py3: __repr__ needs to return str, not bytes

Differential Revision: https://phab.mercurial-scm.org/D8089

mercurial/bundle2.py

@@ -1,2578 +1,2579 b''
 # bundle2.py - generic container format to transmit arbitrary data.
 #
 # Copyright 2013 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.
 """Handling of the new bundle2 format
 
 The goal of bundle2 is to act as an atomically packet to transmit a set of
 payloads in an application agnostic way. It consist in a sequence of "parts"
 that will be handed to and processed by the application layer.
 
 
 General format architecture
 ===========================
 
 The format is architectured as follow
 
  - magic string
  - stream level parameters
  - payload parts (any number)
  - end of stream marker.
 
 the Binary format
 ============================
 
 All numbers are unsigned and big-endian.
 
 stream level parameters
 ------------------------
 
 Binary format is as follow
 
 :params size: int32
 
   The total number of Bytes used by the parameters
 
 :params value: arbitrary number of Bytes
 
   A blob of `params size` containing the serialized version of all stream level
   parameters.
 
   The blob contains a space separated list of parameters. Parameters with value
   are stored in the form `<name>=<value>`. Both name and value are urlquoted.
 
   Empty name are obviously forbidden.
 
   Name MUST start with a letter. If this first letter is lower case, the
   parameter is advisory and can be safely ignored. However when the first
   letter is capital, the parameter is mandatory and the bundling process MUST
   stop if he is not able to proceed it.
 
   Stream parameters use a simple textual format for two main reasons:
 
   - Stream level parameters should remain simple and we want to discourage any
     crazy usage.
   - Textual data allow easy human inspection of a bundle2 header in case of
     troubles.
 
   Any Applicative level options MUST go into a bundle2 part instead.
 
 Payload part
 ------------------------
 
 Binary format is as follow
 
 :header size: int32
 
   The total number of Bytes used by the part header. When the header is empty
   (size = 0) this is interpreted as the end of stream marker.
 
 :header:
 
     The header defines how to interpret the part. It contains two piece of
     data: the part type, and the part parameters.
 
     The part type is used to route an application level handler, that can
     interpret payload.
 
     Part parameters are passed to the application level handler.  They are
     meant to convey information that will help the application level object to
     interpret the part payload.
 
     The binary format of the header is has follow
 
     :typesize: (one byte)
 
     :parttype: alphanumerical part name (restricted to [a-zA-Z0-9_:-]*)
 
     :partid: A 32bits integer (unique in the bundle) that can be used to refer
              to this part.
 
     :parameters:
 
         Part's parameter may have arbitrary content, the binary structure is::
 
             <mandatory-count><advisory-count><param-sizes><param-data>
 
         :mandatory-count: 1 byte, number of mandatory parameters
 
         :advisory-count:  1 byte, number of advisory parameters
 
         :param-sizes:
 
             N couple of bytes, where N is the total number of parameters. Each
             couple contains (<size-of-key>, <size-of-value) for one parameter.
 
         :param-data:
 
             A blob of bytes from which each parameter key and value can be
             retrieved using the list of size couples stored in the previous
             field.
 
             Mandatory parameters comes first, then the advisory ones.
 
             Each parameter's key MUST be unique within the part.
 
 :payload:
 
     payload is a series of `<chunksize><chunkdata>`.
 
     `chunksize` is an int32, `chunkdata` are plain bytes (as much as
     `chunksize` says)` The payload part is concluded by a zero size chunk.
 
     The current implementation always produces either zero or one chunk.
     This is an implementation limitation that will ultimately be lifted.
 
     `chunksize` can be negative to trigger special case processing. No such
     processing is in place yet.
 
 Bundle processing
 ============================
 
 Each part is processed in order using a "part handler". Handler are registered
 for a certain part type.
 
 The matching of a part to its handler is case insensitive. The case of the
 part type is used to know if a part is mandatory or advisory. If the Part type
 contains any uppercase char it is considered mandatory. When no handler is
 known for a Mandatory part, the process is aborted and an exception is raised.
 If the part is advisory and no handler is known, the part is ignored. When the
 process is aborted, the full bundle is still read from the stream to keep the
 channel usable. But none of the part read from an abort are processed. In the
 future, dropping the stream may become an option for channel we do not care to
 preserve.
 """
 
 from __future__ import absolute_import, division
 
 import collections
 import errno
 import os
 import re
 import string
 import struct
 import sys
 
 from .i18n import _
 from . import (
     bookmarks,
     changegroup,
     encoding,
     error,
     node as nodemod,
     obsolete,
     phases,
     pushkey,
     pycompat,
     streamclone,
     tags,
     url,
     util,
 )
 from .utils import stringutil
 
 urlerr = util.urlerr
 urlreq = util.urlreq
 
 _pack = struct.pack
 _unpack = struct.unpack
 
 _fstreamparamsize = b'>i'
 _fpartheadersize = b'>i'
 _fparttypesize = b'>B'
 _fpartid = b'>I'
 _fpayloadsize = b'>i'
 _fpartparamcount = b'>BB'
 
 preferedchunksize = 32768
 
 _parttypeforbidden = re.compile(b'[^a-zA-Z0-9_:-]')
 
 
 def outdebug(ui, message):
     """debug regarding output stream (bundling)"""
     if ui.configbool(b'devel', b'bundle2.debug'):
         ui.debug(b'bundle2-output: %s\n' % message)
 
 
 def indebug(ui, message):
     """debug on input stream (unbundling)"""
     if ui.configbool(b'devel', b'bundle2.debug'):
         ui.debug(b'bundle2-input: %s\n' % message)
 
 
 def validateparttype(parttype):
     """raise ValueError if a parttype contains invalid character"""
     if _parttypeforbidden.search(parttype):
         raise ValueError(parttype)
 
 
 def _makefpartparamsizes(nbparams):
     """return a struct format to read part parameter sizes
 
     The number parameters is variable so we need to build that format
     dynamically.
     """
     return b'>' + (b'BB' * nbparams)
 
 
 parthandlermapping = {}
 
 
 def parthandler(parttype, params=()):
     """decorator that register a function as a bundle2 part handler
 
     eg::
 
         @parthandler('myparttype', ('mandatory', 'param', 'handled'))
         def myparttypehandler(...):
             '''process a part of type "my part".'''
             ...
     """
     validateparttype(parttype)
 
     def _decorator(func):
         lparttype = parttype.lower()  # enforce lower case matching.
         assert lparttype not in parthandlermapping
         parthandlermapping[lparttype] = func
         func.params = frozenset(params)
         return func
 
     return _decorator
 
 
 class unbundlerecords(object):
     """keep record of what happens during and unbundle
 
     New records are added using `records.add('cat', obj)`. Where 'cat' is a
     category of record and obj is an arbitrary object.
 
     `records['cat']` will return all entries of this category 'cat'.
 
     Iterating on the object itself will yield `('category', obj)` tuples
     for all entries.
 
     All iterations happens in chronological order.
     """
 
     def __init__(self):
         self._categories = {}
         self._sequences = []
         self._replies = {}
 
     def add(self, category, entry, inreplyto=None):
         """add a new record of a given category.
 
         The entry can then be retrieved in the list returned by
         self['category']."""
         self._categories.setdefault(category, []).append(entry)
         self._sequences.append((category, entry))
         if inreplyto is not None:
             self.getreplies(inreplyto).add(category, entry)
 
     def getreplies(self, partid):
         """get the records that are replies to a specific part"""
         return self._replies.setdefault(partid, unbundlerecords())
 
     def __getitem__(self, cat):
         return tuple(self._categories.get(cat, ()))
 
     def __iter__(self):
         return iter(self._sequences)
 
     def __len__(self):
         return len(self._sequences)
 
     def __nonzero__(self):
         return bool(self._sequences)
 
     __bool__ = __nonzero__
 
 
 class bundleoperation(object):
     """an object that represents a single bundling process
 
     Its purpose is to carry unbundle-related objects and states.
 
     A new object should be created at the beginning of each bundle processing.
     The object is to be returned by the processing function.
 
     The object has very little content now it will ultimately contain:
     * an access to the repo the bundle is applied to,
     * a ui object,
     * a way to retrieve a transaction to add changes to the repo,
     * a way to record the result of processing each part,
     * a way to construct a bundle response when applicable.
     """
 
     def __init__(self, repo, transactiongetter, captureoutput=True, source=b''):
         self.repo = repo
         self.ui = repo.ui
         self.records = unbundlerecords()
         self.reply = None
         self.captureoutput = captureoutput
         self.hookargs = {}
         self._gettransaction = transactiongetter
         # carries value that can modify part behavior
         self.modes = {}
         self.source = source
 
     def gettransaction(self):
         transaction = self._gettransaction()
 
         if self.hookargs:
             # the ones added to the transaction supercede those added
             # to the operation.
             self.hookargs.update(transaction.hookargs)
             transaction.hookargs = self.hookargs
 
         # mark the hookargs as flushed.  further attempts to add to
         # hookargs will result in an abort.
         self.hookargs = None
 
         return transaction
 
     def addhookargs(self, hookargs):
         if self.hookargs is None:
             raise error.ProgrammingError(
                 b'attempted to add hookargs to '
                 b'operation after transaction started'
             )
         self.hookargs.update(hookargs)
 
 
 class TransactionUnavailable(RuntimeError):
     pass
 
 
 def _notransaction():
     """default method to get a transaction while processing a bundle
 
     Raise an exception to highlight the fact that no transaction was expected
     to be created"""
     raise TransactionUnavailable()
 
 
 def applybundle(repo, unbundler, tr, source, url=None, **kwargs):
     # transform me into unbundler.apply() as soon as the freeze is lifted
     if isinstance(unbundler, unbundle20):
         tr.hookargs[b'bundle2'] = b'1'
         if source is not None and b'source' not in tr.hookargs:
             tr.hookargs[b'source'] = source
         if url is not None and b'url' not in tr.hookargs:
             tr.hookargs[b'url'] = url
         return processbundle(repo, unbundler, lambda: tr, source=source)
     else:
         # the transactiongetter won't be used, but we might as well set it
         op = bundleoperation(repo, lambda: tr, source=source)
         _processchangegroup(op, unbundler, tr, source, url, **kwargs)
         return op
 
 
 class partiterator(object):
     def __init__(self, repo, op, unbundler):
         self.repo = repo
         self.op = op
         self.unbundler = unbundler
         self.iterator = None
         self.count = 0
         self.current = None
 
     def __enter__(self):
         def func():
             itr = enumerate(self.unbundler.iterparts(), 1)
             for count, p in itr:
                 self.count = count
                 self.current = p
                 yield p
                 p.consume()
                 self.current = None
 
         self.iterator = func()
         return self.iterator
 
     def __exit__(self, type, exc, tb):
         if not self.iterator:
             return
 
         # Only gracefully abort in a normal exception situation. User aborts
         # like Ctrl+C throw a KeyboardInterrupt which is not a base Exception,
         # and should not gracefully cleanup.
         if isinstance(exc, Exception):
             # Any exceptions seeking to the end of the bundle at this point are
             # almost certainly related to the underlying stream being bad.
             # And, chances are that the exception we're handling is related to
             # getting in that bad state. So, we swallow the seeking error and
             # re-raise the original error.
             seekerror = False
             try:
                 if self.current:
                     # consume the part content to not corrupt the stream.
                     self.current.consume()
 
                 for part in self.iterator:
                     # consume the bundle content
                     part.consume()
             except Exception:
                 seekerror = True
 
             # Small hack to let caller code distinguish exceptions from bundle2
             # processing from processing the old format. This is mostly needed
             # to handle different return codes to unbundle according to the type
             # of bundle. We should probably clean up or drop this return code
             # craziness in a future version.
             exc.duringunbundle2 = True
             salvaged = []
             replycaps = None
             if self.op.reply is not None:
                 salvaged = self.op.reply.salvageoutput()
                 replycaps = self.op.reply.capabilities
             exc._replycaps = replycaps
             exc._bundle2salvagedoutput = salvaged
 
             # Re-raising from a variable loses the original stack. So only use
             # that form if we need to.
             if seekerror:
                 raise exc
 
         self.repo.ui.debug(
             b'bundle2-input-bundle: %i parts total\n' % self.count
         )
 
 
 def processbundle(repo, unbundler, transactiongetter=None, op=None, source=b''):
     """This function process a bundle, apply effect to/from a repo
 
     It iterates over each part then searches for and uses the proper handling
     code to process the part. Parts are processed in order.
 
     Unknown Mandatory part will abort the process.
 
     It is temporarily possible to provide a prebuilt bundleoperation to the
     function. This is used to ensure output is properly propagated in case of
     an error during the unbundling. This output capturing part will likely be
     reworked and this ability will probably go away in the process.
     """
     if op is None:
         if transactiongetter is None:
             transactiongetter = _notransaction
         op = bundleoperation(repo, transactiongetter, source=source)
     # todo:
     # - replace this is a init function soon.
     # - exception catching
     unbundler.params
     if repo.ui.debugflag:
         msg = [b'bundle2-input-bundle:']
         if unbundler.params:
             msg.append(b' %i params' % len(unbundler.params))
         if op._gettransaction is None or op._gettransaction is _notransaction:
             msg.append(b' no-transaction')
         else:
             msg.append(b' with-transaction')
         msg.append(b'\n')
         repo.ui.debug(b''.join(msg))
 
     processparts(repo, op, unbundler)
 
     return op
 
 
 def processparts(repo, op, unbundler):
     with partiterator(repo, op, unbundler) as parts:
         for part in parts:
             _processpart(op, part)
 
 
 def _processchangegroup(op, cg, tr, source, url, **kwargs):
     ret = cg.apply(op.repo, tr, source, url, **kwargs)
     op.records.add(b'changegroup', {b'return': ret,})
     return ret
 
 
 def _gethandler(op, part):
     status = b'unknown'  # used by debug output
     try:
         handler = parthandlermapping.get(part.type)
         if handler is None:
             status = b'unsupported-type'
             raise error.BundleUnknownFeatureError(parttype=part.type)
         indebug(op.ui, b'found a handler for part %s' % part.type)
         unknownparams = part.mandatorykeys - handler.params
         if unknownparams:
             unknownparams = list(unknownparams)
             unknownparams.sort()
             status = b'unsupported-params (%s)' % b', '.join(unknownparams)
             raise error.BundleUnknownFeatureError(
                 parttype=part.type, params=unknownparams
             )
         status = b'supported'
     except error.BundleUnknownFeatureError as exc:
         if part.mandatory:  # mandatory parts
             raise
         indebug(op.ui, b'ignoring unsupported advisory part %s' % exc)
         return  # skip to part processing
     finally:
         if op.ui.debugflag:
             msg = [b'bundle2-input-part: "%s"' % part.type]
             if not part.mandatory:
                 msg.append(b' (advisory)')
             nbmp = len(part.mandatorykeys)
             nbap = len(part.params) - nbmp
             if nbmp or nbap:
                 msg.append(b' (params:')
                 if nbmp:
                     msg.append(b' %i mandatory' % nbmp)
                 if nbap:
                     msg.append(b' %i advisory' % nbmp)
                 msg.append(b')')
             msg.append(b' %s\n' % status)
             op.ui.debug(b''.join(msg))
 
     return handler
 
 
 def _processpart(op, part):
     """process a single part from a bundle
 
     The part is guaranteed to have been fully consumed when the function exits
     (even if an exception is raised)."""
     handler = _gethandler(op, part)
     if handler is None:
         return
 
     # handler is called outside the above try block so that we don't
     # risk catching KeyErrors from anything other than the
     # parthandlermapping lookup (any KeyError raised by handler()
     # itself represents a defect of a different variety).
     output = None
     if op.captureoutput and op.reply is not None:
         op.ui.pushbuffer(error=True, subproc=True)
         output = b''
     try:
         handler(op, part)
     finally:
         if output is not None:
             output = op.ui.popbuffer()
         if output:
             outpart = op.reply.newpart(b'output', data=output, mandatory=False)
             outpart.addparam(
                 b'in-reply-to', pycompat.bytestr(part.id), mandatory=False
             )
 
 
 def decodecaps(blob):
     """decode a bundle2 caps bytes blob into a dictionary
 
     The blob is a list of capabilities (one per line)
     Capabilities may have values using a line of the form::
 
         capability=value1,value2,value3
 
     The values are always a list."""
     caps = {}
     for line in blob.splitlines():
         if not line:
             continue
         if b'=' not in line:
             key, vals = line, ()
         else:
             key, vals = line.split(b'=', 1)
             vals = vals.split(b',')
         key = urlreq.unquote(key)
         vals = [urlreq.unquote(v) for v in vals]
         caps[key] = vals
     return caps
 
 
 def encodecaps(caps):
     """encode a bundle2 caps dictionary into a bytes blob"""
     chunks = []
     for ca in sorted(caps):
         vals = caps[ca]
         ca = urlreq.quote(ca)
         vals = [urlreq.quote(v) for v in vals]
         if vals:
             ca = b"%s=%s" % (ca, b','.join(vals))
         chunks.append(ca)
     return b'\n'.join(chunks)
 
 
 bundletypes = {
     b"": (b"", b'UN'),  # only when using unbundle on ssh and old http servers
     # since the unification ssh accepts a header but there
     # is no capability signaling it.
     b"HG20": (),  # special-cased below
     b"HG10UN": (b"HG10UN", b'UN'),
     b"HG10BZ": (b"HG10", b'BZ'),
     b"HG10GZ": (b"HG10GZ", b'GZ'),
 }
 
 # hgweb uses this list to communicate its preferred type
 bundlepriority = [b'HG10GZ', b'HG10BZ', b'HG10UN']
 
 
 class bundle20(object):
     """represent an outgoing bundle2 container
 
     Use the `addparam` method to add stream level parameter. and `newpart` to
     populate it. Then call `getchunks` to retrieve all the binary chunks of
     data that compose the bundle2 container."""
 
     _magicstring = b'HG20'
 
     def __init__(self, ui, capabilities=()):
         self.ui = ui
         self._params = []
         self._parts = []
         self.capabilities = dict(capabilities)
         self._compengine = util.compengines.forbundletype(b'UN')
         self._compopts = None
         # If compression is being handled by a consumer of the raw
         # data (e.g. the wire protocol), unsetting this flag tells
         # consumers that the bundle is best left uncompressed.
         self.prefercompressed = True
 
     def setcompression(self, alg, compopts=None):
         """setup core part compression to <alg>"""
         if alg in (None, b'UN'):
             return
         assert not any(n.lower() == b'compression' for n, v in self._params)
         self.addparam(b'Compression', alg)
         self._compengine = util.compengines.forbundletype(alg)
         self._compopts = compopts
 
     @property
     def nbparts(self):
         """total number of parts added to the bundler"""
         return len(self._parts)
 
     # methods used to defines the bundle2 content
     def addparam(self, name, value=None):
         """add a stream level parameter"""
         if not name:
             raise error.ProgrammingError(b'empty parameter name')
         if name[0:1] not in pycompat.bytestr(
             string.ascii_letters  # pytype: disable=wrong-arg-types
         ):
             raise error.ProgrammingError(
                 b'non letter first character: %s' % name
             )
         self._params.append((name, value))
 
     def addpart(self, part):
         """add a new part to the bundle2 container
 
         Parts contains the actual applicative payload."""
         assert part.id is None
         part.id = len(self._parts)  # very cheap counter
         self._parts.append(part)
 
     def newpart(self, typeid, *args, **kwargs):
         """create a new part and add it to the containers
 
         As the part is directly added to the containers. For now, this means
         that any failure to properly initialize the part after calling
         ``newpart`` should result in a failure of the whole bundling process.
 
         You can still fall back to manually create and add if you need better
         control."""
         part = bundlepart(typeid, *args, **kwargs)
         self.addpart(part)
         return part
 
     # methods used to generate the bundle2 stream
     def getchunks(self):
         if self.ui.debugflag:
             msg = [b'bundle2-output-bundle: "%s",' % self._magicstring]
             if self._params:
                 msg.append(b' (%i params)' % len(self._params))
             msg.append(b' %i parts total\n' % len(self._parts))
             self.ui.debug(b''.join(msg))
         outdebug(self.ui, b'start emission of %s stream' % self._magicstring)
         yield self._magicstring
         param = self._paramchunk()
         outdebug(self.ui, b'bundle parameter: %s' % param)
         yield _pack(_fstreamparamsize, len(param))
         if param:
             yield param
         for chunk in self._compengine.compressstream(
             self._getcorechunk(), self._compopts
         ):
             yield chunk
 
     def _paramchunk(self):
         """return a encoded version of all stream parameters"""
         blocks = []
         for par, value in self._params:
             par = urlreq.quote(par)
             if value is not None:
                 value = urlreq.quote(value)
                 par = b'%s=%s' % (par, value)
             blocks.append(par)
         return b' '.join(blocks)
 
     def _getcorechunk(self):
         """yield chunk for the core part of the bundle
 
         (all but headers and parameters)"""
         outdebug(self.ui, b'start of parts')
         for part in self._parts:
             outdebug(self.ui, b'bundle part: "%s"' % part.type)
             for chunk in part.getchunks(ui=self.ui):
                 yield chunk
         outdebug(self.ui, b'end of bundle')
         yield _pack(_fpartheadersize, 0)
 
     def salvageoutput(self):
         """return a list with a copy of all output parts in the bundle
 
         This is meant to be used during error handling to make sure we preserve
         server output"""
         salvaged = []
         for part in self._parts:
             if part.type.startswith(b'output'):
                 salvaged.append(part.copy())
         return salvaged
 
 
 class unpackermixin(object):
     """A mixin to extract bytes and struct data from a stream"""
 
     def __init__(self, fp):
         self._fp = fp
 
     def _unpack(self, format):
         """unpack this struct format from the stream
 
         This method is meant for internal usage by the bundle2 protocol only.
         They directly manipulate the low level stream including bundle2 level
         instruction.
 
         Do not use it to implement higher-level logic or methods."""
         data = self._readexact(struct.calcsize(format))
         return _unpack(format, data)
 
     def _readexact(self, size):
         """read exactly <size> bytes from the stream
 
         This method is meant for internal usage by the bundle2 protocol only.
         They directly manipulate the low level stream including bundle2 level
         instruction.
 
         Do not use it to implement higher-level logic or methods."""
         return changegroup.readexactly(self._fp, size)
 
 
 def getunbundler(ui, fp, magicstring=None):
     """return a valid unbundler object for a given magicstring"""
     if magicstring is None:
         magicstring = changegroup.readexactly(fp, 4)
     magic, version = magicstring[0:2], magicstring[2:4]
     if magic != b'HG':
         ui.debug(
             b"error: invalid magic: %r (version %r), should be 'HG'\n"
             % (magic, version)
         )
         raise error.Abort(_(b'not a Mercurial bundle'))
     unbundlerclass = formatmap.get(version)
     if unbundlerclass is None:
         raise error.Abort(_(b'unknown bundle version %s') % version)
     unbundler = unbundlerclass(ui, fp)
     indebug(ui, b'start processing of %s stream' % magicstring)
     return unbundler
 
 
 class unbundle20(unpackermixin):
     """interpret a bundle2 stream
 
     This class is fed with a binary stream and yields parts through its
     `iterparts` methods."""
 
     _magicstring = b'HG20'
 
     def __init__(self, ui, fp):
         """If header is specified, we do not read it out of the stream."""
         self.ui = ui
         self._compengine = util.compengines.forbundletype(b'UN')
         self._compressed = None
         super(unbundle20, self).__init__(fp)
 
     @util.propertycache
     def params(self):
         """dictionary of stream level parameters"""
         indebug(self.ui, b'reading bundle2 stream parameters')
         params = {}
         paramssize = self._unpack(_fstreamparamsize)[0]
         if paramssize < 0:
             raise error.BundleValueError(
                 b'negative bundle param size: %i' % paramssize
             )
         if paramssize:
             params = self._readexact(paramssize)
             params = self._processallparams(params)
         return params
 
     def _processallparams(self, paramsblock):
         """"""
         params = util.sortdict()
         for p in paramsblock.split(b' '):
             p = p.split(b'=', 1)
             p = [urlreq.unquote(i) for i in p]
             if len(p) < 2:
                 p.append(None)
             self._processparam(*p)
             params[p[0]] = p[1]
         return params
 
     def _processparam(self, name, value):
         """process a parameter, applying its effect if needed
 
         Parameter starting with a lower case letter are advisory and will be
         ignored when unknown.  Those starting with an upper case letter are
         mandatory and will this function will raise a KeyError when unknown.
 
         Note: no option are currently supported. Any input will be either
               ignored or failing.
         """
         if not name:
             raise ValueError('empty parameter name')
         if name[0:1] not in pycompat.bytestr(
             string.ascii_letters  # pytype: disable=wrong-arg-types
         ):
             raise ValueError('non letter first character: %s' % name)
         try:
             handler = b2streamparamsmap[name.lower()]
         except KeyError:
             if name[0:1].islower():
                 indebug(self.ui, b"ignoring unknown parameter %s" % name)
             else:
                 raise error.BundleUnknownFeatureError(params=(name,))
         else:
             handler(self, name, value)
 
     def _forwardchunks(self):
         """utility to transfer a bundle2 as binary
 
         This is made necessary by the fact the 'getbundle' command over 'ssh'
         have no way to know then the reply end, relying on the bundle to be
         interpreted to know its end. This is terrible and we are sorry, but we
         needed to move forward to get general delta enabled.
         """
         yield self._magicstring
         assert 'params' not in vars(self)
         paramssize = self._unpack(_fstreamparamsize)[0]
         if paramssize < 0:
             raise error.BundleValueError(
                 b'negative bundle param size: %i' % paramssize
             )
         if paramssize:
             params = self._readexact(paramssize)
             self._processallparams(params)
             # The payload itself is decompressed below, so drop
             # the compression parameter passed down to compensate.
             outparams = []
             for p in params.split(b' '):
                 k, v = p.split(b'=', 1)
                 if k.lower() != b'compression':
                     outparams.append(p)
             outparams = b' '.join(outparams)
             yield _pack(_fstreamparamsize, len(outparams))
             yield outparams
         else:
             yield _pack(_fstreamparamsize, paramssize)
         # From there, payload might need to be decompressed
         self._fp = self._compengine.decompressorreader(self._fp)
         emptycount = 0
         while emptycount < 2:
             # so we can brainlessly loop
             assert _fpartheadersize == _fpayloadsize
             size = self._unpack(_fpartheadersize)[0]
             yield _pack(_fpartheadersize, size)
             if size:
                 emptycount = 0
             else:
                 emptycount += 1
                 continue
             if size == flaginterrupt:
                 continue
             elif size < 0:
                 raise error.BundleValueError(b'negative chunk size: %i')
             yield self._readexact(size)
 
     def iterparts(self, seekable=False):
         """yield all parts contained in the stream"""
         cls = seekableunbundlepart if seekable else unbundlepart
         # make sure param have been loaded
         self.params
         # From there, payload need to be decompressed
         self._fp = self._compengine.decompressorreader(self._fp)
         indebug(self.ui, b'start extraction of bundle2 parts')
         headerblock = self._readpartheader()
         while headerblock is not None:
             part = cls(self.ui, headerblock, self._fp)
             yield part
             # Ensure part is fully consumed so we can start reading the next
             # part.
             part.consume()
 
             headerblock = self._readpartheader()
         indebug(self.ui, b'end of bundle2 stream')
 
     def _readpartheader(self):
         """reads a part header size and return the bytes blob
 
         returns None if empty"""
         headersize = self._unpack(_fpartheadersize)[0]
         if headersize < 0:
             raise error.BundleValueError(
                 b'negative part header size: %i' % headersize
             )
         indebug(self.ui, b'part header size: %i' % headersize)
         if headersize:
             return self._readexact(headersize)
         return None
 
     def compressed(self):
         self.params  # load params
         return self._compressed
 
     def close(self):
         """close underlying file"""
         if util.safehasattr(self._fp, 'close'):
             return self._fp.close()
 
 
 formatmap = {b'20': unbundle20}
 
 b2streamparamsmap = {}
 
 
 def b2streamparamhandler(name):
     """register a handler for a stream level parameter"""
 
     def decorator(func):
         assert name not in formatmap
         b2streamparamsmap[name] = func
         return func
 
     return decorator
 
 
 @b2streamparamhandler(b'compression')
 def processcompression(unbundler, param, value):
     """read compression parameter and install payload decompression"""
     if value not in util.compengines.supportedbundletypes:
         raise error.BundleUnknownFeatureError(params=(param,), values=(value,))
     unbundler._compengine = util.compengines.forbundletype(value)
     if value is not None:
         unbundler._compressed = True
 
 
 class bundlepart(object):
     """A bundle2 part contains application level payload
 
     The part `type` is used to route the part to the application level
     handler.
 
     The part payload is contained in ``part.data``. It could be raw bytes or a
     generator of byte chunks.
 
     You can add parameters to the part using the ``addparam`` method.
     Parameters can be either mandatory (default) or advisory. Remote side
     should be able to safely ignore the advisory ones.
 
     Both data and parameters cannot be modified after the generation has begun.
     """
 
     def __init__(
         self,
         parttype,
         mandatoryparams=(),
         advisoryparams=(),
         data=b'',
         mandatory=True,
     ):
         validateparttype(parttype)
         self.id = None
         self.type = parttype
         self._data = data
         self._mandatoryparams = list(mandatoryparams)
         self._advisoryparams = list(advisoryparams)
         # checking for duplicated entries
         self._seenparams = set()
         for pname, __ in self._mandatoryparams + self._advisoryparams:
             if pname in self._seenparams:
                 raise error.ProgrammingError(b'duplicated params: %s' % pname)
             self._seenparams.add(pname)
         # status of the part's generation:
         # - None: not started,
         # - False: currently generated,
         # - True: generation done.
         self._generated = None
         self.mandatory = mandatory
 
+    @encoding.strmethod
     def __repr__(self):
         cls = b"%s.%s" % (self.__class__.__module__, self.__class__.__name__)
         return b'<%s object at %x; id: %s; type: %s; mandatory: %s>' % (
             cls,
             id(self),
             self.id,
             self.type,
             self.mandatory,
         )
 
     def copy(self):
         """return a copy of the part
 
         The new part have the very same content but no partid assigned yet.
         Parts with generated data cannot be copied."""
         assert not util.safehasattr(self.data, 'next')
         return self.__class__(
             self.type,
             self._mandatoryparams,
             self._advisoryparams,
             self._data,
             self.mandatory,
         )
 
     # methods used to defines the part content
     @property
     def data(self):
         return self._data
 
     @data.setter
     def data(self, data):
         if self._generated is not None:
             raise error.ReadOnlyPartError(b'part is being generated')
         self._data = data
 
     @property
     def mandatoryparams(self):
         # make it an immutable tuple to force people through ``addparam``
         return tuple(self._mandatoryparams)
 
     @property
     def advisoryparams(self):
         # make it an immutable tuple to force people through ``addparam``
         return tuple(self._advisoryparams)
 
     def addparam(self, name, value=b'', mandatory=True):
         """add a parameter to the part
 
         If 'mandatory' is set to True, the remote handler must claim support
         for this parameter or the unbundling will be aborted.
 
