x in subset or x in _virtualrevs and isinstance(subset, fullreposet):
         return baseset([x])
     return baseset()
 
 
 def rawsmartset(repo, subset, x, order):
     """argument is already a smartset, use that directly"""
     if order == followorder:
         return subset & x
     else:
         return x & subset
 
 
 def rangeset(repo, subset, x, y, order):
     m = getset(repo, fullreposet(repo), x)
     n = getset(repo, fullreposet(repo), y)
 
     if not m or not n:
         return baseset()
     return _makerangeset(repo, subset, m.first(), n.last(), order)
 
 
 def rangeall(repo, subset, x, order):
     assert x is None
     return _makerangeset(repo, subset, 0, repo.changelog.tiprev(), order)
 
 
 def rangepre(repo, subset, y, order):
     # ':y' can't be rewritten to '0:y' since '0' may be hidden
     n = getset(repo, fullreposet(repo), y)
     if not n:
         return baseset()
     return _makerangeset(repo, subset, 0, n.last(), order)
 
 
 def rangepost(repo, subset, x, order):
     m = getset(repo, fullreposet(repo), x)
     if not m:
         return baseset()
     return _makerangeset(
         repo, subset, m.first(), repo.changelog.tiprev(), order
     )
 
 
 def _makerangeset(repo, subset, m, n, order):
     if m == n:
         r = baseset([m])
     elif n == wdirrev:
         r = spanset(repo, m, len(repo)) + baseset([n])
     elif m == wdirrev:
         r = baseset([m]) + spanset(repo, repo.changelog.tiprev(), n - 1)
     elif m < n:
         r = spanset(repo, m, n + 1)
     else:
         r = spanset(repo, m, n - 1)
 
     if order == defineorder:
         return r & subset
     else:
         # carrying the sorting over when possible would be more efficient
         return subset & r
 
 
 def dagrange(repo, subset, x, y, order):
     r = fullreposet(repo)
     xs = dagop.reachableroots(
         repo, getset(repo, r, x), getset(repo, r, y), includepath=True
     )
     return subset & xs
 
 
 def andset(repo, subset, x, y, order):
     if order == anyorder:
         yorder = anyorder
     else:
         yorder = followorder
     return getset(repo, getset(repo, subset, x, order), y, yorder)
 
 
 def andsmallyset(repo, subset, x, y, order):
     # 'andsmally(x, y)' is equivalent to 'and(x, y)', but faster when y is small
     if order == anyorder:
         yorder = anyorder
     else:
         yorder = followorder
     return getset(repo, getset(repo, subset, y, yorder), x, order)
 
 
 def differenceset(repo, subset, x, y, order):
     return getset(repo, subset, x, order) - getset(repo, subset, y, anyorder)
 
 
 def _orsetlist(repo, subset, xs, order):
     assert xs
     if len(xs) == 1:
         return getset(repo, subset, xs[0], order)
     p = len(xs) // 2
     a = _orsetlist(repo, subset, xs[:p], order)
     b = _orsetlist(repo, subset, xs[p:], order)
     return a + b
 
 
 def orset(repo, subset, x, order):
     xs = getlist(x)
     if not xs:
         return baseset()
     if order == followorder:
         # slow path to take the subset order
         return subset & _orsetlist(repo, fullreposet(repo), xs, anyorder)
     else:
         return _orsetlist(repo, subset, xs, order)
 
 
 def notset(repo, subset, x, order):
     return subset - getset(repo, subset, x, anyorder)
 
 
 def relationset(repo, subset, x, y, order):
     # this is pretty basic implementation of 'x#y' operator, still
     # experimental so undocumented. see the wiki for further ideas.
     # https://www.mercurial-scm.org/wiki/RevsetOperatorPlan
     rel = getsymbol(y)
     if rel in relations:
         return relations[rel](repo, subset, x, rel, order)
 
     relnames = [r for r in relations.keys() if len(r) > 1]
     raise error.UnknownIdentifier(rel, relnames)
 
 
 def _splitrange(a, b):
     """Split range with bounds a and b into two ranges at 0 and return two
     tuples of numbers for use as startdepth and stopdepth arguments of
     revancestors and revdescendants.
 
     >>> _splitrange(-10, -5)     # [-10:-5]
     ((5, 11), (None, None))
     >>> _splitrange(5, 10)       # [5:10]
     ((None, None), (5, 11))
     >>> _splitrange(-10, 10)     # [-10:10]
     ((0, 11), (0, 11))
     >>> _splitrange(-10, 0)      # [-10:0]
     ((0, 11), (None, None))
     >>> _splitrange(0, 10)       # [0:10]
     ((None, None), (0, 11))
     >>> _splitrange(0, 0)        # [0:0]
     ((0, 1), (None, None))
     >>> _splitrange(1, -1)       # [1:-1]
     ((None, None), (None, None))
     """
     ancdepths = (None, None)
     descdepths = (None, None)
     if a == b == 0:
         ancdepths = (0, 1)
     if a < 0:
         ancdepths = (-min(b, 0), -a + 1)
     if b > 0:
         descdepths = (max(a, 0), b + 1)
     return ancdepths, descdepths
 
 
 def generationsrel(repo, subset, x, rel, order):
     z = (b'rangeall', None)
     return generationssubrel(repo, subset, x, rel, z, order)
 
 
 def generationssubrel(repo, subset, x, rel, z, order):
     # TODO: rewrite tests, and drop startdepth argument from ancestors() and
     # descendants() predicates
     a, b = getintrange(
         z,
         _(b'relation subscript must be an integer or a range'),
         _(b'relation subscript bounds must be integers'),
         deffirst=-(dagop.maxlogdepth - 1),
         deflast=+(dagop.maxlogdepth - 1),
     )
     (ancstart, ancstop), (descstart, descstop) = _splitrange(a, b)
 
     if ancstart is None and descstart is None:
         return baseset()
 
     revs = getset(repo, fullreposet(repo), x)
     if not revs:
         return baseset()
 
     if ancstart is not None and descstart is not None:
         s = dagop.revancestors(repo, revs, False, ancstart, ancstop)
         s += dagop.revdescendants(repo, revs, False, descstart, descstop)
     elif ancstart is not None:
         s = dagop.revancestors(repo, revs, False, ancstart, ancstop)
     elif descstart is not None:
         s = dagop.revdescendants(repo, revs, False, descstart, descstop)
 
     return subset & s
 
 
 def relsubscriptset(repo, subset, x, y, z, order):
     # this is pretty basic implementation of 'x#y[z]' operator, still
     # experimental so undocumented. see the wiki for further ideas.
     # https://www.mercurial-scm.org/wiki/RevsetOperatorPlan
     rel = getsymbol(y)
     if rel in subscriptrelations:
         return subscriptrelations[rel](repo, subset, x, rel, z, order)
 
     relnames = [r for r in subscriptrelations.keys() if len(r) > 1]
     raise error.UnknownIdentifier(rel, relnames)
 
 
 def subscriptset(repo, subset, x, y, order):
     raise error.ParseError(_(b"can't use a subscript in this context"))
 
 
 def listset(repo, subset, *xs, **opts):
     raise error.ParseError(
         _(b"can't use a list in this context"),
         hint=_(b'see \'hg help "revsets.x or y"\''),
     )
 
 
 def keyvaluepair(repo, subset, k, v, order):
     raise error.ParseError(_(b"can't use a key-value pair in this context"))
 
 
 def func(repo, subset, a, b, order):
     f = getsymbol(a)
     if f in symbols:
         func = symbols[f]
         if getattr(func, '_takeorder', False):
             return func(repo, subset, b, order)
         return func(repo, subset, b)
 
     keep = lambda fn: getattr(fn, '__doc__', None) is not None
 
     syms = [s for (s, fn) in symbols.items() if keep(fn)]
     raise error.UnknownIdentifier(f, syms)
 
 
 # functions
 
 # symbols are callables like:
 #   fn(repo, subset, x)
 # with:
 #   repo - current repository instance
 #   subset - of revisions to be examined
 #   x - argument in tree form
 symbols = revsetlang.symbols
 
 # symbols which can't be used for a DoS attack for any given input
 # (e.g. those which accept regexes as plain strings shouldn't be included)
 # functions that just return a lot of changesets (like all) don't count here
 safesymbols = set()
 
 predicate = registrar.revsetpredicate()
 
 
 @predicate(b'_destupdate')
 def _destupdate(repo, subset, x):
     # experimental revset for update destination
     args = getargsdict(x, b'limit', b'clean')
     return subset & baseset(
         [destutil.destupdate(repo, **pycompat.strkwargs(args))[0]]
     )
 
 
 @predicate(b'_destmerge')
 def _destmerge(repo, subset, x):
     # experimental revset for merge destination
     sourceset = None
     if x is not None:
         sourceset = getset(repo, fullreposet(repo), x)
     return subset & baseset([destutil.destmerge(repo, sourceset=sourceset)])
 
 
 @predicate(b'adds(pattern)', safe=True, weight=30)
 def adds(repo, subset, x):
     """Changesets that add a file matching pattern.
 
     The pattern without explicit kind like ``glob:`` is expected to be
     relative to the current directory and match against a file or a
     directory.
     """
     # i18n: "adds" is a keyword
     pat = getstring(x, _(b"adds requires a pattern"))
     return checkstatus(repo, subset, pat, 'added')
 
 
 @predicate(b'ancestor(*changeset)', safe=True, weight=0.5)
 def ancestor(repo, subset, x):
     """A greatest common ancestor of the changesets.
 
     Accepts 0 or more changesets.
     Will return empty list when passed no args.
     Greatest common ancestor of a single changeset is that changeset.
     """
     reviter = iter(orset(repo, fullreposet(repo), x, order=anyorder))
     try:
         anc = repo[next(reviter)]
     except StopIteration:
         return baseset()
     for r in reviter:
         anc = anc.ancestor(repo[r])
 
     r = scmutil.intrev(anc)
     if r in subset:
         return baseset([r])
     return baseset()
 
 
 def _ancestors(
     repo, subset, x, followfirst=False, startdepth=None, stopdepth=None
 ):
     heads = getset(repo, fullreposet(repo), x)
     if not heads:
         return baseset()
     s = dagop.revancestors(repo, heads, followfirst, startdepth, stopdepth)
     return subset & s
 
 
 @predicate(b'ancestors(set[, depth])', safe=True)
 def ancestors(repo, subset, x):
     """Changesets that are ancestors of changesets in set, including the
     given changesets themselves.
 
     If depth is specified, the result only includes changesets up to
     the specified generation.
     """
     # startdepth is for internal use only until we can decide the UI
     args = getargsdict(x, b'ancestors', b'set depth startdepth')
     if b'set' not in args:
         # i18n: "ancestors" is a keyword
         raise error.ParseError(_(b'ancestors takes at least 1 argument'))
     startdepth = stopdepth = None
     if b'startdepth' in args:
         n = getinteger(
             args[b'startdepth'], b"ancestors expects an integer startdepth"
         )
         if n < 0:
             raise error.ParseError(b"negative startdepth")
         startdepth = n
     if b'depth' in args:
         # i18n: "ancestors" is a keyword
         n = getinteger(args[b'depth'], _(b"ancestors expects an integer depth"))
         if n < 0:
             raise error.ParseError(_(b"negative depth"))
         stopdepth = n + 1
     return _ancestors(
         repo, subset, args[b'set'], startdepth=startdepth, stopdepth=stopdepth
     )
 
 
 @predicate(b'_firstancestors', safe=True)
 def _firstancestors(repo, subset, x):
     # ``_firstancestors(set)``
     # Like ``ancestors(set)`` but follows only the first parents.
     return _ancestors(repo, subset, x, followfirst=True)
 
 
 def _childrenspec(repo, subset, x, n, order):
     """Changesets that are the Nth child of a changeset
     in set.
     """
     cs = set()
     for r in getset(repo, fullreposet(repo), x):
         for i in range(n):
             c = repo[r].children()
             if len(c) == 0:
                 break
             if len(c) > 1:
                 raise error.RepoLookupError(
                     _(b"revision in set has more than one child")
                 )
             r = c[0].rev()
         else:
             cs.add(r)
     return subset & cs
 
 
 def ancestorspec(repo, subset, x, n, order):
     """``set~n``
     Changesets that are the Nth ancestor (first parents only) of a changeset
     in set.
     """
     n = getinteger(n, _(b"~ expects a number"))
     if n < 0:
         # children lookup
         return _childrenspec(repo, subset, x, -n, order)
     ps = set()
     cl = repo.changelog
     for r in getset(repo, fullreposet(repo), x):
         for i in range(n):
             try:
                 r = cl.parentrevs(r)[0]
             except error.WdirUnsupported:
                 r = repo[r].p1().rev()
         ps.add(r)
     return subset & ps
 
 
 @predicate(b'author(string)', safe=True, weight=10)
 def author(repo, subset, x):
     """Alias for ``user(string)``."""
     # i18n: "author" is a keyword
     n = getstring(x, _(b"author requires a string"))
     kind, pattern, matcher = _substringmatcher(n, casesensitive=False)
     return subset.filter(
         lambda x: matcher(repo[x].user()), condrepr=(b'<user %r>', n)
     )
 
 
 @predicate(b'bisect(string)', safe=True)
 def bisect(repo, subset, x):
     """Changesets marked in the specified bisect status:
 
     - ``good``, ``bad``, ``skip``: csets explicitly marked as good/bad/skip
     - ``goods``, ``bads``      : csets topologically good/bad
     - ``range``              : csets taking part in the bisection
     - ``pruned``             : csets that are goods, bads or skipped
     - ``untested``           : csets whose fate is yet unknown
     - ``ignored``            : csets ignored due to DAG topology
     - ``current``            : the cset currently being bisected
     """
     # i18n: "bisect" is a keyword
     status = getstring(x, _(b"bisect requires a string")).lower()
     state = set(hbisect.get(repo, status))
     return subset & state
 
 
 # Backward-compatibility
 # - no help entry so that we do not advertise it any more
 @predicate(b'bisected', safe=True)
 def bisected(repo, subset, x):
     return bisect(repo, subset, x)
 
 
 @predicate(b'bookmark([name])', safe=True)
 def bookmark(repo, subset, x):
     """The named bookmark or all bookmarks.
 
     Pattern matching is supported for `name`. See :hg:`help revisions.patterns`.
     """
     # i18n: "bookmark" is a keyword
     args = getargs(x, 0, 1, _(b'bookmark takes one or no arguments'))
     if args:
         bm = getstring(
             args[0],
             # i18n: "bookmark" is a keyword
             _(b'the argument to bookmark must be a string'),
         )
         kind, pattern, matcher = stringutil.stringmatcher(bm)
         bms = set()
         if kind == b'literal':
             if bm == pattern:
                 pattern = repo._bookmarks.expandname(pattern)
             bmrev = repo._bookmarks.get(pattern, None)
             if not bmrev:
                 raise error.RepoLookupError(
                     _(b"bookmark '%s' does not exist") % pattern
                 )
             bms.add(repo[bmrev].rev())
         else:
             matchrevs = set()
             for name, bmrev in repo._bookmarks.items():
                 if matcher(name):
                     matchrevs.add(bmrev)
             for bmrev in matchrevs:
                 bms.add(repo[bmrev].rev())
     else:
         bms = {repo[r].rev() for r in repo._bookmarks.values()}
     bms -= {nullrev}
     return subset & bms
 
 
 @predicate(b'branch(string or set)', safe=True, weight=10)
 def branch(repo, subset, x):
     """
     All changesets belonging to the given branch or the branches of the given
     changesets.
 
     Pattern matching is supported for `string`. See
     :hg:`help revisions.patterns`.
     """
     getbi = repo.revbranchcache().branchinfo
 
     def getbranch(r):
         try:
             return getbi(r)[0]
         except error.WdirUnsupported:
             return repo[r].branch()
 
     try:
         b = getstring(x, b'')
     except error.ParseError:
         # not a string, but another revspec, e.g. tip()
         pass
     else:
         kind, pattern, matcher = stringutil.stringmatcher(b)
         if kind == b'literal':
             # note: falls through to the revspec case if no branch with
             # this name exists and pattern kind is not specified explicitly
             if repo.branchmap().hasbranch(pattern):
                 return subset.filter(
                     lambda r: matcher(getbranch(r)),
                     condrepr=(b'<branch %r>', b),
                 )
             if b.startswith(b'literal:'):
                 raise error.RepoLookupError(
                     _(b"branch '%s' does not exist") % pattern
                 )
         else:
             return subset.filter(
                 lambda r: matcher(getbranch(r)), condrepr=(b'<branch %r>', b)
             )
 
     s = getset(repo, fullreposet(repo), x)
     b = set()
     for r in s:
         b.add(getbranch(r))
     c = s.__contains__
     return subset.filter(
         lambda r: c(r) or getbranch(r) in b,
         condrepr=lambda: b'<branch %r>' % _sortedb(b),
     )
 
 
 @predicate(b'phasedivergent()', safe=True)
 def phasedivergent(repo, subset, x):
     """Mutable changesets marked as successors of public changesets.
 
     Only non-public and non-obsolete changesets can be `phasedivergent`.
     (EXPERIMENTAL)
     """
     # i18n: "phasedivergent" is a keyword
     getargs(x, 0, 0, _(b"phasedivergent takes no arguments"))
     phasedivergent = obsmod.getrevs(repo, b'phasedivergent')
     return subset & phasedivergent
 
 
 @predicate(b'bundle()', safe=True)
 def bundle(repo, subset, x):
     """Changesets in the bundle.
 
     Bundle must be specified by the -R option."""
 
     try:
         bundlerevs = repo.changelog.bundlerevs
     except AttributeError:
         raise error.Abort(_(b"no bundle provided - specify with -R"))
     return subset & bundlerevs
 
 
 def checkstatus(repo, subset, pat, field):
     """Helper for status-related revsets (adds, removes, modifies).
     The field parameter says which kind is desired.
     """
     hasset = matchmod.patkind(pat) == b'set'
 
     mcache = [None]
 
     def matches(x):
         c = repo[x]
         if not mcache[0] or hasset:
             mcache[0] = matchmod.match(repo.root, repo.getcwd(), [pat], ctx=c)
         m = mcache[0]
         fname = None
 
         assert m is not None  # help pytype
         if not m.anypats() and len(m.files()) == 1:
             fname = m.files()[0]
         if fname is not None:
             if fname not in c.files():
                 return False
         else:
             if not any(m(f) for f in c.files()):
                 return False
         files = getattr(repo.status(c.p1().node(), c.node()), field)
         if fname is not None:
             if fname in files:
                 return True
         else:
             if any(m(f) for f in files):
                 return True
 
     return subset.filter(
         matches, condrepr=(b'<status.%s %r>', pycompat.sysbytes(field), pat)
     )
 
 
 def _children(repo, subset, parentset):
     if not parentset:
         return baseset()
     cs = set()
     pr = repo.changelog.parentrevs
     minrev = parentset.min()
     for r in subset:
         if r <= minrev:
             continue
         p1, p2 = pr(r)
         if p1 in parentset:
             cs.add(r)
         if p2 != nullrev and p2 in parentset:
             cs.add(r)
     return baseset(cs)
 
 
 @predicate(b'children(set)', safe=True)
 def children(repo, subset, x):
     """Child changesets of changesets in set."""
     s = getset(repo, fullreposet(repo), x)
     cs = _children(repo, subset, s)
     return subset & cs
 
 
 @predicate(b'closed()', safe=True, weight=10)
 def closed(repo, subset, x):
     """Changeset is closed."""
     # i18n: "closed" is a keyword
     getargs(x, 0, 0, _(b"closed takes no arguments"))
     return subset.filter(
         lambda r: repo[r].closesbranch(), condrepr=b'<branch closed>'
     )
 
 
 # for internal use
 @predicate(b'_commonancestorheads(set)', safe=True)
 def _commonancestorheads(repo, subset, x):
     # This is an internal method is for quickly calculating "heads(::x and
     # ::y)"
 
     # These greatest common ancestors are the same ones that the consensus bid
     # merge will find.
     startrevs = getset(repo, fullreposet(repo), x, order=anyorder)
 
     ancs = repo.changelog._commonancestorsheads(*list(startrevs))
     return subset & baseset(ancs)
 
 
 @predicate(b'commonancestors(set)', safe=True)
 def commonancestors(repo, subset, x):
     """Changesets that are ancestors of every changeset in set."""
     startrevs = getset(repo, fullreposet(repo), x, order=anyorder)
     if not startrevs:
         return baseset()
     for r in startrevs:
         subset &= dagop.revancestors(repo, baseset([r]))
     return subset
 
 
 @predicate(b'conflictlocal()', safe=True)
 def conflictlocal(repo, subset, x):
     """The local side of the merge, if currently in an unresolved merge.
 
     "merge" here includes merge conflicts from e.g. 'hg rebase' or 'hg graft'.
     """
     getargs(x, 0, 0, _(b"conflictlocal takes no arguments"))
     from . import mergestate as mergestatemod
 
     mergestate = mergestatemod.mergestate.read(repo)
     if mergestate.active() and repo.changelog.hasnode(mergestate.local):
         return subset & {repo.changelog.rev(mergestate.local)}
 
     return baseset()
 
 
 @predicate(b'conflictother()', safe=True)
 def conflictother(repo, subset, x):
     """The other side of the merge, if currently in an unresolved merge.
 
     "merge" here includes merge conflicts from e.g. 'hg rebase' or 'hg graft'.
     """
     getargs(x, 0, 0, _(b"conflictother takes no arguments"))
     from . import mergestate as mergestatemod
 
     mergestate = mergestatemod.mergestate.read(repo)
     if mergestate.active() and repo.changelog.hasnode(mergestate.other):
         return subset & {repo.changelog.rev(mergestate.other)}
 
     return baseset()
 
 
 @predicate(b'contains(pattern)', weight=100)
 def contains(repo, subset, x):
     """The revision's manifest contains a file matching pattern (but might not
     modify it). See :hg:`help patterns` for information about file patterns.
 
     The pattern without explicit kind like ``glob:`` is expected to be
     relative to the current directory and match against a file exactly
     for efficiency.
     """
     # i18n: "contains" is a keyword
     pat = getstring(x, _(b"contains requires a pattern"))
 
     def matches(x):
         if not matchmod.patkind(pat):
             pats = pathutil.canonpath(repo.root, repo.getcwd(), pat)
             if pats in repo[x]:
                 return True
         else:
             c = repo[x]
             m = matchmod.match(repo.root, repo.getcwd(), [pat], ctx=c)
             for f in c.manifest():
                 if m(f):
                     return True
         return False
 
     return subset.filter(matches, condrepr=(b'<contains %r>', pat))
 
 
 @predicate(b'converted([id])', safe=True)
 def converted(repo, subset, x):
     """Changesets converted from the given identifier in the old repository if
     present, or all converted changesets if no identifier is specified.
     """
 
     # There is exactly no chance of resolving the revision, so do a simple
     # string compare and hope for the best
 
     rev = None
     # i18n: "converted" is a keyword
     l = getargs(x, 0, 1, _(b'converted takes one or no arguments'))
     if l:
         # i18n: "converted" is a keyword
         rev = getstring(l[0], _(b'converted requires a revision'))
 
     def _matchvalue(r):
         source = repo[r].extra().get(b'convert_revision', None)
         return source is not None and (rev is None or source.startswith(rev))
 
     return subset.filter(
         lambda r: _matchvalue(r), condrepr=(b'<converted %r>', rev)
     )
 
 
 @predicate(b'date(interval)', safe=True, weight=10)
 def date(repo, subset, x):
     """Changesets within the interval, see :hg:`help dates`."""
     # i18n: "date" is a keyword
     ds = getstring(x, _(b"date requires a string"))
     dm = dateutil.matchdate(ds)
     return subset.filter(
         lambda x: dm(repo[x].date()[0]), condrepr=(b'<date %r>', ds)
     )
 
 
 @predicate(b'desc(string)', safe=True, weight=10)
 def desc(repo, subset, x):
     """Search commit message for string. The match is case-insensitive.
 