         The 'name' and 'value' cannot exceed 255 bytes each.
         """
         if self._generated is not None:
             raise error.ReadOnlyPartError(b'part is being generated')
         if name in self._seenparams:
             raise ValueError(b'duplicated params: %s' % name)
         self._seenparams.add(name)
         params = self._advisoryparams
         if mandatory:
             params = self._mandatoryparams
         params.append((name, value))
 
     # methods used to generates the bundle2 stream
     def getchunks(self, ui):
         if self._generated is not None:
             raise error.ProgrammingError(b'part can only be consumed once')
         self._generated = False
 
         if ui.debugflag:
             msg = [b'bundle2-output-part: "%s"' % self.type]
             if not self.mandatory:
                 msg.append(b' (advisory)')
             nbmp = len(self.mandatoryparams)
             nbap = len(self.advisoryparams)
             if nbmp or nbap:
                 msg.append(b' (params:')
                 if nbmp:
                     msg.append(b' %i mandatory' % nbmp)
                 if nbap:
                     msg.append(b' %i advisory' % nbmp)
                 msg.append(b')')
             if not self.data:
                 msg.append(b' empty payload')
             elif util.safehasattr(self.data, 'next') or util.safehasattr(
                 self.data, b'__next__'
             ):
                 msg.append(b' streamed payload')
             else:
                 msg.append(b' %i bytes payload' % len(self.data))
             msg.append(b'\n')
             ui.debug(b''.join(msg))
 
         #### header
         if self.mandatory:
             parttype = self.type.upper()
         else:
             parttype = self.type.lower()
         outdebug(ui, b'part %s: "%s"' % (pycompat.bytestr(self.id), parttype))
         ## parttype
         header = [
             _pack(_fparttypesize, len(parttype)),
             parttype,
             _pack(_fpartid, self.id),
         ]
         ## parameters
         # count
         manpar = self.mandatoryparams
         advpar = self.advisoryparams
         header.append(_pack(_fpartparamcount, len(manpar), len(advpar)))
         # size
         parsizes = []
         for key, value in manpar:
             parsizes.append(len(key))
             parsizes.append(len(value))
         for key, value in advpar:
             parsizes.append(len(key))
             parsizes.append(len(value))
         paramsizes = _pack(_makefpartparamsizes(len(parsizes) // 2), *parsizes)
         header.append(paramsizes)
         # key, value
         for key, value in manpar:
             header.append(key)
             header.append(value)
         for key, value in advpar:
             header.append(key)
             header.append(value)
         ## finalize header
         try:
             headerchunk = b''.join(header)
         except TypeError:
             raise TypeError(
                 'Found a non-bytes trying to '
                 'build bundle part header: %r' % header
             )
         outdebug(ui, b'header chunk size: %i' % len(headerchunk))
         yield _pack(_fpartheadersize, len(headerchunk))
         yield headerchunk
         ## payload
         try:
             for chunk in self._payloadchunks():
                 outdebug(ui, b'payload chunk size: %i' % len(chunk))
                 yield _pack(_fpayloadsize, len(chunk))
                 yield chunk
         except GeneratorExit:
             # GeneratorExit means that nobody is listening for our
             # results anyway, so just bail quickly rather than trying
             # to produce an error part.
             ui.debug(b'bundle2-generatorexit\n')
             raise
         except BaseException as exc:
             bexc = stringutil.forcebytestr(exc)
             # backup exception data for later
             ui.debug(
                 b'bundle2-input-stream-interrupt: encoding exception %s' % bexc
             )
             tb = sys.exc_info()[2]
             msg = b'unexpected error: %s' % bexc
             interpart = bundlepart(
                 b'error:abort', [(b'message', msg)], mandatory=False
             )
             interpart.id = 0
             yield _pack(_fpayloadsize, -1)
             for chunk in interpart.getchunks(ui=ui):
                 yield chunk
             outdebug(ui, b'closing payload chunk')
             # abort current part payload
             yield _pack(_fpayloadsize, 0)
             pycompat.raisewithtb(exc, tb)
         # end of payload
         outdebug(ui, b'closing payload chunk')
         yield _pack(_fpayloadsize, 0)
         self._generated = True
 
     def _payloadchunks(self):
         """yield chunks of a the part payload
 
         Exists to handle the different methods to provide data to a part."""
         # we only support fixed size data now.
         # This will be improved in the future.
         if util.safehasattr(self.data, 'next') or util.safehasattr(
             self.data, b'__next__'
         ):
             buff = util.chunkbuffer(self.data)
             chunk = buff.read(preferedchunksize)
             while chunk:
                 yield chunk
                 chunk = buff.read(preferedchunksize)
         elif len(self.data):
             yield self.data
 
 
 flaginterrupt = -1
 
 
 class interrupthandler(unpackermixin):
     """read one part and process it with restricted capability
 
     This allows to transmit exception raised on the producer size during part
     iteration while the consumer is reading a part.
 
     Part processed in this manner only have access to a ui object,"""
 
     def __init__(self, ui, fp):
         super(interrupthandler, self).__init__(fp)
         self.ui = ui
 
     def _readpartheader(self):
         """reads a part header size and return the bytes blob
 
         returns None if empty"""
         headersize = self._unpack(_fpartheadersize)[0]
         if headersize < 0:
             raise error.BundleValueError(
                 b'negative part header size: %i' % headersize
             )
         indebug(self.ui, b'part header size: %i\n' % headersize)
         if headersize:
             return self._readexact(headersize)
         return None
 
     def __call__(self):
 
         self.ui.debug(
             b'bundle2-input-stream-interrupt: opening out of band context\n'
         )
         indebug(self.ui, b'bundle2 stream interruption, looking for a part.')
         headerblock = self._readpartheader()
         if headerblock is None:
             indebug(self.ui, b'no part found during interruption.')
             return
         part = unbundlepart(self.ui, headerblock, self._fp)
         op = interruptoperation(self.ui)
         hardabort = False
         try:
             _processpart(op, part)
         except (SystemExit, KeyboardInterrupt):
             hardabort = True
             raise
         finally:
             if not hardabort:
                 part.consume()
         self.ui.debug(
             b'bundle2-input-stream-interrupt: closing out of band context\n'
         )
 
 
 class interruptoperation(object):
     """A limited operation to be use by part handler during interruption
 
     It only have access to an ui object.
     """
 
     def __init__(self, ui):
         self.ui = ui
         self.reply = None
         self.captureoutput = False
 
     @property
     def repo(self):
         raise error.ProgrammingError(b'no repo access from stream interruption')
 
     def gettransaction(self):
         raise TransactionUnavailable(b'no repo access from stream interruption')
 
 
 def decodepayloadchunks(ui, fh):
     """Reads bundle2 part payload data into chunks.
 
     Part payload data consists of framed chunks. This function takes
     a file handle and emits those chunks.
     """
     dolog = ui.configbool(b'devel', b'bundle2.debug')
     debug = ui.debug
 
     headerstruct = struct.Struct(_fpayloadsize)
     headersize = headerstruct.size
     unpack = headerstruct.unpack
 
     readexactly = changegroup.readexactly
     read = fh.read
 
     chunksize = unpack(readexactly(fh, headersize))[0]
     indebug(ui, b'payload chunk size: %i' % chunksize)
 
     # changegroup.readexactly() is inlined below for performance.
     while chunksize:
         if chunksize >= 0:
             s = read(chunksize)
             if len(s) < chunksize:
                 raise error.Abort(
                     _(
                         b'stream ended unexpectedly '
                         b' (got %d bytes, expected %d)'
                     )
                     % (len(s), chunksize)
                 )
 
             yield s
         elif chunksize == flaginterrupt:
             # Interrupt "signal" detected. The regular stream is interrupted
             # and a bundle2 part follows. Consume it.
             interrupthandler(ui, fh)()
         else:
             raise error.BundleValueError(
                 b'negative payload chunk size: %s' % chunksize
             )
 
         s = read(headersize)
         if len(s) < headersize:
             raise error.Abort(
                 _(b'stream ended unexpectedly  (got %d bytes, expected %d)')
                 % (len(s), chunksize)
             )
 
         chunksize = unpack(s)[0]
 
         # indebug() inlined for performance.
         if dolog:
             debug(b'bundle2-input: payload chunk size: %i\n' % chunksize)
 
 
 class unbundlepart(unpackermixin):
     """a bundle part read from a bundle"""
 
     def __init__(self, ui, header, fp):
         super(unbundlepart, self).__init__(fp)
         self._seekable = util.safehasattr(fp, 'seek') and util.safehasattr(
             fp, b'tell'
         )
         self.ui = ui
         # unbundle state attr
         self._headerdata = header
         self._headeroffset = 0
         self._initialized = False
         self.consumed = False
         # part data
         self.id = None
         self.type = None
         self.mandatoryparams = None
         self.advisoryparams = None
         self.params = None
         self.mandatorykeys = ()
         self._readheader()
         self._mandatory = None
         self._pos = 0
 
     def _fromheader(self, size):
         """return the next <size> byte from the header"""
         offset = self._headeroffset
         data = self._headerdata[offset : (offset + size)]
         self._headeroffset = offset + size
         return data
 
     def _unpackheader(self, format):
         """read given format from header
 
         This automatically compute the size of the format to read."""
         data = self._fromheader(struct.calcsize(format))
         return _unpack(format, data)
 
     def _initparams(self, mandatoryparams, advisoryparams):
         """internal function to setup all logic related parameters"""
         # make it read only to prevent people touching it by mistake.
         self.mandatoryparams = tuple(mandatoryparams)
         self.advisoryparams = tuple(advisoryparams)
         # user friendly UI
         self.params = util.sortdict(self.mandatoryparams)
         self.params.update(self.advisoryparams)
         self.mandatorykeys = frozenset(p[0] for p in mandatoryparams)
 
     def _readheader(self):
         """read the header and setup the object"""
         typesize = self._unpackheader(_fparttypesize)[0]
         self.type = self._fromheader(typesize)
         indebug(self.ui, b'part type: "%s"' % self.type)
         self.id = self._unpackheader(_fpartid)[0]
         indebug(self.ui, b'part id: "%s"' % pycompat.bytestr(self.id))
         # extract mandatory bit from type
         self.mandatory = self.type != self.type.lower()
         self.type = self.type.lower()
         ## reading parameters
         # param count
         mancount, advcount = self._unpackheader(_fpartparamcount)
         indebug(self.ui, b'part parameters: %i' % (mancount + advcount))
         # param size
         fparamsizes = _makefpartparamsizes(mancount + advcount)
         paramsizes = self._unpackheader(fparamsizes)
         # make it a list of couple again
         paramsizes = list(zip(paramsizes[::2], paramsizes[1::2]))
         # split mandatory from advisory
         mansizes = paramsizes[:mancount]
         advsizes = paramsizes[mancount:]
         # retrieve param value
         manparams = []
         for key, value in mansizes:
             manparams.append((self._fromheader(key), self._fromheader(value)))
         advparams = []
         for key, value in advsizes:
             advparams.append((self._fromheader(key), self._fromheader(value)))
         self._initparams(manparams, advparams)
         ## part payload
         self._payloadstream = util.chunkbuffer(self._payloadchunks())
         # we read the data, tell it
         self._initialized = True
 
     def _payloadchunks(self):
         """Generator of decoded chunks in the payload."""
         return decodepayloadchunks(self.ui, self._fp)
 
     def consume(self):
         """Read the part payload until completion.
 
         By consuming the part data, the underlying stream read offset will
         be advanced to the next part (or end of stream).
         """
         if self.consumed:
             return
 
         chunk = self.read(32768)
         while chunk:
             self._pos += len(chunk)
             chunk = self.read(32768)
 
     def read(self, size=None):
         """read payload data"""
         if not self._initialized:
             self._readheader()
         if size is None:
             data = self._payloadstream.read()
         else:
             data = self._payloadstream.read(size)
         self._pos += len(data)
         if size is None or len(data) < size:
             if not self.consumed and self._pos:
                 self.ui.debug(
                     b'bundle2-input-part: total payload size %i\n' % self._pos
                 )
             self.consumed = True
         return data
 
 
 class seekableunbundlepart(unbundlepart):
     """A bundle2 part in a bundle that is seekable.
 
     Regular ``unbundlepart`` instances can only be read once. This class
     extends ``unbundlepart`` to enable bi-directional seeking within the
     part.
 
     Bundle2 part data consists of framed chunks. Offsets when seeking
     refer to the decoded data, not the offsets in the underlying bundle2
     stream.
 
     To facilitate quickly seeking within the decoded data, instances of this
     class maintain a mapping between offsets in the underlying stream and
     the decoded payload. This mapping will consume memory in proportion
     to the number of chunks within the payload (which almost certainly
     increases in proportion with the size of the part).
     """
 
     def __init__(self, ui, header, fp):
         # (payload, file) offsets for chunk starts.
         self._chunkindex = []
 
         super(seekableunbundlepart, self).__init__(ui, header, fp)
 
     def _payloadchunks(self, chunknum=0):
         '''seek to specified chunk and start yielding data'''
         if len(self._chunkindex) == 0:
             assert chunknum == 0, b'Must start with chunk 0'
             self._chunkindex.append((0, self._tellfp()))
         else:
             assert chunknum < len(self._chunkindex), (
                 b'Unknown chunk %d' % chunknum
             )
             self._seekfp(self._chunkindex[chunknum][1])
 
         pos = self._chunkindex[chunknum][0]
 
         for chunk in decodepayloadchunks(self.ui, self._fp):
             chunknum += 1
             pos += len(chunk)
             if chunknum == len(self._chunkindex):
                 self._chunkindex.append((pos, self._tellfp()))
 
             yield chunk
 
     def _findchunk(self, pos):
         '''for a given payload position, return a chunk number and offset'''
         for chunk, (ppos, fpos) in enumerate(self._chunkindex):
             if ppos == pos:
                 return chunk, 0
             elif ppos > pos:
                 return chunk - 1, pos - self._chunkindex[chunk - 1][0]
         raise ValueError(b'Unknown chunk')
 
     def tell(self):
         return self._pos
 
     def seek(self, offset, whence=os.SEEK_SET):
         if whence == os.SEEK_SET:
             newpos = offset
         elif whence == os.SEEK_CUR:
             newpos = self._pos + offset
         elif whence == os.SEEK_END:
             if not self.consumed:
                 # Can't use self.consume() here because it advances self._pos.
                 chunk = self.read(32768)
                 while chunk:
                     chunk = self.read(32768)
             newpos = self._chunkindex[-1][0] - offset
         else:
             raise ValueError(b'Unknown whence value: %r' % (whence,))
 
         if newpos > self._chunkindex[-1][0] and not self.consumed:
             # Can't use self.consume() here because it advances self._pos.
             chunk = self.read(32768)
             while chunk:
                 chunk = self.read(32668)
 
         if not 0 <= newpos <= self._chunkindex[-1][0]:
             raise ValueError(b'Offset out of range')
 
         if self._pos != newpos:
             chunk, internaloffset = self._findchunk(newpos)
             self._payloadstream = util.chunkbuffer(self._payloadchunks(chunk))
             adjust = self.read(internaloffset)
             if len(adjust) != internaloffset:
                 raise error.Abort(_(b'Seek failed\n'))
             self._pos = newpos
 
     def _seekfp(self, offset, whence=0):
         """move the underlying file pointer
 
         This method is meant for internal usage by the bundle2 protocol only.
         They directly manipulate the low level stream including bundle2 level
         instruction.
 
         Do not use it to implement higher-level logic or methods."""
         if self._seekable:
             return self._fp.seek(offset, whence)
         else:
             raise NotImplementedError(_(b'File pointer is not seekable'))
 
     def _tellfp(self):
         """return the file offset, or None if file is not seekable
 
         This method is meant for internal usage by the bundle2 protocol only.
         They directly manipulate the low level stream including bundle2 level
         instruction.
 
         Do not use it to implement higher-level logic or methods."""
         if self._seekable:
             try:
                 return self._fp.tell()
             except IOError as e:
                 if e.errno == errno.ESPIPE:
                     self._seekable = False
                 else:
                     raise
         return None
 
 
 # These are only the static capabilities.
 # Check the 'getrepocaps' function for the rest.
 capabilities = {
     b'HG20': (),
     b'bookmarks': (),
     b'error': (b'abort', b'unsupportedcontent', b'pushraced', b'pushkey'),
     b'listkeys': (),
     b'pushkey': (),
     b'digests': tuple(sorted(util.DIGESTS.keys())),
     b'remote-changegroup': (b'http', b'https'),
     b'hgtagsfnodes': (),
     b'rev-branch-cache': (),
     b'phases': (b'heads',),
     b'stream': (b'v2',),
 }
 
 
 def getrepocaps(repo, allowpushback=False, role=None):
     """return the bundle2 capabilities for a given repo
 
     Exists to allow extensions (like evolution) to mutate the capabilities.
 
     The returned value is used for servers advertising their capabilities as
     well as clients advertising their capabilities to servers as part of
     bundle2 requests. The ``role`` argument specifies which is which.
     """
     if role not in (b'client', b'server'):
         raise error.ProgrammingError(b'role argument must be client or server')
 
     caps = capabilities.copy()
     caps[b'changegroup'] = tuple(
         sorted(changegroup.supportedincomingversions(repo))
     )
     if obsolete.isenabled(repo, obsolete.exchangeopt):
         supportedformat = tuple(b'V%i' % v for v in obsolete.formats)
         caps[b'obsmarkers'] = supportedformat
     if allowpushback:
         caps[b'pushback'] = ()
     cpmode = repo.ui.config(b'server', b'concurrent-push-mode')
     if cpmode == b'check-related':
         caps[b'checkheads'] = (b'related',)
     if b'phases' in repo.ui.configlist(b'devel', b'legacy.exchange'):
         caps.pop(b'phases')
 
     # Don't advertise stream clone support in server mode if not configured.
     if role == b'server':
         streamsupported = repo.ui.configbool(
             b'server', b'uncompressed', untrusted=True
         )
         featuresupported = repo.ui.configbool(b'server', b'bundle2.stream')
 
         if not streamsupported or not featuresupported:
             caps.pop(b'stream')
     # Else always advertise support on client, because payload support
     # should always be advertised.
 
     return caps
 
 
 def bundle2caps(remote):
     """return the bundle capabilities of a peer as dict"""
     raw = remote.capable(b'bundle2')
     if not raw and raw != b'':
         return {}
     capsblob = urlreq.unquote(remote.capable(b'bundle2'))
     return decodecaps(capsblob)
 
 
 def obsmarkersversion(caps):
     """extract the list of supported obsmarkers versions from a bundle2caps dict
     """
     obscaps = caps.get(b'obsmarkers', ())
     return [int(c[1:]) for c in obscaps if c.startswith(b'V')]
 
 
 def writenewbundle(
     ui,
     repo,
     source,
     filename,
     bundletype,
     outgoing,
     opts,
     vfs=None,
     compression=None,
     compopts=None,
 ):
     if bundletype.startswith(b'HG10'):
         cg = changegroup.makechangegroup(repo, outgoing, b'01', source)
         return writebundle(
             ui,
             cg,
             filename,
             bundletype,
             vfs=vfs,
             compression=compression,
             compopts=compopts,
         )
     elif not bundletype.startswith(b'HG20'):
         raise error.ProgrammingError(b'unknown bundle type: %s' % bundletype)
 
     caps = {}
     if b'obsolescence' in opts:
         caps[b'obsmarkers'] = (b'V1',)
     bundle = bundle20(ui, caps)
     bundle.setcompression(compression, compopts)
     _addpartsfromopts(ui, repo, bundle, source, outgoing, opts)
     chunkiter = bundle.getchunks()
 
     return changegroup.writechunks(ui, chunkiter, filename, vfs=vfs)
 
 
 def _addpartsfromopts(ui, repo, bundler, source, outgoing, opts):
     # We should eventually reconcile this logic with the one behind
     # 'exchange.getbundle2partsgenerator'.
     #
     # The type of input from 'getbundle' and 'writenewbundle' are a bit
     # different right now. So we keep them separated for now for the sake of
     # simplicity.
 
     # we might not always want a changegroup in such bundle, for example in
     # stream bundles
     if opts.get(b'changegroup', True):
         cgversion = opts.get(b'cg.version')
         if cgversion is None:
             cgversion = changegroup.safeversion(repo)
         cg = changegroup.makechangegroup(repo, outgoing, cgversion, source)
         part = bundler.newpart(b'changegroup', data=cg.getchunks())
         part.addparam(b'version', cg.version)
         if b'clcount' in cg.extras:
             part.addparam(
                 b'nbchanges', b'%d' % cg.extras[b'clcount'], mandatory=False
             )
         if opts.get(b'phases') and repo.revs(
             b'%ln and secret()', outgoing.missingheads
         ):
             part.addparam(
                 b'targetphase', b'%d' % phases.secret, mandatory=False
             )
         if b'exp-sidedata-flag' in repo.requirements:
             part.addparam(b'exp-sidedata', b'1')
 
     if opts.get(b'streamv2', False):
         addpartbundlestream2(bundler, repo, stream=True)
 
     if opts.get(b'tagsfnodescache', True):
         addparttagsfnodescache(repo, bundler, outgoing)
 
     if opts.get(b'revbranchcache', True):
         addpartrevbranchcache(repo, bundler, outgoing)
 
     if opts.get(b'obsolescence', False):
         obsmarkers = repo.obsstore.relevantmarkers(outgoing.missing)
         buildobsmarkerspart(bundler, obsmarkers)
 
     if opts.get(b'phases', False):
         headsbyphase = phases.subsetphaseheads(repo, outgoing.missing)
         phasedata = phases.binaryencode(headsbyphase)
         bundler.newpart(b'phase-heads', data=phasedata)
 
 
 def addparttagsfnodescache(repo, bundler, outgoing):
     # we include the tags fnode cache for the bundle changeset
     # (as an optional parts)
     cache = tags.hgtagsfnodescache(repo.unfiltered())
     chunks = []
 
     # .hgtags fnodes are only relevant for head changesets. While we could
     # transfer values for all known nodes, there will likely be little to
     # no benefit.
     #
     # We don't bother using a generator to produce output data because
     # a) we only have 40 bytes per head and even esoteric numbers of heads
     # consume little memory (1M heads is 40MB) b) we don't want to send the
     # part if we don't have entries and knowing if we have entries requires
     # cache lookups.
     for node in outgoing.missingheads:
         # Don't compute missing, as this may slow down serving.
         fnode = cache.getfnode(node, computemissing=False)
         if fnode is not None:
             chunks.extend([node, fnode])
 
     if chunks:
         bundler.newpart(b'hgtagsfnodes', data=b''.join(chunks))
 
 
 def addpartrevbranchcache(repo, bundler, outgoing):
     # we include the rev branch cache for the bundle changeset
     # (as an optional parts)
     cache = repo.revbranchcache()
     cl = repo.unfiltered().changelog
     branchesdata = collections.defaultdict(lambda: (set(), set()))
     for node in outgoing.missing:
         branch, close = cache.branchinfo(cl.rev(node))
         branchesdata[branch][close].add(node)
 
     def generate():
         for branch, (nodes, closed) in sorted(branchesdata.items()):
             utf8branch = encoding.fromlocal(branch)
             yield rbcstruct.pack(len(utf8branch), len(nodes), len(closed))
             yield utf8branch
             for n in sorted(nodes):
                 yield n
             for n in sorted(closed):
                 yield n
 
     bundler.newpart(b'cache:rev-branch-cache', data=generate(), mandatory=False)
 
 
 def _formatrequirementsspec(requirements):
     requirements = [req for req in requirements if req != b"shared"]
     return urlreq.quote(b','.join(sorted(requirements)))
 
 
 def _formatrequirementsparams(requirements):
     requirements = _formatrequirementsspec(requirements)
     params = b"%s%s" % (urlreq.quote(b"requirements="), requirements)
     return params
 
 
 def addpartbundlestream2(bundler, repo, **kwargs):
     if not kwargs.get('stream', False):
         return
 
     if not streamclone.allowservergeneration(repo):
         raise error.Abort(
             _(
                 b'stream data requested but server does not allow '
                 b'this feature'
             ),
             hint=_(
                 b'well-behaved clients should not be '
                 b'requesting stream data from servers not '
                 b'advertising it; the client may be buggy'
             ),
         )
 
     # Stream clones don't compress well. And compression undermines a
     # goal of stream clones, which is to be fast. Communicate the desire
     # to avoid compression to consumers of the bundle.
     bundler.prefercompressed = False
 
     # get the includes and excludes
     includepats = kwargs.get('includepats')
     excludepats = kwargs.get('excludepats')
 
     narrowstream = repo.ui.configbool(
         b'experimental', b'server.stream-narrow-clones'
     )
 
     if (includepats or excludepats) and not narrowstream:
         raise error.Abort(_(b'server does not support narrow stream clones'))
 
     includeobsmarkers = False
     if repo.obsstore:
         remoteversions = obsmarkersversion(bundler.capabilities)
         if not remoteversions:
             raise error.Abort(
                 _(
                     b'server has obsolescence markers, but client '
                     b'cannot receive them via stream clone'
                 )
             )
         elif repo.obsstore._version in remoteversions:
             includeobsmarkers = True
 
     filecount, bytecount, it = streamclone.generatev2(
         repo, includepats, excludepats, includeobsmarkers
     )
     requirements = _formatrequirementsspec(repo.requirements)
     part = bundler.newpart(b'stream2', data=it)
     part.addparam(b'bytecount', b'%d' % bytecount, mandatory=True)
     part.addparam(b'filecount', b'%d' % filecount, mandatory=True)
     part.addparam(b'requirements', requirements, mandatory=True)
 
 
 def buildobsmarkerspart(bundler, markers):
     """add an obsmarker part to the bundler with <markers>
 
     No part is created if markers is empty.
     Raises ValueError if the bundler doesn't support any known obsmarker format.
     """
     if not markers:
         return None
 
     remoteversions = obsmarkersversion(bundler.capabilities)
     version = obsolete.commonversion(remoteversions)
     if version is None:
         raise ValueError(b'bundler does not support common obsmarker format')
     stream = obsolete.encodemarkers(markers, True, version=version)
     return bundler.newpart(b'obsmarkers', data=stream)
 
 
 def writebundle(
     ui, cg, filename, bundletype, vfs=None, compression=None, compopts=None
 ):
     """Write a bundle file and return its filename.
 
     Existing files will not be overwritten.
     If no filename is specified, a temporary file is created.
     bz2 compression can be turned off.
     The bundle file will be deleted in case of errors.
     """
 
     if bundletype == b"HG20":
         bundle = bundle20(ui)
         bundle.setcompression(compression, compopts)
         part = bundle.newpart(b'changegroup', data=cg.getchunks())
         part.addparam(b'version', cg.version)
         if b'clcount' in cg.extras:
             part.addparam(
                 b'nbchanges', b'%d' % cg.extras[b'clcount'], mandatory=False
             )
         chunkiter = bundle.getchunks()
     else:
         # compression argument is only for the bundle2 case
         assert compression is None
         if cg.version != b'01':
             raise error.Abort(
                 _(b'old bundle types only supports v1 changegroups')
             )
         header, comp = bundletypes[bundletype]
         if comp not in util.compengines.supportedbundletypes:
             raise error.Abort(_(b'unknown stream compression type: %s') % comp)
         compengine = util.compengines.forbundletype(comp)
 
         def chunkiter():
             yield header
             for chunk in compengine.compressstream(cg.getchunks(), compopts):
                 yield chunk
 
         chunkiter = chunkiter()
 
     # parse the changegroup data, otherwise we will block
     # in case of sshrepo because we don't know the end of the stream
     return changegroup.writechunks(ui, chunkiter, filename, vfs=vfs)
 
 
 def combinechangegroupresults(op):
     """logic to combine 0 or more addchangegroup results into one"""
     results = [r.get(b'return', 0) for r in op.records[b'changegroup']]
     changedheads = 0
     result = 1
     for ret in results:
         # If any changegroup result is 0, return 0
         if ret == 0:
             result = 0
             break
         if ret < -1:
             changedheads += ret + 1
         elif ret > 1:
             changedheads += ret - 1
     if changedheads > 0:
         result = 1 + changedheads
     elif changedheads < 0:
         result = -1 + changedheads
     return result
 
 
 @parthandler(
     b'changegroup',
     (
         b'version',
         b'nbchanges',
         b'exp-sidedata',
         b'treemanifest',
         b'targetphase',
     ),
 )
 def handlechangegroup(op, inpart):
     """apply a changegroup part on the repo
 
     This is a very early implementation that will massive rework before being
     inflicted to any end-user.
     """
     from . import localrepo
 
     tr = op.gettransaction()
     unpackerversion = inpart.params.get(b'version', b'01')
     # We should raise an appropriate exception here
     cg = changegroup.getunbundler(unpackerversion, inpart, None)
     # the source and url passed here are overwritten by the one contained in
     # the transaction.hookargs argument. So 'bundle2' is a placeholder
     nbchangesets = None
     if b'nbchanges' in inpart.params:
         nbchangesets = int(inpart.params.get(b'nbchanges'))
     if (
         b'treemanifest' in inpart.params
         and b'treemanifest' not in op.repo.requirements
     ):
         if len(op.repo.changelog) != 0:
             raise error.Abort(
                 _(
                     b"bundle contains tree manifests, but local repo is "
                     b"non-empty and does not use tree manifests"
                 )
             )
         op.repo.requirements.add(b'treemanifest')
         op.repo.svfs.options = localrepo.resolvestorevfsoptions(
             op.repo.ui, op.repo.requirements, op.repo.features
         )
         op.repo._writerequirements()
 
     bundlesidedata = bool(b'exp-sidedata' in inpart.params)
     reposidedata = bool(b'exp-sidedata-flag' in op.repo.requirements)
     if reposidedata and not bundlesidedata:
         msg = b"repository is using sidedata but the bundle source do not"
         hint = b'this is currently unsupported'
         raise error.Abort(msg, hint=hint)
 
     extrakwargs = {}
     targetphase = inpart.params.get(b'targetphase')
     if targetphase is not None:
         extrakwargs['targetphase'] = int(targetphase)
     ret = _processchangegroup(
         op,
         cg,
         tr,
         b'bundle2',
         b'bundle2',
         expectedtotal=nbchangesets,
         **extrakwargs
     )
     if op.reply is not None:
         # This is definitely not the final form of this
         # return. But one need to start somewhere.
         part = op.reply.newpart(b'reply:changegroup', mandatory=False)
         part.addparam(
             b'in-reply-to', pycompat.bytestr(inpart.id), mandatory=False
         )
         part.addparam(b'return', b'%i' % ret, mandatory=False)
     assert not inpart.read()
 
 
 _remotechangegroupparams = tuple(
     [b'url', b'size', b'digests']
     + [b'digest:%s' % k for k in util.DIGESTS.keys()]
 )
 
 
 @parthandler(b'remote-changegroup', _remotechangegroupparams)
 def handleremotechangegroup(op, inpart):
     """apply a bundle10 on the repo, given an url and validation information
 
     All the information about the remote bundle to import are given as
     parameters. The parameters include:
       - url: the url to the bundle10.
       - size: the bundle10 file size. It is used to validate what was
         retrieved by the client matches the server knowledge about the bundle.
       - digests: a space separated list of the digest types provided as
         parameters.
       - digest:<digest-type>: the hexadecimal representation of the digest with
         that name. Like the size, it is used to validate what was retrieved by
         the client matches what the server knows about the bundle.
 