     Pattern matching is supported for `string`. See
     :hg:`help revisions.patterns`.
     """
     # i18n: "desc" is a keyword
     ds = getstring(x, _(b"desc requires a string"))
 
     kind, pattern, matcher = _substringmatcher(ds, casesensitive=False)
 
     return subset.filter(
         lambda r: matcher(repo[r].description()), condrepr=(b'<desc %r>', ds)
     )
 
 
 def _descendants(
     repo, subset, x, followfirst=False, startdepth=None, stopdepth=None
 ):
     roots = getset(repo, fullreposet(repo), x)
     if not roots:
         return baseset()
     s = dagop.revdescendants(repo, roots, followfirst, startdepth, stopdepth)
     return subset & s
 
 
 @predicate(b'descendants(set[, depth])', safe=True)
 def descendants(repo, subset, x):
     """Changesets which are descendants of changesets in set, including the
     given changesets themselves.
 
     If depth is specified, the result only includes changesets up to
     the specified generation.
     """
     # startdepth is for internal use only until we can decide the UI
     args = getargsdict(x, b'descendants', b'set depth startdepth')
     if b'set' not in args:
         # i18n: "descendants" is a keyword
         raise error.ParseError(_(b'descendants takes at least 1 argument'))
     startdepth = stopdepth = None
     if b'startdepth' in args:
         n = getinteger(
             args[b'startdepth'], b"descendants expects an integer startdepth"
         )
         if n < 0:
             raise error.ParseError(b"negative startdepth")
         startdepth = n
     if b'depth' in args:
         # i18n: "descendants" is a keyword
         n = getinteger(
             args[b'depth'], _(b"descendants expects an integer depth")
         )
         if n < 0:
             raise error.ParseError(_(b"negative depth"))
         stopdepth = n + 1
     return _descendants(
         repo, subset, args[b'set'], startdepth=startdepth, stopdepth=stopdepth
     )
 
 
 @predicate(b'_firstdescendants', safe=True)
 def _firstdescendants(repo, subset, x):
     # ``_firstdescendants(set)``
     # Like ``descendants(set)`` but follows only the first parents.
     return _descendants(repo, subset, x, followfirst=True)
 
 
 @predicate(b'destination([set])', safe=True, weight=10)
 def destination(repo, subset, x):
     """Changesets that were created by a graft, transplant or rebase operation,
     with the given revisions specified as the source.  Omitting the optional set
     is the same as passing all().
     """
     if x is not None:
         sources = getset(repo, fullreposet(repo), x)
     else:
         sources = fullreposet(repo)
 
     dests = set()
 
     # subset contains all of the possible destinations that can be returned, so
     # iterate over them and see if their source(s) were provided in the arg set.
     # Even if the immediate src of r is not in the arg set, src's source (or
     # further back) may be.  Scanning back further than the immediate src allows
     # transitive transplants and rebases to yield the same results as transitive
     # grafts.
     for r in subset:
         src = _getrevsource(repo, r)
         lineage = None
 
         while src is not None:
             if lineage is None:
                 lineage = list()
 
             lineage.append(r)
 
             # The visited lineage is a match if the current source is in the arg
             # set.  Since every candidate dest is visited by way of iterating
             # subset, any dests further back in the lineage will be tested by a
             # different iteration over subset.  Likewise, if the src was already
             # selected, the current lineage can be selected without going back
             # further.
             if src in sources or src in dests:
                 dests.update(lineage)
                 break
 
             r = src
             src = _getrevsource(repo, r)
 
     return subset.filter(
         dests.__contains__,
         condrepr=lambda: b'<destination %r>' % _sortedb(dests),
     )
 
 
 @predicate(b'diffcontains(pattern)', weight=110)
 def diffcontains(repo, subset, x):
     """Search revision differences for when the pattern was added or removed.
 
     The pattern may be a substring literal or a regular expression. See
     :hg:`help revisions.patterns`.
     """
     args = getargsdict(x, b'diffcontains', b'pattern')
     if b'pattern' not in args:
         # i18n: "diffcontains" is a keyword
         raise error.ParseError(_(b'diffcontains takes at least 1 argument'))
 
     pattern = getstring(
         args[b'pattern'], _(b'diffcontains requires a string pattern')
     )
     regexp = stringutil.substringregexp(pattern, re.M)
 
     # TODO: add support for file pattern and --follow. For example,
     # diffcontains(pattern[, set]) where set may be file(pattern) or
     # follow(pattern), and we'll eventually add a support for narrowing
     # files by revset?
     fmatch = matchmod.always()
 
     def makefilematcher(ctx):
         return fmatch
 
     # TODO: search in a windowed way
     searcher = grepmod.grepsearcher(repo.ui, repo, regexp, diff=True)
 
     def testdiff(rev):
         # consume the generator to discard revfiles/matches cache
         found = False
         for fn, ctx, pstates, states in searcher.searchfiles(
             baseset([rev]), makefilematcher
         ):
             if next(grepmod.difflinestates(pstates, states), None):
                 found = True
         return found
 
     return subset.filter(testdiff, condrepr=(b'<diffcontains %r>', pattern))
 
 
 @predicate(b'contentdivergent()', safe=True)
 def contentdivergent(repo, subset, x):
     """
     Final successors of changesets with an alternative set of final
     successors. (EXPERIMENTAL)
     """
     # i18n: "contentdivergent" is a keyword
     getargs(x, 0, 0, _(b"contentdivergent takes no arguments"))
     contentdivergent = obsmod.getrevs(repo, b'contentdivergent')
     return subset & contentdivergent
 
 
 @predicate(b'expectsize(set[, size])', safe=True, takeorder=True)
 def expectsize(repo, subset, x, order):
     """Return the given revset if size matches the revset size.
     Abort if the revset doesn't expect given size.
     size can either be an integer range or an integer.
 
     For example, ``expectsize(0:1, 3:5)`` will abort as revset size is 2 and
     2 is not between 3 and 5 inclusive."""
 
     args = getargsdict(x, b'expectsize', b'set size')
     minsize = 0
     maxsize = len(repo) + 1
     err = b''
     if b'size' not in args or b'set' not in args:
         raise error.ParseError(_(b'invalid set of arguments'))
     minsize, maxsize = getintrange(
         args[b'size'],
         _(b'expectsize requires a size range or a positive integer'),
         _(b'size range bounds must be integers'),
         minsize,
         maxsize,
     )
     if minsize < 0 or maxsize < 0:
         raise error.ParseError(_(b'negative size'))
     rev = getset(repo, fullreposet(repo), args[b'set'], order=order)
     if minsize != maxsize and (len(rev) < minsize or len(rev) > maxsize):
         err = _(b'revset size mismatch. expected between %d and %d, got %d') % (
             minsize,
             maxsize,
             len(rev),
         )
     elif minsize == maxsize and len(rev) != minsize:
         err = _(b'revset size mismatch. expected %d, got %d') % (
             minsize,
             len(rev),
         )
     if err:
         raise error.RepoLookupError(err)
     if order == followorder:
         return subset & rev
     else:
         return rev & subset
 
 
 @predicate(b'extdata(source)', safe=False, weight=100)
 def extdata(repo, subset, x):
     """Changesets in the specified extdata source. (EXPERIMENTAL)"""
     # i18n: "extdata" is a keyword
     args = getargsdict(x, b'extdata', b'source')
     source = getstring(
         args.get(b'source'),
         # i18n: "extdata" is a keyword
         _(b'extdata takes at least 1 string argument'),
     )
     data = scmutil.extdatasource(repo, source)
     return subset & baseset(data)
 
 
 @predicate(b'extinct()', safe=True)
 def extinct(repo, subset, x):
     """Obsolete changesets with obsolete descendants only. (EXPERIMENTAL)"""
     # i18n: "extinct" is a keyword
     getargs(x, 0, 0, _(b"extinct takes no arguments"))
     extincts = obsmod.getrevs(repo, b'extinct')
     return subset & extincts
 
 
 @predicate(b'extra(label, [value])', safe=True)
 def extra(repo, subset, x):
     """Changesets with the given label in the extra metadata, with the given
     optional value.
 
     Pattern matching is supported for `value`. See
     :hg:`help revisions.patterns`.
     """
     args = getargsdict(x, b'extra', b'label value')
     if b'label' not in args:
         # i18n: "extra" is a keyword
         raise error.ParseError(_(b'extra takes at least 1 argument'))
     # i18n: "extra" is a keyword
     label = getstring(
         args[b'label'], _(b'first argument to extra must be a string')
     )
     value = None
 
     if b'value' in args:
         # i18n: "extra" is a keyword
         value = getstring(
             args[b'value'], _(b'second argument to extra must be a string')
         )
         kind, value, matcher = stringutil.stringmatcher(value)
 
     def _matchvalue(r):
         extra = repo[r].extra()
         return label in extra and (value is None or matcher(extra[label]))
 
     return subset.filter(
         lambda r: _matchvalue(r), condrepr=(b'<extra[%r] %r>', label, value)
     )
 
 
 @predicate(b'filelog(pattern)', safe=True)
 def filelog(repo, subset, x):
     """Changesets connected to the specified filelog.
 
     For performance reasons, visits only revisions mentioned in the file-level
     filelog, rather than filtering through all changesets (much faster, but
     doesn't include deletes or duplicate changes). For a slower, more accurate
     result, use ``file()``.
 
     The pattern without explicit kind like ``glob:`` is expected to be
     relative to the current directory and match against a file exactly
     for efficiency.
     """
 
     # i18n: "filelog" is a keyword
     pat = getstring(x, _(b"filelog requires a pattern"))
     s = set()
     cl = repo.changelog
 
     if not matchmod.patkind(pat):
         f = pathutil.canonpath(repo.root, repo.getcwd(), pat)
         files = [f]
     else:
         m = matchmod.match(repo.root, repo.getcwd(), [pat], ctx=repo[None])
         files = (f for f in repo[None] if m(f))
 
     for f in files:
         fl = repo.file(f)
         known = {}
         scanpos = 0
         for fr in list(fl):
             fn = fl.node(fr)
             if fn in known:
                 s.add(known[fn])
                 continue
 
             lr = fl.linkrev(fr)
             if lr in cl:
                 s.add(lr)
             elif scanpos is not None:
                 # lowest matching changeset is filtered, scan further
                 # ahead in changelog
                 start = max(lr, scanpos) + 1
                 scanpos = None
                 for r in cl.revs(start):
                     # minimize parsing of non-matching entries
                     if f in cl.revision(r) and f in cl.readfiles(r):
                         try:
                             # try to use manifest delta fastpath
                             n = repo[r].filenode(f)
                             if n not in known:
                                 if n == fn:
                                     s.add(r)
                                     scanpos = r
                                     break
                                 else:
                                     known[n] = r
                         except error.ManifestLookupError:
                             # deletion in changelog
                             continue
 
     return subset & s
 
 
 @predicate(b'first(set, [n])', safe=True, takeorder=True, weight=0)
 def first(repo, subset, x, order):
     """An alias for limit()."""
     return limit(repo, subset, x, order)
 
 
 def _follow(repo, subset, x, name, followfirst=False):
     args = getargsdict(x, name, b'file startrev')
     revs = None
     if b'startrev' in args:
         revs = getset(repo, fullreposet(repo), args[b'startrev'])
     if b'file' in args:
         x = getstring(args[b'file'], _(b"%s expected a pattern") % name)
         if revs is None:
             revs = [None]
         fctxs = []
         for r in revs:
             ctx = mctx = repo[r]
             if r is None:
                 ctx = repo[b'.']
             m = matchmod.match(
                 repo.root, repo.getcwd(), [x], ctx=mctx, default=b'path'
             )
             fctxs.extend(ctx[f].introfilectx() for f in ctx.manifest().walk(m))
         s = dagop.filerevancestors(fctxs, followfirst)
     else:
         if revs is None:
             revs = baseset([repo[b'.'].rev()])
         s = dagop.revancestors(repo, revs, followfirst)
 
     return subset & s
 
 
 @predicate(b'follow([file[, startrev]])', safe=True)
 def follow(repo, subset, x):
     """
     An alias for ``::.`` (ancestors of the working directory's first parent).
     If file pattern is specified, the histories of files matching given
     pattern in the revision given by startrev are followed, including copies.
     """
     return _follow(repo, subset, x, b'follow')
 
 
 @predicate(b'_followfirst', safe=True)
 def _followfirst(repo, subset, x):
     # ``followfirst([file[, startrev]])``
     # Like ``follow([file[, startrev]])`` but follows only the first parent
     # of every revisions or files revisions.
     return _follow(repo, subset, x, b'_followfirst', followfirst=True)
 
 
 @predicate(
     b'followlines(file, fromline:toline[, startrev=., descend=False])',
     safe=True,
 )
 def followlines(repo, subset, x):
     """Changesets modifying `file` in line range ('fromline', 'toline').
 
     Line range corresponds to 'file' content at 'startrev' and should hence be
     consistent with file size. If startrev is not specified, working directory's
     parent is used.
 
     By default, ancestors of 'startrev' are returned. If 'descend' is True,
     descendants of 'startrev' are returned though renames are (currently) not
     followed in this direction.
     """
     args = getargsdict(x, b'followlines', b'file *lines startrev descend')
     if len(args[b'lines']) != 1:
         raise error.ParseError(_(b"followlines requires a line range"))
 
     rev = b'.'
     if b'startrev' in args:
         revs = getset(repo, fullreposet(repo), args[b'startrev'])
         if len(revs) != 1:
             raise error.ParseError(
                 # i18n: "followlines" is a keyword
                 _(b"followlines expects exactly one revision")
             )
         rev = revs.last()
 
     pat = getstring(args[b'file'], _(b"followlines requires a pattern"))
     # i18n: "followlines" is a keyword
     msg = _(b"followlines expects exactly one file")
     fname = scmutil.parsefollowlinespattern(repo, rev, pat, msg)
     fromline, toline = util.processlinerange(
         *getintrange(
             args[b'lines'][0],
             # i18n: "followlines" is a keyword
             _(b"followlines expects a line number or a range"),
             _(b"line range bounds must be integers"),
         )
     )
 
     fctx = repo[rev].filectx(fname)
     descend = False
     if b'descend' in args:
         descend = getboolean(
             args[b'descend'],
             # i18n: "descend" is a keyword
             _(b"descend argument must be a boolean"),
         )
     if descend:
         rs = generatorset(
             (
                 c.rev()
                 for c, _linerange in dagop.blockdescendants(
                     fctx, fromline, toline
                 )
             ),
             iterasc=True,
         )
     else:
         rs = generatorset(
             (
                 c.rev()
                 for c, _linerange in dagop.blockancestors(
                     fctx, fromline, toline
                 )
             ),
             iterasc=False,
         )
     return subset & rs
 
 
 @predicate(b'nodefromfile(path)')
 def nodefromfile(repo, subset, x):
-    """
-    An alias for ``::.`` (ancestors of the working directory's first parent).
-    If file pattern is specified, the histories of files matching given
-    pattern in the revision given by startrev are followed, including copies.
+    """Read a list of nodes from the file at `path`.
+
+    This applies `id(LINE)` to each line of the file.
+
+    This is useful when the amount of nodes you need to specify gets too large
+    for the command line.
     """
     path = getstring(x, _(b"nodefromfile require a file path"))
     listed_rev = set()
     try:
         with pycompat.open(path, 'rb') as f:
             for line in f:
                 n = line.strip()
                 rn = _node(repo, n)
                 if rn is not None:
                     listed_rev.add(rn)
     except IOError as exc:
         m = _(b'cannot open nodes file "%s": %s')
         m %= (path, encoding.strtolocal(exc.strerror))
         raise error.Abort(m)
     return subset & baseset(listed_rev)
 
 
 @predicate(b'all()', safe=True)
 def getall(repo, subset, x):
     """All changesets, the same as ``0:tip``."""
     # i18n: "all" is a keyword
     getargs(x, 0, 0, _(b"all takes no arguments"))
     return subset & spanset(repo)  # drop "null" if any
 
 
 @predicate(b'grep(regex)', weight=10)
 def grep(repo, subset, x):
     """Like ``keyword(string)`` but accepts a regex. Use ``grep(r'...')``
     to ensure special escape characters are handled correctly. Unlike
     ``keyword(string)``, the match is case-sensitive.
     """
     try:
         # i18n: "grep" is a keyword
         gr = re.compile(getstring(x, _(b"grep requires a string")))
     except re.error as e:
         raise error.ParseError(
             _(b'invalid match pattern: %s') % stringutil.forcebytestr(e)
         )
 
     def matches(x):
         c = repo[x]
         for e in c.files() + [c.user(), c.description()]:
             if gr.search(e):
                 return True
         return False
 
     return subset.filter(matches, condrepr=(b'<grep %r>', gr.pattern))
 
 
 @predicate(b'_matchfiles', safe=True)
 def _matchfiles(repo, subset, x):
     # _matchfiles takes a revset list of prefixed arguments:
     #
     #   [p:foo, i:bar, x:baz]
     #
     # builds a match object from them and filters subset. Allowed
     # prefixes are 'p:' for regular patterns, 'i:' for include
     # patterns and 'x:' for exclude patterns. Use 'r:' prefix to pass
     # a revision identifier, or the empty string to reference the
     # working directory, from which the match object is
     # initialized. Use 'd:' to set the default matching mode, default
     # to 'glob'. At most one 'r:' and 'd:' argument can be passed.
 
     l = getargs(x, 1, -1, b"_matchfiles requires at least one argument")
     pats, inc, exc = [], [], []
     rev, default = None, None
     for arg in l:
         s = getstring(arg, b"_matchfiles requires string arguments")
         prefix, value = s[:2], s[2:]
         if prefix == b'p:':
             pats.append(value)
         elif prefix == b'i:':
             inc.append(value)
         elif prefix == b'x:':
             exc.append(value)
         elif prefix == b'r:':
             if rev is not None:
                 raise error.ParseError(
                     b'_matchfiles expected at most one revision'
                 )
             if value == b'':  # empty means working directory
                 rev = wdirrev
             else:
                 rev = value
         elif prefix == b'd:':
             if default is not None:
                 raise error.ParseError(
                     b'_matchfiles expected at most one default mode'
                 )
             default = value
         else:
             raise error.ParseError(b'invalid _matchfiles prefix: %s' % prefix)
     if not default:
         default = b'glob'
     hasset = any(matchmod.patkind(p) == b'set' for p in pats + inc + exc)
 
     mcache = [None]
 
     # This directly read the changelog data as creating changectx for all
     # revisions is quite expensive.
     getfiles = repo.changelog.readfiles
 
     def matches(x):
         if x == wdirrev:
             files = repo[x].files()
         else:
             files = getfiles(x)
 
         if not mcache[0] or (hasset and rev is None):
             r = x if rev is None else rev
             mcache[0] = matchmod.match(
                 repo.root,
                 repo.getcwd(),
                 pats,
                 include=inc,
                 exclude=exc,
                 ctx=repo[r],
                 default=default,
             )
         m = mcache[0]
 
         for f in files:
             if m(f):
                 return True
         return False
 
     return subset.filter(
         matches,
         condrepr=(
             b'<matchfiles patterns=%r, include=%r '
             b'exclude=%r, default=%r, rev=%r>',
             pats,
             inc,
             exc,
             default,
             rev,
         ),
     )
 
 
 @predicate(b'file(pattern)', safe=True, weight=10)
 def hasfile(repo, subset, x):
     """Changesets affecting files matched by pattern.
 
     For a faster but less accurate result, consider using ``filelog()``
     instead.
 
     This predicate uses ``glob:`` as the default kind of pattern.
     """
     # i18n: "file" is a keyword
     pat = getstring(x, _(b"file requires a pattern"))
     return _matchfiles(repo, subset, (b'string', b'p:' + pat))
 
 
 @predicate(b'head()', safe=True)
 def head(repo, subset, x):
     """Changeset is a named branch head."""
     # i18n: "head" is a keyword
     getargs(x, 0, 0, _(b"head takes no arguments"))
     hs = set()
     cl = repo.changelog
     for ls in repo.branchmap().iterheads():
         hs.update(cl.rev(h) for h in ls)
     return subset & baseset(hs)
 
 
 @predicate(b'heads(set)', safe=True, takeorder=True)
 def heads(repo, subset, x, order):
     """Members of set with no children in set."""
     # argument set should never define order
     if order == defineorder:
         order = followorder
     inputset = getset(repo, fullreposet(repo), x, order=order)
     wdirparents = None
     if wdirrev in inputset:
         # a bit slower, but not common so good enough for now
         wdirparents = [p.rev() for p in repo[None].parents()]
         inputset = set(inputset)
         inputset.discard(wdirrev)
     heads = repo.changelog.headrevs(inputset)
     if wdirparents is not None:
         heads.difference_update(wdirparents)
         heads.add(wdirrev)
     heads = baseset(heads)
     return subset & heads
 
 
 @predicate(b'hidden()', safe=True)
 def hidden(repo, subset, x):
     """Hidden changesets."""
     # i18n: "hidden" is a keyword
     getargs(x, 0, 0, _(b"hidden takes no arguments"))
     hiddenrevs = repoview.filterrevs(repo, b'visible')
     return subset & hiddenrevs
 
 
 @predicate(b'keyword(string)', safe=True, weight=10)
 def keyword(repo, subset, x):
     """Search commit message, user name, and names of changed files for
     string. The match is case-insensitive.
 