     When multiple digest types are given, all of them are checked.
     """
     try:
         raw_url = inpart.params[b'url']
     except KeyError:
         raise error.Abort(_(b'remote-changegroup: missing "%s" param') % b'url')
     parsed_url = util.url(raw_url)
     if parsed_url.scheme not in capabilities[b'remote-changegroup']:
         raise error.Abort(
             _(b'remote-changegroup does not support %s urls')
             % parsed_url.scheme
         )
 
     try:
         size = int(inpart.params[b'size'])
     except ValueError:
         raise error.Abort(
             _(b'remote-changegroup: invalid value for param "%s"') % b'size'
         )
     except KeyError:
         raise error.Abort(
             _(b'remote-changegroup: missing "%s" param') % b'size'
         )
 
     digests = {}
     for typ in inpart.params.get(b'digests', b'').split():
         param = b'digest:%s' % typ
         try:
             value = inpart.params[param]
         except KeyError:
             raise error.Abort(
                 _(b'remote-changegroup: missing "%s" param') % param
             )
         digests[typ] = value
 
     real_part = util.digestchecker(url.open(op.ui, raw_url), size, digests)
 
     tr = op.gettransaction()
     from . import exchange
 
     cg = exchange.readbundle(op.repo.ui, real_part, raw_url)
     if not isinstance(cg, changegroup.cg1unpacker):
         raise error.Abort(
             _(b'%s: not a bundle version 1.0') % util.hidepassword(raw_url)
         )
     ret = _processchangegroup(op, cg, tr, b'bundle2', b'bundle2')
     if op.reply is not None:
         # This is definitely not the final form of this
         # return. But one need to start somewhere.
         part = op.reply.newpart(b'reply:changegroup')
         part.addparam(
             b'in-reply-to', pycompat.bytestr(inpart.id), mandatory=False
         )
         part.addparam(b'return', b'%i' % ret, mandatory=False)
     try:
         real_part.validate()
     except error.Abort as e:
         raise error.Abort(
             _(b'bundle at %s is corrupted:\n%s')
             % (util.hidepassword(raw_url), bytes(e))
         )
     assert not inpart.read()
 
 
 @parthandler(b'reply:changegroup', (b'return', b'in-reply-to'))
 def handlereplychangegroup(op, inpart):
     ret = int(inpart.params[b'return'])
     replyto = int(inpart.params[b'in-reply-to'])
     op.records.add(b'changegroup', {b'return': ret}, replyto)
 
 
 @parthandler(b'check:bookmarks')
 def handlecheckbookmarks(op, inpart):
     """check location of bookmarks
 
     This part is to be used to detect push race regarding bookmark, it
     contains binary encoded (bookmark, node) tuple. If the local state does
     not marks the one in the part, a PushRaced exception is raised
     """
     bookdata = bookmarks.binarydecode(inpart)
 
     msgstandard = (
         b'remote repository changed while pushing - please try again '
         b'(bookmark "%s" move from %s to %s)'
     )
     msgmissing = (
         b'remote repository changed while pushing - please try again '
         b'(bookmark "%s" is missing, expected %s)'
     )
     msgexist = (
         b'remote repository changed while pushing - please try again '
         b'(bookmark "%s" set on %s, expected missing)'
     )
     for book, node in bookdata:
         currentnode = op.repo._bookmarks.get(book)
         if currentnode != node:
             if node is None:
                 finalmsg = msgexist % (book, nodemod.short(currentnode))
             elif currentnode is None:
                 finalmsg = msgmissing % (book, nodemod.short(node))
             else:
                 finalmsg = msgstandard % (
                     book,
                     nodemod.short(node),
                     nodemod.short(currentnode),
                 )
             raise error.PushRaced(finalmsg)
 
 
 @parthandler(b'check:heads')
 def handlecheckheads(op, inpart):
     """check that head of the repo did not change
 
     This is used to detect a push race when using unbundle.
     This replaces the "heads" argument of unbundle."""
     h = inpart.read(20)
     heads = []
     while len(h) == 20:
         heads.append(h)
         h = inpart.read(20)
     assert not h
     # Trigger a transaction so that we are guaranteed to have the lock now.
     if op.ui.configbool(b'experimental', b'bundle2lazylocking'):
         op.gettransaction()
     if sorted(heads) != sorted(op.repo.heads()):
         raise error.PushRaced(
             b'remote repository changed while pushing - please try again'
         )
 
 
 @parthandler(b'check:updated-heads')
 def handlecheckupdatedheads(op, inpart):
     """check for race on the heads touched by a push
 
     This is similar to 'check:heads' but focus on the heads actually updated
     during the push. If other activities happen on unrelated heads, it is
     ignored.
 
     This allow server with high traffic to avoid push contention as long as
     unrelated parts of the graph are involved."""
     h = inpart.read(20)
     heads = []
     while len(h) == 20:
         heads.append(h)
         h = inpart.read(20)
     assert not h
     # trigger a transaction so that we are guaranteed to have the lock now.
     if op.ui.configbool(b'experimental', b'bundle2lazylocking'):
         op.gettransaction()
 
     currentheads = set()
     for ls in op.repo.branchmap().iterheads():
         currentheads.update(ls)
 
     for h in heads:
         if h not in currentheads:
             raise error.PushRaced(
                 b'remote repository changed while pushing - '
                 b'please try again'
             )
 
 
 @parthandler(b'check:phases')
 def handlecheckphases(op, inpart):
     """check that phase boundaries of the repository did not change
 
     This is used to detect a push race.
     """
     phasetonodes = phases.binarydecode(inpart)
     unfi = op.repo.unfiltered()
     cl = unfi.changelog
     phasecache = unfi._phasecache
     msg = (
         b'remote repository changed while pushing - please try again '
         b'(%s is %s expected %s)'
     )
     for expectedphase, nodes in enumerate(phasetonodes):
         for n in nodes:
             actualphase = phasecache.phase(unfi, cl.rev(n))
             if actualphase != expectedphase:
                 finalmsg = msg % (
                     nodemod.short(n),
                     phases.phasenames[actualphase],
                     phases.phasenames[expectedphase],
                 )
                 raise error.PushRaced(finalmsg)
 
 
 @parthandler(b'output')
 def handleoutput(op, inpart):
     """forward output captured on the server to the client"""
     for line in inpart.read().splitlines():
         op.ui.status(_(b'remote: %s\n') % line)
 
 
 @parthandler(b'replycaps')
 def handlereplycaps(op, inpart):
     """Notify that a reply bundle should be created
 
     The payload contains the capabilities information for the reply"""
     caps = decodecaps(inpart.read())
     if op.reply is None:
         op.reply = bundle20(op.ui, caps)
 
 
 class AbortFromPart(error.Abort):
     """Sub-class of Abort that denotes an error from a bundle2 part."""
 
 
 @parthandler(b'error:abort', (b'message', b'hint'))
 def handleerrorabort(op, inpart):
     """Used to transmit abort error over the wire"""
     raise AbortFromPart(
         inpart.params[b'message'], hint=inpart.params.get(b'hint')
     )
 
 
 @parthandler(
     b'error:pushkey',
     (b'namespace', b'key', b'new', b'old', b'ret', b'in-reply-to'),
 )
 def handleerrorpushkey(op, inpart):
     """Used to transmit failure of a mandatory pushkey over the wire"""
     kwargs = {}
     for name in (b'namespace', b'key', b'new', b'old', b'ret'):
         value = inpart.params.get(name)
         if value is not None:
             kwargs[name] = value
     raise error.PushkeyFailed(
         inpart.params[b'in-reply-to'], **pycompat.strkwargs(kwargs)
     )
 
 
 @parthandler(b'error:unsupportedcontent', (b'parttype', b'params'))
 def handleerrorunsupportedcontent(op, inpart):
     """Used to transmit unknown content error over the wire"""
     kwargs = {}
     parttype = inpart.params.get(b'parttype')
     if parttype is not None:
         kwargs[b'parttype'] = parttype
     params = inpart.params.get(b'params')
     if params is not None:
         kwargs[b'params'] = params.split(b'\0')
 
     raise error.BundleUnknownFeatureError(**pycompat.strkwargs(kwargs))
 
 
 @parthandler(b'error:pushraced', (b'message',))
 def handleerrorpushraced(op, inpart):
     """Used to transmit push race error over the wire"""
     raise error.ResponseError(_(b'push failed:'), inpart.params[b'message'])
 
 
 @parthandler(b'listkeys', (b'namespace',))
 def handlelistkeys(op, inpart):
     """retrieve pushkey namespace content stored in a bundle2"""
     namespace = inpart.params[b'namespace']
     r = pushkey.decodekeys(inpart.read())
     op.records.add(b'listkeys', (namespace, r))
 
 
 @parthandler(b'pushkey', (b'namespace', b'key', b'old', b'new'))
 def handlepushkey(op, inpart):
     """process a pushkey request"""
     dec = pushkey.decode
     namespace = dec(inpart.params[b'namespace'])
     key = dec(inpart.params[b'key'])
     old = dec(inpart.params[b'old'])
     new = dec(inpart.params[b'new'])
     # Grab the transaction to ensure that we have the lock before performing the
     # pushkey.
     if op.ui.configbool(b'experimental', b'bundle2lazylocking'):
         op.gettransaction()
     ret = op.repo.pushkey(namespace, key, old, new)
     record = {b'namespace': namespace, b'key': key, b'old': old, b'new': new}
     op.records.add(b'pushkey', record)
     if op.reply is not None:
         rpart = op.reply.newpart(b'reply:pushkey')
         rpart.addparam(
             b'in-reply-to', pycompat.bytestr(inpart.id), mandatory=False
         )
         rpart.addparam(b'return', b'%i' % ret, mandatory=False)
     if inpart.mandatory and not ret:
         kwargs = {}
         for key in (b'namespace', b'key', b'new', b'old', b'ret'):
             if key in inpart.params:
                 kwargs[key] = inpart.params[key]
         raise error.PushkeyFailed(
             partid=b'%d' % inpart.id, **pycompat.strkwargs(kwargs)
         )
 
 
 @parthandler(b'bookmarks')
 def handlebookmark(op, inpart):
     """transmit bookmark information
 
     The part contains binary encoded bookmark information.
 
     The exact behavior of this part can be controlled by the 'bookmarks' mode
     on the bundle operation.
 
     When mode is 'apply' (the default) the bookmark information is applied as
     is to the unbundling repository. Make sure a 'check:bookmarks' part is
     issued earlier to check for push races in such update. This behavior is
     suitable for pushing.
 
     When mode is 'records', the information is recorded into the 'bookmarks'
     records of the bundle operation. This behavior is suitable for pulling.
     """
     changes = bookmarks.binarydecode(inpart)
 
     pushkeycompat = op.repo.ui.configbool(
         b'server', b'bookmarks-pushkey-compat'
     )
     bookmarksmode = op.modes.get(b'bookmarks', b'apply')
 
     if bookmarksmode == b'apply':
         tr = op.gettransaction()
         bookstore = op.repo._bookmarks
         if pushkeycompat:
             allhooks = []
             for book, node in changes:
                 hookargs = tr.hookargs.copy()
                 hookargs[b'pushkeycompat'] = b'1'
                 hookargs[b'namespace'] = b'bookmarks'
                 hookargs[b'key'] = book
                 hookargs[b'old'] = nodemod.hex(bookstore.get(book, b''))
                 hookargs[b'new'] = nodemod.hex(
                     node if node is not None else b''
                 )
                 allhooks.append(hookargs)
 
             for hookargs in allhooks:
                 op.repo.hook(
                     b'prepushkey', throw=True, **pycompat.strkwargs(hookargs)
                 )
 
         bookstore.applychanges(op.repo, op.gettransaction(), changes)
 
         if pushkeycompat:
 
             def runhook(unused_success):
                 for hookargs in allhooks:
                     op.repo.hook(b'pushkey', **pycompat.strkwargs(hookargs))
 
             op.repo._afterlock(runhook)
 
     elif bookmarksmode == b'records':
         for book, node in changes:
             record = {b'bookmark': book, b'node': node}
             op.records.add(b'bookmarks', record)
     else:
         raise error.ProgrammingError(
             b'unkown bookmark mode: %s' % bookmarksmode
         )
 
 
 @parthandler(b'phase-heads')
 def handlephases(op, inpart):
     """apply phases from bundle part to repo"""
     headsbyphase = phases.binarydecode(inpart)
     phases.updatephases(op.repo.unfiltered(), op.gettransaction, headsbyphase)
 
 
 @parthandler(b'reply:pushkey', (b'return', b'in-reply-to'))
 def handlepushkeyreply(op, inpart):
     """retrieve the result of a pushkey request"""
     ret = int(inpart.params[b'return'])
     partid = int(inpart.params[b'in-reply-to'])
     op.records.add(b'pushkey', {b'return': ret}, partid)
 
 
 @parthandler(b'obsmarkers')
 def handleobsmarker(op, inpart):
     """add a stream of obsmarkers to the repo"""
     tr = op.gettransaction()
     markerdata = inpart.read()
     if op.ui.config(b'experimental', b'obsmarkers-exchange-debug'):
         op.ui.writenoi18n(
             b'obsmarker-exchange: %i bytes received\n' % len(markerdata)
         )
     # The mergemarkers call will crash if marker creation is not enabled.
     # we want to avoid this if the part is advisory.
     if not inpart.mandatory and op.repo.obsstore.readonly:
         op.repo.ui.debug(
             b'ignoring obsolescence markers, feature not enabled\n'
         )
         return
     new = op.repo.obsstore.mergemarkers(tr, markerdata)
     op.repo.invalidatevolatilesets()
     op.records.add(b'obsmarkers', {b'new': new})
     if op.reply is not None:
         rpart = op.reply.newpart(b'reply:obsmarkers')
         rpart.addparam(
             b'in-reply-to', pycompat.bytestr(inpart.id), mandatory=False
         )
         rpart.addparam(b'new', b'%i' % new, mandatory=False)
 
 
 @parthandler(b'reply:obsmarkers', (b'new', b'in-reply-to'))
 def handleobsmarkerreply(op, inpart):
     """retrieve the result of a pushkey request"""
     ret = int(inpart.params[b'new'])
     partid = int(inpart.params[b'in-reply-to'])
     op.records.add(b'obsmarkers', {b'new': ret}, partid)
 
 
 @parthandler(b'hgtagsfnodes')
 def handlehgtagsfnodes(op, inpart):
     """Applies .hgtags fnodes cache entries to the local repo.
 
     Payload is pairs of 20 byte changeset nodes and filenodes.
     """
     # Grab the transaction so we ensure that we have the lock at this point.
     if op.ui.configbool(b'experimental', b'bundle2lazylocking'):
         op.gettransaction()
     cache = tags.hgtagsfnodescache(op.repo.unfiltered())
 
     count = 0
     while True:
         node = inpart.read(20)
         fnode = inpart.read(20)
         if len(node) < 20 or len(fnode) < 20:
             op.ui.debug(b'ignoring incomplete received .hgtags fnodes data\n')
             break
         cache.setfnode(node, fnode)
         count += 1
 
     cache.write()
     op.ui.debug(b'applied %i hgtags fnodes cache entries\n' % count)
 
 
 rbcstruct = struct.Struct(b'>III')
 
 
 @parthandler(b'cache:rev-branch-cache')
 def handlerbc(op, inpart):
     """receive a rev-branch-cache payload and update the local cache
 
     The payload is a series of data related to each branch
 
     1) branch name length
     2) number of open heads
     3) number of closed heads
     4) open heads nodes
     5) closed heads nodes
     """
     total = 0
     rawheader = inpart.read(rbcstruct.size)
     cache = op.repo.revbranchcache()
     cl = op.repo.unfiltered().changelog
     while rawheader:
         header = rbcstruct.unpack(rawheader)
         total += header[1] + header[2]
         utf8branch = inpart.read(header[0])
         branch = encoding.tolocal(utf8branch)
         for x in pycompat.xrange(header[1]):
             node = inpart.read(20)
             rev = cl.rev(node)
             cache.setdata(branch, rev, node, False)
         for x in pycompat.xrange(header[2]):
             node = inpart.read(20)
             rev = cl.rev(node)
             cache.setdata(branch, rev, node, True)
         rawheader = inpart.read(rbcstruct.size)
     cache.write()
 
 
 @parthandler(b'pushvars')
 def bundle2getvars(op, part):
     '''unbundle a bundle2 containing shellvars on the server'''
     # An option to disable unbundling on server-side for security reasons
     if op.ui.configbool(b'push', b'pushvars.server'):
         hookargs = {}
         for key, value in part.advisoryparams:
             key = key.upper()
             # We want pushed variables to have USERVAR_ prepended so we know
             # they came from the --pushvar flag.
             key = b"USERVAR_" + key
             hookargs[key] = value
         op.addhookargs(hookargs)
 
 
 @parthandler(b'stream2', (b'requirements', b'filecount', b'bytecount'))
 def handlestreamv2bundle(op, part):
 
     requirements = urlreq.unquote(part.params[b'requirements']).split(b',')
     filecount = int(part.params[b'filecount'])
     bytecount = int(part.params[b'bytecount'])
 
     repo = op.repo
     if len(repo):
         msg = _(b'cannot apply stream clone to non empty repository')
         raise error.Abort(msg)
 
     repo.ui.debug(b'applying stream bundle\n')
     streamclone.applybundlev2(repo, part, filecount, bytecount, requirements)
 
 
 def widen_bundle(
     bundler, repo, oldmatcher, newmatcher, common, known, cgversion, ellipses
 ):
     """generates bundle2 for widening a narrow clone
 
     bundler is the bundle to which data should be added
     repo is the localrepository instance
     oldmatcher matches what the client already has
     newmatcher matches what the client needs (including what it already has)
     common is set of common heads between server and client
     known is a set of revs known on the client side (used in ellipses)
     cgversion is the changegroup version to send
     ellipses is boolean value telling whether to send ellipses data or not
 
     returns bundle2 of the data required for extending
     """
     commonnodes = set()
     cl = repo.changelog
     for r in repo.revs(b"::%ln", common):
         commonnodes.add(cl.node(r))
     if commonnodes:
         # XXX: we should only send the filelogs (and treemanifest). user
         # already has the changelog and manifest
         packer = changegroup.getbundler(
             cgversion,
             repo,
             oldmatcher=oldmatcher,
             matcher=newmatcher,
             fullnodes=commonnodes,
         )
         cgdata = packer.generate(
             {nodemod.nullid},
             list(commonnodes),
             False,
             b'narrow_widen',
             changelog=False,
         )
 
         part = bundler.newpart(b'changegroup', data=cgdata)
         part.addparam(b'version', cgversion)
         if b'treemanifest' in repo.requirements:
             part.addparam(b'treemanifest', b'1')
         if b'exp-sidedata-flag' in repo.requirements:
             part.addparam(b'exp-sidedata', b'1')
 
     return bundler

mercurial/linelog.py

@@ -1,467 +1,467 b''
 # linelog - efficient cache for annotate data
 #
 # Copyright 2018 Google LLC.
 #
 # This software may be used and distributed according to the terms of the
 # GNU General Public License version 2 or any later version.
 """linelog is an efficient cache for annotate data inspired by SCCS Weaves.
 
 SCCS Weaves are an implementation of
 https://en.wikipedia.org/wiki/Interleaved_deltas. See
 mercurial/helptext/internals/linelog.txt for an exploration of SCCS weaves
 and how linelog works in detail.
 
 Here's a hacker's summary: a linelog is a program which is executed in
 the context of a revision. Executing the program emits information
 about lines, including the revision that introduced them and the line
 number in the file at the introducing revision. When an insertion or
 deletion is performed on the file, a jump instruction is used to patch
 in a new body of annotate information.
 """
 from __future__ import absolute_import, print_function
 
 import abc
 import struct
 
 from .thirdparty import attr
 from . import pycompat
 
 _llentry = struct.Struct(b'>II')
 
 
 class LineLogError(Exception):
     """Error raised when something bad happens internally in linelog."""
 
 
 @attr.s
 class lineinfo(object):
     # Introducing revision of this line.
     rev = attr.ib()
     # Line number for this line in its introducing revision.
     linenum = attr.ib()
     # Private. Offset in the linelog program of this line. Used internally.
     _offset = attr.ib()
 
 
 @attr.s
 class annotateresult(object):
     rev = attr.ib()
     lines = attr.ib()
     _eof = attr.ib()
 
     def __iter__(self):
         return iter(self.lines)
 
 
 class _llinstruction(object):  # pytype: disable=ignored-metaclass
 
     __metaclass__ = abc.ABCMeta
 
     @abc.abstractmethod
     def __init__(self, op1, op2):
         pass
 
     @abc.abstractmethod
     def __str__(self):
         pass
 
     def __repr__(self):
         return str(self)
 
     @abc.abstractmethod
     def __eq__(self, other):
         pass
 
     @abc.abstractmethod
     def encode(self):
         """Encode this instruction to the binary linelog format."""
 
     @abc.abstractmethod
     def execute(self, rev, pc, emit):
         """Execute this instruction.
 
         Args:
           rev: The revision we're annotating.
           pc: The current offset in the linelog program.
           emit: A function that accepts a single lineinfo object.
 
         Returns:
           The new value of pc. Returns None if exeuction should stop
           (that is, we've found the end of the file.)
         """
 
 
 class _jge(_llinstruction):
     """If the current rev is greater than or equal to op1, jump to op2."""
 
     def __init__(self, op1, op2):
         self._cmprev = op1
         self._target = op2
 
     def __str__(self):
         return 'JGE %d %d' % (self._cmprev, self._target)
 
     def __eq__(self, other):
         return (
             type(self) == type(other)
             and self._cmprev == other._cmprev
             and self._target == other._target
         )
 
     def encode(self):
         return _llentry.pack(self._cmprev << 2, self._target)
 
     def execute(self, rev, pc, emit):
         if rev >= self._cmprev:
             return self._target
         return pc + 1
 
 
 class _jump(_llinstruction):
     """Unconditional jumps are expressed as a JGE with op1 set to 0."""
 
     def __init__(self, op1, op2):
         if op1 != 0:
             raise LineLogError(b"malformed JUMP, op1 must be 0, got %d" % op1)
         self._target = op2
 
     def __str__(self):
         return 'JUMP %d' % (self._target)
 
     def __eq__(self, other):
         return type(self) == type(other) and self._target == other._target
 
     def encode(self):
         return _llentry.pack(0, self._target)
 
     def execute(self, rev, pc, emit):
         return self._target
 
 
 class _eof(_llinstruction):
     """EOF is expressed as a JGE that always jumps to 0."""
 
     def __init__(self, op1, op2):
         if op1 != 0:
             raise LineLogError(b"malformed EOF, op1 must be 0, got %d" % op1)
         if op2 != 0:
             raise LineLogError(b"malformed EOF, op2 must be 0, got %d" % op2)
 
     def __str__(self):
         return r'EOF'
 
     def __eq__(self, other):
         return type(self) == type(other)
 
     def encode(self):
         return _llentry.pack(0, 0)
 
     def execute(self, rev, pc, emit):
         return None
 
 
 class _jl(_llinstruction):
     """If the current rev is less than op1, jump to op2."""
 
     def __init__(self, op1, op2):
         self._cmprev = op1
         self._target = op2
 
     def __str__(self):
         return 'JL %d %d' % (self._cmprev, self._target)
 
     def __eq__(self, other):
         return (
             type(self) == type(other)
             and self._cmprev == other._cmprev
             and self._target == other._target
         )
 
     def encode(self):
         return _llentry.pack(1 | (self._cmprev << 2), self._target)
 
     def execute(self, rev, pc, emit):
         if rev < self._cmprev:
             return self._target
         return pc + 1
 
 
 class _line(_llinstruction):
     """Emit a line."""
 
     def __init__(self, op1, op2):
         # This line was introduced by this revision number.
         self._rev = op1
         # This line had the specified line number in the introducing revision.
         self._origlineno = op2
 
     def __str__(self):
         return 'LINE %d %d' % (self._rev, self._origlineno)
 
     def __eq__(self, other):
         return (
             type(self) == type(other)
             and self._rev == other._rev
             and self._origlineno == other._origlineno
         )
 
     def encode(self):
         return _llentry.pack(2 | (self._rev << 2), self._origlineno)
 
     def execute(self, rev, pc, emit):
         emit(lineinfo(self._rev, self._origlineno, pc))
         return pc + 1
 
 
 def _decodeone(data, offset):
     """Decode a single linelog instruction from an offset in a buffer."""
     try:
         op1, op2 = _llentry.unpack_from(data, offset)
     except struct.error as e:
         raise LineLogError(b'reading an instruction failed: %r' % e)
     opcode = op1 & 0b11
     op1 = op1 >> 2
     if opcode == 0:
         if op1 == 0:
             if op2 == 0:
                 return _eof(op1, op2)
             return _jump(op1, op2)
         return _jge(op1, op2)
     elif opcode == 1:
         return _jl(op1, op2)
     elif opcode == 2:
         return _line(op1, op2)
     raise NotImplementedError(b'Unimplemented opcode %r' % opcode)
 
 
 class linelog(object):
     """Efficient cache for per-line history information."""
 
     def __init__(self, program=None, maxrev=0):
         if program is None:
             # We pad the program with an extra leading EOF so that our
             # offsets will match the C code exactly. This means we can
             # interoperate with the C code.
             program = [_eof(0, 0), _eof(0, 0)]
         self._program = program
         self._lastannotate = None
         self._maxrev = maxrev
 
     def __eq__(self, other):
         return (
             type(self) == type(other)
             and self._program == other._program
             and self._maxrev == other._maxrev
         )
 
     def __repr__(self):
-        return b'<linelog at %s: maxrev=%d size=%d>' % (
+        return '<linelog at %s: maxrev=%d size=%d>' % (
             hex(id(self)),
             self._maxrev,
             len(self._program),
         )
 
     def debugstr(self):
         fmt = '%%%dd %%s' % len(str(len(self._program)))
         return pycompat.sysstr(b'\n').join(
             fmt % (idx, i) for idx, i in enumerate(self._program[1:], 1)
         )
 
     @classmethod
     def fromdata(cls, buf):
         if len(buf) % _llentry.size != 0:
             raise LineLogError(
                 b"invalid linelog buffer size %d (must be a multiple of %d)"
                 % (len(buf), _llentry.size)
             )
         expected = len(buf) / _llentry.size
         fakejge = _decodeone(buf, 0)
         if isinstance(fakejge, _jump):
             maxrev = 0
         elif isinstance(fakejge, (_jge, _jl)):
             maxrev = fakejge._cmprev
         else:
             raise LineLogError(
                 'Expected one of _jump, _jge, or _jl. Got %s.'
                 % type(fakejge).__name__
             )
         assert isinstance(fakejge, (_jump, _jge, _jl))  # help pytype
         numentries = fakejge._target
         if expected != numentries:
             raise LineLogError(
                 b"corrupt linelog data: claimed"
                 b" %d entries but given data for %d entries"
                 % (expected, numentries)
             )
         instructions = [_eof(0, 0)]
         for offset in pycompat.xrange(1, numentries):
             instructions.append(_decodeone(buf, offset * _llentry.size))
         return cls(instructions, maxrev=maxrev)
 
     def encode(self):
         hdr = _jge(self._maxrev, len(self._program)).encode()
         return hdr + b''.join(i.encode() for i in self._program[1:])
 
     def clear(self):
         self._program = []
         self._maxrev = 0
         self._lastannotate = None
 
     def replacelines_vec(self, rev, a1, a2, blines):
         return self.replacelines(
             rev, a1, a2, 0, len(blines), _internal_blines=blines
         )
 
     def replacelines(self, rev, a1, a2, b1, b2, _internal_blines=None):
         """Replace lines [a1, a2) with lines [b1, b2)."""
         if self._lastannotate:
             # TODO(augie): make replacelines() accept a revision at
             # which we're editing as well as a revision to mark
             # responsible for the edits. In hg-experimental it's
             # stateful like this, so we're doing the same thing to
             # retain compatibility with absorb until that's imported.
             ar = self._lastannotate
         else:
             ar = self.annotate(rev)
             #        ar = self.annotate(self._maxrev)
         if a1 > len(ar.lines):
             raise LineLogError(
                 b'%d contains %d lines, tried to access line %d'
                 % (rev, len(ar.lines), a1)
             )
         elif a1 == len(ar.lines):
             # Simulated EOF instruction since we're at EOF, which
             # doesn't have a "real" line.
             a1inst = _eof(0, 0)
             a1info = lineinfo(0, 0, ar._eof)
         else:
             a1info = ar.lines[a1]
             a1inst = self._program[a1info._offset]
         programlen = self._program.__len__
         oldproglen = programlen()
         appendinst = self._program.append
 
         # insert
         blineinfos = []
         bappend = blineinfos.append
         if b1 < b2:
             # Determine the jump target for the JGE at the start of
             # the new block.
             tgt = oldproglen + (b2 - b1 + 1)
             # Jump to skip the insert if we're at an older revision.
             appendinst(_jl(rev, tgt))
             for linenum in pycompat.xrange(b1, b2):
                 if _internal_blines is None:
                     bappend(lineinfo(rev, linenum, programlen()))
                     appendinst(_line(rev, linenum))
                 else:
                     newrev, newlinenum = _internal_blines[linenum]
                     bappend(lineinfo(newrev, newlinenum, programlen()))
                     appendinst(_line(newrev, newlinenum))
         # delete
         if a1 < a2:
             if a2 > len(ar.lines):
                 raise LineLogError(
                     b'%d contains %d lines, tried to access line %d'
                     % (rev, len(ar.lines), a2)
                 )
             elif a2 == len(ar.lines):
                 endaddr = ar._eof
             else:
                 endaddr = ar.lines[a2]._offset
             if a2 > 0 and rev < self._maxrev:
                 # If we're here, we're deleting a chunk of an old
                 # commit, so we need to be careful and not touch
                 # invisible lines between a2-1 and a2 (IOW, lines that
                 # are added later).
                 endaddr = ar.lines[a2 - 1]._offset + 1
             appendinst(_jge(rev, endaddr))
         # copy instruction from a1
         a1instpc = programlen()
         appendinst(a1inst)
         # if a1inst isn't a jump or EOF, then we need to add an unconditional
         # jump back into the program here.
         if not isinstance(a1inst, (_jump, _eof)):
             appendinst(_jump(0, a1info._offset + 1))
         # Patch instruction at a1, which makes our patch live.
         self._program[a1info._offset] = _jump(0, oldproglen)
 
         # Update self._lastannotate in place. This serves as a cache to avoid
         # expensive "self.annotate" in this function, when "replacelines" is
         # used continuously.
         if len(self._lastannotate.lines) > a1:
             self._lastannotate.lines[a1]._offset = a1instpc
         else:
             assert isinstance(a1inst, _eof)
             self._lastannotate._eof = a1instpc
         self._lastannotate.lines[a1:a2] = blineinfos
         self._lastannotate.rev = max(self._lastannotate.rev, rev)
 
         if rev > self._maxrev:
             self._maxrev = rev
 
     def annotate(self, rev):
         pc = 1
         lines = []
         executed = 0
         # Sanity check: if instructions executed exceeds len(program), we
         # hit an infinite loop in the linelog program somehow and we
         # should stop.
         while pc is not None and executed < len(self._program):
             inst = self._program[pc]
             lastpc = pc
             pc = inst.execute(rev, pc, lines.append)
             executed += 1
         if pc is not None:
             raise LineLogError(
                 r'Probably hit an infinite loop in linelog. Program:\n'
                 + self.debugstr()
             )
         ar = annotateresult(rev, lines, lastpc)
         self._lastannotate = ar
         return ar
 
     @property
     def maxrev(self):
         return self._maxrev
 
     # Stateful methods which depend on the value of the last
     # annotation run. This API is for compatiblity with the original
     # linelog, and we should probably consider refactoring it.
     @property
     def annotateresult(self):
         """Return the last annotation result. C linelog code exposed this."""
         return [(l.rev, l.linenum) for l in self._lastannotate.lines]
 
     def getoffset(self, line):
         return self._lastannotate.lines[line]._offset
 
     def getalllines(self, start=0, end=0):
         """Get all lines that ever occurred in [start, end).
 
         Passing start == end == 0 means "all lines ever".
 