     For a regular expression or case sensitive search of these fields, use
     ``grep(regex)``.
     """
     # i18n: "keyword" is a keyword
     kw = encoding.lower(getstring(x, _(b"keyword requires a string")))
 
     def matches(r):
         c = repo[r]
         return any(
             kw in encoding.lower(t)
             for t in c.files() + [c.user(), c.description()]
         )
 
     return subset.filter(matches, condrepr=(b'<keyword %r>', kw))
 
 
 @predicate(b'limit(set[, n[, offset]])', safe=True, takeorder=True, weight=0)
 def limit(repo, subset, x, order):
     """First n members of set, defaulting to 1, starting from offset."""
     args = getargsdict(x, b'limit', b'set n offset')
     if b'set' not in args:
         # i18n: "limit" is a keyword
         raise error.ParseError(_(b"limit requires one to three arguments"))
     # i18n: "limit" is a keyword
     lim = getinteger(args.get(b'n'), _(b"limit expects a number"), default=1)
     if lim < 0:
         raise error.ParseError(_(b"negative number to select"))
     # i18n: "limit" is a keyword
     ofs = getinteger(
         args.get(b'offset'), _(b"limit expects a number"), default=0
     )
     if ofs < 0:
         raise error.ParseError(_(b"negative offset"))
     os = getset(repo, fullreposet(repo), args[b'set'])
     ls = os.slice(ofs, ofs + lim)
     if order == followorder and lim > 1:
         return subset & ls
     return ls & subset
 
 
 @predicate(b'last(set, [n])', safe=True, takeorder=True)
 def last(repo, subset, x, order):
     """Last n members of set, defaulting to 1."""
     # i18n: "last" is a keyword
     l = getargs(x, 1, 2, _(b"last requires one or two arguments"))
     lim = 1
     if len(l) == 2:
         # i18n: "last" is a keyword
         lim = getinteger(l[1], _(b"last expects a number"))
     if lim < 0:
         raise error.ParseError(_(b"negative number to select"))
     os = getset(repo, fullreposet(repo), l[0])
     os.reverse()
     ls = os.slice(0, lim)
     if order == followorder and lim > 1:
         return subset & ls
     ls.reverse()
     return ls & subset
 
 
 @predicate(b'max(set)', safe=True)
 def maxrev(repo, subset, x):
     """Changeset with highest revision number in set."""
     os = getset(repo, fullreposet(repo), x)
     try:
         m = os.max()
         if m in subset:
             return baseset([m], datarepr=(b'<max %r, %r>', subset, os))
     except ValueError:
         # os.max() throws a ValueError when the collection is empty.
         # Same as python's max().
         pass
     return baseset(datarepr=(b'<max %r, %r>', subset, os))
 
 
 @predicate(b'merge()', safe=True)
 def merge(repo, subset, x):
     """Changeset is a merge changeset."""
     # i18n: "merge" is a keyword
     getargs(x, 0, 0, _(b"merge takes no arguments"))
     cl = repo.changelog
 
     def ismerge(r):
         try:
             return cl.parentrevs(r)[1] != nullrev
         except error.WdirUnsupported:
             return bool(repo[r].p2())
 
     return subset.filter(ismerge, condrepr=b'<merge>')
 
 
 @predicate(b'branchpoint()', safe=True)
 def branchpoint(repo, subset, x):
     """Changesets with more than one child."""
     # i18n: "branchpoint" is a keyword
     getargs(x, 0, 0, _(b"branchpoint takes no arguments"))
     cl = repo.changelog
     if not subset:
         return baseset()
     # XXX this should be 'parentset.min()' assuming 'parentset' is a smartset
     # (and if it is not, it should.)
     baserev = min(subset)
     parentscount = [0] * (len(repo) - baserev)
     for r in cl.revs(start=baserev + 1):
         for p in cl.parentrevs(r):
             if p >= baserev:
                 parentscount[p - baserev] += 1
     return subset.filter(
         lambda r: parentscount[r - baserev] > 1, condrepr=b'<branchpoint>'
     )
 
 
 @predicate(b'min(set)', safe=True)
 def minrev(repo, subset, x):
     """Changeset with lowest revision number in set."""
     os = getset(repo, fullreposet(repo), x)
     try:
         m = os.min()
         if m in subset:
             return baseset([m], datarepr=(b'<min %r, %r>', subset, os))
     except ValueError:
         # os.min() throws a ValueError when the collection is empty.
         # Same as python's min().
         pass
     return baseset(datarepr=(b'<min %r, %r>', subset, os))
 
 
 @predicate(b'modifies(pattern)', safe=True, weight=30)
 def modifies(repo, subset, x):
     """Changesets modifying files matched by pattern.
 
     The pattern without explicit kind like ``glob:`` is expected to be
     relative to the current directory and match against a file or a
     directory.
     """
     # i18n: "modifies" is a keyword
     pat = getstring(x, _(b"modifies requires a pattern"))
     return checkstatus(repo, subset, pat, 'modified')
 
 
 @predicate(b'named(namespace)')
 def named(repo, subset, x):
     """The changesets in a given namespace.
 
     Pattern matching is supported for `namespace`. See
     :hg:`help revisions.patterns`.
     """
     # i18n: "named" is a keyword
     args = getargs(x, 1, 1, _(b'named requires a namespace argument'))
 
     ns = getstring(
         args[0],
         # i18n: "named" is a keyword
         _(b'the argument to named must be a string'),
     )
     kind, pattern, matcher = stringutil.stringmatcher(ns)
     namespaces = set()
     if kind == b'literal':
         if pattern not in repo.names:
             raise error.RepoLookupError(
                 _(b"namespace '%s' does not exist") % ns
             )
         namespaces.add(repo.names[pattern])
     else:
         for name, ns in repo.names.items():
             if matcher(name):
                 namespaces.add(ns)
 
     names = set()
     for ns in namespaces:
         for name in ns.listnames(repo):
             if name not in ns.deprecated:
                 names.update(repo[n].rev() for n in ns.nodes(repo, name))
 
     names -= {nullrev}
     return subset & names
 
 
 def _node(repo, n):
     """process a node input"""
     rn = None
     if len(n) == 2 * repo.nodeconstants.nodelen:
         try:
             rn = repo.changelog.rev(bin(n))
         except error.WdirUnsupported:
             rn = wdirrev
         except (binascii.Error, LookupError):
             rn = None
     else:
         try:
             pm = scmutil.resolvehexnodeidprefix(repo, n)
             if pm is not None:
                 rn = repo.changelog.rev(pm)
         except LookupError:
             pass
         except error.WdirUnsupported:
             rn = wdirrev
     return rn
 
 
 @predicate(b'id(string)', safe=True)
 def node_(repo, subset, x):
     """Revision non-ambiguously specified by the given hex string prefix."""
     # i18n: "id" is a keyword
     l = getargs(x, 1, 1, _(b"id requires one argument"))
     # i18n: "id" is a keyword
     n = getstring(l[0], _(b"id requires a string"))
     rn = _node(repo, n)
 
     if rn is None:
         return baseset()
     result = baseset([rn])
     return result & subset
 
 
 @predicate(b'none()', safe=True)
 def none(repo, subset, x):
     """No changesets."""
     # i18n: "none" is a keyword
     getargs(x, 0, 0, _(b"none takes no arguments"))
     return baseset()
 
 
 @predicate(b'obsolete()', safe=True)
 def obsolete(repo, subset, x):
     """Mutable changeset with a newer version. (EXPERIMENTAL)"""
     # i18n: "obsolete" is a keyword
     getargs(x, 0, 0, _(b"obsolete takes no arguments"))
     obsoletes = obsmod.getrevs(repo, b'obsolete')
     return subset & obsoletes
 
 
 @predicate(b'only(set, [set])', safe=True)
 def only(repo, subset, x):
     """Changesets that are ancestors of the first set that are not ancestors
     of any other head in the repo. If a second set is specified, the result
     is ancestors of the first set that are not ancestors of the second set
     (i.e. ::<set1> - ::<set2>).
     """
     cl = repo.changelog
     # i18n: "only" is a keyword
     args = getargs(x, 1, 2, _(b'only takes one or two arguments'))
     include = getset(repo, fullreposet(repo), args[0])
     if len(args) == 1:
         if not include:
             return baseset()
 
         descendants = set(dagop.revdescendants(repo, include, False))
         exclude = [
             rev
             for rev in cl.headrevs()
             if not rev in descendants and not rev in include
         ]
     else:
         exclude = getset(repo, fullreposet(repo), args[1])
 
     results = set(cl.findmissingrevs(common=exclude, heads=include))
     # XXX we should turn this into a baseset instead of a set, smartset may do
     # some optimizations from the fact this is a baseset.
     return subset & results
 
 
 @predicate(b'origin([set])', safe=True)
 def origin(repo, subset, x):
     """
     Changesets that were specified as a source for the grafts, transplants or
     rebases that created the given revisions.  Omitting the optional set is the
     same as passing all().  If a changeset created by these operations is itself
     specified as a source for one of these operations, only the source changeset
     for the first operation is selected.
     """
     if x is not None:
         dests = getset(repo, fullreposet(repo), x)
     else:
         dests = fullreposet(repo)
 
     def _firstsrc(rev):
         src = _getrevsource(repo, rev)
         if src is None:
             return None
 
         while True:
             prev = _getrevsource(repo, src)
 
             if prev is None:
                 return src
             src = prev
 
     o = {_firstsrc(r) for r in dests}
     o -= {None}
     # XXX we should turn this into a baseset instead of a set, smartset may do
     # some optimizations from the fact this is a baseset.
     return subset & o
 
 
 @predicate(b'outgoing([path])', safe=False, weight=10)
 def outgoing(repo, subset, x):
     """Changesets not found in the specified destination repository, or the
     default push location.
 
     If the location resolve to multiple repositories, the union of all
     outgoing changeset will be used.
     """
     # Avoid cycles.
     from . import (
         discovery,
         hg,
     )
 
     # i18n: "outgoing" is a keyword
     l = getargs(x, 0, 1, _(b"outgoing takes one or no arguments"))
     # i18n: "outgoing" is a keyword
     dest = (
         l and getstring(l[0], _(b"outgoing requires a repository path")) or b''
     )
     if dest:
         dests = [dest]
     else:
         dests = []
     missing = set()
     for path in urlutil.get_push_paths(repo, repo.ui, dests):
         dest = path.pushloc or path.loc
         branches = path.branch, []
 
         revs, checkout = hg.addbranchrevs(repo, repo, branches, [])
         if revs:
             revs = [repo.lookup(rev) for rev in revs]
         other = hg.peer(repo, {}, dest)
         try:
             with repo.ui.silent():
                 outgoing = discovery.findcommonoutgoing(
                     repo, other, onlyheads=revs
                 )
         finally:
             other.close()
         missing.update(outgoing.missing)
     cl = repo.changelog
     o = {cl.rev(r) for r in missing}
     return subset & o
 
 
 @predicate(b'p1([set])', safe=True)
 def p1(repo, subset, x):
     """First parent of changesets in set, or the working directory."""
     if x is None:
         p = repo[x].p1().rev()
         if p >= 0:
             return subset & baseset([p])
         return baseset()
 
     ps = set()
     cl = repo.changelog
     for r in getset(repo, fullreposet(repo), x):
         try:
             ps.add(cl.parentrevs(r)[0])
         except error.WdirUnsupported:
             ps.add(repo[r].p1().rev())
     ps -= {nullrev}
     # XXX we should turn this into a baseset instead of a set, smartset may do
     # some optimizations from the fact this is a baseset.
     return subset & ps
 
 
 @predicate(b'p2([set])', safe=True)
 def p2(repo, subset, x):
     """Second parent of changesets in set, or the working directory."""
     if x is None:
         ps = repo[x].parents()
         try:
             p = ps[1].rev()
             if p >= 0:
                 return subset & baseset([p])
             return baseset()
         except IndexError:
             return baseset()
 
     ps = set()
     cl = repo.changelog
     for r in getset(repo, fullreposet(repo), x):
         try:
             ps.add(cl.parentrevs(r)[1])
         except error.WdirUnsupported:
             parents = repo[r].parents()
             if len(parents) == 2:
                 ps.add(parents[1])
     ps -= {nullrev}
     # XXX we should turn this into a baseset instead of a set, smartset may do
     # some optimizations from the fact this is a baseset.
     return subset & ps
 
 
 def parentpost(repo, subset, x, order):
     return p1(repo, subset, x)
 
 
 @predicate(b'parents([set])', safe=True)
 def parents(repo, subset, x):
     """
     The set of all parents for all changesets in set, or the working directory.
     """
     if x is None:
         ps = {p.rev() for p in repo[x].parents()}
     else:
         ps = set()
         cl = repo.changelog
         up = ps.update
         parentrevs = cl.parentrevs
         for r in getset(repo, fullreposet(repo), x):
             try:
                 up(parentrevs(r))
             except error.WdirUnsupported:
                 up(p.rev() for p in repo[r].parents())
     ps -= {nullrev}
     return subset & ps
 
 
 def _phase(repo, subset, *targets):
     """helper to select all rev in <targets> phases"""
     return repo._phasecache.getrevset(repo, targets, subset)
 
 
 @predicate(b'_phase(idx)', safe=True)
 def phase(repo, subset, x):
     l = getargs(x, 1, 1, b"_phase requires one argument")
     target = getinteger(l[0], b"_phase expects a number")
     return _phase(repo, subset, target)
 
 
 @predicate(b'draft()', safe=True)
 def draft(repo, subset, x):
     """Changeset in draft phase."""
     # i18n: "draft" is a keyword
     getargs(x, 0, 0, _(b"draft takes no arguments"))
     target = phases.draft
     return _phase(repo, subset, target)
 
 
 @predicate(b'secret()', safe=True)
 def secret(repo, subset, x):
     """Changeset in secret phase."""
     # i18n: "secret" is a keyword
     getargs(x, 0, 0, _(b"secret takes no arguments"))
     target = phases.secret
     return _phase(repo, subset, target)
 
 
 @predicate(b'stack([revs])', safe=True)
 def stack(repo, subset, x):
     """Experimental revset for the stack of changesets or working directory
     parent. (EXPERIMENTAL)
     """
     if x is None:
         stacks = stackmod.getstack(repo)
     else:
         stacks = smartset.baseset([])
         for revision in getset(repo, fullreposet(repo), x):
             currentstack = stackmod.getstack(repo, revision)
             stacks = stacks + currentstack
 
     return subset & stacks
 
 
 def parentspec(repo, subset, x, n, order):
     """``set^0``
     The set.
     ``set^1`` (or ``set^``), ``set^2``
     First or second parent, respectively, of all changesets in set.
     """
     try:
         n = int(n[1])
         if n not in (0, 1, 2):
             raise ValueError
     except (TypeError, ValueError):
         raise error.ParseError(_(b"^ expects a number 0, 1, or 2"))
     ps = set()
     cl = repo.changelog
     for r in getset(repo, fullreposet(repo), x):
         if n == 0:
             ps.add(r)
         elif n == 1:
             try:
                 ps.add(cl.parentrevs(r)[0])
             except error.WdirUnsupported:
                 ps.add(repo[r].p1().rev())
         else:
             try:
                 parents = cl.parentrevs(r)
                 if parents[1] != nullrev:
                     ps.add(parents[1])
             except error.WdirUnsupported:
                 parents = repo[r].parents()
                 if len(parents) == 2:
                     ps.add(parents[1].rev())
     return subset & ps
 
 
 @predicate(b'present(set)', safe=True, takeorder=True)
 def present(repo, subset, x, order):
     """An empty set, if any revision in set isn't found; otherwise,
     all revisions in set.
 
     If any of specified revisions is not present in the local repository,
     the query is normally aborted. But this predicate allows the query
     to continue even in such cases.
     """
     try:
         return getset(repo, subset, x, order)
     except error.RepoLookupError:
         return baseset()
 
 
 # for internal use
 @predicate(b'_notpublic', safe=True)
 def _notpublic(repo, subset, x):
     getargs(x, 0, 0, b"_notpublic takes no arguments")
     return _phase(repo, subset, phases.draft, phases.secret)
 
 
 # for internal use
 @predicate(b'_phaseandancestors(phasename, set)', safe=True)
 def _phaseandancestors(repo, subset, x):
     # equivalent to (phasename() & ancestors(set)) but more efficient
     # phasename could be one of 'draft', 'secret', or '_notpublic'
     args = getargs(x, 2, 2, b"_phaseandancestors requires two arguments")
     phasename = getsymbol(args[0])
     s = getset(repo, fullreposet(repo), args[1])
 
     draft = phases.draft
     secret = phases.secret
     phasenamemap = {
         b'_notpublic': draft,
         b'draft': draft,  # follow secret's ancestors
         b'secret': secret,
     }
     if phasename not in phasenamemap:
         raise error.ParseError(b'%r is not a valid phasename' % phasename)
 
     minimalphase = phasenamemap[phasename]
     getphase = repo._phasecache.phase
 
     def cutfunc(rev):
         return getphase(repo, rev) < minimalphase
 
     revs = dagop.revancestors(repo, s, cutfunc=cutfunc)
 
     if phasename == b'draft':  # need to remove secret changesets
         revs = revs.filter(lambda r: getphase(repo, r) == draft)
     return subset & revs
 
 
 @predicate(b'public()', safe=True)
 def public(repo, subset, x):
     """Changeset in public phase."""
     # i18n: "public" is a keyword
     getargs(x, 0, 0, _(b"public takes no arguments"))
     return _phase(repo, subset, phases.public)
 
 
 @predicate(b'remote([id [,path]])', safe=False)
 def remote(repo, subset, x):
     """Local revision that corresponds to the given identifier in a
     remote repository, if present. Here, the '.' identifier is a
     synonym for the current local branch.
     """
 
     from . import hg  # avoid start-up nasties
 
     # i18n: "remote" is a keyword
     l = getargs(x, 0, 2, _(b"remote takes zero, one, or two arguments"))
 
     q = b'.'
     if len(l) > 0:
         # i18n: "remote" is a keyword
         q = getstring(l[0], _(b"remote requires a string id"))
     if q == b'.':
         q = repo[b'.'].branch()
 
     dest = b''
     if len(l) > 1:
         # i18n: "remote" is a keyword
         dest = getstring(l[1], _(b"remote requires a repository path"))
     if not dest:
         dest = b'default'
     dest, branches = urlutil.get_unique_pull_path(
         b'remote', repo, repo.ui, dest
     )
 
     other = hg.peer(repo, {}, dest)
     n = other.lookup(q)
     if n in repo:
         r = repo[n].rev()
         if r in subset:
             return baseset([r])
     return baseset()
 
 
 @predicate(b'removes(pattern)', safe=True, weight=30)
 def removes(repo, subset, x):
     """Changesets which remove files matching pattern.
 
     The pattern without explicit kind like ``glob:`` is expected to be
     relative to the current directory and match against a file or a
     directory.
     """
     # i18n: "removes" is a keyword
     pat = getstring(x, _(b"removes requires a pattern"))
     return checkstatus(repo, subset, pat, 'removed')
 
 
 @predicate(b'rev(number)', safe=True)
 def rev(repo, subset, x):
     """Revision with the given numeric identifier."""
     try:
         return _rev(repo, subset, x)
     except error.RepoLookupError:
         return baseset()
 
 
 @predicate(b'_rev(number)', safe=True)
 def _rev(repo, subset, x):
     # internal version of "rev(x)" that raise error if "x" is invalid
     # i18n: "rev" is a keyword
     l = getargs(x, 1, 1, _(b"rev requires one argument"))
     try:
         # i18n: "rev" is a keyword
         l = int(getstring(l[0], _(b"rev requires a number")))
     except (TypeError, ValueError):
         # i18n: "rev" is a keyword
         raise error.ParseError(_(b"rev expects a number"))
     if l not in _virtualrevs:
         try:
             repo.changelog.node(l)  # check that the rev exists
         except IndexError:
             raise error.RepoLookupError(_(b"unknown revision '%d'") % l)
     return subset & baseset([l])
 
 
 @predicate(b'revset(set)', safe=True, takeorder=True)
 def revsetpredicate(repo, subset, x, order):
     """Strictly interpret the content as a revset.
 
     The content of this special predicate will be strictly interpreted as a
     revset. For example, ``revset(id(0))`` will be interpreted as "id(0)"
     without possible ambiguity with a "id(0)" bookmark or tag.
     """
     return getset(repo, subset, x, order)
 
 
 @predicate(b'matching(revision [, field])', safe=True)
 def matching(repo, subset, x):
     """Changesets in which a given set of fields match the set of fields in the
     selected revision or set.
 
     To match more than one field pass the list of fields to match separated
     by spaces (e.g. ``author description``).
 
     Valid fields are most regular revision fields and some special fields.
 
     Regular revision fields are ``description``, ``author``, ``branch``,
     ``date``, ``files``, ``phase``, ``parents``, ``substate``, ``user``
     and ``diff``.
     Note that ``author`` and ``user`` are synonyms. ``diff`` refers to the
     contents of the revision. Two revisions matching their ``diff`` will
     also match their ``files``.
 
     Special fields are ``summary`` and ``metadata``:
     ``summary`` matches the first line of the description.
     ``metadata`` is equivalent to matching ``description user date``
     (i.e. it matches the main metadata fields).
 
     ``metadata`` is the default field which is used when no fields are
     specified. You can match more than one field at a time.
     """
     # i18n: "matching" is a keyword
     l = getargs(x, 1, 2, _(b"matching takes 1 or 2 arguments"))
 
     revs = getset(repo, fullreposet(repo), l[0])
 
     fieldlist = [b'metadata']
     if len(l) > 1:
         fieldlist = getstring(
             l[1],
             # i18n: "matching" is a keyword
             _(b"matching requires a string as its second argument"),
         ).split()
 
     # Make sure that there are no repeated fields,
     # expand the 'special' 'metadata' field type
     # and check the 'files' whenever we check the 'diff'
     fields = []
     for field in fieldlist:
         if field == b'metadata':
             fields += [b'user', b'description', b'date']
         elif field == b'diff':
             # a revision matching the diff must also match the files
             # since matching the diff is very costly, make sure to
             # also match the files first
             fields += [b'files', b'diff']
         else:
             if field == b'author':
                 field = b'user'
             fields.append(field)
     fields = set(fields)
     if b'summary' in fields and b'description' in fields:
         # If a revision matches its description it also matches its summary
         fields.discard(b'summary')
 
     # We may want to match more than one field
     # Not all fields take the same amount of time to be matched
     # Sort the selected fields in order of increasing matching cost
     fieldorder = [
         b'phase',
         b'parents',
         b'user',
         b'date',
         b'branch',
         b'summary',
         b'files',
         b'description',
         b'substate',
         b'diff',
     ]
 
     def fieldkeyfunc(f):
         try:
             return fieldorder.index(f)
         except ValueError:
             # assume an unknown field is very costly
             return len(fieldorder)
 
     fields = list(fields)
     fields.sort(key=fieldkeyfunc)
 
     # Each field will be matched with its own "getfield" function
     # which will be added to the getfieldfuncs array of functions
     getfieldfuncs = []
     _funcs = {
         b'user': lambda r: repo[r].user(),
         b'branch': lambda r: repo[r].branch(),
         b'date': lambda r: repo[r].date(),
         b'description': lambda r: repo[r].description(),
         b'files': lambda r: repo[r].files(),
         b'parents': lambda r: repo[r].parents(),
         b'phase': lambda r: repo[r].phase(),
         b'substate': lambda r: repo[r].substate,
         b'summary': lambda r: repo[r].description().splitlines()[0],
         b'diff': lambda r: list(
             repo[r].diff(opts=diffutil.diffallopts(repo.ui, {b'git': True}))
         ),
     }
     for info in fields:
         getfield = _funcs.get(info, None)
         if getfield is None:
             raise error.ParseError(
                 # i18n: "matching" is a keyword
                 _(b"unexpected field name passed to matching: %s")
                 % info
             )
         getfieldfuncs.append(getfield)
     # convert the getfield array of functions into a "getinfo" function
     # which returns an array of field values (or a single value if there
     # is only one field to match)
     getinfo = lambda r: [f(r) for f in getfieldfuncs]
 
     def matches(x):
         for rev in revs:
             target = getinfo(rev)
             match = True
             for n, f in enumerate(getfieldfuncs):
                 if target[n] != f(x):
                     match = False
             if match:
                 return True
         return False
 
     return subset.filter(matches, condrepr=(b'<matching%r %r>', fields, revs))
 
 
 @predicate(b'reverse(set)', safe=True, takeorder=True, weight=0)
 def reverse(repo, subset, x, order):
     """Reverse order of set."""
     l = getset(repo, subset, x, order)
     if order == defineorder:
         l.reverse()
     return l
 
 
 @predicate(b'roots(set)', safe=True)
 def roots(repo, subset, x):
     """Changesets in set with no parent changeset in set."""
     s = getset(repo, fullreposet(repo), x)
     parents = repo.changelog.parentrevs
 
     def filter(r):
         for p in parents(r):
             if 0 <= p and p in s:
                 return False
         return True
 
     return subset & s.filter(filter, condrepr=b'<roots>')
 
 
 _sortkeyfuncs = {
     b'rev': scmutil.intrev,
     b'branch': lambda c: c.branch(),
     b'desc': lambda c: c.description(),
     b'user': lambda c: c.user(),
     b'author': lambda c: c.user(),
     b'date': lambda c: c.date()[0],
     b'node': scmutil.binnode,
 }
 
 
 def _getsortargs(x):
     """Parse sort options into (set, [(key, reverse)], opts)"""
     args = getargsdict(x, b'sort', b'set keys topo.firstbranch')
     if b'set' not in args:
         # i18n: "sort" is a keyword
         raise error.ParseError(_(b'sort requires one or two arguments'))
     keys = b"rev"
     if b'keys' in args:
         # i18n: "sort" is a keyword
         keys = getstring(args[b'keys'], _(b"sort spec must be a string"))
 
     keyflags = []
     for k in keys.split():
         fk = k
         reverse = k.startswith(b'-')
         if reverse:
             k = k[1:]
         if k not in _sortkeyfuncs and k != b'topo':
             raise error.ParseError(
                 _(b"unknown sort key %r") % pycompat.bytestr(fk)
             )
         keyflags.append((k, reverse))
 
     if len(keyflags) > 1 and any(k == b'topo' for k, reverse in keyflags):
         # i18n: "topo" is a keyword
         raise error.ParseError(
             _(b'topo sort order cannot be combined with other sort keys')
         )
 
     opts = {}
     if b'topo.firstbranch' in args:
         if any(k == b'topo' for k, reverse in keyflags):
             opts[b'topo.firstbranch'] = args[b'topo.firstbranch']
         else:
             # i18n: "topo" and "topo.firstbranch" are keywords
             raise error.ParseError(
                 _(
                     b'topo.firstbranch can only be used '
                     b'when using the topo sort key'
                 )
             )
 
     return args[b'set'], keyflags, opts
 
 
 @predicate(
     b'sort(set[, [-]key... [, ...]])', safe=True, takeorder=True, weight=10
 )
 def sort(repo, subset, x, order):
     """Sort set by keys. The default sort order is ascending, specify a key
     as ``-key`` to sort in descending order.
 