         This works in terms of *internal* program offsets, not line numbers.
         """
         pc = start or 1
         lines = []
         # only take as many steps as there are instructions in the
         # program - if we don't find an EOF or our stop-line before
         # then, something is badly broken.
         for step in pycompat.xrange(len(self._program)):
             inst = self._program[pc]
             nextpc = pc + 1
             if isinstance(inst, _jump):
                 nextpc = inst._target
             elif isinstance(inst, _eof):
                 return lines
             elif isinstance(inst, (_jl, _jge)):
                 pass
             elif isinstance(inst, _line):
                 lines.append((inst._rev, inst._origlineno))
             else:
                 raise LineLogError(b"Illegal instruction %r" % inst)
             if nextpc == end:
                 return lines
             pc = nextpc
         raise LineLogError(b"Failed to perform getalllines")

mercurial/manifest.py

@@ -1,2273 +1,2275 b''
 # manifest.py - manifest revision class for mercurial
 #
 # Copyright 2005-2007 Matt 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.
 
 from __future__ import absolute_import
 
 import heapq
 import itertools
 import struct
 import weakref
 
 from .i18n import _
 from .node import (
     bin,
     hex,
     nullid,
     nullrev,
 )
 from .pycompat import getattr
 from . import (
+    encoding,
     error,
     mdiff,
     pathutil,
     policy,
     pycompat,
     revlog,
     util,
 )
 from .interfaces import (
     repository,
     util as interfaceutil,
 )
 
 parsers = policy.importmod('parsers')
 propertycache = util.propertycache
 
 # Allow tests to more easily test the alternate path in manifestdict.fastdelta()
 FASTDELTA_TEXTDIFF_THRESHOLD = 1000
 
 
 def _parse(data):
     # This method does a little bit of excessive-looking
     # precondition checking. This is so that the behavior of this
     # class exactly matches its C counterpart to try and help
     # prevent surprise breakage for anyone that develops against
     # the pure version.
     if data and data[-1:] != b'\n':
         raise ValueError(b'Manifest did not end in a newline.')
     prev = None
     for l in data.splitlines():
         if prev is not None and prev > l:
             raise ValueError(b'Manifest lines not in sorted order.')
         prev = l
         f, n = l.split(b'\0')
         if len(n) > 40:
             yield f, bin(n[:40]), n[40:]
         else:
             yield f, bin(n), b''
 
 
 def _text(it):
     files = []
     lines = []
     for f, n, fl in it:
         files.append(f)
         # if this is changed to support newlines in filenames,
         # be sure to check the templates/ dir again (especially *-raw.tmpl)
         lines.append(b"%s\0%s%s\n" % (f, hex(n), fl))
 
     _checkforbidden(files)
     return b''.join(lines)
 
 
 class lazymanifestiter(object):
     def __init__(self, lm):
         self.pos = 0
         self.lm = lm
 
     def __iter__(self):
         return self
 
     def next(self):
         try:
             data, pos = self.lm._get(self.pos)
         except IndexError:
             raise StopIteration
         if pos == -1:
             self.pos += 1
             return data[0]
         self.pos += 1
         zeropos = data.find(b'\x00', pos)
         return data[pos:zeropos]
 
     __next__ = next
 
 
 class lazymanifestiterentries(object):
     def __init__(self, lm):
         self.lm = lm
         self.pos = 0
 
     def __iter__(self):
         return self
 
     def next(self):
         try:
             data, pos = self.lm._get(self.pos)
         except IndexError:
             raise StopIteration
         if pos == -1:
             self.pos += 1
             return data
         zeropos = data.find(b'\x00', pos)
         hashval = unhexlify(data, self.lm.extrainfo[self.pos], zeropos + 1, 40)
         flags = self.lm._getflags(data, self.pos, zeropos)
         self.pos += 1
         return (data[pos:zeropos], hashval, flags)
 
     __next__ = next
 
 
 def unhexlify(data, extra, pos, length):
     s = bin(data[pos : pos + length])
     if extra:
         s += chr(extra & 0xFF)
     return s
 
 
 def _cmp(a, b):
     return (a > b) - (a < b)
 
 
 class _lazymanifest(object):
     """A pure python manifest backed by a byte string.  It is supplimented with
     internal lists as it is modified, until it is compacted back to a pure byte
     string.
 
     ``data`` is the initial manifest data.
 
     ``positions`` is a list of offsets, one per manifest entry.  Positive
     values are offsets into ``data``, negative values are offsets into the
     ``extradata`` list.  When an entry is removed, its entry is dropped from
     ``positions``.  The values are encoded such that when walking the list and
     indexing into ``data`` or ``extradata`` as appropriate, the entries are
     sorted by filename.
 
     ``extradata`` is a list of (key, hash, flags) for entries that were added or
     modified since the manifest was created or compacted.
     """
 
     def __init__(
         self,
         data,
         positions=None,
         extrainfo=None,
         extradata=None,
         hasremovals=False,
     ):
         if positions is None:
             self.positions = self.findlines(data)
             self.extrainfo = [0] * len(self.positions)
             self.data = data
             self.extradata = []
             self.hasremovals = False
         else:
             self.positions = positions[:]
             self.extrainfo = extrainfo[:]
             self.extradata = extradata[:]
             self.data = data
             self.hasremovals = hasremovals
 
     def findlines(self, data):
         if not data:
             return []
         pos = data.find(b"\n")
         if pos == -1 or data[-1:] != b'\n':
             raise ValueError(b"Manifest did not end in a newline.")
         positions = [0]
         prev = data[: data.find(b'\x00')]
         while pos < len(data) - 1 and pos != -1:
             positions.append(pos + 1)
             nexts = data[pos + 1 : data.find(b'\x00', pos + 1)]
             if nexts < prev:
                 raise ValueError(b"Manifest lines not in sorted order.")
             prev = nexts
             pos = data.find(b"\n", pos + 1)
         return positions
 
     def _get(self, index):
         # get the position encoded in pos:
         #   positive number is an index in 'data'
         #   negative number is in extrapieces
         pos = self.positions[index]
         if pos >= 0:
             return self.data, pos
         return self.extradata[-pos - 1], -1
 
     def _getkey(self, pos):
         if pos >= 0:
             return self.data[pos : self.data.find(b'\x00', pos + 1)]
         return self.extradata[-pos - 1][0]
 
     def bsearch(self, key):
         first = 0
         last = len(self.positions) - 1
 
         while first <= last:
             midpoint = (first + last) // 2
             nextpos = self.positions[midpoint]
             candidate = self._getkey(nextpos)
             r = _cmp(key, candidate)
             if r == 0:
                 return midpoint
             else:
                 if r < 0:
                     last = midpoint - 1
                 else:
                     first = midpoint + 1
         return -1
 
     def bsearch2(self, key):
         # same as the above, but will always return the position
         # done for performance reasons
         first = 0
         last = len(self.positions) - 1
 
         while first <= last:
             midpoint = (first + last) // 2
             nextpos = self.positions[midpoint]
             candidate = self._getkey(nextpos)
             r = _cmp(key, candidate)
             if r == 0:
                 return (midpoint, True)
             else:
                 if r < 0:
                     last = midpoint - 1
                 else:
                     first = midpoint + 1
         return (first, False)
 
     def __contains__(self, key):
         return self.bsearch(key) != -1
 
     def _getflags(self, data, needle, pos):
         start = pos + 41
         end = data.find(b"\n", start)
         if end == -1:
             end = len(data) - 1
         if start == end:
             return b''
         return self.data[start:end]
 
     def __getitem__(self, key):
         if not isinstance(key, bytes):
             raise TypeError(b"getitem: manifest keys must be a bytes.")
         needle = self.bsearch(key)
         if needle == -1:
             raise KeyError
         data, pos = self._get(needle)
         if pos == -1:
             return (data[1], data[2])
         zeropos = data.find(b'\x00', pos)
         assert 0 <= needle <= len(self.positions)
         assert len(self.extrainfo) == len(self.positions)
         hashval = unhexlify(data, self.extrainfo[needle], zeropos + 1, 40)
         flags = self._getflags(data, needle, zeropos)
         return (hashval, flags)
 
     def __delitem__(self, key):
         needle, found = self.bsearch2(key)
         if not found:
             raise KeyError
         cur = self.positions[needle]
         self.positions = self.positions[:needle] + self.positions[needle + 1 :]
         self.extrainfo = self.extrainfo[:needle] + self.extrainfo[needle + 1 :]
         if cur >= 0:
             # This does NOT unsort the list as far as the search functions are
             # concerned, as they only examine lines mapped by self.positions.
             self.data = self.data[:cur] + b'\x00' + self.data[cur + 1 :]
             self.hasremovals = True
 
     def __setitem__(self, key, value):
         if not isinstance(key, bytes):
             raise TypeError(b"setitem: manifest keys must be a byte string.")
         if not isinstance(value, tuple) or len(value) != 2:
             raise TypeError(
                 b"Manifest values must be a tuple of (node, flags)."
             )
         hashval = value[0]
         if not isinstance(hashval, bytes) or not 20 <= len(hashval) <= 22:
             raise TypeError(b"node must be a 20-byte byte string")
         flags = value[1]
         if len(hashval) == 22:
             hashval = hashval[:-1]
         if not isinstance(flags, bytes) or len(flags) > 1:
             raise TypeError(b"flags must a 0 or 1 byte string, got %r", flags)
         needle, found = self.bsearch2(key)
         if found:
             # put the item
             pos = self.positions[needle]
             if pos < 0:
                 self.extradata[-pos - 1] = (key, hashval, value[1])
             else:
                 # just don't bother
                 self.extradata.append((key, hashval, value[1]))
                 self.positions[needle] = -len(self.extradata)
         else:
             # not found, put it in with extra positions
             self.extradata.append((key, hashval, value[1]))
             self.positions = (
                 self.positions[:needle]
                 + [-len(self.extradata)]
                 + self.positions[needle:]
             )
             self.extrainfo = (
                 self.extrainfo[:needle] + [0] + self.extrainfo[needle:]
             )
 
     def copy(self):
         # XXX call _compact like in C?
         return _lazymanifest(
             self.data,
             self.positions,
             self.extrainfo,
             self.extradata,
             self.hasremovals,
         )
 
     def _compact(self):
         # hopefully not called TOO often
         if len(self.extradata) == 0 and not self.hasremovals:
             return
         l = []
         i = 0
         offset = 0
         self.extrainfo = [0] * len(self.positions)
         while i < len(self.positions):
             if self.positions[i] >= 0:
                 cur = self.positions[i]
                 last_cut = cur
 
                 # Collect all contiguous entries in the buffer at the current
                 # offset, breaking out only for added/modified items held in
                 # extradata, or a deleted line prior to the next position.
                 while True:
                     self.positions[i] = offset
                     i += 1
                     if i == len(self.positions) or self.positions[i] < 0:
                         break
 
                     # A removed file has no positions[] entry, but does have an
                     # overwritten first byte.  Break out and find the end of the
                     # current good entry/entries if there is a removed file
                     # before the next position.
                     if (
                         self.hasremovals
                         and self.data.find(b'\n\x00', cur, self.positions[i])
                         != -1
                     ):
                         break
 
                     offset += self.positions[i] - cur
                     cur = self.positions[i]
                 end_cut = self.data.find(b'\n', cur)
                 if end_cut != -1:
                     end_cut += 1
                 offset += end_cut - cur
                 l.append(self.data[last_cut:end_cut])
             else:
                 while i < len(self.positions) and self.positions[i] < 0:
                     cur = self.positions[i]
                     t = self.extradata[-cur - 1]
                     l.append(self._pack(t))
                     self.positions[i] = offset
                     if len(t[1]) > 20:
                         self.extrainfo[i] = ord(t[1][21])
                     offset += len(l[-1])
                     i += 1
         self.data = b''.join(l)
         self.hasremovals = False
         self.extradata = []
 
     def _pack(self, d):
         return d[0] + b'\x00' + hex(d[1][:20]) + d[2] + b'\n'
 
     def text(self):
         self._compact()
         return self.data
 
     def diff(self, m2, clean=False):
         '''Finds changes between the current manifest and m2.'''
         # XXX think whether efficiency matters here
         diff = {}
 
         for fn, e1, flags in self.iterentries():
             if fn not in m2:
                 diff[fn] = (e1, flags), (None, b'')
             else:
                 e2 = m2[fn]
                 if (e1, flags) != e2:
                     diff[fn] = (e1, flags), e2
                 elif clean:
                     diff[fn] = None
 
         for fn, e2, flags in m2.iterentries():
             if fn not in self:
                 diff[fn] = (None, b''), (e2, flags)
 
         return diff
 
     def iterentries(self):
         return lazymanifestiterentries(self)
 
     def iterkeys(self):
         return lazymanifestiter(self)
 
     def __iter__(self):
         return lazymanifestiter(self)
 
     def __len__(self):
         return len(self.positions)
 
     def filtercopy(self, filterfn):
         # XXX should be optimized
         c = _lazymanifest(b'')
         for f, n, fl in self.iterentries():
             if filterfn(f):
                 c[f] = n, fl
         return c
 
 
 try:
     _lazymanifest = parsers.lazymanifest
 except AttributeError:
     pass
 
 
 @interfaceutil.implementer(repository.imanifestdict)
 class manifestdict(object):
     def __init__(self, data=b''):
         self._lm = _lazymanifest(data)
 
     def __getitem__(self, key):
         return self._lm[key][0]
 
     def find(self, key):
         return self._lm[key]
 
     def __len__(self):
         return len(self._lm)
 
     def __nonzero__(self):
         # nonzero is covered by the __len__ function, but implementing it here
         # makes it easier for extensions to override.
         return len(self._lm) != 0
 
     __bool__ = __nonzero__
 
     def __setitem__(self, key, node):
         self._lm[key] = node, self.flags(key, b'')
 
     def __contains__(self, key):
         if key is None:
             return False
         return key in self._lm
 
     def __delitem__(self, key):
         del self._lm[key]
 
     def __iter__(self):
         return self._lm.__iter__()
 
     def iterkeys(self):
         return self._lm.iterkeys()
 
     def keys(self):
         return list(self.iterkeys())
 
     def filesnotin(self, m2, match=None):
         '''Set of files in this manifest that are not in the other'''
         if match:
             m1 = self.matches(match)
             m2 = m2.matches(match)
             return m1.filesnotin(m2)
         diff = self.diff(m2)
         files = set(
             filepath
             for filepath, hashflags in pycompat.iteritems(diff)
             if hashflags[1][0] is None
         )
         return files
 
     @propertycache
     def _dirs(self):
         return pathutil.dirs(self)
 
     def dirs(self):
         return self._dirs
 
     def hasdir(self, dir):
         return dir in self._dirs
 
     def _filesfastpath(self, match):
         '''Checks whether we can correctly and quickly iterate over matcher
         files instead of over manifest files.'''
         files = match.files()
         return len(files) < 100 and (
             match.isexact()
             or (match.prefix() and all(fn in self for fn in files))
         )
 
     def walk(self, match):
         '''Generates matching file names.
 
         Equivalent to manifest.matches(match).iterkeys(), but without creating
         an entirely new manifest.
 
         It also reports nonexistent files by marking them bad with match.bad().
         '''
         if match.always():
             for f in iter(self):
                 yield f
             return
 
         fset = set(match.files())
 
         # avoid the entire walk if we're only looking for specific files
         if self._filesfastpath(match):
             for fn in sorted(fset):
                 yield fn
             return
 
         for fn in self:
             if fn in fset:
                 # specified pattern is the exact name
                 fset.remove(fn)
             if match(fn):
                 yield fn
 
         # for dirstate.walk, files=[''] means "walk the whole tree".
         # follow that here, too
         fset.discard(b'')
 
         for fn in sorted(fset):
             if not self.hasdir(fn):
                 match.bad(fn, None)
 
     def matches(self, match):
         '''generate a new manifest filtered by the match argument'''
         if match.always():
             return self.copy()
 
         if self._filesfastpath(match):
             m = manifestdict()
             lm = self._lm
             for fn in match.files():
                 if fn in lm:
                     m._lm[fn] = lm[fn]
             return m
 
         m = manifestdict()
         m._lm = self._lm.filtercopy(match)
         return m
 
     def diff(self, m2, match=None, clean=False):
         '''Finds changes between the current manifest and m2.
 
         Args:
           m2: the manifest to which this manifest should be compared.
           clean: if true, include files unchanged between these manifests
                  with a None value in the returned dictionary.
 
         The result is returned as a dict with filename as key and
         values of the form ((n1,fl1),(n2,fl2)), where n1/n2 is the
         nodeid in the current/other manifest and fl1/fl2 is the flag
         in the current/other manifest. Where the file does not exist,
         the nodeid will be None and the flags will be the empty
         string.
         '''
         if match:
             m1 = self.matches(match)
             m2 = m2.matches(match)
             return m1.diff(m2, clean=clean)
         return self._lm.diff(m2._lm, clean)
 
     def setflag(self, key, flag):
         self._lm[key] = self[key], flag
 
     def get(self, key, default=None):
         try:
             return self._lm[key][0]
         except KeyError:
             return default
 
     def flags(self, key, default=b''):
         try:
             return self._lm[key][1]
         except KeyError:
             return default
 
     def copy(self):
         c = manifestdict()
         c._lm = self._lm.copy()
         return c
 
     def items(self):
         return (x[:2] for x in self._lm.iterentries())
 
     def iteritems(self):
         return (x[:2] for x in self._lm.iterentries())
 
     def iterentries(self):
         return self._lm.iterentries()
 
     def text(self):
         # most likely uses native version
         return self._lm.text()
 
     def fastdelta(self, base, changes):
         """Given a base manifest text as a bytearray and a list of changes
         relative to that text, compute a delta that can be used by revlog.
         """
         delta = []
         dstart = None
         dend = None
         dline = [b""]
         start = 0
         # zero copy representation of base as a buffer
         addbuf = util.buffer(base)
 
         changes = list(changes)
         if len(changes) < FASTDELTA_TEXTDIFF_THRESHOLD:
             # start with a readonly loop that finds the offset of
             # each line and creates the deltas
             for f, todelete in changes:
                 # bs will either be the index of the item or the insert point
                 start, end = _msearch(addbuf, f, start)
                 if not todelete:
                     h, fl = self._lm[f]
                     l = b"%s\0%s%s\n" % (f, hex(h), fl)
                 else:
                     if start == end:
                         # item we want to delete was not found, error out
                         raise AssertionError(
                             _(b"failed to remove %s from manifest") % f
                         )
                     l = b""
                 if dstart is not None and dstart <= start and dend >= start:
                     if dend < end:
                         dend = end
                     if l:
                         dline.append(l)
                 else:
                     if dstart is not None:
                         delta.append([dstart, dend, b"".join(dline)])
                     dstart = start
                     dend = end
                     dline = [l]
 
             if dstart is not None:
                 delta.append([dstart, dend, b"".join(dline)])
             # apply the delta to the base, and get a delta for addrevision
             deltatext, arraytext = _addlistdelta(base, delta)
         else:
             # For large changes, it's much cheaper to just build the text and
             # diff it.
             arraytext = bytearray(self.text())
             deltatext = mdiff.textdiff(
                 util.buffer(base), util.buffer(arraytext)
             )
 
         return arraytext, deltatext
 
 
 def _msearch(m, s, lo=0, hi=None):
     '''return a tuple (start, end) that says where to find s within m.
 
     If the string is found m[start:end] are the line containing
     that string.  If start == end the string was not found and
     they indicate the proper sorted insertion point.
 
     m should be a buffer, a memoryview or a byte string.
     s is a byte string'''
 
     def advance(i, c):
         while i < lenm and m[i : i + 1] != c:
             i += 1
         return i
 
     if not s:
         return (lo, lo)
     lenm = len(m)
     if not hi:
         hi = lenm
     while lo < hi:
         mid = (lo + hi) // 2
         start = mid
         while start > 0 and m[start - 1 : start] != b'\n':
             start -= 1
         end = advance(start, b'\0')
         if bytes(m[start:end]) < s:
             # we know that after the null there are 40 bytes of sha1
             # this translates to the bisect lo = mid + 1
             lo = advance(end + 40, b'\n') + 1
         else:
             # this translates to the bisect hi = mid
             hi = start
     end = advance(lo, b'\0')
     found = m[lo:end]
     if s == found:
         # we know that after the null there are 40 bytes of sha1
         end = advance(end + 40, b'\n')
         return (lo, end + 1)
     else:
         return (lo, lo)
 
 
 def _checkforbidden(l):
     """Check filenames for illegal characters."""
     for f in l:
         if b'\n' in f or b'\r' in f:
             raise error.StorageError(
                 _(b"'\\n' and '\\r' disallowed in filenames: %r")
                 % pycompat.bytestr(f)
             )
 
 
 # apply the changes collected during the bisect loop to our addlist
 # return a delta suitable for addrevision
 def _addlistdelta(addlist, x):
     # for large addlist arrays, building a new array is cheaper
     # than repeatedly modifying the existing one
     currentposition = 0
     newaddlist = bytearray()
 
     for start, end, content in x:
         newaddlist += addlist[currentposition:start]
         if content:
             newaddlist += bytearray(content)
 
         currentposition = end
 
     newaddlist += addlist[currentposition:]
 
     deltatext = b"".join(
         struct.pack(b">lll", start, end, len(content)) + content
         for start, end, content in x
     )
     return deltatext, newaddlist
 
 
 def _splittopdir(f):
     if b'/' in f:
         dir, subpath = f.split(b'/', 1)
         return dir + b'/', subpath
     else:
         return b'', f
 
 
 _noop = lambda s: None
 
 
 class treemanifest(object):
     def __init__(self, dir=b'', text=b''):
         self._dir = dir
         self._node = nullid
         self._loadfunc = _noop
         self._copyfunc = _noop
         self._dirty = False
         self._dirs = {}
         self._lazydirs = {}
         # Using _lazymanifest here is a little slower than plain old dicts
         self._files = {}
         self._flags = {}
         if text:
 
             def readsubtree(subdir, subm):
                 raise AssertionError(
                     b'treemanifest constructor only accepts flat manifests'
                 )
 
             self.parse(text, readsubtree)
             self._dirty = True  # Mark flat manifest dirty after parsing
 
     def _subpath(self, path):
         return self._dir + path
 
     def _loadalllazy(self):
         selfdirs = self._dirs
         for d, (path, node, readsubtree, docopy) in pycompat.iteritems(
             self._lazydirs
         ):
             if docopy:
                 selfdirs[d] = readsubtree(path, node).copy()
             else:
                 selfdirs[d] = readsubtree(path, node)
         self._lazydirs = {}
 
     def _loadlazy(self, d):
         v = self._lazydirs.get(d)
         if v:
             path, node, readsubtree, docopy = v
             if docopy:
                 self._dirs[d] = readsubtree(path, node).copy()
             else:
                 self._dirs[d] = readsubtree(path, node)
             del self._lazydirs[d]
 
     def _loadchildrensetlazy(self, visit):
         if not visit:
             return None
         if visit == b'all' or visit == b'this':
             self._loadalllazy()
             return None
 
         loadlazy = self._loadlazy
         for k in visit:
             loadlazy(k + b'/')
         return visit
 
     def _loaddifflazy(self, t1, t2):
         """load items in t1 and t2 if they're needed for diffing.
 
         The criteria currently is:
         - if it's not present in _lazydirs in either t1 or t2, load it in the
           other (it may already be loaded or it may not exist, doesn't matter)
         - if it's present in _lazydirs in both, compare the nodeid; if it
           differs, load it in both
         """
         toloadlazy = []
         for d, v1 in pycompat.iteritems(t1._lazydirs):
             v2 = t2._lazydirs.get(d)
             if not v2 or v2[1] != v1[1]:
                 toloadlazy.append(d)
         for d, v1 in pycompat.iteritems(t2._lazydirs):
             if d not in t1._lazydirs:
                 toloadlazy.append(d)
 
         for d in toloadlazy:
             t1._loadlazy(d)
             t2._loadlazy(d)
 
     def __len__(self):
         self._load()
         size = len(self._files)
         self._loadalllazy()
         for m in self._dirs.values():
             size += m.__len__()
         return size
 
     def __nonzero__(self):
         # Faster than "__len() != 0" since it avoids loading sub-manifests
         return not self._isempty()
 
     __bool__ = __nonzero__
 
     def _isempty(self):
         self._load()  # for consistency; already loaded by all callers
         # See if we can skip loading everything.
         if self._files or (
             self._dirs and any(not m._isempty() for m in self._dirs.values())
         ):
             return False
         self._loadalllazy()
         return not self._dirs or all(m._isempty() for m in self._dirs.values())
 
+    @encoding.strmethod
     def __repr__(self):
         return (
-            b'<treemanifest dir=%s, node=%s, loaded=%s, dirty=%s at 0x%x>'
+            b'<treemanifest dir=%s, node=%s, loaded=%r, dirty=%r at 0x%x>'
             % (
                 self._dir,
                 hex(self._node),
                 bool(self._loadfunc is _noop),
                 self._dirty,
                 id(self),
             )
         )
 
     def dir(self):
         '''The directory that this tree manifest represents, including a
         trailing '/'. Empty string for the repo root directory.'''
         return self._dir
 
     def node(self):
         '''This node of this instance. nullid for unsaved instances. Should
         be updated when the instance is read or written from a revlog.
         '''
         assert not self._dirty
         return self._node
 
     def setnode(self, node):
         self._node = node
         self._dirty = False
 
     def iterentries(self):
         self._load()
         self._loadalllazy()
         for p, n in sorted(
             itertools.chain(self._dirs.items(), self._files.items())
         ):
             if p in self._files:
                 yield self._subpath(p), n, self._flags.get(p, b'')
             else:
                 for x in n.iterentries():
                     yield x
 
     def items(self):
         self._load()
         self._loadalllazy()
         for p, n in sorted(
             itertools.chain(self._dirs.items(), self._files.items())
         ):
             if p in self._files:
                 yield self._subpath(p), n
             else:
                 for f, sn in pycompat.iteritems(n):
                     yield f, sn
 
     iteritems = items
 
     def iterkeys(self):
         self._load()
         self._loadalllazy()
         for p in sorted(itertools.chain(self._dirs, self._files)):
             if p in self._files:
                 yield self._subpath(p)
             else:
                 for f in self._dirs[p]:
                     yield f
 
     def keys(self):
         return list(self.iterkeys())
 
     def __iter__(self):
         return self.iterkeys()
 
     def __contains__(self, f):
         if f is None:
             return False
         self._load()
         dir, subpath = _splittopdir(f)
         if dir:
             self._loadlazy(dir)
 
             if dir not in self._dirs:
                 return False
 
             return self._dirs[dir].__contains__(subpath)
         else:
             return f in self._files
 
     def get(self, f, default=None):
         self._load()
         dir, subpath = _splittopdir(f)
         if dir:
             self._loadlazy(dir)
 
             if dir not in self._dirs:
                 return default
             return self._dirs[dir].get(subpath, default)
         else:
             return self._files.get(f, default)
 
     def __getitem__(self, f):
         self._load()
         dir, subpath = _splittopdir(f)
         if dir:
             self._loadlazy(dir)
 
             return self._dirs[dir].__getitem__(subpath)
         else:
             return self._files[f]
 
     def flags(self, f):
         self._load()
         dir, subpath = _splittopdir(f)
         if dir:
             self._loadlazy(dir)
 
             if dir not in self._dirs:
                 return b''
             return self._dirs[dir].flags(subpath)
         else:
             if f in self._lazydirs or f in self._dirs:
                 return b''
             return self._flags.get(f, b'')
 
     def find(self, f):
         self._load()
         dir, subpath = _splittopdir(f)
         if dir:
             self._loadlazy(dir)
 
             return self._dirs[dir].find(subpath)
         else:
             return self._files[f], self._flags.get(f, b'')
 
     def __delitem__(self, f):
         self._load()
         dir, subpath = _splittopdir(f)
         if dir:
             self._loadlazy(dir)
 
             self._dirs[dir].__delitem__(subpath)
             # If the directory is now empty, remove it
             if self._dirs[dir]._isempty():
                 del self._dirs[dir]
         else:
             del self._files[f]
             if f in self._flags:
                 del self._flags[f]
         self._dirty = True
 
     def __setitem__(self, f, n):
         assert n is not None
         self._load()
         dir, subpath = _splittopdir(f)
         if dir:
             self._loadlazy(dir)
             if dir not in self._dirs:
                 self._dirs[dir] = treemanifest(self._subpath(dir))
             self._dirs[dir].__setitem__(subpath, n)
         else:
             self._files[f] = n[:21]  # to match manifestdict's behavior
         self._dirty = True
 
     def _load(self):
         if self._loadfunc is not _noop:
             lf, self._loadfunc = self._loadfunc, _noop
             lf(self)
         elif self._copyfunc is not _noop:
             cf, self._copyfunc = self._copyfunc, _noop
             cf(self)
 
     def setflag(self, f, flags):
         """Set the flags (symlink, executable) for path f."""
         self._load()
         dir, subpath = _splittopdir(f)
         if dir:
             self._loadlazy(dir)
             if dir not in self._dirs:
                 self._dirs[dir] = treemanifest(self._subpath(dir))
             self._dirs[dir].setflag(subpath, flags)
         else:
             self._flags[f] = flags
         self._dirty = True
 
     def copy(self):
         copy = treemanifest(self._dir)
         copy._node = self._node
         copy._dirty = self._dirty
         if self._copyfunc is _noop:
 
             def _copyfunc(s):
                 self._load()
                 s._lazydirs = {
                     d: (p, n, r, True)
                     for d, (p, n, r, c) in pycompat.iteritems(self._lazydirs)
                 }
                 sdirs = s._dirs
                 for d, v in pycompat.iteritems(self._dirs):
                     sdirs[d] = v.copy()
                 s._files = dict.copy(self._files)
                 s._flags = dict.copy(self._flags)
 
             if self._loadfunc is _noop:
                 _copyfunc(copy)
             else:
                 copy._copyfunc = _copyfunc
         else:
             copy._copyfunc = self._copyfunc
         return copy
 
     def filesnotin(self, m2, match=None):
         '''Set of files in this manifest that are not in the other'''
         if match and not match.always():
             m1 = self.matches(match)
             m2 = m2.matches(match)
             return m1.filesnotin(m2)
 
         files = set()
 
         def _filesnotin(t1, t2):
             if t1._node == t2._node and not t1._dirty and not t2._dirty:
                 return
             t1._load()
             t2._load()
             self._loaddifflazy(t1, t2)
             for d, m1 in pycompat.iteritems(t1._dirs):
                 if d in t2._dirs:
                     m2 = t2._dirs[d]
                     _filesnotin(m1, m2)
                 else:
                     files.update(m1.iterkeys())
 
             for fn in t1._files:
                 if fn not in t2._files:
                     files.add(t1._subpath(fn))
 
         _filesnotin(self, m2)
         return files
 
     @propertycache
     def _alldirs(self):
         return pathutil.dirs(self)
 
     def dirs(self):
         return self._alldirs
 
     def hasdir(self, dir):
         self._load()
         topdir, subdir = _splittopdir(dir)
         if topdir:
             self._loadlazy(topdir)
             if topdir in self._dirs:
                 return self._dirs[topdir].hasdir(subdir)
             return False
         dirslash = dir + b'/'
         return dirslash in self._dirs or dirslash in self._lazydirs
 
     def walk(self, match):
         '''Generates matching file names.
 
         Equivalent to manifest.matches(match).iterkeys(), but without creating
         an entirely new manifest.
 