     The keys can be:
 
     - ``rev`` for the revision number,
     - ``branch`` for the branch name,
     - ``desc`` for the commit message (description),
     - ``user`` for user name (``author`` can be used as an alias),
     - ``date`` for the commit date
     - ``topo`` for a reverse topographical sort
     - ``node`` the nodeid of the revision
 
     The ``topo`` sort order cannot be combined with other sort keys. This sort
     takes one optional argument, ``topo.firstbranch``, which takes a revset that
     specifies what topographical branches to prioritize in the sort.
 
     """
     s, keyflags, opts = _getsortargs(x)
     revs = getset(repo, subset, s, order)
 
     if not keyflags or order != defineorder:
         return revs
     if len(keyflags) == 1 and keyflags[0][0] == b"rev":
         revs.sort(reverse=keyflags[0][1])
         return revs
     elif keyflags[0][0] == b"topo":
         firstbranch = ()
         if b'topo.firstbranch' in opts:
             firstbranch = getset(repo, subset, opts[b'topo.firstbranch'])
         revs = baseset(
             dagop.toposort(revs, repo.changelog.parentrevs, firstbranch),
             istopo=True,
         )
         if keyflags[0][1]:
             revs.reverse()
         return revs
 
     # sort() is guaranteed to be stable
     ctxs = [repo[r] for r in revs]
     for k, reverse in reversed(keyflags):
         ctxs.sort(key=_sortkeyfuncs[k], reverse=reverse)
     return baseset([c.rev() for c in ctxs])
 
 
 @predicate(b'subrepo([pattern])')
 def subrepo(repo, subset, x):
     """Changesets that add, modify or remove the given subrepo.  If no subrepo
     pattern is named, any subrepo changes are returned.
     """
     # i18n: "subrepo" is a keyword
     args = getargs(x, 0, 1, _(b'subrepo takes at most one argument'))
     pat = None
     if len(args) != 0:
         pat = getstring(args[0], _(b"subrepo requires a pattern"))
 
     m = matchmod.exact([b'.hgsubstate'])
 
     def submatches(names):
         k, p, m = stringutil.stringmatcher(pat)
         for name in names:
             if m(name):
                 yield name
 
     def matches(x):
         c = repo[x]
         s = repo.status(c.p1().node(), c.node(), match=m)
 
         if pat is None:
             return s.added or s.modified or s.removed
 
         if s.added:
             return any(submatches(c.substate.keys()))
 
         if s.modified:
             subs = set(c.p1().substate.keys())
             subs.update(c.substate.keys())
 
             for path in submatches(subs):
                 if c.p1().substate.get(path) != c.substate.get(path):
                     return True
 
         if s.removed:
             return any(submatches(c.p1().substate.keys()))
 
         return False
 
     return subset.filter(matches, condrepr=(b'<subrepo %r>', pat))
 
 
 def _mapbynodefunc(repo, s, f):
     """(repo, smartset, [node] -> [node]) -> smartset
 
     Helper method to map a smartset to another smartset given a function only
     talking about nodes. Handles converting between rev numbers and nodes, and
     filtering.
     """
     cl = repo.unfiltered().changelog
     torev = cl.index.get_rev
     tonode = cl.node
     result = {torev(n) for n in f(tonode(r) for r in s)}
     result.discard(None)
     return smartset.baseset(result - repo.changelog.filteredrevs)
 
 
 @predicate(b'successors(set)', safe=True)
 def successors(repo, subset, x):
     """All successors for set, including the given set themselves.
     (EXPERIMENTAL)"""
     s = getset(repo, fullreposet(repo), x)
     f = lambda nodes: obsutil.allsuccessors(repo.obsstore, nodes)
     d = _mapbynodefunc(repo, s, f)
     return subset & d
 
 
 def _substringmatcher(pattern, casesensitive=True):
     kind, pattern, matcher = stringutil.stringmatcher(
         pattern, casesensitive=casesensitive
     )
     if kind == b'literal':
         if not casesensitive:
             pattern = encoding.lower(pattern)
             matcher = lambda s: pattern in encoding.lower(s)
         else:
             matcher = lambda s: pattern in s
     return kind, pattern, matcher
 
 
 @predicate(b'tag([name])', safe=True)
 def tag(repo, subset, x):
     """The specified tag by name, or all tagged revisions if no name is given.
 
     Pattern matching is supported for `name`. See
     :hg:`help revisions.patterns`.
     """
     # i18n: "tag" is a keyword
     args = getargs(x, 0, 1, _(b"tag takes one or no arguments"))
     cl = repo.changelog
     if args:
         pattern = getstring(
             args[0],
             # i18n: "tag" is a keyword
             _(b'the argument to tag must be a string'),
         )
         kind, pattern, matcher = stringutil.stringmatcher(pattern)
         if kind == b'literal':
             # avoid resolving all tags
             tn = repo._tagscache.tags.get(pattern, None)
             if tn is None:
                 raise error.RepoLookupError(
                     _(b"tag '%s' does not exist") % pattern
                 )
             s = {repo[tn].rev()}
         else:
             s = {cl.rev(n) for t, n in repo.tagslist() if matcher(t)}
     else:
         s = {cl.rev(n) for t, n in repo.tagslist() if t != b'tip'}
     return subset & s
 
 
 @predicate(b'tagged', safe=True)
 def tagged(repo, subset, x):
     return tag(repo, subset, x)
 
 
 @predicate(b'orphan()', safe=True)
 def orphan(repo, subset, x):
     """Non-obsolete changesets with obsolete ancestors. (EXPERIMENTAL)"""
     # i18n: "orphan" is a keyword
     getargs(x, 0, 0, _(b"orphan takes no arguments"))
     orphan = obsmod.getrevs(repo, b'orphan')
     return subset & orphan
 
 
 @predicate(b'unstable()', safe=True)
 def unstable(repo, subset, x):
     """Changesets with instabilities. (EXPERIMENTAL)"""
     # i18n: "unstable" is a keyword
     getargs(x, 0, 0, b'unstable takes no arguments')
     _unstable = set()
     _unstable.update(obsmod.getrevs(repo, b'orphan'))
     _unstable.update(obsmod.getrevs(repo, b'phasedivergent'))
     _unstable.update(obsmod.getrevs(repo, b'contentdivergent'))
     return subset & baseset(_unstable)
 
 
 @predicate(b'user(string)', safe=True, weight=10)
 def user(repo, subset, x):
     """User name contains string. The match is case-insensitive.
 
     Pattern matching is supported for `string`. See
     :hg:`help revisions.patterns`.
     """
     return author(repo, subset, x)
 
 
 @predicate(b'wdir()', safe=True, weight=0)
 def wdir(repo, subset, x):
     """Working directory. (EXPERIMENTAL)"""
     # i18n: "wdir" is a keyword
     getargs(x, 0, 0, _(b"wdir takes no arguments"))
     if wdirrev in subset or isinstance(subset, fullreposet):
         return baseset([wdirrev])
     return baseset()
 
 
 def _orderedlist(repo, subset, x):
     s = getstring(x, b"internal error")
     if not s:
         return baseset()
     # remove duplicates here. it's difficult for caller to deduplicate sets
     # because different symbols can point to the same rev.
     cl = repo.changelog
     ls = []
     seen = set()
     for t in s.split(b'\0'):
         try:
             # fast path for integer revision
             r = int(t)
             if (b'%d' % r) != t or r not in cl:
                 raise ValueError
             revs = [r]
         except ValueError:
             revs = stringset(repo, subset, t, defineorder)
 
         for r in revs:
             if r in seen:
                 continue
             if (
                 r in subset
                 or r in _virtualrevs
                 and isinstance(subset, fullreposet)
             ):
                 ls.append(r)
             seen.add(r)
     return baseset(ls)
 
 
 # for internal use
 @predicate(b'_list', safe=True, takeorder=True)
 def _list(repo, subset, x, order):
     if order == followorder:
         # slow path to take the subset order
         return subset & _orderedlist(repo, fullreposet(repo), x)
     else:
         return _orderedlist(repo, subset, x)
 
 
 def _orderedintlist(repo, subset, x):
     s = getstring(x, b"internal error")
     if not s:
         return baseset()
     ls = [int(r) for r in s.split(b'\0')]
     s = subset
     return baseset([r for r in ls if r in s])
 
 
 # for internal use
 @predicate(b'_intlist', safe=True, takeorder=True, weight=0)
 def _intlist(repo, subset, x, order):
     if order == followorder:
         # slow path to take the subset order
         return subset & _orderedintlist(repo, fullreposet(repo), x)
     else:
         return _orderedintlist(repo, subset, x)
 
 
 def _orderedhexlist(repo, subset, x):
     s = getstring(x, b"internal error")
     if not s:
         return baseset()
     cl = repo.changelog
     ls = [cl.rev(bin(r)) for r in s.split(b'\0')]
     s = subset
     return baseset([r for r in ls if r in s])
 
 
 # for internal use
 @predicate(b'_hexlist', safe=True, takeorder=True)
 def _hexlist(repo, subset, x, order):
     if order == followorder:
         # slow path to take the subset order
         return subset & _orderedhexlist(repo, fullreposet(repo), x)
     else:
         return _orderedhexlist(repo, subset, x)
 
 
 methods = {
     b"range": rangeset,
     b"rangeall": rangeall,
     b"rangepre": rangepre,
     b"rangepost": rangepost,
     b"dagrange": dagrange,
     b"string": stringset,
     b"symbol": stringset,
     b"and": andset,
     b"andsmally": andsmallyset,
     b"or": orset,
     b"not": notset,
     b"difference": differenceset,
     b"relation": relationset,
     b"relsubscript": relsubscriptset,
     b"subscript": subscriptset,
     b"list": listset,
     b"keyvalue": keyvaluepair,
     b"func": func,
     b"ancestor": ancestorspec,
     b"parent": parentspec,
     b"parentpost": parentpost,
     b"smartset": rawsmartset,
 }
 
 relations = {
     b"g": generationsrel,
     b"generations": generationsrel,
 }
 
 subscriptrelations = {
     b"g": generationssubrel,
     b"generations": generationssubrel,
 }
 
 
 def lookupfn(repo):
     def fn(symbol):
         try:
             return scmutil.isrevsymbol(repo, symbol)
         except error.AmbiguousPrefixLookupError:
             raise error.InputError(
                 b'ambiguous revision identifier: %s' % symbol
             )
 
     return fn
 
 
 def match(ui, spec, lookup=None):
     """Create a matcher for a single revision spec"""
     return matchany(ui, [spec], lookup=lookup)
 
 
 def matchany(ui, specs, lookup=None, localalias=None):
     """Create a matcher that will include any revisions matching one of the
     given specs
 
     If lookup function is not None, the parser will first attempt to handle
     old-style ranges, which may contain operator characters.
 
     If localalias is not None, it is a dict {name: definitionstring}. It takes
     precedence over [revsetalias] config section.
     """
     if not specs:
 
         def mfunc(repo, subset=None):
             return baseset()
 
         return mfunc
     if not all(specs):
         raise error.ParseError(_(b"empty query"))
     if len(specs) == 1:
         tree = revsetlang.parse(specs[0], lookup)
     else:
         tree = (
             b'or',
             (b'list',) + tuple(revsetlang.parse(s, lookup) for s in specs),
         )
 
     aliases = []
     warn = None
     if ui:
         aliases.extend(ui.configitems(b'revsetalias'))
         warn = ui.warn
     if localalias:
         aliases.extend(localalias.items())
     if aliases:
         tree = revsetlang.expandaliases(tree, aliases, warn=warn)
     tree = revsetlang.foldconcat(tree)
     tree = revsetlang.analyze(tree)
     tree = revsetlang.optimize(tree)
     return makematcher(tree)
 
 
 def makematcher(tree):
     """Create a matcher from an evaluatable tree"""
 
     def mfunc(repo, subset=None, order=None):
         if order is None:
             if subset is None:
                 order = defineorder  # 'x'
             else:
                 order = followorder  # 'subset & x'
         if subset is None:
             subset = fullreposet(repo)
         return getset(repo, subset, tree, order)
 
     return mfunc
 
 
 def loadpredicate(ui, extname, registrarobj):
     """Load revset predicates from specified registrarobj"""
     for name, func in registrarobj._table.items():
         symbols[name] = func
         if func._safe:
             safesymbols.add(name)
 
 
 # load built-in predicates explicitly to setup safesymbols
 loadpredicate(None, None, predicate)
 
 # tell hggettext to extract docstrings from these functions:
 i18nfunctions = symbols.values()

mercurial/url.py

@@ -1,723 +1,651 b''
 # url.py - HTTP handling for mercurial
 #
 # Copyright 2005, 2006, 2007, 2008 Olivia Mackall <[e-mail]>
 # Copyright 2006, 2007 Alexis S. L. Carvalho <[e-mail]>
 # Copyright 2006 Vadim Gelfer <[e-mail]>
 #
 # This software may be used and distributed according to the terms of the
 # GNU General Public License version 2 or any later version.
 
 
 import base64
 import socket
 
 from .i18n import _
-from .pycompat import getattr
 from . import (
     encoding,
     error,
     httpconnection as httpconnectionmod,
     keepalive,
     pycompat,
     sslutil,
     urllibcompat,
     util,
 )
 from .utils import (
     stringutil,
     urlutil,
 )
 
 httplib = util.httplib
 stringio = util.stringio
 urlerr = util.urlerr
 urlreq = util.urlreq
 
 
 def escape(s, quote=None):
     """Replace special characters "&", "<" and ">" to HTML-safe sequences.
     If the optional flag quote is true, the quotation mark character (")
     is also translated.
 
     This is the same as cgi.escape in Python, but always operates on
     bytes, whereas cgi.escape in Python 3 only works on unicodes.
     """
     s = s.replace(b"&", b"&amp;")
     s = s.replace(b"<", b"&lt;")
     s = s.replace(b">", b"&gt;")
     if quote:
         s = s.replace(b'"', b"&quot;")
     return s
 
 
 class passwordmgr:
     def __init__(self, ui, passwddb):
         self.ui = ui
         self.passwddb = passwddb
 
     def add_password(self, realm, uri, user, passwd):
         return self.passwddb.add_password(realm, uri, user, passwd)
 
     def find_user_password(self, realm, authuri):
         assert isinstance(realm, (type(None), str))
         assert isinstance(authuri, str)
         authinfo = self.passwddb.find_user_password(realm, authuri)
         user, passwd = authinfo
         user, passwd = pycompat.bytesurl(user), pycompat.bytesurl(passwd)
         if user and passwd:
             self._writedebug(user, passwd)
             return (pycompat.strurl(user), pycompat.strurl(passwd))
 
         if not user or not passwd:
             res = httpconnectionmod.readauthforuri(self.ui, authuri, user)
             if res:
                 group, auth = res
                 user, passwd = auth.get(b'username'), auth.get(b'password')
                 self.ui.debug(b"using auth.%s.* for authentication\n" % group)
         if not user or not passwd:
             u = urlutil.url(pycompat.bytesurl(authuri))
             u.query = None
             if not self.ui.interactive():
                 raise error.Abort(
                     _(b'http authorization required for %s')
                     % urlutil.hidepassword(bytes(u))
                 )
 
             self.ui.write(
                 _(b"http authorization required for %s\n")
                 % urlutil.hidepassword(bytes(u))
             )
             self.ui.write(_(b"realm: %s\n") % pycompat.bytesurl(realm))
             if user:
                 self.ui.write(_(b"user: %s\n") % user)
             else:
                 user = self.ui.prompt(_(b"user:"), default=None)
 
             if not passwd:
                 passwd = self.ui.getpass()
 
         # As of Python 3.8, the default implementation of
         # AbstractBasicAuthHandler.retry_http_basic_auth() assumes the user
         # is set if pw is not None. This means (None, str) is not a valid
         # return type of find_user_password().
         if user is None:
             return None, None
 
         self.passwddb.add_password(realm, authuri, user, passwd)
         self._writedebug(user, passwd)
         return (pycompat.strurl(user), pycompat.strurl(passwd))
 
     def _writedebug(self, user, passwd):
         msg = _(b'http auth: user %s, password %s\n')
         self.ui.debug(msg % (user, passwd and b'*' * len(passwd) or b'not set'))
 
     def find_stored_password(self, authuri):
         return self.passwddb.find_user_password(None, authuri)
 
 
 class proxyhandler(urlreq.proxyhandler):
     def __init__(self, ui):
         proxyurl = ui.config(b"http_proxy", b"host") or encoding.environ.get(
             b'http_proxy'
         )
         # XXX proxyauthinfo = None
 
         if proxyurl:
             # proxy can be proper url or host[:port]
             if not (
                 proxyurl.startswith(b'http:') or proxyurl.startswith(b'https:')
             ):
                 proxyurl = b'http://' + proxyurl + b'/'
             proxy = urlutil.url(proxyurl)
             if not proxy.user:
                 proxy.user = ui.config(b"http_proxy", b"user")
                 proxy.passwd = ui.config(b"http_proxy", b"passwd")
 
             # see if we should use a proxy for this url
             no_list = [b"localhost", b"127.0.0.1"]
             no_list.extend(
                 [p.lower() for p in ui.configlist(b"http_proxy", b"no")]
             )
             no_list.extend(
                 [
                     p.strip().lower()
                     for p in encoding.environ.get(b"no_proxy", b'').split(b',')
                     if p.strip()
                 ]
             )
             # "http_proxy.always" config is for running tests on localhost
             if ui.configbool(b"http_proxy", b"always"):
                 self.no_list = []
             else:
                 self.no_list = no_list
 
             # Keys and values need to be str because the standard library
             # expects them to be.
             proxyurl = str(proxy)
             proxies = {'http': proxyurl, 'https': proxyurl}
             ui.debug(
                 b'proxying through %s\n' % urlutil.hidepassword(bytes(proxy))
             )
         else:
             proxies = {}
 
         urlreq.proxyhandler.__init__(self, proxies)
         self.ui = ui
 
     def proxy_open(self, req, proxy, type_):
         host = pycompat.bytesurl(urllibcompat.gethost(req)).split(b':')[0]
         for e in self.no_list:
             if host == e:
                 return None
             if e.startswith(b'*.') and host.endswith(e[2:]):
                 return None
             if e.startswith(b'.') and host.endswith(e[1:]):
                 return None
 
         return urlreq.proxyhandler.proxy_open(self, req, proxy, type_)
 
 
 def _gen_sendfile(orgsend):
     def _sendfile(self, data):
         # send a file
         if isinstance(data, httpconnectionmod.httpsendfile):
             # if auth required, some data sent twice, so rewind here
             data.seek(0)
             for chunk in util.filechunkiter(data):
                 orgsend(self, chunk)
         else:
             orgsend(self, data)
 
     return _sendfile
 
 
 has_https = util.safehasattr(urlreq, b'httpshandler')
 
 
 class httpconnection(keepalive.HTTPConnection):
     # must be able to send big bundle as stream.
     send = _gen_sendfile(keepalive.HTTPConnection.send)
 
-    def getresponse(self):
-        proxyres = getattr(self, 'proxyres', None)
-        if proxyres:
-            if proxyres.will_close:
-                self.close()
-            self.proxyres = None
-            return proxyres
-        return keepalive.HTTPConnection.getresponse(self)
-
 
 # Large parts of this function have their origin from before Python 2.6
 # and could potentially be removed.
 def _generic_start_transaction(handler, h, req):
     tunnel_host = req._tunnel_host
     if tunnel_host:
         if tunnel_host[:7] not in ['http://', 'https:/']:
             tunnel_host = 'https://' + tunnel_host
         new_tunnel = True
     else:
         tunnel_host = urllibcompat.getselector(req)
         new_tunnel = False
 
     if new_tunnel or tunnel_host == urllibcompat.getfullurl(req):  # has proxy
         u = urlutil.url(pycompat.bytesurl(tunnel_host))
         if new_tunnel or u.scheme == b'https':  # only use CONNECT for HTTPS
             h.realhostport = b':'.join([u.host, (u.port or b'443')])
             h.headers = req.headers.copy()
             h.headers.update(handler.parent.addheaders)
             return
 
     h.realhostport = None
     h.headers = None
 
 
 def _generic_proxytunnel(self):
     proxyheaders = {
         pycompat.bytestr(x): pycompat.bytestr(self.headers[x])
         for x in self.headers
         if x.lower().startswith('proxy-')
     }
     self.send(b'CONNECT %s HTTP/1.0\r\n' % self.realhostport)
     for header in proxyheaders.items():
         self.send(b'%s: %s\r\n' % header)
     self.send(b'\r\n')
 
     # majority of the following code is duplicated from
     # httplib.HTTPConnection as there are no adequate places to
     # override functions to provide the needed functionality.
 
     res = self.response_class(self.sock, method=self._method)
 
     while True:
         version, status, reason = res._read_status()
         if status != httplib.CONTINUE:
             break
         # skip lines that are all whitespace
         list(iter(lambda: res.fp.readline().strip(), b''))
-    res.status = status
-    res.reason = reason.strip()
 
-    if res.status == 200:
+    if status == 200:
         # skip lines until we find a blank line
         list(iter(res.fp.readline, b'\r\n'))
-        return True
-
-    if version == b'HTTP/1.0':
-        res.version = 10
-    elif version.startswith(b'HTTP/1.'):
-        res.version = 11
-    elif version == b'HTTP/0.9':
-        res.version = 9
     else:
-        raise httplib.UnknownProtocol(version)
-
-    if res.version == 9:
-        res.length = None
-        res.chunked = 0
-        res.will_close = 1
-        res.msg = httplib.HTTPMessage(stringio())
-        return False
-
-    res.msg = httplib.HTTPMessage(res.fp)
-    res.msg.fp = None
-
-    # are we using the chunked-style of transfer encoding?
-    trenc = res.msg.getheader(b'transfer-encoding')
-    if trenc and trenc.lower() == b"chunked":
-        res.chunked = 1
-        res.chunk_left = None
-    else:
-        res.chunked = 0
-
-    # will the connection close at the end of the response?
-    res.will_close = res._check_close()
-
-    # do we have a Content-Length?
-    # NOTE: RFC 2616, section 4.4, #3 says we ignore this if
-    # transfer-encoding is "chunked"
-    length = res.msg.getheader(b'content-length')
-    if length and not res.chunked:
-        try:
-            res.length = int(length)
-        except ValueError:
-            res.length = None
-        else:
-            if res.length < 0:  # ignore nonsensical negative lengths
-                res.length = None
-    else:
-        res.length = None
-
-    # does the body have a fixed length? (of zero)
-    if (
-        status == httplib.NO_CONTENT
-        or status == httplib.NOT_MODIFIED
-        or 100 <= status < 200
-        or res._method == b'HEAD'  # 1xx codes
-    ):
-        res.length = 0
-
-    # if the connection remains open, and we aren't using chunked, and
-    # a content-length was not provided, then assume that the connection
-    # WILL close.
-    if not res.will_close and not res.chunked and res.length is None:
-        res.will_close = 1
-
-    self.proxyres = res
-
-    return False
+        self.close()
+        raise socket.error(
+            "Tunnel connection failed: %d %s" % (status, reason.strip())
+        )
 
 
 class httphandler(keepalive.HTTPHandler):
     def http_open(self, req):
         return self.do_open(httpconnection, req)
 
     def _start_transaction(self, h, req):
         _generic_start_transaction(self, h, req)
         return keepalive.HTTPHandler._start_transaction(self, h, req)
 
 
 class logginghttpconnection(keepalive.HTTPConnection):
     def __init__(self, createconn, *args, **kwargs):
         keepalive.HTTPConnection.__init__(self, *args, **kwargs)
         self._create_connection = createconn
 
 
 class logginghttphandler(httphandler):
     """HTTP handler that logs socket I/O."""
 
     def __init__(self, logfh, name, observeropts, timeout=None):
         super(logginghttphandler, self).__init__(timeout=timeout)
 
         self._logfh = logfh
         self._logname = name
         self._observeropts = observeropts
 
     # do_open() calls the passed class to instantiate an HTTPConnection. We
     # pass in a callable method that creates a custom HTTPConnection instance
     # whose callback to create the socket knows how to proxy the socket.
     def http_open(self, req):
         return self.do_open(self._makeconnection, req)
 
     def _makeconnection(self, *args, **kwargs):
         def createconnection(*args, **kwargs):
             sock = socket.create_connection(*args, **kwargs)
             return util.makeloggingsocket(
                 self._logfh, sock, self._logname, **self._observeropts
             )
 
         return logginghttpconnection(createconnection, *args, **kwargs)
 
 
 if has_https:
 
     class httpsconnection(keepalive.HTTPConnection):
         response_class = keepalive.HTTPResponse
         default_port = httplib.HTTPS_PORT
         # must be able to send big bundle as stream.
         send = _gen_sendfile(keepalive.safesend)
         getresponse = keepalive.wrapgetresponse(httplib.HTTPConnection)
 
         def __init__(
             self,
             host,
             port=None,
             key_file=None,
             cert_file=None,
             *args,
             **kwargs
         ):
             keepalive.HTTPConnection.__init__(self, host, port, *args, **kwargs)
             self.key_file = key_file
             self.cert_file = cert_file
 
         def connect(self):
             self.sock = socket.create_connection((self.host, self.port))
 
             host = self.host
             if self.realhostport:  # use CONNECT proxy
                 _generic_proxytunnel(self)
                 host = self.realhostport.rsplit(b':', 1)[0]
             self.sock = sslutil.wrapsocket(
                 self.sock,
                 self.key_file,
                 self.cert_file,
                 ui=self.ui,
                 serverhostname=host,
             )
             sslutil.validatesocket(self.sock)
 
     class httpshandler(keepalive.KeepAliveHandler, urlreq.httpshandler):
         def __init__(self, ui, timeout=None):
             keepalive.KeepAliveHandler.__init__(self, timeout=timeout)
             urlreq.httpshandler.__init__(self)
             self.ui = ui
             self.pwmgr = passwordmgr(self.ui, self.ui.httppasswordmgrdb)
 
         def _start_transaction(self, h, req):
             _generic_start_transaction(self, h, req)
             return keepalive.KeepAliveHandler._start_transaction(self, h, req)
 
         def https_open(self, req):
             # urllibcompat.getfullurl() does not contain credentials
             # and we may need them to match the certificates.
             url = urllibcompat.getfullurl(req)
             user, password = self.pwmgr.find_stored_password(url)
             res = httpconnectionmod.readauthforuri(self.ui, url, user)
             if res:
                 group, auth = res
                 self.auth = auth
                 self.ui.debug(b"using auth.%s.* for authentication\n" % group)
             else:
                 self.auth = None
             return self.do_open(self._makeconnection, req)
 
         def _makeconnection(self, host, port=None, *args, **kwargs):
             keyfile = None
             certfile = None
 
             if len(args) >= 1:  # key_file
                 keyfile = args[0]
             if len(args) >= 2:  # cert_file
                 certfile = args[1]
             args = args[2:]
 
             # if the user has specified different key/cert files in
             # hgrc, we prefer these
             if self.auth and b'key' in self.auth and b'cert' in self.auth:
                 keyfile = self.auth[b'key']
                 certfile = self.auth[b'cert']
 
             conn = httpsconnection(
                 host, port, keyfile, certfile, *args, **kwargs
             )
             conn.ui = self.ui
             return conn
 
 
 class httpdigestauthhandler(urlreq.httpdigestauthhandler):
     def __init__(self, *args, **kwargs):
         urlreq.httpdigestauthhandler.__init__(self, *args, **kwargs)
         self.retried_req = None
 
     def reset_retry_count(self):
         # Python 2.6.5 will call this on 401 or 407 errors and thus loop
         # forever. We disable reset_retry_count completely and reset in
         # http_error_auth_reqed instead.
         pass
 
     def http_error_auth_reqed(self, auth_header, host, req, headers):
         # Reset the retry counter once for each request.
         if req is not self.retried_req:
             self.retried_req = req
             self.retried = 0
         return urlreq.httpdigestauthhandler.http_error_auth_reqed(
             self, auth_header, host, req, headers
         )
 
 
 class httpbasicauthhandler(urlreq.httpbasicauthhandler):
     def __init__(self, *args, **kwargs):
         self.auth = None
         urlreq.httpbasicauthhandler.__init__(self, *args, **kwargs)
         self.retried_req = None
 
     def http_request(self, request):
         if self.auth:
             request.add_unredirected_header(self.auth_header, self.auth)
 
         return request
 
     def https_request(self, request):
         if self.auth:
             request.add_unredirected_header(self.auth_header, self.auth)
 
         return request
 
     def reset_retry_count(self):
         # Python 2.6.5 will call this on 401 or 407 errors and thus loop
         # forever. We disable reset_retry_count completely and reset in
         # http_error_auth_reqed instead.
         pass
 
     def http_error_auth_reqed(self, auth_header, host, req, headers):
         # Reset the retry counter once for each request.
         if req is not self.retried_req:
             self.retried_req = req
             self.retried = 0
         return urlreq.httpbasicauthhandler.http_error_auth_reqed(
             self, auth_header, host, req, headers
         )
 
     def retry_http_basic_auth(self, host, req, realm):
         user, pw = self.passwd.find_user_password(
             realm, urllibcompat.getfullurl(req)
         )
         if pw is not None:
             raw = b"%s:%s" % (pycompat.bytesurl(user), pycompat.bytesurl(pw))
             auth = 'Basic %s' % pycompat.strurl(base64.b64encode(raw).strip())
             if req.get_header(self.auth_header, None) == auth:
                 return None
             self.auth = auth
             req.add_unredirected_header(self.auth_header, auth)
             return self.parent.open(req)
         else:
             return None
 
 
 class cookiehandler(urlreq.basehandler):
     def __init__(self, ui):
         self.cookiejar = None
 
         cookiefile = ui.config(b'auth', b'cookiefile')
         if not cookiefile:
             return
 
         cookiefile = util.expandpath(cookiefile)
         try:
             cookiejar = util.cookielib.MozillaCookieJar(
                 pycompat.fsdecode(cookiefile)
             )
             cookiejar.load()
             self.cookiejar = cookiejar
         except util.cookielib.LoadError as e:
             ui.warn(
                 _(
                     b'(error loading cookie file %s: %s; continuing without '
                     b'cookies)\n'
                 )
                 % (cookiefile, stringutil.forcebytestr(e))
             )
 
     def http_request(self, request):
         if self.cookiejar:
             self.cookiejar.add_cookie_header(request)
 
         return request
 
     def https_request(self, request):
         if self.cookiejar:
             self.cookiejar.add_cookie_header(request)
 
         return request
 
 
 handlerfuncs = []
 
 
 def opener(
     ui,
     authinfo=None,
     useragent=None,
     loggingfh=None,
     loggingname=b's',
     loggingopts=None,
     sendaccept=True,
 ):
     """
     construct an opener suitable for urllib2
     authinfo will be added to the password manager
 
     The opener can be configured to log socket events if the various
     ``logging*`` arguments are specified.
 
     ``loggingfh`` denotes a file object to log events to.
     ``loggingname`` denotes the name of the to print when logging.
     ``loggingopts`` is a dict of keyword arguments to pass to the constructed
     ``util.socketobserver`` instance.
 
     ``sendaccept`` allows controlling whether the ``Accept`` request header
     is sent. The header is sent by default.
     """
     timeout = ui.configwith(float, b'http', b'timeout')
     handlers = []
 
     if loggingfh:
         handlers.append(
             logginghttphandler(
                 loggingfh, loggingname, loggingopts or {}, timeout=timeout
             )
         )
         # We don't yet support HTTPS when logging I/O. If we attempt to open
         # an HTTPS URL, we'll likely fail due to unknown protocol.
 
     else:
         handlers.append(httphandler(timeout=timeout))
         if has_https:
             handlers.append(httpshandler(ui, timeout=timeout))
 
     handlers.append(proxyhandler(ui))
 
     passmgr = passwordmgr(ui, ui.httppasswordmgrdb)
     if authinfo is not None:
         realm, uris, user, passwd = authinfo
         saveduser, savedpass = passmgr.find_stored_password(uris[0])
         if user != saveduser or passwd:
             passmgr.add_password(realm, uris, user, passwd)
         ui.debug(
             b'http auth: user %s, password %s\n'
             % (user, passwd and b'*' * len(passwd) or b'not set')
         )
 
     handlers.extend(
         (httpbasicauthhandler(passmgr), httpdigestauthhandler(passmgr))
     )
     handlers.extend([h(ui, passmgr) for h in handlerfuncs])
     handlers.append(cookiehandler(ui))
     opener = urlreq.buildopener(*handlers)
 
     # keepalive.py's handlers will populate these attributes if they exist.
     opener.requestscount = 0
     opener.sentbytescount = 0
     opener.receivedbytescount = 0
 
     # The user agent should should *NOT* be used by servers for e.g.
     # protocol detection or feature negotiation: there are other
     # facilities for that.
     #
     # "mercurial/proto-1.0" was the original user agent string and
     # exists for backwards compatibility reasons.
     #
     # The "(Mercurial %s)" string contains the distribution
     # name and version. Other client implementations should choose their
     # own distribution name. Since servers should not be using the user
     # agent string for anything, clients should be able to define whatever
     # user agent they deem appropriate.
     #
     # The custom user agent is for lfs, because unfortunately some servers
     # do look at this value.
     if not useragent:
         agent = b'mercurial/proto-1.0 (Mercurial %s)' % util.version()
         opener.addheaders = [('User-agent', pycompat.sysstr(agent))]
     else:
         opener.addheaders = [('User-agent', pycompat.sysstr(useragent))]
 
     # This header should only be needed by wire protocol requests. But it has
     # been sent on all requests since forever. We keep sending it for backwards
     # compatibility reasons. Modern versions of the wire protocol use
     # X-HgProto-<N> for advertising client support.
     if sendaccept:
         opener.addheaders.append(('Accept', 'application/mercurial-0.1'))
 
     return opener
 
 
 def open(ui, url_, data=None, sendaccept=True):
     u = urlutil.url(url_)
     if u.scheme:
         u.scheme = u.scheme.lower()
         url_, authinfo = u.authinfo()
     else:
         path = util.normpath(util.abspath(url_))
         url_ = b'file://' + pycompat.bytesurl(
             urlreq.pathname2url(pycompat.fsdecode(path))
         )
         authinfo = None
     return opener(ui, authinfo, sendaccept=sendaccept).open(
         pycompat.strurl(url_), data
     )
 
 
 def wrapresponse(resp):
     """Wrap a response object with common error handlers.
 
     This ensures that any I/O from any consumer raises the appropriate
     error and messaging.
     """
     origread = resp.read
 
     class readerproxy(resp.__class__):
         def read(self, size=None):
             try:
                 return origread(size)
             except httplib.IncompleteRead as e:
                 # e.expected is an integer if length known or None otherwise.
                 if e.expected:
                     got = len(e.partial)
                     total = e.expected + got
                     msg = _(
                         b'HTTP request error (incomplete response; '
                         b'expected %d bytes got %d)'
                     ) % (total, got)
                 else:
                     msg = _(b'HTTP request error (incomplete response)')
 
                 raise error.PeerTransportError(
                     msg,
                     hint=_(
                         b'this may be an intermittent network failure; '
                         b'if the error persists, consider contacting the '
                         b'network or server operator'
                     ),
                 )
             except httplib.HTTPException as e:
                 raise error.PeerTransportError(
                     _(b'HTTP request error (%s)') % e,
                     hint=_(
                         b'this may be an intermittent network failure; '
                         b'if the error persists, consider contacting the '
                         b'network or server operator'
                     ),
                 )
 
     resp.__class__ = readerproxy

rust/hg-core/Cargo.toml

@@ -1,48 +1,48 b''
 [package]
 name = "hg-core"
 version = "0.1.0"
 authors = ["Georges Racinet <[e-mail]>"]
 description = "Mercurial pure Rust core library, with no assumption on Python bindings (FFI)"
 edition = "2018"
 
 [lib]
 name = "hg"
 
 [dependencies]
 bitflags = "1.3.2"
 bytes-cast = "0.2.0"
 byteorder = "1.4.3"
 derive_more = "0.99.17"
 hashbrown = { version = "0.9.1", features = ["rayon"] }
 home = "0.5.3"
-im-rc = "15.0.0"
+im-rc = "15.0"
 itertools = "0.10.3"
 lazy_static = "1.4.0"
 libc = "0.2"
 ouroboros = "0.15.0"
 rand = "0.8.4"
 rand_pcg = "0.3.1"
 rand_distr = "0.4.3"
 rayon = "1.5.1"
 regex = "1.5.5"
 sha-1 = "0.10.0"
 twox-hash = "1.6.2"
 same-file = "1.0.6"
 tempfile = "3.1.0"
 crossbeam-channel = "0.4"
 micro-timer = "0.3.0"
 log = "0.4.8"
 memmap2 = {version = "0.4", features = ["stable_deref_trait"]}
 zstd = "0.5.3"
 format-bytes = "0.3.0"
 
 # We don't use the `miniz-oxide` backend to not change rhg benchmarks and until
 # we have a clearer view of which backend is the fastest.
 [dependencies.flate2]
 version = "1.0.22"
 features = ["zlib"]
 default-features = false
 
 [dev-dependencies]
 clap = "2.34.0"
 pretty_assertions = "1.1.0"

tests/hghave.py

@@ -1,1170 +1,1170 b''
 import distutils.version
 import os
 import re
 import socket
 import stat
 import subprocess
 import sys
 import tempfile
 
 tempprefix = 'hg-hghave-'
 
 checks = {
     "true": (lambda: True, "yak shaving"),
     "false": (lambda: False, "nail clipper"),
     "known-bad-output": (lambda: True, "use for currently known bad output"),
     "missing-correct-output": (lambda: False, "use for missing good output"),
 }
 
 try:
     import msvcrt
 
     msvcrt.setmode(sys.stdout.fileno(), os.O_BINARY)
     msvcrt.setmode(sys.stderr.fileno(), os.O_BINARY)
 except ImportError:
     pass
 
 stdout = getattr(sys.stdout, 'buffer', sys.stdout)
 stderr = getattr(sys.stderr, 'buffer', sys.stderr)
 
 is_not_python2 = sys.version_info[0] >= 3
 if is_not_python2:
 
     def _sys2bytes(p):
         if p is None:
             return p
         return p.encode('utf-8')
 
     def _bytes2sys(p):
         if p is None:
             return p
         return p.decode('utf-8')
 
 
 else:
 
     def _sys2bytes(p):
         return p
 
     _bytes2sys = _sys2bytes
 
 
 def check(name, desc):
     """Registers a check function for a feature."""
 
     def decorator(func):
         checks[name] = (func, desc)
         return func
 
     return decorator
 
 
 def checkvers(name, desc, vers):
     """Registers a check function for each of a series of versions.
 
     vers can be a list or an iterator.
 
     Produces a series of feature checks that have the form <name><vers> without
     any punctuation (even if there's punctuation in 'vers'; i.e. this produces
     'py38', not 'py3.8' or 'py-38')."""
 
     def decorator(func):
         def funcv(v):
             def f():
                 return func(v)
 
             return f
 
         for v in vers:
             v = str(v)
             f = funcv(v)
             checks['%s%s' % (name, v.replace('.', ''))] = (f, desc % v)
         return func
 
     return decorator
 
 
 def checkfeatures(features):
     result = {
         'error': [],
         'missing': [],
         'skipped': [],
     }
 
     for feature in features:
         negate = feature.startswith('no-')
         if negate:
             feature = feature[3:]
 
         if feature not in checks:
             result['missing'].append(feature)
             continue
 
         check, desc = checks[feature]
         try:
             available = check()
         except Exception as e:
             result['error'].append('hghave check %s failed: %r' % (feature, e))
             continue
 
         if not negate and not available:
             result['skipped'].append('missing feature: %s' % desc)
         elif negate and available:
             result['skipped'].append('system supports %s' % desc)
 
     return result
 
 
 def require(features):
     """Require that features are available, exiting if not."""
     result = checkfeatures(features)
 
     for missing in result['missing']:
         stderr.write(
             ('skipped: unknown feature: %s\n' % missing).encode('utf-8')
         )
     for msg in result['skipped']:
         stderr.write(('skipped: %s\n' % msg).encode('utf-8'))
     for msg in result['error']:
         stderr.write(('%s\n' % msg).encode('utf-8'))
 
     if result['missing']:
         sys.exit(2)
 
     if result['skipped'] or result['error']:
         sys.exit(1)
 
 
 def matchoutput(cmd, regexp, ignorestatus=False):
     """Return the match object if cmd executes successfully and its output
     is matched by the supplied regular expression.
     """
 
     # Tests on Windows have to fake USERPROFILE to point to the test area so
     # that `~` is properly expanded on py3.8+.  However, some tools like black
     # make calls that need the real USERPROFILE in order to run `foo --version`.
     env = os.environ
     if os.name == 'nt':
         env = os.environ.copy()
         env['USERPROFILE'] = env['REALUSERPROFILE']
 
     r = re.compile(regexp)
     p = subprocess.Popen(
         cmd,
         shell=True,
         stdout=subprocess.PIPE,
         stderr=subprocess.STDOUT,
         env=env,
     )
     s = p.communicate()[0]
     ret = p.returncode
     return (ignorestatus or not ret) and r.search(s)
 
 
 @check("baz", "GNU Arch baz client")
 def has_baz():
     return matchoutput('baz --version 2>&1', br'baz Bazaar version')
 
 
 @check("bzr", "Breezy library and executable version >= 3.1")
 def has_bzr():
     if not is_not_python2:
         return False
     try:
         # Test the Breezy python lib
         import breezy
         import breezy.bzr.bzrdir
         import breezy.errors
         import breezy.revision
         import breezy.revisionspec
 
         breezy.revisionspec.RevisionSpec
         if breezy.__doc__ is None or breezy.version_info[:2] < (3, 1):
             return False
     except (AttributeError, ImportError):
         return False
     # Test the executable
     return matchoutput('brz --version 2>&1', br'Breezy \(brz\) ')
 
 
 @check("chg", "running with chg")
 def has_chg():
     return 'CHG_INSTALLED_AS_HG' in os.environ
 
 
 @check("rhg", "running with rhg as 'hg'")
 def has_rhg():
     return 'RHG_INSTALLED_AS_HG' in os.environ
 
 
 @check("pyoxidizer", "running with pyoxidizer build as 'hg'")
 def has_rhg():
     return 'PYOXIDIZED_INSTALLED_AS_HG' in os.environ
 
 
 @check("cvs", "cvs client/server")
 def has_cvs():
     re = br'Concurrent Versions System.*?server'
     return matchoutput('cvs --version 2>&1', re) and not has_msys()
 
 
 @check("cvs112", "cvs client/server 1.12.* (not cvsnt)")
 def has_cvs112():
     re = br'Concurrent Versions System \(CVS\) 1.12.*?server'
     return matchoutput('cvs --version 2>&1', re) and not has_msys()
 
 
 @check("cvsnt", "cvsnt client/server")
 def has_cvsnt():
     re = br'Concurrent Versions System \(CVSNT\) (\d+).(\d+).*\(client/server\)'
     return matchoutput('cvsnt --version 2>&1', re)
 
 
 @check("darcs", "darcs client")
 def has_darcs():
     return matchoutput('darcs --version', br'\b2\.([2-9]|\d{2})', True)
 
 
 @check("mtn", "monotone client (>= 1.0)")
 def has_mtn():
     return matchoutput('mtn --version', br'monotone', True) and not matchoutput(
         'mtn --version', br'monotone 0\.', True
     )
 
 
 @check("eol-in-paths", "end-of-lines in paths")
 def has_eol_in_paths():
     try:
         fd, path = tempfile.mkstemp(dir='.', prefix=tempprefix, suffix='\n\r')
         os.close(fd)
         os.remove(path)
         return True
     except (IOError, OSError):
         return False
 
 
 @check("execbit", "executable bit")
 def has_executablebit():
     try:
         EXECFLAGS = stat.S_IXUSR | stat.S_IXGRP | stat.S_IXOTH
         fh, fn = tempfile.mkstemp(dir='.', prefix=tempprefix)
         try:
             os.close(fh)
             m = os.stat(fn).st_mode & 0o777
             new_file_has_exec = m & EXECFLAGS
             os.chmod(fn, m ^ EXECFLAGS)
             exec_flags_cannot_flip = (os.stat(fn).st_mode & 0o777) == m
         finally:
             os.unlink(fn)
     except (IOError, OSError):
         # we don't care, the user probably won't be able to commit anyway
         return False
     return not (new_file_has_exec or exec_flags_cannot_flip)
 
 
 @check("suidbit", "setuid and setgid bit")
 def has_suidbit():
     if (
         getattr(os, "statvfs", None) is None
         or getattr(os, "ST_NOSUID", None) is None
     ):
         return False
     return bool(os.statvfs('.').f_flag & os.ST_NOSUID)
 
 
 @check("icasefs", "case insensitive file system")
 def has_icasefs():
     # Stolen from mercurial.util
     fd, path = tempfile.mkstemp(dir='.', prefix=tempprefix)
     os.close(fd)
     try:
         s1 = os.stat(path)
         d, b = os.path.split(path)
         p2 = os.path.join(d, b.upper())
         if path == p2:
             p2 = os.path.join(d, b.lower())
         try:
             s2 = os.stat(p2)
             return s2 == s1
         except OSError:
             return False
     finally:
         os.remove(path)
 
 
 @check("fifo", "named pipes")
 def has_fifo():
     if getattr(os, "mkfifo", None) is None:
         return False
     name = tempfile.mktemp(dir='.', prefix=tempprefix)
     try:
         os.mkfifo(name)
         os.unlink(name)
         return True
     except OSError:
         return False
 