         It also reports nonexistent files by marking them bad with match.bad().
         '''
         if match.always():
             for f in iter(self):
                 yield f
             return
 
         fset = set(match.files())
 
         for fn in self._walk(match):
             if fn in fset:
                 # specified pattern is the exact name
                 fset.remove(fn)
             yield fn
 
         # for dirstate.walk, files=[''] means "walk the whole tree".
         # follow that here, too
         fset.discard(b'')
 
         for fn in sorted(fset):
             if not self.hasdir(fn):
                 match.bad(fn, None)
 
     def _walk(self, match):
         '''Recursively generates matching file names for walk().'''
         visit = match.visitchildrenset(self._dir[:-1])
         if not visit:
             return
 
         # yield this dir's files and walk its submanifests
         self._load()
         visit = self._loadchildrensetlazy(visit)
         for p in sorted(list(self._dirs) + list(self._files)):
             if p in self._files:
                 fullp = self._subpath(p)
                 if match(fullp):
                     yield fullp
             else:
                 if not visit or p[:-1] in visit:
                     for f in self._dirs[p]._walk(match):
                         yield f
 
     def matches(self, match):
         '''generate a new manifest filtered by the match argument'''
         if match.always():
             return self.copy()
 
         return self._matches(match)
 
     def _matches(self, match):
         '''recursively generate a new manifest filtered by the match argument.
         '''
 
         visit = match.visitchildrenset(self._dir[:-1])
         if visit == b'all':
             return self.copy()
         ret = treemanifest(self._dir)
         if not visit:
             return ret
 
         self._load()
         for fn in self._files:
             # While visitchildrenset *usually* lists only subdirs, this is
             # actually up to the matcher and may have some files in the set().
             # If visit == 'this', we should obviously look at the files in this
             # directory; if visit is a set, and fn is in it, we should inspect
             # fn (but no need to inspect things not in the set).
             if visit != b'this' and fn not in visit:
                 continue
             fullp = self._subpath(fn)
             # visitchildrenset isn't perfect, we still need to call the regular
             # matcher code to further filter results.
             if not match(fullp):
                 continue
             ret._files[fn] = self._files[fn]
             if fn in self._flags:
                 ret._flags[fn] = self._flags[fn]
 
         visit = self._loadchildrensetlazy(visit)
         for dir, subm in pycompat.iteritems(self._dirs):
             if visit and dir[:-1] not in visit:
                 continue
             m = subm._matches(match)
             if not m._isempty():
                 ret._dirs[dir] = m
 
         if not ret._isempty():
             ret._dirty = True
         return ret
 
     def diff(self, m2, match=None, clean=False):
         '''Finds changes between the current manifest and m2.
 
         Args:
           m2: the manifest to which this manifest should be compared.
           clean: if true, include files unchanged between these manifests
                  with a None value in the returned dictionary.
 
         The result is returned as a dict with filename as key and
         values of the form ((n1,fl1),(n2,fl2)), where n1/n2 is the
         nodeid in the current/other manifest and fl1/fl2 is the flag
         in the current/other manifest. Where the file does not exist,
         the nodeid will be None and the flags will be the empty
         string.
         '''
         if match and not match.always():
             m1 = self.matches(match)
             m2 = m2.matches(match)
             return m1.diff(m2, clean=clean)
         result = {}
         emptytree = treemanifest()
 
         def _iterativediff(t1, t2, stack):
             """compares two tree manifests and append new tree-manifests which
             needs to be compared to stack"""
             if t1._node == t2._node and not t1._dirty and not t2._dirty:
                 return
             t1._load()
             t2._load()
             self._loaddifflazy(t1, t2)
 
             for d, m1 in pycompat.iteritems(t1._dirs):
                 m2 = t2._dirs.get(d, emptytree)
                 stack.append((m1, m2))
 
             for d, m2 in pycompat.iteritems(t2._dirs):
                 if d not in t1._dirs:
                     stack.append((emptytree, m2))
 
             for fn, n1 in pycompat.iteritems(t1._files):
                 fl1 = t1._flags.get(fn, b'')
                 n2 = t2._files.get(fn, None)
                 fl2 = t2._flags.get(fn, b'')
                 if n1 != n2 or fl1 != fl2:
                     result[t1._subpath(fn)] = ((n1, fl1), (n2, fl2))
                 elif clean:
                     result[t1._subpath(fn)] = None
 
             for fn, n2 in pycompat.iteritems(t2._files):
                 if fn not in t1._files:
                     fl2 = t2._flags.get(fn, b'')
                     result[t2._subpath(fn)] = ((None, b''), (n2, fl2))
 
         stackls = []
         _iterativediff(self, m2, stackls)
         while stackls:
             t1, t2 = stackls.pop()
             # stackls is populated in the function call
             _iterativediff(t1, t2, stackls)
         return result
 
     def unmodifiedsince(self, m2):
         return not self._dirty and not m2._dirty and self._node == m2._node
 
     def parse(self, text, readsubtree):
         selflazy = self._lazydirs
         subpath = self._subpath
         for f, n, fl in _parse(text):
             if fl == b't':
                 f = f + b'/'
                 # False below means "doesn't need to be copied" and can use the
                 # cached value from readsubtree directly.
                 selflazy[f] = (subpath(f), n, readsubtree, False)
             elif b'/' in f:
                 # This is a flat manifest, so use __setitem__ and setflag rather
                 # than assigning directly to _files and _flags, so we can
                 # assign a path in a subdirectory, and to mark dirty (compared
                 # to nullid).
                 self[f] = n
                 if fl:
                     self.setflag(f, fl)
             else:
                 # Assigning to _files and _flags avoids marking as dirty,
                 # and should be a little faster.
                 self._files[f] = n
                 if fl:
                     self._flags[f] = fl
 
     def text(self):
         """Get the full data of this manifest as a bytestring."""
         self._load()
         return _text(self.iterentries())
 
     def dirtext(self):
         """Get the full data of this directory as a bytestring. Make sure that
         any submanifests have been written first, so their nodeids are correct.
         """
         self._load()
         flags = self.flags
         lazydirs = [
             (d[:-1], v[1], b't') for d, v in pycompat.iteritems(self._lazydirs)
         ]
         dirs = [(d[:-1], self._dirs[d]._node, b't') for d in self._dirs]
         files = [(f, self._files[f], flags(f)) for f in self._files]
         return _text(sorted(dirs + files + lazydirs))
 
     def read(self, gettext, readsubtree):
         def _load_for_read(s):
             s.parse(gettext(), readsubtree)
             s._dirty = False
 
         self._loadfunc = _load_for_read
 
     def writesubtrees(self, m1, m2, writesubtree, match):
         self._load()  # for consistency; should never have any effect here
         m1._load()
         m2._load()
         emptytree = treemanifest()
 
         def getnode(m, d):
             ld = m._lazydirs.get(d)
             if ld:
                 return ld[1]
             return m._dirs.get(d, emptytree)._node
 
         # let's skip investigating things that `match` says we do not need.
         visit = match.visitchildrenset(self._dir[:-1])
         visit = self._loadchildrensetlazy(visit)
         if visit == b'this' or visit == b'all':
             visit = None
         for d, subm in pycompat.iteritems(self._dirs):
             if visit and d[:-1] not in visit:
                 continue
             subp1 = getnode(m1, d)
             subp2 = getnode(m2, d)
             if subp1 == nullid:
                 subp1, subp2 = subp2, subp1
             writesubtree(subm, subp1, subp2, match)
 
     def walksubtrees(self, matcher=None):
         """Returns an iterator of the subtrees of this manifest, including this
         manifest itself.
 
         If `matcher` is provided, it only returns subtrees that match.
         """
         if matcher and not matcher.visitdir(self._dir[:-1]):
             return
         if not matcher or matcher(self._dir[:-1]):
             yield self
 
         self._load()
         # OPT: use visitchildrenset to avoid loading everything.
         self._loadalllazy()
         for d, subm in pycompat.iteritems(self._dirs):
             for subtree in subm.walksubtrees(matcher=matcher):
                 yield subtree
 
 
 class manifestfulltextcache(util.lrucachedict):
     """File-backed LRU cache for the manifest cache
 
     File consists of entries, up to EOF:
 
     - 20 bytes node, 4 bytes length, <length> manifest data
 
     These are written in reverse cache order (oldest to newest).
 
     """
 
     _file = b'manifestfulltextcache'
 
     def __init__(self, max):
         super(manifestfulltextcache, self).__init__(max)
         self._dirty = False
         self._read = False
         self._opener = None
 
     def read(self):
         if self._read or self._opener is None:
             return
 
         try:
             with self._opener(self._file) as fp:
                 set = super(manifestfulltextcache, self).__setitem__
                 # ignore trailing data, this is a cache, corruption is skipped
                 while True:
                     node = fp.read(20)
                     if len(node) < 20:
                         break
                     try:
                         size = struct.unpack(b'>L', fp.read(4))[0]
                     except struct.error:
                         break
                     value = bytearray(fp.read(size))
                     if len(value) != size:
                         break
                     set(node, value)
         except IOError:
             # the file is allowed to be missing
             pass
 
         self._read = True
         self._dirty = False
 
     def write(self):
         if not self._dirty or self._opener is None:
             return
         # rotate backwards to the first used node
         with self._opener(
             self._file, b'w', atomictemp=True, checkambig=True
         ) as fp:
             node = self._head.prev
             while True:
                 if node.key in self._cache:
                     fp.write(node.key)
                     fp.write(struct.pack(b'>L', len(node.value)))
                     fp.write(node.value)
                 if node is self._head:
                     break
                 node = node.prev
 
     def __len__(self):
         if not self._read:
             self.read()
         return super(manifestfulltextcache, self).__len__()
 
     def __contains__(self, k):
         if not self._read:
             self.read()
         return super(manifestfulltextcache, self).__contains__(k)
 
     def __iter__(self):
         if not self._read:
             self.read()
         return super(manifestfulltextcache, self).__iter__()
 
     def __getitem__(self, k):
         if not self._read:
             self.read()
         # the cache lru order can change on read
         setdirty = self._cache.get(k) is not self._head
         value = super(manifestfulltextcache, self).__getitem__(k)
         if setdirty:
             self._dirty = True
         return value
 
     def __setitem__(self, k, v):
         if not self._read:
             self.read()
         super(manifestfulltextcache, self).__setitem__(k, v)
         self._dirty = True
 
     def __delitem__(self, k):
         if not self._read:
             self.read()
         super(manifestfulltextcache, self).__delitem__(k)
         self._dirty = True
 
     def get(self, k, default=None):
         if not self._read:
             self.read()
         return super(manifestfulltextcache, self).get(k, default=default)
 
     def clear(self, clear_persisted_data=False):
         super(manifestfulltextcache, self).clear()
         if clear_persisted_data:
             self._dirty = True
             self.write()
         self._read = False
 
 
 # and upper bound of what we expect from compression
 # (real live value seems to be "3")
 MAXCOMPRESSION = 3
 
 
 @interfaceutil.implementer(repository.imanifeststorage)
 class manifestrevlog(object):
     '''A revlog that stores manifest texts. This is responsible for caching the
     full-text manifest contents.
     '''
 
     def __init__(
         self,
         opener,
         tree=b'',
         dirlogcache=None,
         indexfile=None,
         treemanifest=False,
     ):
         """Constructs a new manifest revlog
 
         `indexfile` - used by extensions to have two manifests at once, like
         when transitioning between flatmanifeset and treemanifests.
 
         `treemanifest` - used to indicate this is a tree manifest revlog. Opener
         options can also be used to make this a tree manifest revlog. The opener
         option takes precedence, so if it is set to True, we ignore whatever
         value is passed in to the constructor.
         """
         # During normal operations, we expect to deal with not more than four
         # revs at a time (such as during commit --amend). When rebasing large
         # stacks of commits, the number can go up, hence the config knob below.
         cachesize = 4
         optiontreemanifest = False
         opts = getattr(opener, 'options', None)
         if opts is not None:
             cachesize = opts.get(b'manifestcachesize', cachesize)
             optiontreemanifest = opts.get(b'treemanifest', False)
 
         self._treeondisk = optiontreemanifest or treemanifest
 
         self._fulltextcache = manifestfulltextcache(cachesize)
 
         if tree:
             assert self._treeondisk, b'opts is %r' % opts
 
         if indexfile is None:
             indexfile = b'00manifest.i'
             if tree:
                 indexfile = b"meta/" + tree + indexfile
 
         self.tree = tree
 
         # The dirlogcache is kept on the root manifest log
         if tree:
             self._dirlogcache = dirlogcache
         else:
             self._dirlogcache = {b'': self}
 
         self._revlog = revlog.revlog(
             opener,
             indexfile,
             # only root indexfile is cached
             checkambig=not bool(tree),
             mmaplargeindex=True,
             upperboundcomp=MAXCOMPRESSION,
         )
 
         self.index = self._revlog.index
         self.version = self._revlog.version
         self._generaldelta = self._revlog._generaldelta
 
     def _setupmanifestcachehooks(self, repo):
         """Persist the manifestfulltextcache on lock release"""
         if not util.safehasattr(repo, b'_wlockref'):
             return
 
         self._fulltextcache._opener = repo.wcachevfs
         if repo._currentlock(repo._wlockref) is None:
             return
 
         reporef = weakref.ref(repo)
         manifestrevlogref = weakref.ref(self)
 
         def persistmanifestcache(success):
             # Repo is in an unknown state, do not persist.
             if not success:
                 return
 
             repo = reporef()
             self = manifestrevlogref()
             if repo is None or self is None:
                 return
             if repo.manifestlog.getstorage(b'') is not self:
                 # there's a different manifest in play now, abort
                 return
             self._fulltextcache.write()
 
         repo._afterlock(persistmanifestcache)
 
     @property
     def fulltextcache(self):
         return self._fulltextcache
 
     def clearcaches(self, clear_persisted_data=False):
         self._revlog.clearcaches()
         self._fulltextcache.clear(clear_persisted_data=clear_persisted_data)
         self._dirlogcache = {self.tree: self}
 
     def dirlog(self, d):
         if d:
             assert self._treeondisk
         if d not in self._dirlogcache:
             mfrevlog = manifestrevlog(
                 self.opener, d, self._dirlogcache, treemanifest=self._treeondisk
             )
             self._dirlogcache[d] = mfrevlog
         return self._dirlogcache[d]
 
     def add(
         self,
         m,
         transaction,
         link,
         p1,
         p2,
         added,
         removed,
         readtree=None,
         match=None,
     ):
         if p1 in self.fulltextcache and util.safehasattr(m, b'fastdelta'):
             # If our first parent is in the manifest cache, we can
             # compute a delta here using properties we know about the
             # manifest up-front, which may save time later for the
             # revlog layer.
 
             _checkforbidden(added)
             # combine the changed lists into one sorted iterator
             work = heapq.merge(
                 [(x, False) for x in sorted(added)],
                 [(x, True) for x in sorted(removed)],
             )
 
             arraytext, deltatext = m.fastdelta(self.fulltextcache[p1], work)
             cachedelta = self._revlog.rev(p1), deltatext
             text = util.buffer(arraytext)
             n = self._revlog.addrevision(
                 text, transaction, link, p1, p2, cachedelta
             )
         else:
             # The first parent manifest isn't already loaded, so we'll
             # just encode a fulltext of the manifest and pass that
             # through to the revlog layer, and let it handle the delta
             # process.
             if self._treeondisk:
                 assert readtree, b"readtree must be set for treemanifest writes"
                 assert match, b"match must be specified for treemanifest writes"
                 m1 = readtree(self.tree, p1)
                 m2 = readtree(self.tree, p2)
                 n = self._addtree(
                     m, transaction, link, m1, m2, readtree, match=match
                 )
                 arraytext = None
             else:
                 text = m.text()
                 n = self._revlog.addrevision(text, transaction, link, p1, p2)
                 arraytext = bytearray(text)
 
         if arraytext is not None:
             self.fulltextcache[n] = arraytext
 
         return n
 
     def _addtree(self, m, transaction, link, m1, m2, readtree, match):
         # If the manifest is unchanged compared to one parent,
         # don't write a new revision
         if self.tree != b'' and (
             m.unmodifiedsince(m1) or m.unmodifiedsince(m2)
         ):
             return m.node()
 
         def writesubtree(subm, subp1, subp2, match):
             sublog = self.dirlog(subm.dir())
             sublog.add(
                 subm,
                 transaction,
                 link,
                 subp1,
                 subp2,
                 None,
                 None,
                 readtree=readtree,
                 match=match,
             )
 
         m.writesubtrees(m1, m2, writesubtree, match)
         text = m.dirtext()
         n = None
         if self.tree != b'':
             # Double-check whether contents are unchanged to one parent
             if text == m1.dirtext():
                 n = m1.node()
             elif text == m2.dirtext():
                 n = m2.node()
 
         if not n:
             n = self._revlog.addrevision(
                 text, transaction, link, m1.node(), m2.node()
             )
 
         # Save nodeid so parent manifest can calculate its nodeid
         m.setnode(n)
         return n
 
     def __len__(self):
         return len(self._revlog)
 
     def __iter__(self):
         return self._revlog.__iter__()
 
     def rev(self, node):
         return self._revlog.rev(node)
 
     def node(self, rev):
         return self._revlog.node(rev)
 
     def lookup(self, value):
         return self._revlog.lookup(value)
 
     def parentrevs(self, rev):
         return self._revlog.parentrevs(rev)
 
     def parents(self, node):
         return self._revlog.parents(node)
 
     def linkrev(self, rev):
         return self._revlog.linkrev(rev)
 
     def checksize(self):
         return self._revlog.checksize()
 
     def revision(self, node, _df=None, raw=False):
         return self._revlog.revision(node, _df=_df, raw=raw)
 
     def rawdata(self, node, _df=None):
         return self._revlog.rawdata(node, _df=_df)
 
     def revdiff(self, rev1, rev2):
         return self._revlog.revdiff(rev1, rev2)
 
     def cmp(self, node, text):
         return self._revlog.cmp(node, text)
 
     def deltaparent(self, rev):
         return self._revlog.deltaparent(rev)
 
     def emitrevisions(
         self,
         nodes,
         nodesorder=None,
         revisiondata=False,
         assumehaveparentrevisions=False,
         deltamode=repository.CG_DELTAMODE_STD,
     ):
         return self._revlog.emitrevisions(
             nodes,
             nodesorder=nodesorder,
             revisiondata=revisiondata,
             assumehaveparentrevisions=assumehaveparentrevisions,
             deltamode=deltamode,
         )
 
     def addgroup(self, deltas, linkmapper, transaction, addrevisioncb=None):
         return self._revlog.addgroup(
             deltas, linkmapper, transaction, addrevisioncb=addrevisioncb
         )
 
     def rawsize(self, rev):
         return self._revlog.rawsize(rev)
 
     def getstrippoint(self, minlink):
         return self._revlog.getstrippoint(minlink)
 
     def strip(self, minlink, transaction):
         return self._revlog.strip(minlink, transaction)
 
     def files(self):
         return self._revlog.files()
 
     def clone(self, tr, destrevlog, **kwargs):
         if not isinstance(destrevlog, manifestrevlog):
             raise error.ProgrammingError(b'expected manifestrevlog to clone()')
 
         return self._revlog.clone(tr, destrevlog._revlog, **kwargs)
 
     def storageinfo(
         self,
         exclusivefiles=False,
         sharedfiles=False,
         revisionscount=False,
         trackedsize=False,
         storedsize=False,
     ):
         return self._revlog.storageinfo(
             exclusivefiles=exclusivefiles,
             sharedfiles=sharedfiles,
             revisionscount=revisionscount,
             trackedsize=trackedsize,
             storedsize=storedsize,
         )
 
     @property
     def indexfile(self):
         return self._revlog.indexfile
 
     @indexfile.setter
     def indexfile(self, value):
         self._revlog.indexfile = value
 
     @property
     def opener(self):
         return self._revlog.opener
 
     @opener.setter
     def opener(self, value):
         self._revlog.opener = value
 
 
 @interfaceutil.implementer(repository.imanifestlog)
 class manifestlog(object):
     """A collection class representing the collection of manifest snapshots
     referenced by commits in the repository.
 
     In this situation, 'manifest' refers to the abstract concept of a snapshot
     of the list of files in the given commit. Consumers of the output of this
     class do not care about the implementation details of the actual manifests
     they receive (i.e. tree or flat or lazily loaded, etc)."""
 
     def __init__(self, opener, repo, rootstore, narrowmatch):
         usetreemanifest = False
         cachesize = 4
 
         opts = getattr(opener, 'options', None)
         if opts is not None:
             usetreemanifest = opts.get(b'treemanifest', usetreemanifest)
             cachesize = opts.get(b'manifestcachesize', cachesize)
 
         self._treemanifests = usetreemanifest
 
         self._rootstore = rootstore
         self._rootstore._setupmanifestcachehooks(repo)
         self._narrowmatch = narrowmatch
 
         # A cache of the manifestctx or treemanifestctx for each directory
         self._dirmancache = {}
         self._dirmancache[b''] = util.lrucachedict(cachesize)
 
         self._cachesize = cachesize
 
     def __getitem__(self, node):
         """Retrieves the manifest instance for the given node. Throws a
         LookupError if not found.
         """
         return self.get(b'', node)
 
     def get(self, tree, node, verify=True):
         """Retrieves the manifest instance for the given node. Throws a
         LookupError if not found.
 
         `verify` - if True an exception will be thrown if the node is not in
                    the revlog
         """
         if node in self._dirmancache.get(tree, ()):
             return self._dirmancache[tree][node]
 
         if not self._narrowmatch.always():
             if not self._narrowmatch.visitdir(tree[:-1]):
                 return excludeddirmanifestctx(tree, node)
         if tree:
             if self._rootstore._treeondisk:
                 if verify:
                     # Side-effect is LookupError is raised if node doesn't
                     # exist.
                     self.getstorage(tree).rev(node)
 
                 m = treemanifestctx(self, tree, node)
             else:
                 raise error.Abort(
                     _(
                         b"cannot ask for manifest directory '%s' in a flat "
                         b"manifest"
                     )
                     % tree
                 )
         else:
             if verify:
                 # Side-effect is LookupError is raised if node doesn't exist.
                 self._rootstore.rev(node)
 
             if self._treemanifests:
                 m = treemanifestctx(self, b'', node)
             else:
                 m = manifestctx(self, node)
 
         if node != nullid:
             mancache = self._dirmancache.get(tree)
             if not mancache:
                 mancache = util.lrucachedict(self._cachesize)
                 self._dirmancache[tree] = mancache
             mancache[node] = m
         return m
 
     def getstorage(self, tree):
         return self._rootstore.dirlog(tree)
 
     def clearcaches(self, clear_persisted_data=False):
         self._dirmancache.clear()
         self._rootstore.clearcaches(clear_persisted_data=clear_persisted_data)
 
     def rev(self, node):
         return self._rootstore.rev(node)
 
 
 @interfaceutil.implementer(repository.imanifestrevisionwritable)
 class memmanifestctx(object):
     def __init__(self, manifestlog):
         self._manifestlog = manifestlog
         self._manifestdict = manifestdict()
 
     def _storage(self):
         return self._manifestlog.getstorage(b'')
 
     def new(self):
         return memmanifestctx(self._manifestlog)
 
     def copy(self):
         memmf = memmanifestctx(self._manifestlog)
         memmf._manifestdict = self.read().copy()
         return memmf
 
     def read(self):
         return self._manifestdict
 
     def write(self, transaction, link, p1, p2, added, removed, match=None):
         return self._storage().add(
             self._manifestdict,
             transaction,
             link,
             p1,
             p2,
             added,
             removed,
             match=match,
         )
 
 
 @interfaceutil.implementer(repository.imanifestrevisionstored)
 class manifestctx(object):
     """A class representing a single revision of a manifest, including its
     contents, its parent revs, and its linkrev.
     """
 
     def __init__(self, manifestlog, node):
         self._manifestlog = manifestlog
         self._data = None
 
         self._node = node
 
         # TODO: We eventually want p1, p2, and linkrev exposed on this class,
         # but let's add it later when something needs it and we can load it
         # lazily.
         # self.p1, self.p2 = store.parents(node)
         # rev = store.rev(node)
         # self.linkrev = store.linkrev(rev)
 
     def _storage(self):
         return self._manifestlog.getstorage(b'')
 
     def node(self):
         return self._node
 
     def new(self):
         return memmanifestctx(self._manifestlog)
 
     def copy(self):
         memmf = memmanifestctx(self._manifestlog)
         memmf._manifestdict = self.read().copy()
         return memmf
 
     @propertycache
     def parents(self):
         return self._storage().parents(self._node)
 
     def read(self):
         if self._data is None:
             if self._node == nullid:
                 self._data = manifestdict()
             else:
                 store = self._storage()
                 if self._node in store.fulltextcache:
                     text = pycompat.bytestr(store.fulltextcache[self._node])
                 else:
                     text = store.revision(self._node)
                     arraytext = bytearray(text)
                     store.fulltextcache[self._node] = arraytext
                 self._data = manifestdict(text)
         return self._data
 
     def readfast(self, shallow=False):
         '''Calls either readdelta or read, based on which would be less work.
         readdelta is called if the delta is against the p1, and therefore can be
         read quickly.
 
         If `shallow` is True, nothing changes since this is a flat manifest.
         '''
         store = self._storage()
         r = store.rev(self._node)
         deltaparent = store.deltaparent(r)
         if deltaparent != nullrev and deltaparent in store.parentrevs(r):
             return self.readdelta()
         return self.read()
 
     def readdelta(self, shallow=False):
         '''Returns a manifest containing just the entries that are present
         in this manifest, but not in its p1 manifest. This is efficient to read
         if the revlog delta is already p1.
 
         Changing the value of `shallow` has no effect on flat manifests.
         '''
         store = self._storage()
         r = store.rev(self._node)
         d = mdiff.patchtext(store.revdiff(store.deltaparent(r), r))
         return manifestdict(d)
 
     def find(self, key):
         return self.read().find(key)
 
 
 @interfaceutil.implementer(repository.imanifestrevisionwritable)
 class memtreemanifestctx(object):
     def __init__(self, manifestlog, dir=b''):
         self._manifestlog = manifestlog
         self._dir = dir
         self._treemanifest = treemanifest()
 
     def _storage(self):
         return self._manifestlog.getstorage(b'')
 
     def new(self, dir=b''):
         return memtreemanifestctx(self._manifestlog, dir=dir)
 
     def copy(self):
         memmf = memtreemanifestctx(self._manifestlog, dir=self._dir)
         memmf._treemanifest = self._treemanifest.copy()
         return memmf
 
     def read(self):
         return self._treemanifest
 
     def write(self, transaction, link, p1, p2, added, removed, match=None):
         def readtree(dir, node):
             return self._manifestlog.get(dir, node).read()
 
         return self._storage().add(
             self._treemanifest,
             transaction,
             link,
             p1,
             p2,
             added,
             removed,
             readtree=readtree,
             match=match,
         )
 
 
 @interfaceutil.implementer(repository.imanifestrevisionstored)
 class treemanifestctx(object):
     def __init__(self, manifestlog, dir, node):
         self._manifestlog = manifestlog
         self._dir = dir
         self._data = None
 
         self._node = node
 
         # TODO: Load p1/p2/linkrev lazily. They need to be lazily loaded so that
         # we can instantiate treemanifestctx objects for directories we don't
         # have on disk.
         # self.p1, self.p2 = store.parents(node)
         # rev = store.rev(node)
         # self.linkrev = store.linkrev(rev)
 
     def _storage(self):
         narrowmatch = self._manifestlog._narrowmatch
         if not narrowmatch.always():
             if not narrowmatch.visitdir(self._dir[:-1]):
                 return excludedmanifestrevlog(self._dir)
         return self._manifestlog.getstorage(self._dir)
 
     def read(self):
         if self._data is None:
             store = self._storage()
             if self._node == nullid:
                 self._data = treemanifest()
             # TODO accessing non-public API
             elif store._treeondisk:
                 m = treemanifest(dir=self._dir)
 
                 def gettext():
                     return store.revision(self._node)
 
                 def readsubtree(dir, subm):
                     # Set verify to False since we need to be able to create
                     # subtrees for trees that don't exist on disk.
                     return self._manifestlog.get(dir, subm, verify=False).read()
 
                 m.read(gettext, readsubtree)
                 m.setnode(self._node)
                 self._data = m
             else:
                 if self._node in store.fulltextcache:
                     text = pycompat.bytestr(store.fulltextcache[self._node])
                 else:
                     text = store.revision(self._node)
                     arraytext = bytearray(text)
                     store.fulltextcache[self._node] = arraytext
                 self._data = treemanifest(dir=self._dir, text=text)
 
         return self._data
 
     def node(self):
         return self._node
 
     def new(self, dir=b''):
         return memtreemanifestctx(self._manifestlog, dir=dir)
 
     def copy(self):
         memmf = memtreemanifestctx(self._manifestlog, dir=self._dir)
         memmf._treemanifest = self.read().copy()
         return memmf
 
     @propertycache
     def parents(self):
         return self._storage().parents(self._node)
 
     def readdelta(self, shallow=False):
         '''Returns a manifest containing just the entries that are present
         in this manifest, but not in its p1 manifest. This is efficient to read
         if the revlog delta is already p1.
 
         If `shallow` is True, this will read the delta for this directory,
         without recursively reading subdirectory manifests. Instead, any
         subdirectory entry will be reported as it appears in the manifest, i.e.
         the subdirectory will be reported among files and distinguished only by
         its 't' flag.
         '''
         store = self._storage()
         if shallow:
             r = store.rev(self._node)
             d = mdiff.patchtext(store.revdiff(store.deltaparent(r), r))
             return manifestdict(d)
         else:
             # Need to perform a slow delta
             r0 = store.deltaparent(store.rev(self._node))
             m0 = self._manifestlog.get(self._dir, store.node(r0)).read()
             m1 = self.read()
             md = treemanifest(dir=self._dir)
             for f, ((n0, fl0), (n1, fl1)) in pycompat.iteritems(m0.diff(m1)):
                 if n1:
                     md[f] = n1
                     if fl1:
                         md.setflag(f, fl1)
             return md
 
     def readfast(self, shallow=False):
         '''Calls either readdelta or read, based on which would be less work.
         readdelta is called if the delta is against the p1, and therefore can be
         read quickly.
 
         If `shallow` is True, it only returns the entries from this manifest,
         and not any submanifests.
         '''
         store = self._storage()
         r = store.rev(self._node)
         deltaparent = store.deltaparent(r)
         if deltaparent != nullrev and deltaparent in store.parentrevs(r):
             return self.readdelta(shallow=shallow)
 
         if shallow:
             return manifestdict(store.revision(self._node))
         else:
             return self.read()
 
     def find(self, key):
         return self.read().find(key)
 
 
 class excludeddir(treemanifest):
     """Stand-in for a directory that is excluded from the repository.
 
     With narrowing active on a repository that uses treemanifests,
     some of the directory revlogs will be excluded from the resulting
     clone. This is a huge storage win for clients, but means we need
     some sort of pseudo-manifest to surface to internals so we can
     detect a merge conflict outside the narrowspec. That's what this
     class is: it stands in for a directory whose node is known, but
     whose contents are unknown.
     """
 
     def __init__(self, dir, node):
         super(excludeddir, self).__init__(dir)
         self._node = node
         # Add an empty file, which will be included by iterators and such,
         # appearing as the directory itself (i.e. something like "dir/")
         self._files[b''] = node
         self._flags[b''] = b't'
 
     # Manifests outside the narrowspec should never be modified, so avoid
     # copying. This makes a noticeable difference when there are very many
     # directories outside the narrowspec. Also, it makes sense for the copy to
     # be of the same type as the original, which would not happen with the
     # super type's copy().
     def copy(self):
         return self
 
 
 class excludeddirmanifestctx(treemanifestctx):
     """context wrapper for excludeddir - see that docstring for rationale"""
 
     def __init__(self, dir, node):
         self._dir = dir
         self._node = node
 
     def read(self):
         return excludeddir(self._dir, self._node)
 
     def write(self, *args):
         raise error.ProgrammingError(
             b'attempt to write manifest from excluded dir %s' % self._dir
         )
 
 
 class excludedmanifestrevlog(manifestrevlog):
     """Stand-in for excluded treemanifest revlogs.
 