 
 @check("killdaemons", 'killdaemons.py support')
 def has_killdaemons():
     return True
 
 
 @check("cacheable", "cacheable filesystem")
 def has_cacheable_fs():
     from mercurial import util
 
     fd, path = tempfile.mkstemp(dir='.', prefix=tempprefix)
     os.close(fd)
     try:
         return util.cachestat(path).cacheable()
     finally:
         os.remove(path)
 
 
 @check("lsprof", "python lsprof module")
 def has_lsprof():
     try:
         import _lsprof
 
         _lsprof.Profiler  # silence unused import warning
         return True
     except ImportError:
         return False
 
 
 def _gethgversion():
     m = matchoutput('hg --version --quiet 2>&1', br'(\d+)\.(\d+)')
     if not m:
         return (0, 0)
     return (int(m.group(1)), int(m.group(2)))
 
 
 _hgversion = None
 
 
 def gethgversion():
     global _hgversion
     if _hgversion is None:
         _hgversion = _gethgversion()
     return _hgversion
 
 
 @checkvers(
     "hg", "Mercurial >= %s", list([(1.0 * x) / 10 for x in range(9, 99)])
 )
 def has_hg_range(v):
     major, minor = v.split('.')[0:2]
     return gethgversion() >= (int(major), int(minor))
 
 
 @check("rust", "Using the Rust extensions")
 def has_rust():
     """Check is the mercurial currently running is using some rust code"""
     cmd = 'hg debuginstall --quiet 2>&1'
     match = br'checking module policy \(([^)]+)\)'
     policy = matchoutput(cmd, match)
     if not policy:
         return False
     return b'rust' in policy.group(1)
 
 
 @check("hg08", "Mercurial >= 0.8")
 def has_hg08():
     if checks["hg09"][0]():
         return True
     return matchoutput('hg help annotate 2>&1', '--date')
 
 
 @check("hg07", "Mercurial >= 0.7")
 def has_hg07():
     if checks["hg08"][0]():
         return True
     return matchoutput('hg --version --quiet 2>&1', 'Mercurial Distributed SCM')
 
 
 @check("hg06", "Mercurial >= 0.6")
 def has_hg06():
     if checks["hg07"][0]():
         return True
     return matchoutput('hg --version --quiet 2>&1', 'Mercurial version')
 
 
 @check("gettext", "GNU Gettext (msgfmt)")
 def has_gettext():
     return matchoutput('msgfmt --version', br'GNU gettext-tools')
 
 
 @check("git", "git command line client")
 def has_git():
     return matchoutput('git --version 2>&1', br'^git version')
 
 
 def getgitversion():
     m = matchoutput('git --version 2>&1', br'git version (\d+)\.(\d+)')
     if not m:
         return (0, 0)
     return (int(m.group(1)), int(m.group(2)))
 
 
 @check("pygit2", "pygit2 Python library")
 def has_git():
     try:
         import pygit2
 
         pygit2.Oid  # silence unused import
         return True
     except ImportError:
         return False
 
 
 # https://github.com/git-lfs/lfs-test-server
 @check("lfs-test-server", "git-lfs test server")
 def has_lfsserver():
     exe = 'lfs-test-server'
     if has_windows():
         exe = 'lfs-test-server.exe'
     return any(
         os.access(os.path.join(path, exe), os.X_OK)
         for path in os.environ["PATH"].split(os.pathsep)
     )
 
 
 @checkvers("git", "git client (with ext::sh support) version >= %s", (1.9,))
 def has_git_range(v):
     major, minor = v.split('.')[0:2]
     return getgitversion() >= (int(major), int(minor))
 
 
 @check("docutils", "Docutils text processing library")
 def has_docutils():
     try:
         import docutils.core
 
         docutils.core.publish_cmdline  # silence unused import
         return True
     except ImportError:
         return False
 
 
 def getsvnversion():
     m = matchoutput('svn --version --quiet 2>&1', br'^(\d+)\.(\d+)')
     if not m:
         return (0, 0)
     return (int(m.group(1)), int(m.group(2)))
 
 
 @checkvers("svn", "subversion client and admin tools >= %s", (1.3, 1.5))
 def has_svn_range(v):
     major, minor = v.split('.')[0:2]
     return getsvnversion() >= (int(major), int(minor))
 
 
 @check("svn", "subversion client and admin tools")
 def has_svn():
     return matchoutput('svn --version 2>&1', br'^svn, version') and matchoutput(
         'svnadmin --version 2>&1', br'^svnadmin, version'
     )
 
 
 @check("svn-bindings", "subversion python bindings")
 def has_svn_bindings():
     try:
         import svn.core
 
         version = svn.core.SVN_VER_MAJOR, svn.core.SVN_VER_MINOR
         if version < (1, 4):
             return False
         return True
     except ImportError:
         return False
 
 
 @check("p4", "Perforce server and client")
 def has_p4():
     return matchoutput('p4 -V', br'Rev\. P4/') and matchoutput(
         'p4d -V', br'Rev\. P4D/'
     )
 
 
 @check("symlink", "symbolic links")
 def has_symlink():
     # mercurial.windows.checklink() is a hard 'no' at the moment
     if os.name == 'nt' or getattr(os, "symlink", None) is None:
         return False
     name = tempfile.mktemp(dir='.', prefix=tempprefix)
     try:
         os.symlink(".", name)
         os.unlink(name)
         return True
     except (OSError, AttributeError):
         return False
 
 
 @check("hardlink", "hardlinks")
 def has_hardlink():
     from mercurial import util
 
     fh, fn = tempfile.mkstemp(dir='.', prefix=tempprefix)
     os.close(fh)
     name = tempfile.mktemp(dir='.', prefix=tempprefix)
     try:
         util.oslink(_sys2bytes(fn), _sys2bytes(name))
         os.unlink(name)
         return True
     except OSError:
         return False
     finally:
         os.unlink(fn)
 
 
 @check("hardlink-whitelisted", "hardlinks on whitelisted filesystems")
 def has_hardlink_whitelisted():
     from mercurial import util
 
     try:
         fstype = util.getfstype(b'.')
     except OSError:
         return False
     return fstype in util._hardlinkfswhitelist
 
 
 @check("rmcwd", "can remove current working directory")
 def has_rmcwd():
     ocwd = os.getcwd()
     temp = tempfile.mkdtemp(dir='.', prefix=tempprefix)
     try:
         os.chdir(temp)
         # On Linux, 'rmdir .' isn't allowed, but the other names are okay.
         # On Solaris and Windows, the cwd can't be removed by any names.
         os.rmdir(os.getcwd())
         return True
     except OSError:
         return False
     finally:
         os.chdir(ocwd)
         # clean up temp dir on platforms where cwd can't be removed
         try:
             os.rmdir(temp)
         except OSError:
             pass
 
 
 @check("tla", "GNU Arch tla client")
 def has_tla():
     return matchoutput('tla --version 2>&1', br'The GNU Arch Revision')
 
 
 @check("gpg", "gpg client")
 def has_gpg():
     return matchoutput('gpg --version 2>&1', br'GnuPG')
 
 
 @check("gpg2", "gpg client v2")
 def has_gpg2():
     return matchoutput('gpg --version 2>&1', br'GnuPG[^0-9]+2\.')
 
 
 @check("gpg21", "gpg client v2.1+")
 def has_gpg21():
     return matchoutput('gpg --version 2>&1', br'GnuPG[^0-9]+2\.(?!0)')
 
 
 @check("unix-permissions", "unix-style permissions")
 def has_unix_permissions():
     d = tempfile.mkdtemp(dir='.', prefix=tempprefix)
     try:
         fname = os.path.join(d, 'foo')
         for umask in (0o77, 0o07, 0o22):
             os.umask(umask)
             f = open(fname, 'w')
             f.close()
             mode = os.stat(fname).st_mode
             os.unlink(fname)
             if mode & 0o777 != ~umask & 0o666:
                 return False
         return True
     finally:
         os.rmdir(d)
 
 
 @check("unix-socket", "AF_UNIX socket family")
 def has_unix_socket():
     return getattr(socket, 'AF_UNIX', None) is not None
 
 
 @check("root", "root permissions")
 def has_root():
     return getattr(os, 'geteuid', None) and os.geteuid() == 0
 
 
 @check("pyflakes", "Pyflakes python linter")
 def has_pyflakes():
     try:
         import pyflakes
 
         pyflakes.__version__
     except ImportError:
         return False
     else:
         return True
 
 
 @check("pylint", "Pylint python linter")
 def has_pylint():
     return matchoutput("pylint --help", br"Usage:[ ]+pylint", True)
 
 
 @check("clang-format", "clang-format C code formatter (>= 11)")
 def has_clang_format():
     m = matchoutput('clang-format --version', br'clang-format version (\d+)')
     # style changed somewhere between 10.x and 11.x
     if m:
         return int(m.group(1)) >= 11
     # Assist Googler contributors, they have a centrally-maintained version of
     # clang-format that is generally very fresh, but unlike most builds (both
     # official and unofficial), it does *not* include a version number.
     return matchoutput(
         'clang-format --version', br'clang-format .*google3-trunk \([0-9a-f]+\)'
     )
 
 
 @check("jshint", "JSHint static code analysis tool")
 def has_jshint():
     return matchoutput("jshint --version 2>&1", br"jshint v")
 
 
 @check("pygments", "Pygments source highlighting library")
 def has_pygments():
     try:
         import pygments
 
         pygments.highlight  # silence unused import warning
         return True
     except ImportError:
         return False
 
 
 @check("pygments25", "Pygments version >= 2.5")
 def pygments25():
     try:
         import pygments
 
         v = pygments.__version__
     except ImportError:
         return False
 
     parts = v.split(".")
     major = int(parts[0])
     minor = int(parts[1])
 
     return (major, minor) >= (2, 5)
 
 
 @check("pygments211", "Pygments version >= 2.11")
 def pygments211():
     try:
         import pygments
 
         v = pygments.__version__
     except ImportError:
         return False
 
     parts = v.split(".")
     major = int(parts[0])
     minor = int(parts[1])
 
     return (major, minor) >= (2, 11)
 
 
 @check("outer-repo", "outer repo")
 def has_outer_repo():
     # failing for other reasons than 'no repo' imply that there is a repo
     return not matchoutput('hg root 2>&1', br'abort: no repository found', True)
 
 
 @check("ssl", "ssl module available")
 def has_ssl():
     try:
         import ssl
 
         ssl.CERT_NONE
         return True
     except ImportError:
         return False
 
 
 @check("defaultcacertsloaded", "detected presence of loaded system CA certs")
 def has_defaultcacertsloaded():
     import ssl
     from mercurial import sslutil, ui as uimod
 
     ui = uimod.ui.load()
     cafile = sslutil._defaultcacerts(ui)
     ctx = ssl.create_default_context()
     if cafile:
         ctx.load_verify_locations(cafile=cafile)
     else:
         ctx.load_default_certs()
 
     return len(ctx.get_ca_certs()) > 0
 
 
 @check("tls1.2", "TLS 1.2 protocol support")
 def has_tls1_2():
     from mercurial import sslutil
 
     return b'tls1.2' in sslutil.supportedprotocols
 
 
 @check("windows", "Windows")
 def has_windows():
     return os.name == 'nt'
 
 
 @check("system-sh", "system() uses sh")
 def has_system_sh():
     return os.name != 'nt'
 
 
 @check("serve", "platform and python can manage 'hg serve -d'")
 def has_serve():
     return True
 
 
 @check("setprocname", "whether osutil.setprocname is available or not")
 def has_setprocname():
     try:
         from mercurial.utils import procutil
 
         procutil.setprocname
         return True
     except AttributeError:
         return False
 
 
 @check("test-repo", "running tests from repository")
 def has_test_repo():
     t = os.environ["TESTDIR"]
     return os.path.isdir(os.path.join(t, "..", ".hg"))
 
 
 @check("network-io", "whether tests are allowed to access 3rd party services")
 def has_test_repo():
     t = os.environ.get("HGTESTS_ALLOW_NETIO")
     return t == "1"
 
 
 @check("curses", "terminfo compiler and curses module")
 def has_curses():
     try:
         import curses
 
         curses.COLOR_BLUE
 
         # Windows doesn't have a `tic` executable, but the windows_curses
         # package is sufficient to run the tests without it.
         if os.name == 'nt':
             return True
 
         return has_tic()
 
     except (ImportError, AttributeError):
         return False
 
 
 @check("tic", "terminfo compiler")
 def has_tic():
     return matchoutput('test -x "`which tic`"', br'')
 
 
 @check("xz", "xz compression utility")
 def has_xz():
     # When Windows invokes a subprocess in shell mode, it uses `cmd.exe`, which
     # only knows `where`, not `which`.  So invoke MSYS shell explicitly.
     return matchoutput("sh -c 'test -x \"`which xz`\"'", b'')
 
 
 @check("msys", "Windows with MSYS")
 def has_msys():
     return os.getenv('MSYSTEM')
 
 
 @check("aix", "AIX")
 def has_aix():
     return sys.platform.startswith("aix")
 
 
 @check("osx", "OS X")
 def has_osx():
     return sys.platform == 'darwin'
 
 
 @check("osxpackaging", "OS X packaging tools")
 def has_osxpackaging():
     try:
         return (
             matchoutput('pkgbuild', br'Usage: pkgbuild ', ignorestatus=1)
             and matchoutput(
                 'productbuild', br'Usage: productbuild ', ignorestatus=1
             )
             and matchoutput('lsbom', br'Usage: lsbom', ignorestatus=1)
             and matchoutput('xar --help', br'Usage: xar', ignorestatus=1)
         )
     except ImportError:
         return False
 
 
 @check('linuxormacos', 'Linux or MacOS')
 def has_linuxormacos():
     # This isn't a perfect test for MacOS. But it is sufficient for our needs.
     return sys.platform.startswith(('linux', 'darwin'))
 
 
 @check("docker", "docker support")
 def has_docker():
     pat = br'A self-sufficient runtime for'
     if matchoutput('docker --help', pat):
         if 'linux' not in sys.platform:
             # TODO: in theory we should be able to test docker-based
             # package creation on non-linux using boot2docker, but in
             # practice that requires extra coordination to make sure
             # $TESTTEMP is going to be visible at the same path to the
             # boot2docker VM. If we figure out how to verify that, we
             # can use the following instead of just saying False:
             # return 'DOCKER_HOST' in os.environ
             return False
 
         return True
     return False
 
 
 @check("debhelper", "debian packaging tools")
 def has_debhelper():
     # Some versions of dpkg say `dpkg', some say 'dpkg' (` vs ' on the first
     # quote), so just accept anything in that spot.
     dpkg = matchoutput(
         'dpkg --version', br"Debian .dpkg' package management program"
     )
     dh = matchoutput(
         'dh --help', br'dh is a part of debhelper.', ignorestatus=True
     )
     dh_py2 = matchoutput(
         'dh_python2 --help', br'other supported Python versions'
     )
     # debuild comes from the 'devscripts' package, though you might want
     # the 'build-debs' package instead, which has a dependency on devscripts.
     debuild = matchoutput(
         'debuild --help', br'to run debian/rules with given parameter'
     )
     return dpkg and dh and dh_py2 and debuild
 
 
 @check(
     "debdeps", "debian build dependencies (run dpkg-checkbuilddeps in contrib/)"
 )
 def has_debdeps():
     # just check exit status (ignoring output)
     path = '%s/../contrib/packaging/debian/control' % os.environ['TESTDIR']
     return matchoutput('dpkg-checkbuilddeps %s' % path, br'')
 
 
 @check("demandimport", "demandimport enabled")
 def has_demandimport():
     # chg disables demandimport intentionally for performance wins.
     return (not has_chg()) and os.environ.get('HGDEMANDIMPORT') != 'disable'
 
 
 # Add "py27", "py35", ... as possible feature checks. Note that there's no
 # punctuation here.
 @checkvers("py", "Python >= %s", (2.7, 3.5, 3.6, 3.7, 3.8, 3.9))
 def has_python_range(v):
     major, minor = v.split('.')[0:2]
     py_major, py_minor = sys.version_info.major, sys.version_info.minor
 
     return (py_major, py_minor) >= (int(major), int(minor))
 
 
 @check("py3", "running with Python 3.x")
 def has_py3():
     return 3 == sys.version_info[0]
 
 
 @check("py3exe", "a Python 3.x interpreter is available")
 def has_python3exe():
     py = 'python3'
     if os.name == 'nt':
         py = 'py -3'
     return matchoutput('%s -V' % py, br'^Python 3.(5|6|7|8|9)')
 
 
 @check("pure", "running with pure Python code")
 def has_pure():
     return any(
         [
             os.environ.get("HGMODULEPOLICY") == "py",
             os.environ.get("HGTEST_RUN_TESTS_PURE") == "--pure",
         ]
     )
 
 
 @check("slow", "allow slow tests (use --allow-slow-tests)")
 def has_slow():
     return os.environ.get('HGTEST_SLOW') == 'slow'
 
 
 @check("hypothesis", "Hypothesis automated test generation")
 def has_hypothesis():
     try:
         import hypothesis
 
         hypothesis.given
         return True
     except ImportError:
         return False
 
 
 @check("unziplinks", "unzip(1) understands and extracts symlinks")
 def unzip_understands_symlinks():
     return matchoutput('unzip --help', br'Info-ZIP')
 
 
 @check("zstd", "zstd Python module available")
 def has_zstd():
     try:
         import mercurial.zstd
 
         mercurial.zstd.__version__
         return True
     except ImportError:
         return False
 
 
 @check("devfull", "/dev/full special file")
 def has_dev_full():
     return os.path.exists('/dev/full')
 
 
 @check("ensurepip", "ensurepip module")
 def has_ensurepip():
     try:
         import ensurepip
 
         ensurepip.bootstrap
         return True
     except ImportError:
         return False
 
 
 @check("virtualenv", "virtualenv support")
 def has_virtualenv():
     try:
         import virtualenv
 
         # --no-site-package became the default in 1.7 (Nov 2011), and the
         # argument was removed in 20.0 (Feb 2020).  Rather than make the
         # script complicated, just ignore ancient versions.
         return int(virtualenv.__version__.split('.')[0]) > 1
     except (AttributeError, ImportError, IndexError):
         return False
 
 
 @check("fsmonitor", "running tests with fsmonitor")
 def has_fsmonitor():
     return 'HGFSMONITOR_TESTS' in os.environ
 
 
 @check("fuzzywuzzy", "Fuzzy string matching library")
 def has_fuzzywuzzy():
     try:
         import fuzzywuzzy
 
         fuzzywuzzy.__version__
         return True
     except ImportError:
         return False
 
 
 @check("clang-libfuzzer", "clang new enough to include libfuzzer")
 def has_clang_libfuzzer():
     mat = matchoutput('clang --version', br'clang version (\d)')
     if mat:
         # libfuzzer is new in clang 6
         return int(mat.group(1)) > 5
     return False
 
 
 @check("clang-6.0", "clang 6.0 with version suffix (libfuzzer included)")
 def has_clang60():
     return matchoutput('clang-6.0 --version', br'clang version 6\.')
 
 
 @check("xdiff", "xdiff algorithm")
 def has_xdiff():
     try:
         from mercurial import policy
 
         bdiff = policy.importmod('bdiff')
         return bdiff.xdiffblocks(b'', b'') == [(0, 0, 0, 0)]
     except (ImportError, AttributeError):
         return False
 
 
 @check('extraextensions', 'whether tests are running with extra extensions')
 def has_extraextensions():
     return 'HGTESTEXTRAEXTENSIONS' in os.environ
 
 
 def getrepofeatures():
     """Obtain set of repository features in use.
 
     HGREPOFEATURES can be used to define or remove features. It contains
     a space-delimited list of feature strings. Strings beginning with ``-``
     mean to remove.
     """
     # Default list provided by core.
     features = {
         'bundlerepo',
         'revlogstore',
         'fncache',
     }
 
     # Features that imply other features.
     implies = {
         'simplestore': ['-revlogstore', '-bundlerepo', '-fncache'],
     }
 
     for override in os.environ.get('HGREPOFEATURES', '').split(' '):
         if not override:
             continue
 
         if override.startswith('-'):
             if override[1:] in features:
                 features.remove(override[1:])
         else:
             features.add(override)
 
             for imply in implies.get(override, []):
                 if imply.startswith('-'):
                     if imply[1:] in features:
                         features.remove(imply[1:])
                 else:
                     features.add(imply)
 
     return features
 
 
 @check('reporevlogstore', 'repository using the default revlog store')
 def has_reporevlogstore():
     return 'revlogstore' in getrepofeatures()
 
 
 @check('reposimplestore', 'repository using simple storage extension')
 def has_reposimplestore():
     return 'simplestore' in getrepofeatures()
 
 
 @check('repobundlerepo', 'whether we can open bundle files as repos')
 def has_repobundlerepo():
     return 'bundlerepo' in getrepofeatures()
 
 
 @check('repofncache', 'repository has an fncache')
 def has_repofncache():
     return 'fncache' in getrepofeatures()
 
 
 @check('dirstate-v2', 'using the v2 format of .hg/dirstate')
 def has_dirstate_v2():
     # Keep this logic in sync with `newreporequirements()` in `mercurial/localrepo.py`
     return has_rust() and matchoutput(
         'hg config format.exp-rc-dirstate-v2', b'(?i)1|yes|true|on|always'
     )
 
 
 @check('sqlite', 'sqlite3 module and matching cli is available')
 def has_sqlite():
     try:
         import sqlite3
 
         version = sqlite3.sqlite_version_info
     except ImportError:
         return False
 
     if version < (3, 8, 3):
         # WITH clause not supported
         return False
 
     return matchoutput('sqlite3 -version', br'^3\.\d+')
 
 
 @check('vcr', 'vcr http mocking library (pytest-vcr)')
 def has_vcr():
     try:
         import vcr
 
         vcr.VCR
         return True
     except (ImportError, AttributeError):
         pass
     return False
 
 
 @check('emacs', 'GNU Emacs')
 def has_emacs():
     # Our emacs lisp uses `with-eval-after-load` which is new in emacs
     # 24.4, so we allow emacs 24.4, 24.5, and 25+ (24.5 was the last
     # 24 release)
     return matchoutput('emacs --version', b'GNU Emacs 2(4.4|4.5|5|6|7|8|9)')
 
 
 @check('black', 'the black formatter for python (>= 20.8b1)')
 def has_black():
     blackcmd = 'black --version'
-    version_regex = b'black, version ([0-9a-b.]+)'
+    version_regex = b'black, (?:version )?([0-9a-b.]+)'
     version = matchoutput(blackcmd, version_regex)
     sv = distutils.version.StrictVersion
     return version and sv(_bytes2sys(version.group(1))) >= sv('20.8b1')
 
 
 @check('pytype', 'the pytype type checker')
 def has_pytype():
     pytypecmd = 'pytype --version'
     version = matchoutput(pytypecmd, b'[0-9a-b.]+')
     sv = distutils.version.StrictVersion
     return version and sv(_bytes2sys(version.group(0))) >= sv('2019.10.17')
 
 
 @check("rustfmt", "rustfmt tool at version nightly-2021-11-02")
 def has_rustfmt():
     # We use Nightly's rustfmt due to current unstable config options.
     return matchoutput(
         '`rustup which --toolchain nightly-2021-11-02 rustfmt` --version',
         b'rustfmt',
     )
 
 
 @check("cargo", "cargo tool")
 def has_cargo():
     return matchoutput('`rustup which cargo` --version', b'cargo')
 
 
 @check("lzma", "python lzma module")
 def has_lzma():
     try:
         import _lzma
 
         _lzma.FORMAT_XZ
         return True
     except ImportError:
         return False
 
 
 @check("bash", "bash shell")
 def has_bash():
     return matchoutput("bash -c 'echo hi'", b'^hi$')
 
 
 @check("bigendian", "big-endian CPU")
 def has_bigendian():
     return sys.byteorder == 'big'

tests/test-http-proxy.t

@@ -1,122 +1,131 b''
 #require serve
 
   $ hg init a
   $ cd a
   $ echo a > a
   $ hg ci -Ama -d '1123456789 0'
   adding a
   $ hg serve --config server.uncompressed=True -p $HGPORT -d --pid-file=hg.pid
   $ cat hg.pid >> $DAEMON_PIDS
   $ cd ..
   $ tinyproxy.py $HGPORT1 localhost 2>proxy.log >/dev/null </dev/null &
   $ while [ ! -f proxy.pid ]; do sleep 0; done
   $ cat proxy.pid >> $DAEMON_PIDS
 
 url for proxy, stream
 
   $ http_proxy=http://localhost:$HGPORT1/ hg --config http_proxy.always=True clone --stream http://localhost:$HGPORT/ b
   streaming all changes
   6 files to transfer, 412 bytes of data (reporevlogstore !)
   4 files to transfer, 330 bytes of data (reposimplestore !)
   transferred * bytes in * seconds (*/sec) (glob)
   updating to branch default
   1 files updated, 0 files merged, 0 files removed, 0 files unresolved
   $ cd b
   $ hg verify
   checking changesets
   checking manifests
   crosschecking files in changesets and manifests
   checking files
   checked 1 changesets with 1 changes to 1 files
   $ cd ..
 
 url for proxy, pull
 
   $ http_proxy=http://localhost:$HGPORT1/ hg --config http_proxy.always=True clone http://localhost:$HGPORT/ b-pull
   requesting all changes
   adding changesets
   adding manifests
   adding file changes
   added 1 changesets with 1 changes to 1 files
   new changesets 83180e7845de
   updating to branch default
   1 files updated, 0 files merged, 0 files removed, 0 files unresolved
   $ cd b-pull
   $ hg verify
   checking changesets
   checking manifests
   crosschecking files in changesets and manifests
   checking files
   checked 1 changesets with 1 changes to 1 files
   $ cd ..
 
 host:port for proxy
 
   $ http_proxy=localhost:$HGPORT1 hg clone --config http_proxy.always=True http://localhost:$HGPORT/ c
   requesting all changes
   adding changesets
   adding manifests
   adding file changes
   added 1 changesets with 1 changes to 1 files
   new changesets 83180e7845de
   updating to branch default
   1 files updated, 0 files merged, 0 files removed, 0 files unresolved
 
 proxy url with user name and password
 
   $ http_proxy=http://user:passwd@localhost:$HGPORT1 hg clone --config http_proxy.always=True http://localhost:$HGPORT/ d
   requesting all changes
   adding changesets
   adding manifests
   adding file changes
   added 1 changesets with 1 changes to 1 files
   new changesets 83180e7845de
   updating to branch default
   1 files updated, 0 files merged, 0 files removed, 0 files unresolved
 
 url with user name and password
 
   $ http_proxy=http://user:passwd@localhost:$HGPORT1 hg clone --config http_proxy.always=True http://user:passwd@localhost:$HGPORT/ e
   requesting all changes
   adding changesets
   adding manifests
   adding file changes
   added 1 changesets with 1 changes to 1 files
   new changesets 83180e7845de
   updating to branch default
   1 files updated, 0 files merged, 0 files removed, 0 files unresolved
 
 bad host:port for proxy ("Protocol not supported" can happen on
 misconfigured hosts)
 
   $ http_proxy=localhost:$HGPORT2 hg clone --config http_proxy.always=True http://localhost:$HGPORT/ f
   abort: error: (Connection refused|Protocol not supported|.* actively refused it|\$EADDRNOTAVAIL\$|No route to host) (re)
   [100]
 
 do not use the proxy if it is in the no list
 
   $ http_proxy=localhost:$HGPORT1 hg clone --config http_proxy.no=localhost http://localhost:$HGPORT/ g
   requesting all changes
   adding changesets
   adding manifests
   adding file changes
   added 1 changesets with 1 changes to 1 files
   new changesets 83180e7845de
   updating to branch default
   1 files updated, 0 files merged, 0 files removed, 0 files unresolved
+
+proxy can't connect to server
+
+  $ http_proxy=http://localhost:$HGPORT1/ hg --config http_proxy.always=True clone http://localhost:$HGPORT2/ h
+  abort: HTTP Error 404: Connection refused
+  [100]
+
   $ cat proxy.log
   * - - [*] "GET http://localhost:$HGPORT/?cmd=capabilities HTTP/1.1" - - (glob)
   $LOCALIP - - [$LOGDATE$] "GET http://localhost:$HGPORT/?cmd=batch HTTP/1.1" - - x-hgarg-1:cmds=heads+%3Bknown+nodes%3D x-hgproto-1:0.1 0.2 comp=$USUAL_COMPRESSIONS$ partial-pull (glob)
   $LOCALIP - - [$LOGDATE$] "GET http://localhost:$HGPORT/?cmd=getbundle HTTP/1.1" - - x-hgarg-1:bookmarks=1&$USUAL_BUNDLE_CAPS$&cg=0&common=0000000000000000000000000000000000000000&heads=83180e7845de420a1bb46896fd5fe05294f8d629&listkeys=bookmarks&stream=1 x-hgproto-1:0.1 0.2 comp=$USUAL_COMPRESSIONS$ partial-pull (glob)
   * - - [*] "GET http://localhost:$HGPORT/?cmd=capabilities HTTP/1.1" - - (glob)
   $LOCALIP - - [$LOGDATE$] "GET http://localhost:$HGPORT/?cmd=batch HTTP/1.1" - - x-hgarg-1:cmds=heads+%3Bknown+nodes%3D x-hgproto-1:0.1 0.2 comp=$USUAL_COMPRESSIONS$ partial-pull (glob)
   $LOCALIP - - [$LOGDATE$] "GET http://localhost:$HGPORT/?cmd=getbundle HTTP/1.1" - - x-hgarg-1:bookmarks=1&$USUAL_BUNDLE_CAPS$&cg=1&common=0000000000000000000000000000000000000000&heads=83180e7845de420a1bb46896fd5fe05294f8d629&listkeys=bookmarks&phases=1 x-hgproto-1:0.1 0.2 comp=$USUAL_COMPRESSIONS$ partial-pull (glob)
   * - - [*] "GET http://localhost:$HGPORT/?cmd=capabilities HTTP/1.1" - - (glob)
   $LOCALIP - - [$LOGDATE$] "GET http://localhost:$HGPORT/?cmd=batch HTTP/1.1" - - x-hgarg-1:cmds=heads+%3Bknown+nodes%3D x-hgproto-1:0.1 0.2 comp=$USUAL_COMPRESSIONS$ partial-pull (glob)
   $LOCALIP - - [$LOGDATE$] "GET http://localhost:$HGPORT/?cmd=getbundle HTTP/1.1" - - x-hgarg-1:bookmarks=1&$USUAL_BUNDLE_CAPS$&cg=1&common=0000000000000000000000000000000000000000&heads=83180e7845de420a1bb46896fd5fe05294f8d629&listkeys=bookmarks&phases=1 x-hgproto-1:0.1 0.2 comp=$USUAL_COMPRESSIONS$ partial-pull (glob)
   * - - [*] "GET http://localhost:$HGPORT/?cmd=capabilities HTTP/1.1" - - (glob)
   $LOCALIP - - [$LOGDATE$] "GET http://localhost:$HGPORT/?cmd=batch HTTP/1.1" - - x-hgarg-1:cmds=heads+%3Bknown+nodes%3D x-hgproto-1:0.1 0.2 comp=$USUAL_COMPRESSIONS$ partial-pull (glob)
   $LOCALIP - - [$LOGDATE$] "GET http://localhost:$HGPORT/?cmd=getbundle HTTP/1.1" - - x-hgarg-1:bookmarks=1&$USUAL_BUNDLE_CAPS$&cg=1&common=0000000000000000000000000000000000000000&heads=83180e7845de420a1bb46896fd5fe05294f8d629&listkeys=bookmarks&phases=1 x-hgproto-1:0.1 0.2 comp=$USUAL_COMPRESSIONS$ partial-pull (glob)
   * - - [*] "GET http://localhost:$HGPORT/?cmd=capabilities HTTP/1.1" - - (glob)
   $LOCALIP - - [$LOGDATE$] "GET http://localhost:$HGPORT/?cmd=batch HTTP/1.1" - - x-hgarg-1:cmds=heads+%3Bknown+nodes%3D x-hgproto-1:0.1 0.2 comp=$USUAL_COMPRESSIONS$ partial-pull (glob)
   $LOCALIP - - [$LOGDATE$] "GET http://localhost:$HGPORT/?cmd=getbundle HTTP/1.1" - - x-hgarg-1:bookmarks=1&$USUAL_BUNDLE_CAPS$&cg=1&common=0000000000000000000000000000000000000000&heads=83180e7845de420a1bb46896fd5fe05294f8d629&listkeys=bookmarks&phases=1 x-hgproto-1:0.1 0.2 comp=$USUAL_COMPRESSIONS$ partial-pull (glob)
+  * - - [*] code 404, message Connection refused (glob)
+  $LOCALIP - - [$LOGDATE$] "GET http://localhost:$HGPORT2/?cmd=capabilities HTTP/1.1" 404 - (glob)

tests/test-https.t

@@ -1,555 +1,562 b''
 #require serve ssl
 
 Proper https client requires the built-in ssl from Python 2.6.
 
 Disable the system configuration which may set stricter TLS requirements.
 This test expects that legacy TLS versions are supported.
 
   $ OPENSSL_CONF=
   $ export OPENSSL_CONF
 
 Make server certificates:
 
   $ CERTSDIR="$TESTDIR/sslcerts"
   $ cat "$CERTSDIR/priv.pem" "$CERTSDIR/pub.pem" >> server.pem
   $ PRIV=`pwd`/server.pem
   $ cat "$CERTSDIR/priv.pem" "$CERTSDIR/pub-not-yet.pem" > server-not-yet.pem
   $ cat "$CERTSDIR/priv.pem" "$CERTSDIR/pub-expired.pem" > server-expired.pem
 
   $ hg init test
   $ cd test
   $ echo foo>foo
   $ mkdir foo.d foo.d/bAr.hg.d foo.d/baR.d.hg
   $ echo foo>foo.d/foo
   $ echo bar>foo.d/bAr.hg.d/BaR
   $ echo bar>foo.d/baR.d.hg/bAR
   $ hg commit -A -m 1
   adding foo
   adding foo.d/bAr.hg.d/BaR
   adding foo.d/baR.d.hg/bAR
   adding foo.d/foo
   $ hg serve -p $HGPORT -d --pid-file=../hg0.pid --certificate=$PRIV
   $ cat ../hg0.pid >> $DAEMON_PIDS
 
 cacert not found
 
   $ hg in --config web.cacerts=no-such.pem https://localhost:$HGPORT/
   abort: could not find web.cacerts: no-such.pem
   [255]
 
 Test server address cannot be reused
 
   $ hg serve -p $HGPORT --certificate=$PRIV 2>&1
   abort: cannot start server at 'localhost:$HGPORT': $EADDRINUSE$
   [255]
 
   $ cd ..
 
 Our test cert is not signed by a trusted CA. It should fail to verify if
 we are able to load CA certs.
 
 #if no-defaultcacertsloaded
   $ hg clone https://localhost:$HGPORT/ copy-pull
   (an attempt was made to load CA certificates but none were loaded; see https://mercurial-scm.org/wiki/SecureConnections for how to configure Mercurial to avoid this error)
   abort: error: *certificate verify failed* (glob)
   [100]
 #endif
 
 #if defaultcacertsloaded
   $ hg clone https://localhost:$HGPORT/ copy-pull
   (the full certificate chain may not be available locally; see "hg help debugssl") (windows !)
   abort: error: *certificate verify failed* (glob)
   [100]
 #endif
 
 Specifying a per-host certificate file that doesn't exist will abort.  The full
 C:/path/to/msysroot will print on Windows.
 
   $ hg --config hostsecurity.localhost:verifycertsfile=/does/not/exist clone https://localhost:$HGPORT/
   abort: path specified by hostsecurity.localhost:verifycertsfile does not exist: */does/not/exist (glob)
   [255]
 
 A malformed per-host certificate file will raise an error
 
   $ echo baddata > badca.pem
   $ hg --config hostsecurity.localhost:verifycertsfile=badca.pem clone https://localhost:$HGPORT/
   abort: error loading CA file badca.pem: * (glob)
   (file is empty or malformed?)
   [255]
 
 A per-host certificate mismatching the server will fail verification
 
 (modern ssl is able to discern whether the loaded cert is a CA cert)
   $ hg --config hostsecurity.localhost:verifycertsfile="$CERTSDIR/client-cert.pem" clone https://localhost:$HGPORT/
   (an attempt was made to load CA certificates but none were loaded; see https://mercurial-scm.org/wiki/SecureConnections for how to configure Mercurial to avoid this error)
   (the full certificate chain may not be available locally; see "hg help debugssl") (windows !)
   abort: error: *certificate verify failed* (glob)
   [100]
 
 A per-host certificate matching the server's cert will be accepted
 
   $ hg --config hostsecurity.localhost:verifycertsfile="$CERTSDIR/pub.pem" clone -U https://localhost:$HGPORT/ perhostgood1
   requesting all changes
   adding changesets
   adding manifests
   adding file changes
   added 1 changesets with 4 changes to 4 files
   new changesets 8b6053c928fe
 
 A per-host certificate with multiple certs and one matching will be accepted
 
   $ cat "$CERTSDIR/client-cert.pem" "$CERTSDIR/pub.pem" > perhost.pem
   $ hg --config hostsecurity.localhost:verifycertsfile=perhost.pem clone -U https://localhost:$HGPORT/ perhostgood2
   requesting all changes
   adding changesets
   adding manifests
   adding file changes
   added 1 changesets with 4 changes to 4 files
   new changesets 8b6053c928fe
 
 Defining both per-host certificate and a fingerprint will print a warning
 
   $ hg --config hostsecurity.localhost:verifycertsfile="$CERTSDIR/pub.pem" --config hostsecurity.localhost:fingerprints=sha1:ecd87cd6b386d04fc1b8b41c9d8f5e168eef1c03 clone -U https://localhost:$HGPORT/ caandfingerwarning
   (hostsecurity.localhost:verifycertsfile ignored when host fingerprints defined; using host fingerprints for verification)
   requesting all changes
   adding changesets
   adding manifests
   adding file changes
   added 1 changesets with 4 changes to 4 files
   new changesets 8b6053c928fe
 
   $ DISABLECACERTS="--config devel.disableloaddefaultcerts=true"
 
 Inability to verify peer certificate will result in abort
 
   $ hg clone https://localhost:$HGPORT/ copy-pull $DISABLECACERTS
   abort: unable to verify security of localhost (no loaded CA certificates); refusing to connect
   (see https://mercurial-scm.org/wiki/SecureConnections for how to configure Mercurial to avoid this error or set hostsecurity.localhost:fingerprints=sha256:20:de:b3:ad:b4:cd:a5:42:f0:74:41:1c:a2:70:1e:da:6e:c0:5c:16:9e:e7:22:0f:f1:b7:e5:6e:e4:92:af:7e to trust this server)
   [150]
 
   $ hg clone --insecure https://localhost:$HGPORT/ copy-pull
   warning: connection security to localhost is disabled per current settings; communication is susceptible to eavesdropping and tampering
   requesting all changes
   adding changesets
   adding manifests
   adding file changes
   added 1 changesets with 4 changes to 4 files
   new changesets 8b6053c928fe
   updating to branch default
   4 files updated, 0 files merged, 0 files removed, 0 files unresolved
   $ hg verify -R copy-pull
   checking changesets
   checking manifests
   crosschecking files in changesets and manifests
   checking files
   checked 1 changesets with 4 changes to 4 files
   $ cd test
   $ echo bar > bar
   $ hg commit -A -d '1 0' -m 2
   adding bar
   $ cd ..
 
 pull without cacert
 
   $ cd copy-pull
   $ cat >> .hg/hgrc <<EOF
   > [hooks]
   > changegroup = sh -c "printenv.py --line changegroup"
   > EOF
   $ hg pull $DISABLECACERTS
   pulling from https://localhost:$HGPORT/
   abort: unable to verify security of localhost (no loaded CA certificates); refusing to connect
   (see https://mercurial-scm.org/wiki/SecureConnections for how to configure Mercurial to avoid this error or set hostsecurity.localhost:fingerprints=sha256:20:de:b3:ad:b4:cd:a5:42:f0:74:41:1c:a2:70:1e:da:6e:c0:5c:16:9e:e7:22:0f:f1:b7:e5:6e:e4:92:af:7e to trust this server)
   [150]
 
   $ hg pull --insecure
   pulling from https://localhost:$HGPORT/
   warning: connection security to localhost is disabled per current settings; communication is susceptible to eavesdropping and tampering
   searching for changes
   adding changesets
   adding manifests
   adding file changes
   added 1 changesets with 1 changes to 1 files
   new changesets 5fed3813f7f5
   changegroup hook: HG_HOOKNAME=changegroup
   HG_HOOKTYPE=changegroup
   HG_NODE=5fed3813f7f5e1824344fdc9cf8f63bb662c292d
   HG_NODE_LAST=5fed3813f7f5e1824344fdc9cf8f63bb662c292d
   HG_SOURCE=pull
   HG_TXNID=TXN:$ID$
   HG_TXNNAME=pull
   https://localhost:$HGPORT/
   HG_URL=https://localhost:$HGPORT/
   
   (run 'hg update' to get a working copy)
   $ cd ..
 
 cacert configured in local repo
 
   $ cp copy-pull/.hg/hgrc copy-pull/.hg/hgrc.bu
   $ echo "[web]" >> copy-pull/.hg/hgrc
   $ echo "cacerts=$CERTSDIR/pub.pem" >> copy-pull/.hg/hgrc
   $ hg -R copy-pull pull
   pulling from https://localhost:$HGPORT/
   searching for changes
   no changes found
   $ mv copy-pull/.hg/hgrc.bu copy-pull/.hg/hgrc
 
 cacert configured globally, also testing expansion of environment
 variables in the filename
 
   $ echo "[web]" >> $HGRCPATH
   $ echo 'cacerts=$P/pub.pem' >> $HGRCPATH
   $ P="$CERTSDIR" hg -R copy-pull pull
   pulling from https://localhost:$HGPORT/
   searching for changes
   no changes found
   $ P="$CERTSDIR" hg -R copy-pull pull --insecure
   pulling from https://localhost:$HGPORT/
   warning: connection security to localhost is disabled per current settings; communication is susceptible to eavesdropping and tampering
   searching for changes
   no changes found
 
 empty cacert file
 
   $ touch emptycafile
 
   $ hg --config web.cacerts=emptycafile -R copy-pull pull
   pulling from https://localhost:$HGPORT/
   abort: error loading CA file emptycafile: * (glob)
   (file is empty or malformed?)
   [255]
 
 cacert mismatch
 
   $ hg -R copy-pull pull --config web.cacerts="$CERTSDIR/pub.pem" \
   > https://$LOCALIP:$HGPORT/
   pulling from https://*:$HGPORT/ (glob)
   abort: $LOCALIP certificate error: certificate is for localhost (glob)
   (set hostsecurity.$LOCALIP:certfingerprints=sha256:20:de:b3:ad:b4:cd:a5:42:f0:74:41:1c:a2:70:1e:da:6e:c0:5c:16:9e:e7:22:0f:f1:b7:e5:6e:e4:92:af:7e config setting or use --insecure to connect insecurely)
   [150]
   $ hg -R copy-pull pull --config web.cacerts="$CERTSDIR/pub.pem" \
   > https://$LOCALIP:$HGPORT/ --insecure
   pulling from https://*:$HGPORT/ (glob)
   warning: connection security to $LOCALIP is disabled per current settings; communication is susceptible to eavesdropping and tampering (glob)
   searching for changes
   no changes found
   $ hg -R copy-pull pull --config web.cacerts="$CERTSDIR/pub-other.pem"
   pulling from https://localhost:$HGPORT/
   (the full certificate chain may not be available locally; see "hg help debugssl") (windows !)
   abort: error: *certificate verify failed* (glob)
   [100]
   $ hg -R copy-pull pull --config web.cacerts="$CERTSDIR/pub-other.pem" \
   > --insecure
   pulling from https://localhost:$HGPORT/
   warning: connection security to localhost is disabled per current settings; communication is susceptible to eavesdropping and tampering
   searching for changes
   no changes found
 
 Test server cert which isn't valid yet
 
   $ hg serve -R test -p $HGPORT1 -d --pid-file=hg1.pid --certificate=server-not-yet.pem
   $ cat hg1.pid >> $DAEMON_PIDS
   $ hg -R copy-pull pull --config web.cacerts="$CERTSDIR/pub-not-yet.pem" \
   > https://localhost:$HGPORT1/
   pulling from https://localhost:$HGPORT1/
   (the full certificate chain may not be available locally; see "hg help debugssl") (windows !)
   abort: error: *certificate verify failed* (glob)
   [100]
 
 Test server cert which no longer is valid
 
   $ hg serve -R test -p $HGPORT2 -d --pid-file=hg2.pid --certificate=server-expired.pem
   $ cat hg2.pid >> $DAEMON_PIDS
   $ hg -R copy-pull pull --config web.cacerts="$CERTSDIR/pub-expired.pem" \
   > https://localhost:$HGPORT2/
   pulling from https://localhost:$HGPORT2/
   (the full certificate chain may not be available locally; see "hg help debugssl") (windows !)
   abort: error: *certificate verify failed* (glob)
   [100]
 
 Setting ciphers to an invalid value aborts
   $ P="$CERTSDIR" hg --config hostsecurity.ciphers=invalid -R copy-pull id https://localhost:$HGPORT/
   abort: could not set ciphers: No cipher can be selected.
   (change cipher string (invalid) in config)
   [255]
 
   $ P="$CERTSDIR" hg --config hostsecurity.localhost:ciphers=invalid -R copy-pull id https://localhost:$HGPORT/
   abort: could not set ciphers: No cipher can be selected.
   (change cipher string (invalid) in config)
   [255]
 
 Changing the cipher string works
 
   $ P="$CERTSDIR" hg --config hostsecurity.ciphers=HIGH -R copy-pull id https://localhost:$HGPORT/
   5fed3813f7f5
 
 Fingerprints
 
 - works without cacerts (hostfingerprints)
   $ hg -R copy-pull id https://localhost:$HGPORT/ --insecure --config hostfingerprints.localhost=ec:d8:7c:d6:b3:86:d0:4f:c1:b8:b4:1c:9d:8f:5e:16:8e:ef:1c:03
   (SHA-1 fingerprint for localhost found in legacy [hostfingerprints] section; if you trust this fingerprint, remove the old SHA-1 fingerprint from [hostfingerprints] and add the following entry to the new [hostsecurity] section: localhost:fingerprints=sha256:20:de:b3:ad:b4:cd:a5:42:f0:74:41:1c:a2:70:1e:da:6e:c0:5c:16:9e:e7:22:0f:f1:b7:e5:6e:e4:92:af:7e)
   5fed3813f7f5
 
 - works without cacerts (hostsecurity)
   $ hg -R copy-pull id https://localhost:$HGPORT/ --config hostsecurity.localhost:fingerprints=sha1:ecd87cd6b386d04fc1b8b41c9d8f5e168eef1c03
   5fed3813f7f5
 