     When narrowing is active on a treemanifest repository, we'll have
     references to directories we can't see due to the revlog being
     skipped. This class exists to conform to the manifestrevlog
     interface for those directories and proactively prevent writes to
     outside the narrowspec.
     """
 
     def __init__(self, dir):
         self._dir = dir
 
     def __len__(self):
         raise error.ProgrammingError(
             b'attempt to get length of excluded dir %s' % self._dir
         )
 
     def rev(self, node):
         raise error.ProgrammingError(
             b'attempt to get rev from excluded dir %s' % self._dir
         )
 
     def linkrev(self, node):
         raise error.ProgrammingError(
             b'attempt to get linkrev from excluded dir %s' % self._dir
         )
 
     def node(self, rev):
         raise error.ProgrammingError(
             b'attempt to get node from excluded dir %s' % self._dir
         )
 
     def add(self, *args, **kwargs):
         # We should never write entries in dirlogs outside the narrow clone.
         # However, the method still gets called from writesubtree() in
         # _addtree(), so we need to handle it. We should possibly make that
         # avoid calling add() with a clean manifest (_dirty is always False
         # in excludeddir instances).
         pass

mercurial/patch.py

@@ -1,3226 +1,3227 b''
 # patch.py - patch file parsing routines
 #
 # Copyright 2006 Brendan Cully <[e-mail]>
 # Copyright 2007 Chris Mason <[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.
 
 from __future__ import absolute_import, print_function
 
 import collections
 import contextlib
 import copy
 import errno
 import os
 import re
 import shutil
 import zlib
 
 from .i18n import _
 from .node import (
     hex,
     short,
 )
 from .pycompat import open
 from . import (
     copies,
     diffhelper,
     diffutil,
     encoding,
     error,
     mail,
     mdiff,
     pathutil,
     pycompat,
     scmutil,
     similar,
     util,
     vfs as vfsmod,
 )
 from .utils import (
     dateutil,
     hashutil,
     procutil,
     stringutil,
 )
 
 stringio = util.stringio
 
 gitre = re.compile(br'diff --git a/(.*) b/(.*)')
 tabsplitter = re.compile(br'(\t+|[^\t]+)')
 wordsplitter = re.compile(
     br'(\t+| +|[a-zA-Z0-9_\x80-\xff]+|[^ \ta-zA-Z0-9_\x80-\xff])'
 )
 
 PatchError = error.PatchError
 
 # public functions
 
 
 def split(stream):
     '''return an iterator of individual patches from a stream'''
 
     def isheader(line, inheader):
         if inheader and line.startswith((b' ', b'\t')):
             # continuation
             return True
         if line.startswith((b' ', b'-', b'+')):
             # diff line - don't check for header pattern in there
             return False
         l = line.split(b': ', 1)
         return len(l) == 2 and b' ' not in l[0]
 
     def chunk(lines):
         return stringio(b''.join(lines))
 
     def hgsplit(stream, cur):
         inheader = True
 
         for line in stream:
             if not line.strip():
                 inheader = False
             if not inheader and line.startswith(b'# HG changeset patch'):
                 yield chunk(cur)
                 cur = []
                 inheader = True
 
             cur.append(line)
 
         if cur:
             yield chunk(cur)
 
     def mboxsplit(stream, cur):
         for line in stream:
             if line.startswith(b'From '):
                 for c in split(chunk(cur[1:])):
                     yield c
                 cur = []
 
             cur.append(line)
 
         if cur:
             for c in split(chunk(cur[1:])):
                 yield c
 
     def mimesplit(stream, cur):
         def msgfp(m):
             fp = stringio()
             g = mail.Generator(fp, mangle_from_=False)
             g.flatten(m)
             fp.seek(0)
             return fp
 
         for line in stream:
             cur.append(line)
         c = chunk(cur)
 
         m = mail.parse(c)
         if not m.is_multipart():
             yield msgfp(m)
         else:
             ok_types = (b'text/plain', b'text/x-diff', b'text/x-patch')
             for part in m.walk():
                 ct = part.get_content_type()
                 if ct not in ok_types:
                     continue
                 yield msgfp(part)
 
     def headersplit(stream, cur):
         inheader = False
 
         for line in stream:
             if not inheader and isheader(line, inheader):
                 yield chunk(cur)
                 cur = []
                 inheader = True
             if inheader and not isheader(line, inheader):
                 inheader = False
 
             cur.append(line)
 
         if cur:
             yield chunk(cur)
 
     def remainder(cur):
         yield chunk(cur)
 
     class fiter(object):
         def __init__(self, fp):
             self.fp = fp
 
         def __iter__(self):
             return self
 
         def next(self):
             l = self.fp.readline()
             if not l:
                 raise StopIteration
             return l
 
         __next__ = next
 
     inheader = False
     cur = []
 
     mimeheaders = [b'content-type']
 
     if not util.safehasattr(stream, b'next'):
         # http responses, for example, have readline but not next
         stream = fiter(stream)
 
     for line in stream:
         cur.append(line)
         if line.startswith(b'# HG changeset patch'):
             return hgsplit(stream, cur)
         elif line.startswith(b'From '):
             return mboxsplit(stream, cur)
         elif isheader(line, inheader):
             inheader = True
             if line.split(b':', 1)[0].lower() in mimeheaders:
                 # let email parser handle this
                 return mimesplit(stream, cur)
         elif line.startswith(b'--- ') and inheader:
             # No evil headers seen by diff start, split by hand
             return headersplit(stream, cur)
         # Not enough info, keep reading
 
     # if we are here, we have a very plain patch
     return remainder(cur)
 
 
 ## Some facility for extensible patch parsing:
 # list of pairs ("header to match", "data key")
 patchheadermap = [
     (b'Date', b'date'),
     (b'Branch', b'branch'),
     (b'Node ID', b'nodeid'),
 ]
 
 
 @contextlib.contextmanager
 def extract(ui, fileobj):
     '''extract patch from data read from fileobj.
 
     patch can be a normal patch or contained in an email message.
 
     return a dictionary. Standard keys are:
       - filename,
       - message,
       - user,
       - date,
       - branch,
       - node,
       - p1,
       - p2.
     Any item can be missing from the dictionary. If filename is missing,
     fileobj did not contain a patch. Caller must unlink filename when done.'''
 
     fd, tmpname = pycompat.mkstemp(prefix=b'hg-patch-')
     tmpfp = os.fdopen(fd, 'wb')
     try:
         yield _extract(ui, fileobj, tmpname, tmpfp)
     finally:
         tmpfp.close()
         os.unlink(tmpname)
 
 
 def _extract(ui, fileobj, tmpname, tmpfp):
 
     # attempt to detect the start of a patch
     # (this heuristic is borrowed from quilt)
     diffre = re.compile(
         br'^(?:Index:[ \t]|diff[ \t]-|RCS file: |'
         br'retrieving revision [0-9]+(\.[0-9]+)*$|'
         br'---[ \t].*?^\+\+\+[ \t]|'
         br'\*\*\*[ \t].*?^---[ \t])',
         re.MULTILINE | re.DOTALL,
     )
 
     data = {}
 
     msg = mail.parse(fileobj)
 
     subject = msg['Subject'] and mail.headdecode(msg['Subject'])
     data[b'user'] = msg['From'] and mail.headdecode(msg['From'])
     if not subject and not data[b'user']:
         # Not an email, restore parsed headers if any
         subject = (
             b'\n'.join(
                 b': '.join(map(encoding.strtolocal, h)) for h in msg.items()
             )
             + b'\n'
         )
 
     # should try to parse msg['Date']
     parents = []
 
     nodeid = msg['X-Mercurial-Node']
     if nodeid:
         data[b'nodeid'] = nodeid = mail.headdecode(nodeid)
         ui.debug(b'Node ID: %s\n' % nodeid)
 
     if subject:
         if subject.startswith(b'[PATCH'):
             pend = subject.find(b']')
             if pend >= 0:
                 subject = subject[pend + 1 :].lstrip()
         subject = re.sub(br'\n[ \t]+', b' ', subject)
         ui.debug(b'Subject: %s\n' % subject)
     if data[b'user']:
         ui.debug(b'From: %s\n' % data[b'user'])
     diffs_seen = 0
     ok_types = (b'text/plain', b'text/x-diff', b'text/x-patch')
     message = b''
     for part in msg.walk():
         content_type = pycompat.bytestr(part.get_content_type())
         ui.debug(b'Content-Type: %s\n' % content_type)
         if content_type not in ok_types:
             continue
         payload = part.get_payload(decode=True)
         m = diffre.search(payload)
         if m:
             hgpatch = False
             hgpatchheader = False
             ignoretext = False
 
             ui.debug(b'found patch at byte %d\n' % m.start(0))
             diffs_seen += 1
             cfp = stringio()
             for line in payload[: m.start(0)].splitlines():
                 if line.startswith(b'# HG changeset patch') and not hgpatch:
                     ui.debug(b'patch generated by hg export\n')
                     hgpatch = True
                     hgpatchheader = True
                     # drop earlier commit message content
                     cfp.seek(0)
                     cfp.truncate()
                     subject = None
                 elif hgpatchheader:
                     if line.startswith(b'# User '):
                         data[b'user'] = line[7:]
                         ui.debug(b'From: %s\n' % data[b'user'])
                     elif line.startswith(b"# Parent "):
                         parents.append(line[9:].lstrip())
                     elif line.startswith(b"# "):
                         for header, key in patchheadermap:
                             prefix = b'# %s ' % header
                             if line.startswith(prefix):
                                 data[key] = line[len(prefix) :]
                                 ui.debug(b'%s: %s\n' % (header, data[key]))
                     else:
                         hgpatchheader = False
                 elif line == b'---':
                     ignoretext = True
                 if not hgpatchheader and not ignoretext:
                     cfp.write(line)
                     cfp.write(b'\n')
             message = cfp.getvalue()
             if tmpfp:
                 tmpfp.write(payload)
                 if not payload.endswith(b'\n'):
                     tmpfp.write(b'\n')
         elif not diffs_seen and message and content_type == b'text/plain':
             message += b'\n' + payload
 
     if subject and not message.startswith(subject):
         message = b'%s\n%s' % (subject, message)
     data[b'message'] = message
     tmpfp.close()
     if parents:
         data[b'p1'] = parents.pop(0)
         if parents:
             data[b'p2'] = parents.pop(0)
 
     if diffs_seen:
         data[b'filename'] = tmpname
 
     return data
 
 
 class patchmeta(object):
     """Patched file metadata
 
     'op' is the performed operation within ADD, DELETE, RENAME, MODIFY
     or COPY.  'path' is patched file path. 'oldpath' is set to the
     origin file when 'op' is either COPY or RENAME, None otherwise. If
     file mode is changed, 'mode' is a tuple (islink, isexec) where
     'islink' is True if the file is a symlink and 'isexec' is True if
     the file is executable. Otherwise, 'mode' is None.
     """
 
     def __init__(self, path):
         self.path = path
         self.oldpath = None
         self.mode = None
         self.op = b'MODIFY'
         self.binary = False
 
     def setmode(self, mode):
         islink = mode & 0o20000
         isexec = mode & 0o100
         self.mode = (islink, isexec)
 
     def copy(self):
         other = patchmeta(self.path)
         other.oldpath = self.oldpath
         other.mode = self.mode
         other.op = self.op
         other.binary = self.binary
         return other
 
     def _ispatchinga(self, afile):
         if afile == b'/dev/null':
             return self.op == b'ADD'
         return afile == b'a/' + (self.oldpath or self.path)
 
     def _ispatchingb(self, bfile):
         if bfile == b'/dev/null':
             return self.op == b'DELETE'
         return bfile == b'b/' + self.path
 
     def ispatching(self, afile, bfile):
         return self._ispatchinga(afile) and self._ispatchingb(bfile)
 
     def __repr__(self):
         return "<patchmeta %s %r>" % (self.op, self.path)
 
 
 def readgitpatch(lr):
     """extract git-style metadata about patches from <patchname>"""
 
     # Filter patch for git information
     gp = None
     gitpatches = []
     for line in lr:
         line = line.rstrip(b' \r\n')
         if line.startswith(b'diff --git a/'):
             m = gitre.match(line)
             if m:
                 if gp:
                     gitpatches.append(gp)
                 dst = m.group(2)
                 gp = patchmeta(dst)
         elif gp:
             if line.startswith(b'--- '):
                 gitpatches.append(gp)
                 gp = None
                 continue
             if line.startswith(b'rename from '):
                 gp.op = b'RENAME'
                 gp.oldpath = line[12:]
             elif line.startswith(b'rename to '):
                 gp.path = line[10:]
             elif line.startswith(b'copy from '):
                 gp.op = b'COPY'
                 gp.oldpath = line[10:]
             elif line.startswith(b'copy to '):
                 gp.path = line[8:]
             elif line.startswith(b'deleted file'):
                 gp.op = b'DELETE'
             elif line.startswith(b'new file mode '):
                 gp.op = b'ADD'
                 gp.setmode(int(line[-6:], 8))
             elif line.startswith(b'new mode '):
                 gp.setmode(int(line[-6:], 8))
             elif line.startswith(b'GIT binary patch'):
                 gp.binary = True
     if gp:
         gitpatches.append(gp)
 
     return gitpatches
 
 
 class linereader(object):
     # simple class to allow pushing lines back into the input stream
     def __init__(self, fp):
         self.fp = fp
         self.buf = []
 
     def push(self, line):
         if line is not None:
             self.buf.append(line)
 
     def readline(self):
         if self.buf:
             l = self.buf[0]
             del self.buf[0]
             return l
         return self.fp.readline()
 
     def __iter__(self):
         return iter(self.readline, b'')
 
 
 class abstractbackend(object):
     def __init__(self, ui):
         self.ui = ui
 
     def getfile(self, fname):
         """Return target file data and flags as a (data, (islink,
         isexec)) tuple. Data is None if file is missing/deleted.
         """
         raise NotImplementedError
 
     def setfile(self, fname, data, mode, copysource):
         """Write data to target file fname and set its mode. mode is a
         (islink, isexec) tuple. If data is None, the file content should
         be left unchanged. If the file is modified after being copied,
         copysource is set to the original file name.
         """
         raise NotImplementedError
 
     def unlink(self, fname):
         """Unlink target file."""
         raise NotImplementedError
 
     def writerej(self, fname, failed, total, lines):
         """Write rejected lines for fname. total is the number of hunks
         which failed to apply and total the total number of hunks for this
         files.
         """
 
     def exists(self, fname):
         raise NotImplementedError
 
     def close(self):
         raise NotImplementedError
 
 
 class fsbackend(abstractbackend):
     def __init__(self, ui, basedir):
         super(fsbackend, self).__init__(ui)
         self.opener = vfsmod.vfs(basedir)
 
     def getfile(self, fname):
         if self.opener.islink(fname):
             return (self.opener.readlink(fname), (True, False))
 
         isexec = False
         try:
             isexec = self.opener.lstat(fname).st_mode & 0o100 != 0
         except OSError as e:
             if e.errno != errno.ENOENT:
                 raise
         try:
             return (self.opener.read(fname), (False, isexec))
         except IOError as e:
             if e.errno != errno.ENOENT:
                 raise
             return None, None
 
     def setfile(self, fname, data, mode, copysource):
         islink, isexec = mode
         if data is None:
             self.opener.setflags(fname, islink, isexec)
             return
         if islink:
             self.opener.symlink(data, fname)
         else:
             self.opener.write(fname, data)
             if isexec:
                 self.opener.setflags(fname, False, True)
 
     def unlink(self, fname):
         rmdir = self.ui.configbool(b'experimental', b'removeemptydirs')
         self.opener.unlinkpath(fname, ignoremissing=True, rmdir=rmdir)
 
     def writerej(self, fname, failed, total, lines):
         fname = fname + b".rej"
         self.ui.warn(
             _(b"%d out of %d hunks FAILED -- saving rejects to file %s\n")
             % (failed, total, fname)
         )
         fp = self.opener(fname, b'w')
         fp.writelines(lines)
         fp.close()
 
     def exists(self, fname):
         return self.opener.lexists(fname)
 
 
 class workingbackend(fsbackend):
     def __init__(self, ui, repo, similarity):
         super(workingbackend, self).__init__(ui, repo.root)
         self.repo = repo
         self.similarity = similarity
         self.removed = set()
         self.changed = set()
         self.copied = []
 
     def _checkknown(self, fname):
         if self.repo.dirstate[fname] == b'?' and self.exists(fname):
             raise PatchError(_(b'cannot patch %s: file is not tracked') % fname)
 
     def setfile(self, fname, data, mode, copysource):
         self._checkknown(fname)
         super(workingbackend, self).setfile(fname, data, mode, copysource)
         if copysource is not None:
             self.copied.append((copysource, fname))
         self.changed.add(fname)
 
     def unlink(self, fname):
         self._checkknown(fname)
         super(workingbackend, self).unlink(fname)
         self.removed.add(fname)
         self.changed.add(fname)
 
     def close(self):
         wctx = self.repo[None]
         changed = set(self.changed)
         for src, dst in self.copied:
             scmutil.dirstatecopy(self.ui, self.repo, wctx, src, dst)
         if self.removed:
             wctx.forget(sorted(self.removed))
             for f in self.removed:
                 if f not in self.repo.dirstate:
                     # File was deleted and no longer belongs to the
                     # dirstate, it was probably marked added then
                     # deleted, and should not be considered by
                     # marktouched().
                     changed.discard(f)
         if changed:
             scmutil.marktouched(self.repo, changed, self.similarity)
         return sorted(self.changed)
 
 
 class filestore(object):
     def __init__(self, maxsize=None):
         self.opener = None
         self.files = {}
         self.created = 0
         self.maxsize = maxsize
         if self.maxsize is None:
             self.maxsize = 4 * (2 ** 20)
         self.size = 0
         self.data = {}
 
     def setfile(self, fname, data, mode, copied=None):
         if self.maxsize < 0 or (len(data) + self.size) <= self.maxsize:
             self.data[fname] = (data, mode, copied)
             self.size += len(data)
         else:
             if self.opener is None:
                 root = pycompat.mkdtemp(prefix=b'hg-patch-')
                 self.opener = vfsmod.vfs(root)
             # Avoid filename issues with these simple names
             fn = b'%d' % self.created
             self.opener.write(fn, data)
             self.created += 1
             self.files[fname] = (fn, mode, copied)
 
     def getfile(self, fname):
         if fname in self.data:
             return self.data[fname]
         if not self.opener or fname not in self.files:
             return None, None, None
         fn, mode, copied = self.files[fname]
         return self.opener.read(fn), mode, copied
 
     def close(self):
         if self.opener:
             shutil.rmtree(self.opener.base)
 
 
 class repobackend(abstractbackend):
     def __init__(self, ui, repo, ctx, store):
         super(repobackend, self).__init__(ui)
         self.repo = repo
         self.ctx = ctx
         self.store = store
         self.changed = set()
         self.removed = set()
         self.copied = {}
 
     def _checkknown(self, fname):
         if fname not in self.ctx:
             raise PatchError(_(b'cannot patch %s: file is not tracked') % fname)
 
     def getfile(self, fname):
         try:
             fctx = self.ctx[fname]
         except error.LookupError:
             return None, None
         flags = fctx.flags()
         return fctx.data(), (b'l' in flags, b'x' in flags)
 
     def setfile(self, fname, data, mode, copysource):
         if copysource:
             self._checkknown(copysource)
         if data is None:
             data = self.ctx[fname].data()
         self.store.setfile(fname, data, mode, copysource)
         self.changed.add(fname)
         if copysource:
             self.copied[fname] = copysource
 
     def unlink(self, fname):
         self._checkknown(fname)
         self.removed.add(fname)
 
     def exists(self, fname):
         return fname in self.ctx
 
     def close(self):
         return self.changed | self.removed
 
 
 # @@ -start,len +start,len @@ or @@ -start +start @@ if len is 1
 unidesc = re.compile(br'@@ -(\d+)(?:,(\d+))? \+(\d+)(?:,(\d+))? @@')
 contextdesc = re.compile(br'(?:---|\*\*\*) (\d+)(?:,(\d+))? (?:---|\*\*\*)')
 eolmodes = [b'strict', b'crlf', b'lf', b'auto']
 
 
 class patchfile(object):
     def __init__(self, ui, gp, backend, store, eolmode=b'strict'):
         self.fname = gp.path
         self.eolmode = eolmode
         self.eol = None
         self.backend = backend
         self.ui = ui
         self.lines = []
         self.exists = False
         self.missing = True
         self.mode = gp.mode
         self.copysource = gp.oldpath
         self.create = gp.op in (b'ADD', b'COPY', b'RENAME')
         self.remove = gp.op == b'DELETE'
         if self.copysource is None:
             data, mode = backend.getfile(self.fname)
         else:
             data, mode = store.getfile(self.copysource)[:2]
         if data is not None:
             self.exists = self.copysource is None or backend.exists(self.fname)
             self.missing = False
             if data:
                 self.lines = mdiff.splitnewlines(data)
             if self.mode is None:
                 self.mode = mode
             if self.lines:
                 # Normalize line endings
                 if self.lines[0].endswith(b'\r\n'):
                     self.eol = b'\r\n'
                 elif self.lines[0].endswith(b'\n'):
                     self.eol = b'\n'
                 if eolmode != b'strict':
                     nlines = []
                     for l in self.lines:
                         if l.endswith(b'\r\n'):
                             l = l[:-2] + b'\n'
                         nlines.append(l)
                     self.lines = nlines
         else:
             if self.create:
                 self.missing = False
             if self.mode is None:
                 self.mode = (False, False)
         if self.missing:
             self.ui.warn(_(b"unable to find '%s' for patching\n") % self.fname)
             self.ui.warn(
                 _(
                     b"(use '--prefix' to apply patch relative to the "
                     b"current directory)\n"
                 )
             )
 
         self.hash = {}
         self.dirty = 0
         self.offset = 0
         self.skew = 0
         self.rej = []
         self.fileprinted = False
         self.printfile(False)
         self.hunks = 0
 
     def writelines(self, fname, lines, mode):
         if self.eolmode == b'auto':
             eol = self.eol
         elif self.eolmode == b'crlf':
             eol = b'\r\n'
         else:
             eol = b'\n'
 
         if self.eolmode != b'strict' and eol and eol != b'\n':
             rawlines = []
             for l in lines:
                 if l and l.endswith(b'\n'):
                     l = l[:-1] + eol
                 rawlines.append(l)
             lines = rawlines
 
         self.backend.setfile(fname, b''.join(lines), mode, self.copysource)
 
     def printfile(self, warn):
         if self.fileprinted:
             return
         if warn or self.ui.verbose:
             self.fileprinted = True
         s = _(b"patching file %s\n") % self.fname
         if warn:
             self.ui.warn(s)
         else:
             self.ui.note(s)
 
     def findlines(self, l, linenum):
         # looks through the hash and finds candidate lines.  The
         # result is a list of line numbers sorted based on distance
         # from linenum
 
         cand = self.hash.get(l, [])
         if len(cand) > 1:
             # resort our list of potentials forward then back.
             cand.sort(key=lambda x: abs(x - linenum))
         return cand
 
     def write_rej(self):
         # our rejects are a little different from patch(1).  This always
         # creates rejects in the same form as the original patch.  A file
         # header is inserted so that you can run the reject through patch again
         # without having to type the filename.
         if not self.rej:
             return
         base = os.path.basename(self.fname)
         lines = [b"--- %s\n+++ %s\n" % (base, base)]
         for x in self.rej:
             for l in x.hunk:
                 lines.append(l)
                 if l[-1:] != b'\n':
                     lines.append(b"\n\\ No newline at end of file\n")
         self.backend.writerej(self.fname, len(self.rej), self.hunks, lines)
 
     def apply(self, h):
         if not h.complete():
             raise PatchError(
                 _(b"bad hunk #%d %s (%d %d %d %d)")
                 % (h.number, h.desc, len(h.a), h.lena, len(h.b), h.lenb)
             )
 
         self.hunks += 1
 
         if self.missing:
             self.rej.append(h)
             return -1
 
         if self.exists and self.create:
             if self.copysource:
                 self.ui.warn(
                     _(b"cannot create %s: destination already exists\n")
                     % self.fname
                 )
             else:
                 self.ui.warn(_(b"file %s already exists\n") % self.fname)
             self.rej.append(h)
             return -1
 
         if isinstance(h, binhunk):
             if self.remove:
                 self.backend.unlink(self.fname)
             else:
                 l = h.new(self.lines)
                 self.lines[:] = l
                 self.offset += len(l)
                 self.dirty = True
             return 0
 
         horig = h
         if (
             self.eolmode in (b'crlf', b'lf')
             or self.eolmode == b'auto'
             and self.eol
         ):
             # If new eols are going to be normalized, then normalize
             # hunk data before patching. Otherwise, preserve input
             # line-endings.
             h = h.getnormalized()
 
         # fast case first, no offsets, no fuzz
         old, oldstart, new, newstart = h.fuzzit(0, False)
         oldstart += self.offset
         orig_start = oldstart
         # if there's skew we want to emit the "(offset %d lines)" even
         # when the hunk cleanly applies at start + skew, so skip the
         # fast case code
         if self.skew == 0 and diffhelper.testhunk(old, self.lines, oldstart):
             if self.remove:
                 self.backend.unlink(self.fname)
             else:
                 self.lines[oldstart : oldstart + len(old)] = new
                 self.offset += len(new) - len(old)
                 self.dirty = True
             return 0
 
         # ok, we couldn't match the hunk. Lets look for offsets and fuzz it
         self.hash = {}
         for x, s in enumerate(self.lines):
             self.hash.setdefault(s, []).append(x)
 
         for fuzzlen in pycompat.xrange(
             self.ui.configint(b"patch", b"fuzz") + 1
         ):
             for toponly in [True, False]:
                 old, oldstart, new, newstart = h.fuzzit(fuzzlen, toponly)
                 oldstart = oldstart + self.offset + self.skew
                 oldstart = min(oldstart, len(self.lines))
                 if old:
                     cand = self.findlines(old[0][1:], oldstart)
                 else:
                     # Only adding lines with no or fuzzed context, just
                     # take the skew in account
                     cand = [oldstart]
 
                 for l in cand:
                     if not old or diffhelper.testhunk(old, self.lines, l):
                         self.lines[l : l + len(old)] = new
                         self.offset += len(new) - len(old)
                         self.skew = l - orig_start
                         self.dirty = True
                         offset = l - orig_start - fuzzlen
                         if fuzzlen:
                             msg = _(
                                 b"Hunk #%d succeeded at %d "
                                 b"with fuzz %d "
                                 b"(offset %d lines).\n"
                             )
                             self.printfile(True)
                             self.ui.warn(
                                 msg % (h.number, l + 1, fuzzlen, offset)
                             )
                         else:
                             msg = _(
                                 b"Hunk #%d succeeded at %d "
                                 b"(offset %d lines).\n"
                             )
                             self.ui.note(msg % (h.number, l + 1, offset))
                         return fuzzlen
         self.printfile(True)
         self.ui.warn(_(b"Hunk #%d FAILED at %d\n") % (h.number, orig_start))
         self.rej.append(horig)
         return -1
 
     def close(self):
         if self.dirty:
             self.writelines(self.fname, self.lines, self.mode)
         self.write_rej()
         return len(self.rej)
 
 
 class header(object):
     """patch header
     """
 
     diffgit_re = re.compile(b'diff --git a/(.*) b/(.*)$')
     diff_re = re.compile(b'diff -r .* (.*)$')
     allhunks_re = re.compile(b'(?:index|deleted file) ')
     pretty_re = re.compile(b'(?:new file|deleted file) ')
     special_re = re.compile(b'(?:index|deleted|copy|rename|new mode) ')
     newfile_re = re.compile(b'(?:new file|copy to|rename to)')
 
     def __init__(self, header):
         self.header = header
         self.hunks = []
 
     def binary(self):
         return any(h.startswith(b'index ') for h in self.header)
 
     def pretty(self, fp):
         for h in self.header:
             if h.startswith(b'index '):
                 fp.write(_(b'this modifies a binary file (all or nothing)\n'))
                 break
             if self.pretty_re.match(h):
                 fp.write(h)
                 if self.binary():
                     fp.write(_(b'this is a binary file\n'))
                 break
             if h.startswith(b'---'):
                 fp.write(
                     _(b'%d hunks, %d lines changed\n')
                     % (
                         len(self.hunks),
                         sum([max(h.added, h.removed) for h in self.hunks]),
                     )
                 )
                 break
             fp.write(h)
 
     def write(self, fp):
         fp.write(b''.join(self.header))
 
     def allhunks(self):
         return any(self.allhunks_re.match(h) for h in self.header)
 
     def files(self):
         match = self.diffgit_re.match(self.header[0])
         if match:
             fromfile, tofile = match.groups()
             if fromfile == tofile:
                 return [fromfile]
             return [fromfile, tofile]
         else:
             return self.diff_re.match(self.header[0]).groups()
 
     def filename(self):
         return self.files()[-1]
 
     def __repr__(self):
         return '<header %s>' % (
             ' '.join(pycompat.rapply(pycompat.fsdecode, self.files()))
         )
 
     def isnewfile(self):
         return any(self.newfile_re.match(h) for h in self.header)
 
     def special(self):
         # Special files are shown only at the header level and not at the hunk
         # level for example a file that has been deleted is a special file.
         # The user cannot change the content of the operation, in the case of
         # the deleted file he has to take the deletion or not take it, he
         # cannot take some of it.
         # Newly added files are special if they are empty, they are not special
         # if they have some content as we want to be able to change it
         nocontent = len(self.header) == 2
         emptynewfile = self.isnewfile() and nocontent
         return emptynewfile or any(
             self.special_re.match(h) for h in self.header
         )
 
 
 class recordhunk(object):
     """patch hunk
 
     XXX shouldn't we merge this with the other hunk class?
     """
 
     def __init__(
         self,
         header,
         fromline,
         toline,
         proc,
         before,
         hunk,
         after,
         maxcontext=None,
     ):
         def trimcontext(lines, reverse=False):
             if maxcontext is not None:
                 delta = len(lines) - maxcontext
                 if delta > 0:
                     if reverse:
                         return delta, lines[delta:]
                     else:
                         return delta, lines[:maxcontext]
             return 0, lines
 
         self.header = header
         trimedbefore, self.before = trimcontext(before, True)
         self.fromline = fromline + trimedbefore
         self.toline = toline + trimedbefore
         _trimedafter, self.after = trimcontext(after, False)
         self.proc = proc
         self.hunk = hunk
         self.added, self.removed = self.countchanges(self.hunk)
 
     def __eq__(self, v):
         if not isinstance(v, recordhunk):
             return False
 
         return (
             (v.hunk == self.hunk)
             and (v.proc == self.proc)
             and (self.fromline == v.fromline)
             and (self.header.files() == v.header.files())
         )
 
     def __hash__(self):
         return hash(
             (
                 tuple(self.hunk),
                 tuple(self.header.files()),
                 self.fromline,
                 self.proc,
             )
         )
 
     def countchanges(self, hunk):
         """hunk -> (n+,n-)"""
         add = len([h for h in hunk if h.startswith(b'+')])
         rem = len([h for h in hunk if h.startswith(b'-')])
         return add, rem
 
     def reversehunk(self):
         """return another recordhunk which is the reverse of the hunk
 
         If this hunk is diff(A, B), the returned hunk is diff(B, A). To do
         that, swap fromline/toline and +/- signs while keep other things
         unchanged.
         """
         m = {b'+': b'-', b'-': b'+', b'\\': b'\\'}
         hunk = [b'%s%s' % (m[l[0:1]], l[1:]) for l in self.hunk]
         return recordhunk(
             self.header,
             self.toline,
             self.fromline,
             self.proc,
             self.before,
             hunk,
             self.after,
         )
 
     def write(self, fp):
         delta = len(self.before) + len(self.after)
         if self.after and self.after[-1] == b'\\ No newline at end of file\n':
             delta -= 1
         fromlen = delta + self.removed
         tolen = delta + self.added
         fp.write(
             b'@@ -%d,%d +%d,%d @@%s\n'
             % (
                 self.fromline,
                 fromlen,
                 self.toline,
                 tolen,
                 self.proc and (b' ' + self.proc),
             )
         )
         fp.write(b''.join(self.before + self.hunk + self.after))
 
     pretty = write
 
     def filename(self):
         return self.header.filename()
 
+    @encoding.strmethod
     def __repr__(self):
         return b'<hunk %r@%d>' % (self.filename(), self.fromline)
 
 
 def getmessages():
     return {
         b'multiple': {
             b'apply': _(b"apply change %d/%d to '%s'?"),
             b'discard': _(b"discard change %d/%d to '%s'?"),
             b'keep': _(b"keep change %d/%d to '%s'?"),
             b'record': _(b"record change %d/%d to '%s'?"),
         },
         b'single': {
             b'apply': _(b"apply this change to '%s'?"),
             b'discard': _(b"discard this change to '%s'?"),
             b'keep': _(b"keep this change to '%s'?"),
             b'record': _(b"record this change to '%s'?"),
         },
         b'help': {
             b'apply': _(
                 b'[Ynesfdaq?]'
                 b'$$ &Yes, apply this change'
                 b'$$ &No, skip this change'
                 b'$$ &Edit this change manually'
                 b'$$ &Skip remaining changes to this file'
                 b'$$ Apply remaining changes to this &file'
                 b'$$ &Done, skip remaining changes and files'
                 b'$$ Apply &all changes to all remaining files'
                 b'$$ &Quit, applying no changes'
                 b'$$ &? (display help)'
             ),
             b'discard': _(
                 b'[Ynesfdaq?]'
                 b'$$ &Yes, discard this change'
                 b'$$ &No, skip this change'
                 b'$$ &Edit this change manually'
                 b'$$ &Skip remaining changes to this file'
                 b'$$ Discard remaining changes to this &file'
                 b'$$ &Done, skip remaining changes and files'
                 b'$$ Discard &all changes to all remaining files'
                 b'$$ &Quit, discarding no changes'
                 b'$$ &? (display help)'
             ),
             b'keep': _(
                 b'[Ynesfdaq?]'
                 b'$$ &Yes, keep this change'
                 b'$$ &No, skip this change'
                 b'$$ &Edit this change manually'
                 b'$$ &Skip remaining changes to this file'
                 b'$$ Keep remaining changes to this &file'
                 b'$$ &Done, skip remaining changes and files'
                 b'$$ Keep &all changes to all remaining files'
                 b'$$ &Quit, keeping all changes'
                 b'$$ &? (display help)'
             ),
             b'record': _(
                 b'[Ynesfdaq?]'
                 b'$$ &Yes, record this change'
                 b'$$ &No, skip this change'
                 b'$$ &Edit this change manually'
                 b'$$ &Skip remaining changes to this file'
                 b'$$ Record remaining changes to this &file'
                 b'$$ &Done, skip remaining changes and files'
                 b'$$ Record &all changes to all remaining files'
                 b'$$ &Quit, recording no changes'
                 b'$$ &? (display help)'
             ),
         },
     }
 
 
 def filterpatch(ui, headers, match, operation=None):
     """Interactively filter patch chunks into applied-only chunks"""
     messages = getmessages()
 
     if operation is None:
         operation = b'record'
 
     def prompt(skipfile, skipall, query, chunk):
         """prompt query, and process base inputs
 
         - y/n for the rest of file
         - y/n for the rest
         - ? (help)
         - q (quit)
 
         Return True/False and possibly updated skipfile and skipall.
         """
         newpatches = None
         if skipall is not None:
             return skipall, skipfile, skipall, newpatches
         if skipfile is not None:
             return skipfile, skipfile, skipall, newpatches
         while True:
             resps = messages[b'help'][operation]
             # IMPORTANT: keep the last line of this prompt short (<40 english
             # chars is a good target) because of issue6158.
             r = ui.promptchoice(b"%s\n(enter ? for help) %s" % (query, resps))
             ui.write(b"\n")
             if r == 8:  # ?
                 for c, t in ui.extractchoices(resps)[1]:
                     ui.write(b'%s - %s\n' % (c, encoding.lower(t)))
                 continue
             elif r == 0:  # yes
                 ret = True
             elif r == 1:  # no
                 ret = False
             elif r == 2:  # Edit patch
                 if chunk is None:
                     ui.write(_(b'cannot edit patch for whole file'))
                     ui.write(b"\n")
                     continue
                 if chunk.header.binary():
                     ui.write(_(b'cannot edit patch for binary file'))
                     ui.write(b"\n")
                     continue
                 # Patch comment based on the Git one (based on comment at end of
                 # https://mercurial-scm.org/wiki/RecordExtension)
                 phelp = b'---' + _(
                     """
 To remove '-' lines, make them ' ' lines (context).
 To remove '+' lines, delete them.
 Lines starting with # will be removed from the patch.
 