   $ hg -R copy-pull id https://localhost:$HGPORT/ --config hostsecurity.localhost:fingerprints=sha256:20:de:b3:ad:b4:cd:a5:42:f0:74:41:1c:a2:70:1e:da:6e:c0:5c:16:9e:e7:22:0f:f1:b7:e5:6e:e4:92:af:7e
   5fed3813f7f5
 
 - multiple fingerprints specified and first matches
   $ hg --config 'hostfingerprints.localhost=ecd87cd6b386d04fc1b8b41c9d8f5e168eef1c03, deadbeefdeadbeefdeadbeefdeadbeefdeadbeef' -R copy-pull id https://localhost:$HGPORT/ --insecure
   (SHA-1 fingerprint for localhost found in legacy [hostfingerprints] section; if you trust this fingerprint, remove the old SHA-1 fingerprint from [hostfingerprints] and add the following entry to the new [hostsecurity] section: localhost:fingerprints=sha256:20:de:b3:ad:b4:cd:a5:42:f0:74:41:1c:a2:70:1e:da:6e:c0:5c:16:9e:e7:22:0f:f1:b7:e5:6e:e4:92:af:7e)
   5fed3813f7f5
 
   $ hg --config 'hostsecurity.localhost:fingerprints=sha1:ecd87cd6b386d04fc1b8b41c9d8f5e168eef1c03, sha1:deadbeefdeadbeefdeadbeefdeadbeefdeadbeef' -R copy-pull id https://localhost:$HGPORT/
   5fed3813f7f5
 
 - multiple fingerprints specified and last matches
   $ hg --config 'hostfingerprints.localhost=deadbeefdeadbeefdeadbeefdeadbeefdeadbeef, ecd87cd6b386d04fc1b8b41c9d8f5e168eef1c03' -R copy-pull id https://localhost:$HGPORT/ --insecure
   (SHA-1 fingerprint for localhost found in legacy [hostfingerprints] section; if you trust this fingerprint, remove the old SHA-1 fingerprint from [hostfingerprints] and add the following entry to the new [hostsecurity] section: localhost:fingerprints=sha256:20:de:b3:ad:b4:cd:a5:42:f0:74:41:1c:a2:70:1e:da:6e:c0:5c:16:9e:e7:22:0f:f1:b7:e5:6e:e4:92:af:7e)
   5fed3813f7f5
 
   $ hg --config 'hostsecurity.localhost:fingerprints=sha1:deadbeefdeadbeefdeadbeefdeadbeefdeadbeef, sha1:ecd87cd6b386d04fc1b8b41c9d8f5e168eef1c03' -R copy-pull id https://localhost:$HGPORT/
   5fed3813f7f5
 
 - multiple fingerprints specified and none match
 
   $ hg --config 'hostfingerprints.localhost=deadbeefdeadbeefdeadbeefdeadbeefdeadbeef, aeadbeefdeadbeefdeadbeefdeadbeefdeadbeef' -R copy-pull id https://localhost:$HGPORT/ --insecure
   abort: certificate for localhost has unexpected fingerprint ec:d8:7c:d6:b3:86:d0:4f:c1:b8:b4:1c:9d:8f:5e:16:8e:ef:1c:03
   (check hostfingerprint configuration)
   [150]
 
   $ hg --config 'hostsecurity.localhost:fingerprints=sha1:deadbeefdeadbeefdeadbeefdeadbeefdeadbeef, sha1:aeadbeefdeadbeefdeadbeefdeadbeefdeadbeef' -R copy-pull id https://localhost:$HGPORT/
   abort: certificate for localhost has unexpected fingerprint sha1:ec:d8:7c:d6:b3:86:d0:4f:c1:b8:b4:1c:9d:8f:5e:16:8e:ef:1c:03
   (check hostsecurity configuration)
   [150]
 
 - fails when cert doesn't match hostname (port is ignored)
   $ hg -R copy-pull id https://localhost:$HGPORT1/ --config hostfingerprints.localhost=ecd87cd6b386d04fc1b8b41c9d8f5e168eef1c03
   abort: certificate for localhost has unexpected fingerprint f4:2f:5a:0c:3e:52:5b:db:e7:24:a8:32:1d:18:97:6d:69:b5:87:84
   (check hostfingerprint configuration)
   [150]
 
 
 - ignores that certificate doesn't match hostname
   $ hg -R copy-pull id https://$LOCALIP:$HGPORT/ --config hostfingerprints.$LOCALIP=ecd87cd6b386d04fc1b8b41c9d8f5e168eef1c03
   (SHA-1 fingerprint for $LOCALIP found in legacy [hostfingerprints] section; if you trust this fingerprint, remove the old SHA-1 fingerprint from [hostfingerprints] and add the following entry to the new [hostsecurity] section: $LOCALIP:fingerprints=sha256:20:de:b3:ad:b4:cd:a5:42:f0:74:41:1c:a2:70:1e:da:6e:c0:5c:16:9e:e7:22:0f:f1:b7:e5:6e:e4:92:af:7e)
   5fed3813f7f5
 
 Ports used by next test. Kill servers.
 
   $ killdaemons.py hg0.pid
   $ killdaemons.py hg1.pid
   $ killdaemons.py hg2.pid
 
 #if tls1.2
 Start servers running supported TLS versions
 
   $ cd test
   $ hg serve -p $HGPORT -d --pid-file=../hg0.pid --certificate=$PRIV \
   > --config devel.serverexactprotocol=tls1.0
   $ cat ../hg0.pid >> $DAEMON_PIDS
   $ hg serve -p $HGPORT1 -d --pid-file=../hg1.pid --certificate=$PRIV \
   > --config devel.serverexactprotocol=tls1.1
   $ cat ../hg1.pid >> $DAEMON_PIDS
   $ hg serve -p $HGPORT2 -d --pid-file=../hg2.pid --certificate=$PRIV \
   > --config devel.serverexactprotocol=tls1.2
   $ cat ../hg2.pid >> $DAEMON_PIDS
   $ cd ..
 
 Clients talking same TLS versions work
 
   $ P="$CERTSDIR" hg --config hostsecurity.minimumprotocol=tls1.0 --config hostsecurity.ciphers=DEFAULT id https://localhost:$HGPORT/
   5fed3813f7f5
   $ P="$CERTSDIR" hg --config hostsecurity.minimumprotocol=tls1.1 --config hostsecurity.ciphers=DEFAULT id https://localhost:$HGPORT1/
   5fed3813f7f5
   $ P="$CERTSDIR" hg --config hostsecurity.minimumprotocol=tls1.2 id https://localhost:$HGPORT2/
   5fed3813f7f5
 
 Clients requiring newer TLS version than what server supports fail
 
   $ P="$CERTSDIR" hg id https://localhost:$HGPORT/
   (could not negotiate a common security protocol (tls1.1+) with localhost; the likely cause is Mercurial is configured to be more secure than the server can support)
   (consider contacting the operator of this server and ask them to support modern TLS protocol versions; or, set hostsecurity.localhost:minimumprotocol=tls1.0 to allow use of legacy, less secure protocols when communicating with this server)
   (see https://mercurial-scm.org/wiki/SecureConnections for more info)
   abort: error: .*(unsupported protocol|wrong ssl version|alert protocol version).* (re)
   [100]
 
   $ P="$CERTSDIR" hg --config hostsecurity.minimumprotocol=tls1.1 id https://localhost:$HGPORT/
   (could not negotiate a common security protocol (tls1.1+) with localhost; the likely cause is Mercurial is configured to be more secure than the server can support)
   (consider contacting the operator of this server and ask them to support modern TLS protocol versions; or, set hostsecurity.localhost:minimumprotocol=tls1.0 to allow use of legacy, less secure protocols when communicating with this server)
   (see https://mercurial-scm.org/wiki/SecureConnections for more info)
   abort: error: .*(unsupported protocol|wrong ssl version|alert protocol version).* (re)
   [100]
   $ P="$CERTSDIR" hg --config hostsecurity.minimumprotocol=tls1.2 id https://localhost:$HGPORT/
   (could not negotiate a common security protocol (tls1.2+) with localhost; the likely cause is Mercurial is configured to be more secure than the server can support)
   (consider contacting the operator of this server and ask them to support modern TLS protocol versions; or, set hostsecurity.localhost:minimumprotocol=tls1.0 to allow use of legacy, less secure protocols when communicating with this server)
   (see https://mercurial-scm.org/wiki/SecureConnections for more info)
   abort: error: .*(unsupported protocol|wrong ssl version|alert protocol version).* (re)
   [100]
   $ P="$CERTSDIR" hg --config hostsecurity.minimumprotocol=tls1.2 id https://localhost:$HGPORT1/
   (could not negotiate a common security protocol (tls1.2+) with localhost; the likely cause is Mercurial is configured to be more secure than the server can support)
   (consider contacting the operator of this server and ask them to support modern TLS protocol versions; or, set hostsecurity.localhost:minimumprotocol=tls1.0 to allow use of legacy, less secure protocols when communicating with this server)
   (see https://mercurial-scm.org/wiki/SecureConnections for more info)
   abort: error: .*(unsupported protocol|wrong ssl version|alert protocol version).* (re)
   [100]
 
 --insecure will allow TLS 1.0 connections and override configs
 
   $ hg --config hostsecurity.minimumprotocol=tls1.2 id --insecure https://localhost:$HGPORT1/
   warning: connection security to localhost is disabled per current settings; communication is susceptible to eavesdropping and tampering
   5fed3813f7f5
 
 The per-host config option overrides the default
 
   $ P="$CERTSDIR" hg id https://localhost:$HGPORT/ \
   > --config hostsecurity.ciphers=DEFAULT \
   > --config hostsecurity.minimumprotocol=tls1.2 \
   > --config hostsecurity.localhost:minimumprotocol=tls1.0
   5fed3813f7f5
 
 The per-host config option by itself works
 
   $ P="$CERTSDIR" hg id https://localhost:$HGPORT/ \
   > --config hostsecurity.localhost:minimumprotocol=tls1.2
   (could not negotiate a common security protocol (tls1.2+) with localhost; the likely cause is Mercurial is configured to be more secure than the server can support)
   (consider contacting the operator of this server and ask them to support modern TLS protocol versions; or, set hostsecurity.localhost:minimumprotocol=tls1.0 to allow use of legacy, less secure protocols when communicating with this server)
   (see https://mercurial-scm.org/wiki/SecureConnections for more info)
   abort: error: .*(unsupported protocol|wrong ssl version|alert protocol version).* (re)
   [100]
 
 .hg/hgrc file [hostsecurity] settings are applied to remote ui instances (issue5305)
 
   $ cat >> copy-pull/.hg/hgrc << EOF
   > [hostsecurity]
   > localhost:minimumprotocol=tls1.2
   > EOF
   $ P="$CERTSDIR" hg -R copy-pull id https://localhost:$HGPORT/
   (could not negotiate a common security protocol (tls1.2+) with localhost; the likely cause is Mercurial is configured to be more secure than the server can support)
   (consider contacting the operator of this server and ask them to support modern TLS protocol versions; or, set hostsecurity.localhost:minimumprotocol=tls1.0 to allow use of legacy, less secure protocols when communicating with this server)
   (see https://mercurial-scm.org/wiki/SecureConnections for more info)
   abort: error: .*(unsupported protocol|wrong ssl version|alert protocol version).* (re)
   [100]
 
   $ killdaemons.py hg0.pid
   $ killdaemons.py hg1.pid
   $ killdaemons.py hg2.pid
 #endif
 
 Prepare for connecting through proxy
 
   $ hg serve -R test -p $HGPORT -d --pid-file=hg0.pid --certificate=$PRIV
   $ cat hg0.pid >> $DAEMON_PIDS
   $ hg serve -R test -p $HGPORT2 -d --pid-file=hg2.pid --certificate=server-expired.pem
   $ cat hg2.pid >> $DAEMON_PIDS
 tinyproxy.py doesn't fully detach, so killing it may result in extra output
 from the shell. So don't kill it.
   $ tinyproxy.py $HGPORT1 localhost >proxy.log </dev/null 2>&1 &
   $ while [ ! -f proxy.pid ]; do sleep 0; done
   $ cat proxy.pid >> $DAEMON_PIDS
 
   $ echo "[http_proxy]" >> copy-pull/.hg/hgrc
   $ echo "always=True" >> copy-pull/.hg/hgrc
   $ echo "[hostfingerprints]" >> copy-pull/.hg/hgrc
   $ echo "localhost =" >> copy-pull/.hg/hgrc
 
 Test unvalidated https through proxy
 
   $ http_proxy=http://localhost:$HGPORT1/ hg -R copy-pull pull --insecure
   pulling from https://localhost:$HGPORT/
   warning: connection security to localhost is disabled per current settings; communication is susceptible to eavesdropping and tampering
   searching for changes
   no changes found
 
 Test https with cacert and fingerprint through proxy
 
   $ http_proxy=http://localhost:$HGPORT1/ hg -R copy-pull pull \
   > --config web.cacerts="$CERTSDIR/pub.pem"
   pulling from https://localhost:$HGPORT/
   searching for changes
   no changes found
   $ http_proxy=http://localhost:$HGPORT1/ hg -R copy-pull pull https://localhost:$HGPORT/ --config hostfingerprints.localhost=ecd87cd6b386d04fc1b8b41c9d8f5e168eef1c03 --trace
   pulling from https://*:$HGPORT/ (glob)
   (SHA-1 fingerprint for localhost found in legacy [hostfingerprints] section; if you trust this fingerprint, remove the old SHA-1 fingerprint from [hostfingerprints] and add the following entry to the new [hostsecurity] section: localhost:fingerprints=sha256:20:de:b3:ad:b4:cd:a5:42:f0:74:41:1c:a2:70:1e:da:6e:c0:5c:16:9e:e7:22:0f:f1:b7:e5:6e:e4:92:af:7e)
   searching for changes
   no changes found
 
 Test https with cert problems through proxy
 
   $ http_proxy=http://localhost:$HGPORT1/ hg -R copy-pull pull \
   > --config web.cacerts="$CERTSDIR/pub-other.pem"
   pulling from https://localhost:$HGPORT/
   (the full certificate chain may not be available locally; see "hg help debugssl") (windows !)
   abort: error: *certificate verify failed* (glob)
   [100]
   $ http_proxy=http://localhost:$HGPORT1/ hg -R copy-pull pull \
   > --config web.cacerts="$CERTSDIR/pub-expired.pem" https://localhost:$HGPORT2/
   pulling from https://localhost:$HGPORT2/
   (the full certificate chain may not be available locally; see "hg help debugssl") (windows !)
   abort: error: *certificate verify failed* (glob)
   [100]
 
+Test when proxy can't connect to server
+
+  $ http_proxy=http://localhost:$HGPORT1/ hg -R copy-pull pull --insecure https://localhost:0/
+  pulling from https://localhost:0/
+  abort: error: Tunnel connection failed: 404 Connection refused
+  [100]
+
 
   $ killdaemons.py hg0.pid
 
   $ cd test
 
 Missing certificate file(s) are detected
 
   $ hg serve -p $HGPORT --certificate=/missing/certificate \
   > --config devel.servercafile=$PRIV --config devel.serverrequirecert=true
   abort: referenced certificate file (*/missing/certificate) does not exist (glob)
   [255]
 
   $ hg serve -p $HGPORT --certificate=$PRIV \
   > --config devel.servercafile=/missing/cafile --config devel.serverrequirecert=true
   abort: referenced certificate file (*/missing/cafile) does not exist (glob)
   [255]
 
 Start hgweb that requires client certificates:
 
   $ hg serve -p $HGPORT -d --pid-file=../hg0.pid --certificate=$PRIV \
   > --config devel.servercafile=$PRIV --config devel.serverrequirecert=true
   $ cat ../hg0.pid >> $DAEMON_PIDS
   $ cd ..
 
 without client certificate:
 
   $ P="$CERTSDIR" hg id https://localhost:$HGPORT/
   abort: error: .*(\$ECONNRESET\$|certificate required|handshake failure|EOF occurred).* (re)
   [100]
 
 with client certificate:
 
   $ cat << EOT >> $HGRCPATH
   > [auth]
   > l.prefix = localhost
   > l.cert = $CERTSDIR/client-cert.pem
   > l.key = $CERTSDIR/client-key.pem
   > EOT
 
   $ P="$CERTSDIR" hg id https://localhost:$HGPORT/ \
   > --config auth.l.key="$CERTSDIR/client-key-decrypted.pem"
   5fed3813f7f5
 
   $ printf '1234\n' | env P="$CERTSDIR" hg id https://localhost:$HGPORT/ \
   > --config ui.interactive=True --config ui.nontty=True
   passphrase for */client-key.pem: 5fed3813f7f5 (glob)
 
   $ env P="$CERTSDIR" hg id https://localhost:$HGPORT/
   abort: error: * (glob)
   [100]
 
 Missing certficate and key files result in error
 
   $ hg id https://localhost:$HGPORT/ --config auth.l.cert=/missing/cert
   abort: certificate file (*/missing/cert) does not exist; cannot connect to localhost (glob)
   (restore missing file or fix references in Mercurial config)
   [255]
 
   $ hg id https://localhost:$HGPORT/ --config auth.l.key=/missing/key
   abort: certificate file (*/missing/key) does not exist; cannot connect to localhost (glob)
   (restore missing file or fix references in Mercurial config)
   [255]

tests/tinyproxy.py

@@ -1,222 +1,218 b''
 #!/usr/bin/env python
 
 
 __doc__ = """Tiny HTTP Proxy.
 
 This module implements GET, HEAD, POST, PUT and DELETE methods
 on BaseHTTPServer, and behaves as an HTTP proxy.  The CONNECT
 method is also implemented experimentally, but has not been
 tested yet.
 
 Any help will be greatly appreciated.           SUZUKI Hisao
 """
 
 __version__ = "0.2.1"
 
 import optparse
 import os
 import select
 import socket
 import sys
 
 from mercurial import (
     pycompat,
     util,
 )
 
 httpserver = util.httpserver
 socketserver = util.socketserver
 urlreq = util.urlreq
 
 if os.environ.get('HGIPV6', '0') == '1':
     family = socket.AF_INET6
 else:
     family = socket.AF_INET
 
 
 class ProxyHandler(httpserver.basehttprequesthandler):
     __base = httpserver.basehttprequesthandler
     __base_handle = __base.handle
 
     server_version = "TinyHTTPProxy/" + __version__
     rbufsize = 0  # self.rfile Be unbuffered
 
     def handle(self):
         (ip, port) = self.client_address
         allowed = getattr(self, 'allowed_clients', None)
         if allowed is not None and ip not in allowed:
             self.raw_requestline = self.rfile.readline()
             if self.parse_request():
                 self.send_error(403)
         else:
             self.__base_handle()
 
     def log_request(self, code='-', size='-'):
         xheaders = [h for h in self.headers.items() if h[0].startswith('x-')]
         self.log_message(
             '"%s" %s %s%s',
             self.requestline,
             str(code),
             str(size),
             ''.join([' %s:%s' % h for h in sorted(xheaders)]),
         )
         # Flush for Windows, so output isn't lost on TerminateProcess()
         sys.stdout.flush()
         sys.stderr.flush()
 
     def _connect_to(self, netloc, soc):
         i = netloc.find(':')
         if i >= 0:
             host_port = netloc[:i], int(netloc[i + 1 :])
         else:
             host_port = netloc, 80
         print("\t" "connect to %s:%d" % host_port)
         try:
             soc.connect(host_port)
-        except socket.error as arg:
-            try:
-                msg = arg[1]
-            except (IndexError, TypeError):
-                msg = arg
-            self.send_error(404, msg)
+        except socket.error as e:
+            self.send_error(404, e.strerror)
             return 0
         return 1
 
     def do_CONNECT(self):
         soc = socket.socket(family, socket.SOCK_STREAM)
         try:
             if self._connect_to(self.path, soc):
                 self.log_request(200)
                 self.wfile.write(
                     pycompat.bytestr(self.protocol_version)
                     + b" 200 Connection established\r\n"
                 )
                 self.wfile.write(
                     b"Proxy-agent: %s\r\n"
                     % pycompat.bytestr(self.version_string())
                 )
                 self.wfile.write(b"\r\n")
                 self._read_write(soc, 300)
         finally:
             print("\t" "bye")
             soc.close()
             self.connection.close()
 
     def do_GET(self):
         (scm, netloc, path, params, query, fragment) = urlreq.urlparse(
             self.path, 'http'
         )
         if scm != 'http' or fragment or not netloc:
             self.send_error(400, "bad url %s" % self.path)
             return
         soc = socket.socket(family, socket.SOCK_STREAM)
         try:
             if self._connect_to(netloc, soc):
                 self.log_request()
                 url = urlreq.urlunparse(('', '', path, params, query, ''))
                 soc.send(
                     b"%s %s %s\r\n"
                     % (
                         pycompat.bytestr(self.command),
                         pycompat.bytestr(url),
                         pycompat.bytestr(self.request_version),
                     )
                 )
                 self.headers['Connection'] = 'close'
                 del self.headers['Proxy-Connection']
                 for key, val in self.headers.items():
                     soc.send(
                         b"%s: %s\r\n"
                         % (pycompat.bytestr(key), pycompat.bytestr(val))
                     )
                 soc.send(b"\r\n")
                 self._read_write(soc)
         finally:
             print("\t" "bye")
             soc.close()
             self.connection.close()
 
     def _read_write(self, soc, max_idling=20):
         iw = [self.connection, soc]
         ow = []
         count = 0
         while True:
             count += 1
             (ins, _, exs) = select.select(iw, ow, iw, 3)
             if exs:
                 break
             if ins:
                 for i in ins:
                     if i is soc:
                         out = self.connection
                     else:
                         out = soc
                     try:
                         data = i.recv(8192)
                     except socket.error:
                         break
                     if data:
                         out.send(data)
                         count = 0
             else:
                 print("\t" "idle", count)
             if count == max_idling:
                 break
 
     do_HEAD = do_GET
     do_POST = do_GET
     do_PUT = do_GET
     do_DELETE = do_GET
 
 
 class ThreadingHTTPServer(socketserver.ThreadingMixIn, httpserver.httpserver):
     def __init__(self, *args, **kwargs):
         httpserver.httpserver.__init__(self, *args, **kwargs)
         a = open("proxy.pid", "w")
         a.write(str(os.getpid()) + "\n")
         a.close()
 
 
 def runserver(port=8000, bind=""):
     server_address = (bind, port)
     ProxyHandler.protocol_version = "HTTP/1.0"
     httpd = ThreadingHTTPServer(server_address, ProxyHandler)
     sa = httpd.socket.getsockname()
     print("Serving HTTP on", sa[0], "port", sa[1], "...")
     try:
         httpd.serve_forever()
     except KeyboardInterrupt:
         print("\nKeyboard interrupt received, exiting.")
         httpd.server_close()
         sys.exit(0)
 
 
 if __name__ == '__main__':
     argv = sys.argv
     if argv[1:] and argv[1] in ('-h', '--help'):
         print(argv[0], "[port [allowed_client_name ...]]")
     else:
         if argv[2:]:
             allowed = []
             for name in argv[2:]:
                 client = socket.gethostbyname(name)
                 allowed.append(client)
                 print("Accept: %s (%s)" % (client, name))
             ProxyHandler.allowed_clients = allowed
             del argv[2:]
         else:
             print("Any clients will be served...")
 
         parser = optparse.OptionParser()
         parser.add_option(
             '-b',
             '--bind',
             metavar='ADDRESS',
             help='Specify alternate bind address ' '[default: all interfaces]',
             default='',
         )
         (options, args) = parser.parse_args()
         port = 8000
         if len(args) == 1:
             port = int(args[0])
         runserver(port, options.bind)