 If the patch applies cleanly, the edited hunk will immediately be
 added to the record list. If it does not apply cleanly, a rejects
 file will be generated: you can use that when you try again. If
 all lines of the hunk are removed, then the edit is aborted and
 the hunk is left unchanged.
 """
                 )
                 (patchfd, patchfn) = pycompat.mkstemp(
                     prefix=b"hg-editor-", suffix=b".diff"
                 )
                 ncpatchfp = None
                 try:
                     # Write the initial patch
                     f = util.nativeeolwriter(os.fdopen(patchfd, 'wb'))
                     chunk.header.write(f)
                     chunk.write(f)
                     f.write(
                         b''.join(
                             [b'# ' + i + b'\n' for i in phelp.splitlines()]
                         )
                     )
                     f.close()
                     # Start the editor and wait for it to complete
                     editor = ui.geteditor()
                     ret = ui.system(
                         b"%s \"%s\"" % (editor, patchfn),
                         environ={b'HGUSER': ui.username()},
                         blockedtag=b'filterpatch',
                     )
                     if ret != 0:
                         ui.warn(_(b"editor exited with exit code %d\n") % ret)
                         continue
                     # Remove comment lines
                     patchfp = open(patchfn, 'rb')
                     ncpatchfp = stringio()
                     for line in util.iterfile(patchfp):
                         line = util.fromnativeeol(line)
                         if not line.startswith(b'#'):
                             ncpatchfp.write(line)
                     patchfp.close()
                     ncpatchfp.seek(0)
                     newpatches = parsepatch(ncpatchfp)
                 finally:
                     os.unlink(patchfn)
                     del ncpatchfp
                 # Signal that the chunk shouldn't be applied as-is, but
                 # provide the new patch to be used instead.
                 ret = False
             elif r == 3:  # Skip
                 ret = skipfile = False
             elif r == 4:  # file (Record remaining)
                 ret = skipfile = True
             elif r == 5:  # done, skip remaining
                 ret = skipall = False
             elif r == 6:  # all
                 ret = skipall = True
             elif r == 7:  # quit
                 raise error.Abort(_(b'user quit'))
             return ret, skipfile, skipall, newpatches
 
     seen = set()
     applied = {}  # 'filename' -> [] of chunks
     skipfile, skipall = None, None
     pos, total = 1, sum(len(h.hunks) for h in headers)
     for h in headers:
         pos += len(h.hunks)
         skipfile = None
         fixoffset = 0
         hdr = b''.join(h.header)
         if hdr in seen:
             continue
         seen.add(hdr)
         if skipall is None:
             h.pretty(ui)
         files = h.files()
         msg = _(b'examine changes to %s?') % _(b' and ').join(
             b"'%s'" % f for f in files
         )
         if all(match.exact(f) for f in files):
             r, skipall, np = True, None, None
         else:
             r, skipfile, skipall, np = prompt(skipfile, skipall, msg, None)
         if not r:
             continue
         applied[h.filename()] = [h]
         if h.allhunks():
             applied[h.filename()] += h.hunks
             continue
         for i, chunk in enumerate(h.hunks):
             if skipfile is None and skipall is None:
                 chunk.pretty(ui)
             if total == 1:
                 msg = messages[b'single'][operation] % chunk.filename()
             else:
                 idx = pos - len(h.hunks) + i
                 msg = messages[b'multiple'][operation] % (
                     idx,
                     total,
                     chunk.filename(),
                 )
             r, skipfile, skipall, newpatches = prompt(
                 skipfile, skipall, msg, chunk
             )
             if r:
                 if fixoffset:
                     chunk = copy.copy(chunk)
                     chunk.toline += fixoffset
                 applied[chunk.filename()].append(chunk)
             elif newpatches is not None:
                 for newpatch in newpatches:
                     for newhunk in newpatch.hunks:
                         if fixoffset:
                             newhunk.toline += fixoffset
                         applied[newhunk.filename()].append(newhunk)
             else:
                 fixoffset += chunk.removed - chunk.added
     return (
         sum(
             [
                 h
                 for h in pycompat.itervalues(applied)
                 if h[0].special() or len(h) > 1
             ],
             [],
         ),
         {},
     )
 
 
 class hunk(object):
     def __init__(self, desc, num, lr, context):
         self.number = num
         self.desc = desc
         self.hunk = [desc]
         self.a = []
         self.b = []
         self.starta = self.lena = None
         self.startb = self.lenb = None
         if lr is not None:
             if context:
                 self.read_context_hunk(lr)
             else:
                 self.read_unified_hunk(lr)
 
     def getnormalized(self):
         """Return a copy with line endings normalized to LF."""
 
         def normalize(lines):
             nlines = []
             for line in lines:
                 if line.endswith(b'\r\n'):
                     line = line[:-2] + b'\n'
                 nlines.append(line)
             return nlines
 
         # Dummy object, it is rebuilt manually
         nh = hunk(self.desc, self.number, None, None)
         nh.number = self.number
         nh.desc = self.desc
         nh.hunk = self.hunk
         nh.a = normalize(self.a)
         nh.b = normalize(self.b)
         nh.starta = self.starta
         nh.startb = self.startb
         nh.lena = self.lena
         nh.lenb = self.lenb
         return nh
 
     def read_unified_hunk(self, lr):
         m = unidesc.match(self.desc)
         if not m:
             raise PatchError(_(b"bad hunk #%d") % self.number)
         self.starta, self.lena, self.startb, self.lenb = m.groups()
         if self.lena is None:
             self.lena = 1
         else:
             self.lena = int(self.lena)
         if self.lenb is None:
             self.lenb = 1
         else:
             self.lenb = int(self.lenb)
         self.starta = int(self.starta)
         self.startb = int(self.startb)
         try:
             diffhelper.addlines(
                 lr, self.hunk, self.lena, self.lenb, self.a, self.b
             )
         except error.ParseError as e:
             raise PatchError(_(b"bad hunk #%d: %s") % (self.number, e))
         # if we hit eof before finishing out the hunk, the last line will
         # be zero length.  Lets try to fix it up.
         while len(self.hunk[-1]) == 0:
             del self.hunk[-1]
             del self.a[-1]
             del self.b[-1]
             self.lena -= 1
             self.lenb -= 1
         self._fixnewline(lr)
 
     def read_context_hunk(self, lr):
         self.desc = lr.readline()
         m = contextdesc.match(self.desc)
         if not m:
             raise PatchError(_(b"bad hunk #%d") % self.number)
         self.starta, aend = m.groups()
         self.starta = int(self.starta)
         if aend is None:
             aend = self.starta
         self.lena = int(aend) - self.starta
         if self.starta:
             self.lena += 1
         for x in pycompat.xrange(self.lena):
             l = lr.readline()
             if l.startswith(b'---'):
                 # lines addition, old block is empty
                 lr.push(l)
                 break
             s = l[2:]
             if l.startswith(b'- ') or l.startswith(b'! '):
                 u = b'-' + s
             elif l.startswith(b'  '):
                 u = b' ' + s
             else:
                 raise PatchError(
                     _(b"bad hunk #%d old text line %d") % (self.number, x)
                 )
             self.a.append(u)
             self.hunk.append(u)
 
         l = lr.readline()
         if l.startswith(br'\ '):
             s = self.a[-1][:-1]
             self.a[-1] = s
             self.hunk[-1] = s
             l = lr.readline()
         m = contextdesc.match(l)
         if not m:
             raise PatchError(_(b"bad hunk #%d") % self.number)
         self.startb, bend = m.groups()
         self.startb = int(self.startb)
         if bend is None:
             bend = self.startb
         self.lenb = int(bend) - self.startb
         if self.startb:
             self.lenb += 1
         hunki = 1
         for x in pycompat.xrange(self.lenb):
             l = lr.readline()
             if l.startswith(br'\ '):
                 # XXX: the only way to hit this is with an invalid line range.
                 # The no-eol marker is not counted in the line range, but I
                 # guess there are diff(1) out there which behave differently.
                 s = self.b[-1][:-1]
                 self.b[-1] = s
                 self.hunk[hunki - 1] = s
                 continue
             if not l:
                 # line deletions, new block is empty and we hit EOF
                 lr.push(l)
                 break
             s = l[2:]
             if l.startswith(b'+ ') or l.startswith(b'! '):
                 u = b'+' + s
             elif l.startswith(b'  '):
                 u = b' ' + s
             elif len(self.b) == 0:
                 # line deletions, new block is empty
                 lr.push(l)
                 break
             else:
                 raise PatchError(
                     _(b"bad hunk #%d old text line %d") % (self.number, x)
                 )
             self.b.append(s)
             while True:
                 if hunki >= len(self.hunk):
                     h = b""
                 else:
                     h = self.hunk[hunki]
                 hunki += 1
                 if h == u:
                     break
                 elif h.startswith(b'-'):
                     continue
                 else:
                     self.hunk.insert(hunki - 1, u)
                     break
 
         if not self.a:
             # this happens when lines were only added to the hunk
             for x in self.hunk:
                 if x.startswith(b'-') or x.startswith(b' '):
                     self.a.append(x)
         if not self.b:
             # this happens when lines were only deleted from the hunk
             for x in self.hunk:
                 if x.startswith(b'+') or x.startswith(b' '):
                     self.b.append(x[1:])
         # @@ -start,len +start,len @@
         self.desc = b"@@ -%d,%d +%d,%d @@\n" % (
             self.starta,
             self.lena,
             self.startb,
             self.lenb,
         )
         self.hunk[0] = self.desc
         self._fixnewline(lr)
 
     def _fixnewline(self, lr):
         l = lr.readline()
         if l.startswith(br'\ '):
             diffhelper.fixnewline(self.hunk, self.a, self.b)
         else:
             lr.push(l)
 
     def complete(self):
         return len(self.a) == self.lena and len(self.b) == self.lenb
 
     def _fuzzit(self, old, new, fuzz, toponly):
         # this removes context lines from the top and bottom of list 'l'.  It
         # checks the hunk to make sure only context lines are removed, and then
         # returns a new shortened list of lines.
         fuzz = min(fuzz, len(old))
         if fuzz:
             top = 0
             bot = 0
             hlen = len(self.hunk)
             for x in pycompat.xrange(hlen - 1):
                 # the hunk starts with the @@ line, so use x+1
                 if self.hunk[x + 1].startswith(b' '):
                     top += 1
                 else:
                     break
             if not toponly:
                 for x in pycompat.xrange(hlen - 1):
                     if self.hunk[hlen - bot - 1].startswith(b' '):
                         bot += 1
                     else:
                         break
 
             bot = min(fuzz, bot)
             top = min(fuzz, top)
             return old[top : len(old) - bot], new[top : len(new) - bot], top
         return old, new, 0
 
     def fuzzit(self, fuzz, toponly):
         old, new, top = self._fuzzit(self.a, self.b, fuzz, toponly)
         oldstart = self.starta + top
         newstart = self.startb + top
         # zero length hunk ranges already have their start decremented
         if self.lena and oldstart > 0:
             oldstart -= 1
         if self.lenb and newstart > 0:
             newstart -= 1
         return old, oldstart, new, newstart
 
 
 class binhunk(object):
     """A binary patch file."""
 
     def __init__(self, lr, fname):
         self.text = None
         self.delta = False
         self.hunk = [b'GIT binary patch\n']
         self._fname = fname
         self._read(lr)
 
     def complete(self):
         return self.text is not None
 
     def new(self, lines):
         if self.delta:
             return [applybindelta(self.text, b''.join(lines))]
         return [self.text]
 
     def _read(self, lr):
         def getline(lr, hunk):
             l = lr.readline()
             hunk.append(l)
             return l.rstrip(b'\r\n')
 
         while True:
             line = getline(lr, self.hunk)
             if not line:
                 raise PatchError(
                     _(b'could not extract "%s" binary data') % self._fname
                 )
             if line.startswith(b'literal '):
                 size = int(line[8:].rstrip())
                 break
             if line.startswith(b'delta '):
                 size = int(line[6:].rstrip())
                 self.delta = True
                 break
         dec = []
         line = getline(lr, self.hunk)
         while len(line) > 1:
             l = line[0:1]
             if l <= b'Z' and l >= b'A':
                 l = ord(l) - ord(b'A') + 1
             else:
                 l = ord(l) - ord(b'a') + 27
             try:
                 dec.append(util.b85decode(line[1:])[:l])
             except ValueError as e:
                 raise PatchError(
                     _(b'could not decode "%s" binary patch: %s')
                     % (self._fname, stringutil.forcebytestr(e))
                 )
             line = getline(lr, self.hunk)
         text = zlib.decompress(b''.join(dec))
         if len(text) != size:
             raise PatchError(
                 _(b'"%s" length is %d bytes, should be %d')
                 % (self._fname, len(text), size)
             )
         self.text = text
 
 
 def parsefilename(str):
     # --- filename \t|space stuff
     s = str[4:].rstrip(b'\r\n')
     i = s.find(b'\t')
     if i < 0:
         i = s.find(b' ')
         if i < 0:
             return s
     return s[:i]
 
 
 def reversehunks(hunks):
     '''reverse the signs in the hunks given as argument
 
     This function operates on hunks coming out of patch.filterpatch, that is
     a list of the form: [header1, hunk1, hunk2, header2...]. Example usage:
 
     >>> rawpatch = b"""diff --git a/folder1/g b/folder1/g
     ... --- a/folder1/g
     ... +++ b/folder1/g
     ... @@ -1,7 +1,7 @@
     ... +firstline
     ...  c
     ...  1
     ...  2
     ... + 3
     ... -4
     ...  5
     ...  d
     ... +lastline"""
     >>> hunks = parsepatch([rawpatch])
     >>> hunkscomingfromfilterpatch = []
     >>> for h in hunks:
     ...     hunkscomingfromfilterpatch.append(h)
     ...     hunkscomingfromfilterpatch.extend(h.hunks)
 
     >>> reversedhunks = reversehunks(hunkscomingfromfilterpatch)
     >>> from . import util
     >>> fp = util.stringio()
     >>> for c in reversedhunks:
     ...      c.write(fp)
     >>> fp.seek(0) or None
     >>> reversedpatch = fp.read()
     >>> print(pycompat.sysstr(reversedpatch))
     diff --git a/folder1/g b/folder1/g
     --- a/folder1/g
     +++ b/folder1/g
     @@ -1,4 +1,3 @@
     -firstline
      c
      1
      2
     @@ -2,6 +1,6 @@
      c
      1
      2
     - 3
     +4
      5
      d
     @@ -6,3 +5,2 @@
      5
      d
     -lastline
 
     '''
 
     newhunks = []
     for c in hunks:
         if util.safehasattr(c, b'reversehunk'):
             c = c.reversehunk()
         newhunks.append(c)
     return newhunks
 
 
 def parsepatch(originalchunks, maxcontext=None):
     """patch -> [] of headers -> [] of hunks
 
     If maxcontext is not None, trim context lines if necessary.
 
     >>> rawpatch = b'''diff --git a/folder1/g b/folder1/g
     ... --- a/folder1/g
     ... +++ b/folder1/g
     ... @@ -1,8 +1,10 @@
     ...  1
     ...  2
     ... -3
     ...  4
     ...  5
     ...  6
     ... +6.1
     ... +6.2
     ...  7
     ...  8
     ... +9'''
     >>> out = util.stringio()
     >>> headers = parsepatch([rawpatch], maxcontext=1)
     >>> for header in headers:
     ...     header.write(out)
     ...     for hunk in header.hunks:
     ...         hunk.write(out)
     >>> print(pycompat.sysstr(out.getvalue()))
     diff --git a/folder1/g b/folder1/g
     --- a/folder1/g
     +++ b/folder1/g
     @@ -2,3 +2,2 @@
      2
     -3
      4
     @@ -6,2 +5,4 @@
      6
     +6.1
     +6.2
      7
     @@ -8,1 +9,2 @@
      8
     +9
     """
 
     class parser(object):
         """patch parsing state machine"""
 
         def __init__(self):
             self.fromline = 0
             self.toline = 0
             self.proc = b''
             self.header = None
             self.context = []
             self.before = []
             self.hunk = []
             self.headers = []
 
         def addrange(self, limits):
             self.addcontext([])
             fromstart, fromend, tostart, toend, proc = limits
             self.fromline = int(fromstart)
             self.toline = int(tostart)
             self.proc = proc
 
         def addcontext(self, context):
             if self.hunk:
                 h = recordhunk(
                     self.header,
                     self.fromline,
                     self.toline,
                     self.proc,
                     self.before,
                     self.hunk,
                     context,
                     maxcontext,
                 )
                 self.header.hunks.append(h)
                 self.fromline += len(self.before) + h.removed
                 self.toline += len(self.before) + h.added
                 self.before = []
                 self.hunk = []
             self.context = context
 
         def addhunk(self, hunk):
             if self.context:
                 self.before = self.context
                 self.context = []
             if self.hunk:
                 self.addcontext([])
             self.hunk = hunk
 
         def newfile(self, hdr):
             self.addcontext([])
             h = header(hdr)
             self.headers.append(h)
             self.header = h
 
         def addother(self, line):
             pass  # 'other' lines are ignored
 
         def finished(self):
             self.addcontext([])
             return self.headers
 
         transitions = {
             b'file': {
                 b'context': addcontext,
                 b'file': newfile,
                 b'hunk': addhunk,
                 b'range': addrange,
             },
             b'context': {
                 b'file': newfile,
                 b'hunk': addhunk,
                 b'range': addrange,
                 b'other': addother,
             },
             b'hunk': {
                 b'context': addcontext,
                 b'file': newfile,
                 b'range': addrange,
             },
             b'range': {b'context': addcontext, b'hunk': addhunk},
             b'other': {b'other': addother},
         }
 
     p = parser()
     fp = stringio()
     fp.write(b''.join(originalchunks))
     fp.seek(0)
 
     state = b'context'
     for newstate, data in scanpatch(fp):
         try:
             p.transitions[state][newstate](p, data)
         except KeyError:
             raise PatchError(
                 b'unhandled transition: %s -> %s' % (state, newstate)
             )
         state = newstate
     del fp
     return p.finished()
 
 
 def pathtransform(path, strip, prefix):
     '''turn a path from a patch into a path suitable for the repository
 
     prefix, if not empty, is expected to be normalized with a / at the end.
 
     Returns (stripped components, path in repository).
 
     >>> pathtransform(b'a/b/c', 0, b'')
     ('', 'a/b/c')
     >>> pathtransform(b'   a/b/c   ', 0, b'')
     ('', '   a/b/c')
     >>> pathtransform(b'   a/b/c   ', 2, b'')
     ('a/b/', 'c')
     >>> pathtransform(b'a/b/c', 0, b'd/e/')
     ('', 'd/e/a/b/c')
     >>> pathtransform(b'   a//b/c   ', 2, b'd/e/')
     ('a//b/', 'd/e/c')
     >>> pathtransform(b'a/b/c', 3, b'')
     Traceback (most recent call last):
     PatchError: unable to strip away 1 of 3 dirs from a/b/c
     '''
     pathlen = len(path)
     i = 0
     if strip == 0:
         return b'', prefix + path.rstrip()
     count = strip
     while count > 0:
         i = path.find(b'/', i)
         if i == -1:
             raise PatchError(
                 _(b"unable to strip away %d of %d dirs from %s")
                 % (count, strip, path)
             )
         i += 1
         # consume '//' in the path
         while i < pathlen - 1 and path[i : i + 1] == b'/':
             i += 1
         count -= 1
     return path[:i].lstrip(), prefix + path[i:].rstrip()
 
 
 def makepatchmeta(backend, afile_orig, bfile_orig, hunk, strip, prefix):
     nulla = afile_orig == b"/dev/null"
     nullb = bfile_orig == b"/dev/null"
     create = nulla and hunk.starta == 0 and hunk.lena == 0
     remove = nullb and hunk.startb == 0 and hunk.lenb == 0
     abase, afile = pathtransform(afile_orig, strip, prefix)
     gooda = not nulla and backend.exists(afile)
     bbase, bfile = pathtransform(bfile_orig, strip, prefix)
     if afile == bfile:
         goodb = gooda
     else:
         goodb = not nullb and backend.exists(bfile)
     missing = not goodb and not gooda and not create
 
     # some diff programs apparently produce patches where the afile is
     # not /dev/null, but afile starts with bfile
     abasedir = afile[: afile.rfind(b'/') + 1]
     bbasedir = bfile[: bfile.rfind(b'/') + 1]
     if (
         missing
         and abasedir == bbasedir
         and afile.startswith(bfile)
         and hunk.starta == 0
         and hunk.lena == 0
     ):
         create = True
         missing = False
 
     # If afile is "a/b/foo" and bfile is "a/b/foo.orig" we assume the
     # diff is between a file and its backup. In this case, the original
     # file should be patched (see original mpatch code).
     isbackup = abase == bbase and bfile.startswith(afile)
     fname = None
     if not missing:
         if gooda and goodb:
             if isbackup:
                 fname = afile
             else:
                 fname = bfile
         elif gooda:
             fname = afile
 
     if not fname:
         if not nullb:
             if isbackup:
                 fname = afile
             else:
                 fname = bfile
         elif not nulla:
             fname = afile
         else:
             raise PatchError(_(b"undefined source and destination files"))
 
     gp = patchmeta(fname)
     if create:
         gp.op = b'ADD'
     elif remove:
         gp.op = b'DELETE'
     return gp
 
 
 def scanpatch(fp):
     """like patch.iterhunks, but yield different events
 
     - ('file',    [header_lines + fromfile + tofile])
     - ('context', [context_lines])
     - ('hunk',    [hunk_lines])
     - ('range',   (-start,len, +start,len, proc))
     """
     lines_re = re.compile(br'@@ -(\d+),(\d+) \+(\d+),(\d+) @@\s*(.*)')
     lr = linereader(fp)
 
     def scanwhile(first, p):
         """scan lr while predicate holds"""
         lines = [first]
         for line in iter(lr.readline, b''):
             if p(line):
                 lines.append(line)
             else:
                 lr.push(line)
                 break
         return lines
 
     for line in iter(lr.readline, b''):
         if line.startswith(b'diff --git a/') or line.startswith(b'diff -r '):
 
             def notheader(line):
                 s = line.split(None, 1)
                 return not s or s[0] not in (b'---', b'diff')
 
             header = scanwhile(line, notheader)
             fromfile = lr.readline()
             if fromfile.startswith(b'---'):
                 tofile = lr.readline()
                 header += [fromfile, tofile]
             else:
                 lr.push(fromfile)
             yield b'file', header
         elif line.startswith(b' '):
             cs = (b' ', b'\\')
             yield b'context', scanwhile(line, lambda l: l.startswith(cs))
         elif line.startswith((b'-', b'+')):
             cs = (b'-', b'+', b'\\')
             yield b'hunk', scanwhile(line, lambda l: l.startswith(cs))
         else:
             m = lines_re.match(line)
             if m:
                 yield b'range', m.groups()
             else:
                 yield b'other', line
 
 
 def scangitpatch(lr, firstline):
     """
     Git patches can emit:
     - rename a to b
     - change b
     - copy a to c
     - change c
 
     We cannot apply this sequence as-is, the renamed 'a' could not be
     found for it would have been renamed already. And we cannot copy
     from 'b' instead because 'b' would have been changed already. So
     we scan the git patch for copy and rename commands so we can
     perform the copies ahead of time.
     """
     pos = 0
     try:
         pos = lr.fp.tell()
         fp = lr.fp
     except IOError:
         fp = stringio(lr.fp.read())
     gitlr = linereader(fp)
     gitlr.push(firstline)
     gitpatches = readgitpatch(gitlr)
     fp.seek(pos)
     return gitpatches
 
 
 def iterhunks(fp):
     """Read a patch and yield the following events:
     - ("file", afile, bfile, firsthunk): select a new target file.
     - ("hunk", hunk): a new hunk is ready to be applied, follows a
     "file" event.
     - ("git", gitchanges): current diff is in git format, gitchanges
     maps filenames to gitpatch records. Unique event.
     """
     afile = b""
     bfile = b""
     state = None
     hunknum = 0
     emitfile = newfile = False
     gitpatches = None
 
     # our states
     BFILE = 1
     context = None
     lr = linereader(fp)
 
     for x in iter(lr.readline, b''):
         if state == BFILE and (
             (not context and x.startswith(b'@'))
             or (context is not False and x.startswith(b'***************'))
             or x.startswith(b'GIT binary patch')
         ):
             gp = None
             if gitpatches and gitpatches[-1].ispatching(afile, bfile):
                 gp = gitpatches.pop()
             if x.startswith(b'GIT binary patch'):
                 h = binhunk(lr, gp.path)
             else:
                 if context is None and x.startswith(b'***************'):
                     context = True
                 h = hunk(x, hunknum + 1, lr, context)
             hunknum += 1
             if emitfile:
                 emitfile = False
                 yield b'file', (afile, bfile, h, gp and gp.copy() or None)
             yield b'hunk', h
         elif x.startswith(b'diff --git a/'):
             m = gitre.match(x.rstrip(b' \r\n'))
             if not m:
                 continue
             if gitpatches is None:
                 # scan whole input for git metadata
                 gitpatches = scangitpatch(lr, x)
                 yield b'git', [
                     g.copy() for g in gitpatches if g.op in (b'COPY', b'RENAME')
                 ]
                 gitpatches.reverse()
             afile = b'a/' + m.group(1)
             bfile = b'b/' + m.group(2)
             while gitpatches and not gitpatches[-1].ispatching(afile, bfile):
                 gp = gitpatches.pop()
                 yield b'file', (
                     b'a/' + gp.path,
                     b'b/' + gp.path,
                     None,
                     gp.copy(),
                 )
             if not gitpatches:
                 raise PatchError(
                     _(b'failed to synchronize metadata for "%s"') % afile[2:]
                 )
             newfile = True
         elif x.startswith(b'---'):
             # check for a unified diff
             l2 = lr.readline()
             if not l2.startswith(b'+++'):
                 lr.push(l2)
                 continue
             newfile = True
             context = False
             afile = parsefilename(x)
             bfile = parsefilename(l2)
         elif x.startswith(b'***'):
             # check for a context diff
             l2 = lr.readline()
             if not l2.startswith(b'---'):
                 lr.push(l2)
                 continue
             l3 = lr.readline()
             lr.push(l3)
             if not l3.startswith(b"***************"):
                 lr.push(l2)
                 continue
             newfile = True
             context = True
             afile = parsefilename(x)
             bfile = parsefilename(l2)
 
         if newfile:
             newfile = False
             emitfile = True
             state = BFILE
             hunknum = 0
 
     while gitpatches:
         gp = gitpatches.pop()
         yield b'file', (b'a/' + gp.path, b'b/' + gp.path, None, gp.copy())
 
 
 def applybindelta(binchunk, data):
     """Apply a binary delta hunk
     The algorithm used is the algorithm from git's patch-delta.c
     """
 
     def deltahead(binchunk):
         i = 0
         for c in pycompat.bytestr(binchunk):
             i += 1
             if not (ord(c) & 0x80):
                 return i
         return i
 
     out = b""
     s = deltahead(binchunk)
     binchunk = binchunk[s:]
     s = deltahead(binchunk)
     binchunk = binchunk[s:]
     i = 0
     while i < len(binchunk):
         cmd = ord(binchunk[i : i + 1])
         i += 1
         if cmd & 0x80:
             offset = 0
             size = 0
             if cmd & 0x01:
                 offset = ord(binchunk[i : i + 1])
                 i += 1
             if cmd & 0x02:
                 offset |= ord(binchunk[i : i + 1]) << 8
                 i += 1
             if cmd & 0x04:
                 offset |= ord(binchunk[i : i + 1]) << 16
                 i += 1
             if cmd & 0x08:
                 offset |= ord(binchunk[i : i + 1]) << 24
                 i += 1
             if cmd & 0x10:
                 size = ord(binchunk[i : i + 1])
                 i += 1
             if cmd & 0x20:
                 size |= ord(binchunk[i : i + 1]) << 8
                 i += 1
             if cmd & 0x40:
                 size |= ord(binchunk[i : i + 1]) << 16
                 i += 1
             if size == 0:
                 size = 0x10000
             offset_end = offset + size
             out += data[offset:offset_end]
         elif cmd != 0:
             offset_end = i + cmd
             out += binchunk[i:offset_end]
             i += cmd
         else:
             raise PatchError(_(b'unexpected delta opcode 0'))
     return out
 
 
 def applydiff(ui, fp, backend, store, strip=1, prefix=b'', eolmode=b'strict'):
     """Reads a patch from fp and tries to apply it.
 
     Returns 0 for a clean patch, -1 if any rejects were found and 1 if
     there was any fuzz.
 
     If 'eolmode' is 'strict', the patch content and patched file are
     read in binary mode. Otherwise, line endings are ignored when
     patching then normalized according to 'eolmode'.
     """
     return _applydiff(
         ui,
         fp,
         patchfile,
         backend,
         store,
         strip=strip,
         prefix=prefix,
         eolmode=eolmode,
     )
 
 
 def _canonprefix(repo, prefix):
     if prefix:
         prefix = pathutil.canonpath(repo.root, repo.getcwd(), prefix)
         if prefix != b'':
             prefix += b'/'
     return prefix
 
 
 def _applydiff(
     ui, fp, patcher, backend, store, strip=1, prefix=b'', eolmode=b'strict'
 ):
     prefix = _canonprefix(backend.repo, prefix)
 
     def pstrip(p):
         return pathtransform(p, strip - 1, prefix)[1]
 
     rejects = 0
     err = 0
     current_file = None
 
     for state, values in iterhunks(fp):
         if state == b'hunk':
             if not current_file:
                 continue
             ret = current_file.apply(values)
             if ret > 0:
                 err = 1
         elif state == b'file':
             if current_file:
                 rejects += current_file.close()
                 current_file = None
             afile, bfile, first_hunk, gp = values
             if gp:
                 gp.path = pstrip(gp.path)
                 if gp.oldpath:
                     gp.oldpath = pstrip(gp.oldpath)
             else:
                 gp = makepatchmeta(
                     backend, afile, bfile, first_hunk, strip, prefix
                 )
             if gp.op == b'RENAME':
                 backend.unlink(gp.oldpath)
             if not first_hunk:
                 if gp.op == b'DELETE':
                     backend.unlink(gp.path)
                     continue
                 data, mode = None, None
                 if gp.op in (b'RENAME', b'COPY'):
                     data, mode = store.getfile(gp.oldpath)[:2]
                     if data is None:
                         # This means that the old path does not exist
                         raise PatchError(
                             _(b"source file '%s' does not exist") % gp.oldpath
                         )
                 if gp.mode:
                     mode = gp.mode
                     if gp.op == b'ADD':
                         # Added files without content have no hunk and
                         # must be created
                         data = b''
                 if data or mode:
                     if gp.op in (b'ADD', b'RENAME', b'COPY') and backend.exists(
                         gp.path
                     ):
                         raise PatchError(
                             _(
                                 b"cannot create %s: destination "
                                 b"already exists"
                             )
                             % gp.path
                         )
                     backend.setfile(gp.path, data, mode, gp.oldpath)
                 continue
             try:
                 current_file = patcher(ui, gp, backend, store, eolmode=eolmode)
             except PatchError as inst:
                 ui.warn(stringutil.forcebytestr(inst) + b'\n')
                 current_file = None
                 rejects += 1
                 continue
         elif state == b'git':
             for gp in values:
                 path = pstrip(gp.oldpath)
                 data, mode = backend.getfile(path)
                 if data is None:
                     # The error ignored here will trigger a getfile()
                     # error in a place more appropriate for error
                     # handling, and will not interrupt the patching
                     # process.
                     pass
                 else:
                     store.setfile(path, data, mode)
         else:
             raise error.Abort(_(b'unsupported parser state: %s') % state)
 
     if current_file:
         rejects += current_file.close()
 
     if rejects:
         return -1
     return err
 
 
 def _externalpatch(ui, repo, patcher, patchname, strip, files, similarity):
     """use <patcher> to apply <patchname> to the working directory.
     returns whether patch was applied with fuzz factor."""
 
     fuzz = False
     args = []
     cwd = repo.root
     if cwd:
         args.append(b'-d %s' % procutil.shellquote(cwd))
     cmd = b'%s %s -p%d < %s' % (
         patcher,
         b' '.join(args),
         strip,
         procutil.shellquote(patchname),
     )
     ui.debug(b'Using external patch tool: %s\n' % cmd)
     fp = procutil.popen(cmd, b'rb')
     try:
         for line in util.iterfile(fp):
             line = line.rstrip()
             ui.note(line + b'\n')
             if line.startswith(b'patching file '):
                 pf = util.parsepatchoutput(line)
                 printed_file = False
                 files.add(pf)
             elif line.find(b'with fuzz') >= 0:
                 fuzz = True
                 if not printed_file:
                     ui.warn(pf + b'\n')
                     printed_file = True
                 ui.warn(line + b'\n')
             elif line.find(b'saving rejects to file') >= 0:
                 ui.warn(line + b'\n')
             elif line.find(b'FAILED') >= 0:
                 if not printed_file:
                     ui.warn(pf + b'\n')
                     printed_file = True
                 ui.warn(line + b'\n')
     finally:
         if files:
             scmutil.marktouched(repo, files, similarity)
     code = fp.close()
     if code:
         raise PatchError(
             _(b"patch command failed: %s") % procutil.explainexit(code)
         )
     return fuzz
 
 
 def patchbackend(
     ui, backend, patchobj, strip, prefix, files=None, eolmode=b'strict'
 ):
     if files is None:
         files = set()
     if eolmode is None:
         eolmode = ui.config(b'patch', b'eol')
     if eolmode.lower() not in eolmodes:
         raise error.Abort(_(b'unsupported line endings type: %s') % eolmode)
     eolmode = eolmode.lower()
 
     store = filestore()
     try:
         fp = open(patchobj, b'rb')
     except TypeError:
         fp = patchobj
     try:
         ret = applydiff(
             ui, fp, backend, store, strip=strip, prefix=prefix, eolmode=eolmode
         )
     finally:
         if fp != patchobj:
             fp.close()
         files.update(backend.close())
         store.close()
     if ret < 0:
         raise PatchError(_(b'patch failed to apply'))
     return ret > 0
 
 
 def internalpatch(
     ui,
     repo,
     patchobj,
     strip,
     prefix=b'',
     files=None,
     eolmode=b'strict',
     similarity=0,
 ):
     """use builtin patch to apply <patchobj> to the working directory.
     returns whether patch was applied with fuzz factor."""
     backend = workingbackend(ui, repo, similarity)
     return patchbackend(ui, backend, patchobj, strip, prefix, files, eolmode)
 
 
 def patchrepo(
     ui, repo, ctx, store, patchobj, strip, prefix, files=None, eolmode=b'strict'
 ):
     backend = repobackend(ui, repo, ctx, store)
     return patchbackend(ui, backend, patchobj, strip, prefix, files, eolmode)
 
 
 def patch(
     ui,
     repo,
     patchname,
     strip=1,
     prefix=b'',
     files=None,
     eolmode=b'strict',
     similarity=0,
 ):
     """Apply <patchname> to the working directory.
 
     'eolmode' specifies how end of lines should be handled. It can be:
     - 'strict': inputs are read in binary mode, EOLs are preserved
     - 'crlf': EOLs are ignored when patching and reset to CRLF
     - 'lf': EOLs are ignored when patching and reset to LF
     - None: get it from user settings, default to 'strict'
     'eolmode' is ignored when using an external patcher program.
 
     Returns whether patch was applied with fuzz factor.
     """
     patcher = ui.config(b'ui', b'patch')
     if files is None:
         files = set()
     if patcher:
         return _externalpatch(
             ui, repo, patcher, patchname, strip, files, similarity
         )
     return internalpatch(
         ui, repo, patchname, strip, prefix, files, eolmode, similarity
     )
 
 
 def changedfiles(ui, repo, patchpath, strip=1, prefix=b''):
     backend = fsbackend(ui, repo.root)
     prefix = _canonprefix(repo, prefix)
     with open(patchpath, b'rb') as fp:
         changed = set()
         for state, values in iterhunks(fp):
             if state == b'file':
                 afile, bfile, first_hunk, gp = values
                 if gp:
                     gp.path = pathtransform(gp.path, strip - 1, prefix)[1]
                     if gp.oldpath:
                         gp.oldpath = pathtransform(
                             gp.oldpath, strip - 1, prefix
                         )[1]
                 else:
                     gp = makepatchmeta(
                         backend, afile, bfile, first_hunk, strip, prefix
                     )
                 changed.add(gp.path)
                 if gp.op == b'RENAME':
                     changed.add(gp.oldpath)
             elif state not in (b'hunk', b'git'):
                 raise error.Abort(_(b'unsupported parser state: %s') % state)
         return changed
 
 
 class GitDiffRequired(Exception):
     pass
 
 
 diffopts = diffutil.diffallopts
 diffallopts = diffutil.diffallopts
 difffeatureopts = diffutil.difffeatureopts
 
 
 def diff(
     repo,
     node1=None,
     node2=None,
     match=None,
     changes=None,
     opts=None,
     losedatafn=None,
     pathfn=None,
     copy=None,
     copysourcematch=None,
     hunksfilterfn=None,
 ):
     '''yields diff of changes to files between two nodes, or node and
     working directory.
 
     if node1 is None, use first dirstate parent instead.
     if node2 is None, compare node1 with working directory.
 
     losedatafn(**kwarg) is a callable run when opts.upgrade=True and
     every time some change cannot be represented with the current
     patch format. Return False to upgrade to git patch format, True to
     accept the loss or raise an exception to abort the diff. It is
     called with the name of current file being diffed as 'fn'. If set
     to None, patches will always be upgraded to git format when
     necessary.
 
     prefix is a filename prefix that is prepended to all filenames on
     display (used for subrepos).
 
     relroot, if not empty, must be normalized with a trailing /. Any match
     patterns that fall outside it will be ignored.
 
     copy, if not empty, should contain mappings {dst@y: src@x} of copy
     information.
 
     if copysourcematch is not None, then copy sources will be filtered by this
     matcher
 
     hunksfilterfn, if not None, should be a function taking a filectx and
     hunks generator that may yield filtered hunks.
     '''
     if not node1 and not node2:
         node1 = repo.dirstate.p1()
 
     ctx1 = repo[node1]
     ctx2 = repo[node2]
 
     for fctx1, fctx2, hdr, hunks in diffhunks(
         repo,
         ctx1=ctx1,
         ctx2=ctx2,
         match=match,
         changes=changes,
         opts=opts,
         losedatafn=losedatafn,
         pathfn=pathfn,
         copy=copy,
         copysourcematch=copysourcematch,
     ):
         if hunksfilterfn is not None:
             # If the file has been removed, fctx2 is None; but this should
             # not occur here since we catch removed files early in
             # logcmdutil.getlinerangerevs() for 'hg log -L'.
             assert (
                 fctx2 is not None
             ), b'fctx2 unexpectly None in diff hunks filtering'
             hunks = hunksfilterfn(fctx2, hunks)
         text = b''.join(sum((list(hlines) for hrange, hlines in hunks), []))
         if hdr and (text or len(hdr) > 1):
             yield b'\n'.join(hdr) + b'\n'
         if text:
             yield text
 
 
 def diffhunks(
     repo,
     ctx1,
     ctx2,
     match=None,
     changes=None,
     opts=None,
     losedatafn=None,
     pathfn=None,
     copy=None,
     copysourcematch=None,
 ):
     """Yield diff of changes to files in the form of (`header`, `hunks`) tuples
     where `header` is a list of diff headers and `hunks` is an iterable of
     (`hunkrange`, `hunklines`) tuples.
 
     See diff() for the meaning of parameters.
     """
 
     if opts is None:
         opts = mdiff.defaultopts
 
     def lrugetfilectx():
         cache = {}
         order = collections.deque()
 
         def getfilectx(f, ctx):
             fctx = ctx.filectx(f, filelog=cache.get(f))
             if f not in cache:
                 if len(cache) > 20:
                     del cache[order.popleft()]
                 cache[f] = fctx.filelog()
             else:
                 order.remove(f)
             order.append(f)
             return fctx
 
         return getfilectx
 
     getfilectx = lrugetfilectx()
 
     if not changes:
         changes = ctx1.status(ctx2, match=match)
     if isinstance(changes, list):
         modified, added, removed = changes[:3]
     else:
         modified, added, removed = (
             changes.modified,
             changes.added,
             changes.removed,
         )
 
     if not modified and not added and not removed:
         return []
 
     if repo.ui.debugflag:
         hexfunc = hex
     else:
         hexfunc = short
     revs = [hexfunc(node) for node in [ctx1.node(), ctx2.node()] if node]
 
     if copy is None:
         copy = {}
         if opts.git or opts.upgrade:
             copy = copies.pathcopies(ctx1, ctx2, match=match)
 
     if copysourcematch:
         # filter out copies where source side isn't inside the matcher
         # (copies.pathcopies() already filtered out the destination)
         copy = {
             dst: src
             for dst, src in pycompat.iteritems(copy)
             if copysourcematch(src)
         }
 
     modifiedset = set(modified)
     addedset = set(added)
     removedset = set(removed)
     for f in modified:
         if f not in ctx1:
             # Fix up added, since merged-in additions appear as
             # modifications during merges
             modifiedset.remove(f)
             addedset.add(f)
     for f in removed:
         if f not in ctx1:
             # Merged-in additions that are then removed are reported as removed.
             # They are not in ctx1, so We don't want to show them in the diff.
             removedset.remove(f)
     modified = sorted(modifiedset)
     added = sorted(addedset)
     removed = sorted(removedset)
     for dst, src in list(copy.items()):
         if src not in ctx1:
             # Files merged in during a merge and then copied/renamed are
             # reported as copies. We want to show them in the diff as additions.
             del copy[dst]
 
     prefetchmatch = scmutil.matchfiles(
         repo, list(modifiedset | addedset | removedset)
     )
     scmutil.prefetchfiles(repo, [ctx1.rev(), ctx2.rev()], prefetchmatch)
 
     def difffn(opts, losedata):
         return trydiff(
             repo,
             revs,
             ctx1,
             ctx2,
             modified,
             added,
             removed,
             copy,
             getfilectx,
             opts,
             losedata,
             pathfn,
         )
 
     if opts.upgrade and not opts.git:
         try:
 
             def losedata(fn):
                 if not losedatafn or not losedatafn(fn=fn):
                     raise GitDiffRequired
 
             # Buffer the whole output until we are sure it can be generated
             return list(difffn(opts.copy(git=False), losedata))
         except GitDiffRequired:
             return difffn(opts.copy(git=True), None)
     else:
         return difffn(opts, None)
 
 
 def diffsinglehunk(hunklines):
     """yield tokens for a list of lines in a single hunk"""
     for line in hunklines:
         # chomp
         chompline = line.rstrip(b'\r\n')
         # highlight tabs and trailing whitespace
         stripline = chompline.rstrip()
         if line.startswith(b'-'):
             label = b'diff.deleted'
         elif line.startswith(b'+'):
             label = b'diff.inserted'
         else:
             raise error.ProgrammingError(b'unexpected hunk line: %s' % line)
         for token in tabsplitter.findall(stripline):
             if token.startswith(b'\t'):
                 yield (token, b'diff.tab')
             else:
                 yield (token, label)
 
         if chompline != stripline:
             yield (chompline[len(stripline) :], b'diff.trailingwhitespace')
         if chompline != line:
             yield (line[len(chompline) :], b'')
 
 
 def diffsinglehunkinline(hunklines):
     """yield tokens for a list of lines in a single hunk, with inline colors"""
     # prepare deleted, and inserted content
     a = b''
     b = b''
     for line in hunklines:
         if line[0:1] == b'-':
             a += line[1:]
         elif line[0:1] == b'+':
             b += line[1:]
         else:
             raise error.ProgrammingError(b'unexpected hunk line: %s' % line)
     # fast path: if either side is empty, use diffsinglehunk
     if not a or not b:
         for t in diffsinglehunk(hunklines):
             yield t
         return
     # re-split the content into words
     al = wordsplitter.findall(a)
     bl = wordsplitter.findall(b)
     # re-arrange the words to lines since the diff algorithm is line-based
     aln = [s if s == b'\n' else s + b'\n' for s in al]
     bln = [s if s == b'\n' else s + b'\n' for s in bl]
     an = b''.join(aln)
     bn = b''.join(bln)
     # run the diff algorithm, prepare atokens and btokens
     atokens = []
     btokens = []
     blocks = mdiff.allblocks(an, bn, lines1=aln, lines2=bln)
     for (a1, a2, b1, b2), btype in blocks:
         changed = btype == b'!'
         for token in mdiff.splitnewlines(b''.join(al[a1:a2])):
             atokens.append((changed, token))
         for token in mdiff.splitnewlines(b''.join(bl[b1:b2])):
             btokens.append((changed, token))
 
     # yield deleted tokens, then inserted ones
     for prefix, label, tokens in [
         (b'-', b'diff.deleted', atokens),
         (b'+', b'diff.inserted', btokens),
     ]:
         nextisnewline = True
         for changed, token in tokens:
             if nextisnewline:
                 yield (prefix, label)
                 nextisnewline = False
             # special handling line end
             isendofline = token.endswith(b'\n')
             if isendofline:
                 chomp = token[:-1]  # chomp
                 if chomp.endswith(b'\r'):
                     chomp = chomp[:-1]
                 endofline = token[len(chomp) :]
                 token = chomp.rstrip()  # detect spaces at the end
                 endspaces = chomp[len(token) :]
             # scan tabs
             for maybetab in tabsplitter.findall(token):
                 if b'\t' == maybetab[0:1]:
                     currentlabel = b'diff.tab'
                 else:
                     if changed:
                         currentlabel = label + b'.changed'
                     else:
                         currentlabel = label + b'.unchanged'
                 yield (maybetab, currentlabel)
             if isendofline:
                 if endspaces:
                     yield (endspaces, b'diff.trailingwhitespace')
                 yield (endofline, b'')
                 nextisnewline = True
 
 
 def difflabel(func, *args, **kw):
     '''yields 2-tuples of (output, label) based on the output of func()'''
     if kw.get('opts') and kw['opts'].worddiff:
         dodiffhunk = diffsinglehunkinline
     else:
         dodiffhunk = diffsinglehunk
     headprefixes = [
         (b'diff', b'diff.diffline'),
         (b'copy', b'diff.extended'),
         (b'rename', b'diff.extended'),
         (b'old', b'diff.extended'),
         (b'new', b'diff.extended'),
         (b'deleted', b'diff.extended'),
         (b'index', b'diff.extended'),
         (b'similarity', b'diff.extended'),
         (b'---', b'diff.file_a'),
         (b'+++', b'diff.file_b'),
     ]
     textprefixes = [
         (b'@', b'diff.hunk'),
         # - and + are handled by diffsinglehunk
     ]
     head = False
 
     # buffers a hunk, i.e. adjacent "-", "+" lines without other changes.
     hunkbuffer = []
 
     def consumehunkbuffer():
         if hunkbuffer:
             for token in dodiffhunk(hunkbuffer):
                 yield token
             hunkbuffer[:] = []
 
     for chunk in func(*args, **kw):
         lines = chunk.split(b'\n')
         linecount = len(lines)
         for i, line in enumerate(lines):
             if head:
                 if line.startswith(b'@'):
                     head = False
             else:
                 if line and not line.startswith(
                     (b' ', b'+', b'-', b'@', b'\\')
                 ):
                     head = True
             diffline = False
             if not head and line and line.startswith((b'+', b'-')):
                 diffline = True
 
             prefixes = textprefixes
             if head:
                 prefixes = headprefixes
             if diffline:
                 # buffered
                 bufferedline = line
                 if i + 1 < linecount:
                     bufferedline += b"\n"
                 hunkbuffer.append(bufferedline)
             else:
                 # unbuffered
                 for token in consumehunkbuffer():
                     yield token
                 stripline = line.rstrip()
                 for prefix, label in prefixes:
                     if stripline.startswith(prefix):
                         yield (stripline, label)
                         if line != stripline:
                             yield (
                                 line[len(stripline) :],
                                 b'diff.trailingwhitespace',
                             )
                         break
                 else:
                     yield (line, b'')
                 if i + 1 < linecount:
                     yield (b'\n', b'')
         for token in consumehunkbuffer():
             yield token
 
 
 def diffui(*args, **kw):
     '''like diff(), but yields 2-tuples of (output, label) for ui.write()'''
     return difflabel(diff, *args, **kw)
 
 
 def _filepairs(modified, added, removed, copy, opts):
     '''generates tuples (f1, f2, copyop), where f1 is the name of the file
     before and f2 is the the name after. For added files, f1 will be None,
     and for removed files, f2 will be None. copyop may be set to None, 'copy'
     or 'rename' (the latter two only if opts.git is set).'''
     gone = set()
 
     copyto = dict([(v, k) for k, v in copy.items()])
 
     addedset, removedset = set(added), set(removed)
 
     for f in sorted(modified + added + removed):
         copyop = None
         f1, f2 = f, f
         if f in addedset:
             f1 = None
             if f in copy:
                 if opts.git:
                     f1 = copy[f]
                     if f1 in removedset and f1 not in gone:
                         copyop = b'rename'
                         gone.add(f1)
                     else:
                         copyop = b'copy'
         elif f in removedset:
             f2 = None
             if opts.git:
                 # have we already reported a copy above?
                 if (
                     f in copyto
                     and copyto[f] in addedset
                     and copy[copyto[f]] == f
                 ):
                     continue
         yield f1, f2, copyop
 
 
 def trydiff(
     repo,
     revs,
     ctx1,
     ctx2,
     modified,
     added,
     removed,
     copy,
     getfilectx,
     opts,
     losedatafn,
     pathfn,
 ):
     '''given input data, generate a diff and yield it in blocks
 
     If generating a diff would lose data like flags or binary data and
     losedatafn is not None, it will be called.
 
     pathfn is applied to every path in the diff output.
     '''
 
     def gitindex(text):
         if not text:
             text = b""
         l = len(text)
         s = hashutil.sha1(b'blob %d\0' % l)
         s.update(text)
         return hex(s.digest())
 
     if opts.noprefix:
         aprefix = bprefix = b''
     else:
         aprefix = b'a/'
         bprefix = b'b/'
 
     def diffline(f, revs):
         revinfo = b' '.join([b"-r %s" % rev for rev in revs])
         return b'diff %s %s' % (revinfo, f)
 
     def isempty(fctx):
         return fctx is None or fctx.size() == 0
 
     date1 = dateutil.datestr(ctx1.date())
     date2 = dateutil.datestr(ctx2.date())
 
     gitmode = {b'l': b'120000', b'x': b'100755', b'': b'100644'}
 
     if not pathfn:
         pathfn = lambda f: f
 
     for f1, f2, copyop in _filepairs(modified, added, removed, copy, opts):
         content1 = None
         content2 = None
         fctx1 = None
         fctx2 = None
         flag1 = None
         flag2 = None
         if f1:
             fctx1 = getfilectx(f1, ctx1)
             if opts.git or losedatafn:
                 flag1 = ctx1.flags(f1)
         if f2:
             fctx2 = getfilectx(f2, ctx2)
             if opts.git or losedatafn:
                 flag2 = ctx2.flags(f2)
         # if binary is True, output "summary" or "base85", but not "text diff"
         if opts.text:
             binary = False
         else:
             binary = any(f.isbinary() for f in [fctx1, fctx2] if f is not None)
 
         if losedatafn and not opts.git:
             if (
                 binary
                 or
                 # copy/rename
                 f2 in copy
                 or
                 # empty file creation
                 (not f1 and isempty(fctx2))
                 or
                 # empty file deletion
                 (isempty(fctx1) and not f2)
                 or
                 # create with flags
                 (not f1 and flag2)
                 or
                 # change flags
                 (f1 and f2 and flag1 != flag2)
             ):
                 losedatafn(f2 or f1)
 
         path1 = pathfn(f1 or f2)
         path2 = pathfn(f2 or f1)
         header = []
         if opts.git:
             header.append(
                 b'diff --git %s%s %s%s' % (aprefix, path1, bprefix, path2)
             )
             if not f1:  # added
                 header.append(b'new file mode %s' % gitmode[flag2])
             elif not f2:  # removed
                 header.append(b'deleted file mode %s' % gitmode[flag1])
             else:  # modified/copied/renamed
                 mode1, mode2 = gitmode[flag1], gitmode[flag2]
                 if mode1 != mode2:
                     header.append(b'old mode %s' % mode1)
                     header.append(b'new mode %s' % mode2)
                 if copyop is not None:
                     if opts.showsimilarity:
                         sim = similar.score(ctx1[path1], ctx2[path2]) * 100
                         header.append(b'similarity index %d%%' % sim)
                     header.append(b'%s from %s' % (copyop, path1))
                     header.append(b'%s to %s' % (copyop, path2))
         elif revs:
             header.append(diffline(path1, revs))
 
         #  fctx.is  | diffopts                | what to   | is fctx.data()
         #  binary() | text nobinary git index | output?   | outputted?
         # ------------------------------------|----------------------------
         #  yes      | no   no       no  *     | summary   | no
         #  yes      | no   no       yes *     | base85    | yes
         #  yes      | no   yes      no  *     | summary   | no
         #  yes      | no   yes      yes 0     | summary   | no
         #  yes      | no   yes      yes >0    | summary   | semi [1]
         #  yes      | yes  *        *   *     | text diff | yes
         #  no       | *    *        *   *     | text diff | yes
         # [1]: hash(fctx.data()) is outputted. so fctx.data() cannot be faked
         if binary and (
             not opts.git or (opts.git and opts.nobinary and not opts.index)
         ):
             # fast path: no binary content will be displayed, content1 and
             # content2 are only used for equivalent test. cmp() could have a
             # fast path.
             if fctx1 is not None:
                 content1 = b'\0'
             if fctx2 is not None:
                 if fctx1 is not None and not fctx1.cmp(fctx2):
                     content2 = b'\0'  # not different
                 else:
                     content2 = b'\0\0'
         else:
             # normal path: load contents
             if fctx1 is not None:
                 content1 = fctx1.data()
             if fctx2 is not None:
                 content2 = fctx2.data()
 
         if binary and opts.git and not opts.nobinary:
             text = mdiff.b85diff(content1, content2)
             if text:
                 header.append(
                     b'index %s..%s' % (gitindex(content1), gitindex(content2))
                 )
             hunks = ((None, [text]),)
         else:
             if opts.git and opts.index > 0:
                 flag = flag1
                 if flag is None:
                     flag = flag2
                 header.append(
                     b'index %s..%s %s'
                     % (
                         gitindex(content1)[0 : opts.index],
                         gitindex(content2)[0 : opts.index],
                         gitmode[flag],
                     )
                 )
 
             uheaders, hunks = mdiff.unidiff(
                 content1,
                 date1,
                 content2,
                 date2,
                 path1,
                 path2,
                 binary=binary,
                 opts=opts,
             )
             header.extend(uheaders)
         yield fctx1, fctx2, header, hunks
 
 
 def diffstatsum(stats):
     maxfile, maxtotal, addtotal, removetotal, binary = 0, 0, 0, 0, False
     for f, a, r, b in stats:
         maxfile = max(maxfile, encoding.colwidth(f))
         maxtotal = max(maxtotal, a + r)
         addtotal += a
         removetotal += r
         binary = binary or b
 
     return maxfile, maxtotal, addtotal, removetotal, binary
 
 
 def diffstatdata(lines):
     diffre = re.compile(br'^diff .*-r [a-z0-9]+\s(.*)$')
 
     results = []
     filename, adds, removes, isbinary = None, 0, 0, False
 
     def addresult():
         if filename:
             results.append((filename, adds, removes, isbinary))
 
     # inheader is used to track if a line is in the
     # header portion of the diff.  This helps properly account
     # for lines that start with '--' or '++'
     inheader = False
 
     for line in lines:
         if line.startswith(b'diff'):
             addresult()
             # starting a new file diff
             # set numbers to 0 and reset inheader
             inheader = True
             adds, removes, isbinary = 0, 0, False
             if line.startswith(b'diff --git a/'):
                 filename = gitre.search(line).group(2)
             elif line.startswith(b'diff -r'):
                 # format: "diff -r ... -r ... filename"
                 filename = diffre.search(line).group(1)
         elif line.startswith(b'@@'):
             inheader = False
         elif line.startswith(b'+') and not inheader:
             adds += 1
         elif line.startswith(b'-') and not inheader:
             removes += 1
         elif line.startswith(b'GIT binary patch') or line.startswith(
             b'Binary file'
         ):
             isbinary = True
         elif line.startswith(b'rename from'):
             filename = line[12:]
         elif line.startswith(b'rename to'):
             filename += b' => %s' % line[10:]
     addresult()
     return results
 
 
 def diffstat(lines, width=80):
     output = []
     stats = diffstatdata(lines)
     maxname, maxtotal, totaladds, totalremoves, hasbinary = diffstatsum(stats)
 
     countwidth = len(str(maxtotal))
     if hasbinary and countwidth < 3:
         countwidth = 3
     graphwidth = width - countwidth - maxname - 6
     if graphwidth < 10:
         graphwidth = 10
 
     def scale(i):
         if maxtotal <= graphwidth:
             return i
         # If diffstat runs out of room it doesn't print anything,
         # which isn't very useful, so always print at least one + or -
         # if there were at least some changes.
         return max(i * graphwidth // maxtotal, int(bool(i)))
 
     for filename, adds, removes, isbinary in stats:
         if isbinary:
             count = b'Bin'
         else:
             count = b'%d' % (adds + removes)
         pluses = b'+' * scale(adds)
         minuses = b'-' * scale(removes)
         output.append(
             b' %s%s |  %*s %s%s\n'
             % (
                 filename,
                 b' ' * (maxname - encoding.colwidth(filename)),
                 countwidth,
                 count,
                 pluses,
                 minuses,
             )
         )
 
     if stats:
         output.append(
             _(b' %d files changed, %d insertions(+), %d deletions(-)\n')
             % (len(stats), totaladds, totalremoves)
         )
 
     return b''.join(output)
 
 
 def diffstatui(*args, **kw):
     '''like diffstat(), but yields 2-tuples of (output, label) for
     ui.write()
     '''
 
     for line in diffstat(*args, **kw).splitlines():
         if line and line[-1] in b'+-':
             name, graph = line.rsplit(b' ', 1)
             yield (name + b' ', b'')
             m = re.search(br'\++', graph)
             if m:
                 yield (m.group(0), b'diffstat.inserted')
             m = re.search(br'-+', graph)
             if m:
                 yield (m.group(0), b'diffstat.deleted')
         else:
             yield (line, b'')
         yield (b'\n', b'')