revlog: decouple caching from addrevision callback for addgroup

For changesets, it is useful to cache the content as it will almost
always be processed afterwards. For manifests on the other hand, the
content is often not used directly as there is a fast path for deltas.
Explicitly disable the cache in exchangev2's manifest handling for that
reason.

Differential Revision: https://phab.mercurial-scm.org/D9847

mercurial/changegroup.py

@@ -1,1706 +1,1707 b''
 # changegroup.py - Mercurial changegroup manipulation functions
 #
 #  Copyright 2006 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 os
 import struct
 import weakref
 
 from .i18n import _
 from .node import (
     hex,
     nullid,
     nullrev,
     short,
 )
 from .pycompat import open
 
 from . import (
     error,
     match as matchmod,
     mdiff,
     phases,
     pycompat,
     requirements,
     scmutil,
     util,
 )
 
 from .interfaces import repository
 
 _CHANGEGROUPV1_DELTA_HEADER = struct.Struct(b"20s20s20s20s")
 _CHANGEGROUPV2_DELTA_HEADER = struct.Struct(b"20s20s20s20s20s")
 _CHANGEGROUPV3_DELTA_HEADER = struct.Struct(b">20s20s20s20s20sH")
 
 LFS_REQUIREMENT = b'lfs'
 
 readexactly = util.readexactly
 
 
 def getchunk(stream):
     """return the next chunk from stream as a string"""
     d = readexactly(stream, 4)
     l = struct.unpack(b">l", d)[0]
     if l <= 4:
         if l:
             raise error.Abort(_(b"invalid chunk length %d") % l)
         return b""
     return readexactly(stream, l - 4)
 
 
 def chunkheader(length):
     """return a changegroup chunk header (string)"""
     return struct.pack(b">l", length + 4)
 
 
 def closechunk():
     """return a changegroup chunk header (string) for a zero-length chunk"""
     return struct.pack(b">l", 0)
 
 
 def _fileheader(path):
     """Obtain a changegroup chunk header for a named path."""
     return chunkheader(len(path)) + path
 
 
 def writechunks(ui, chunks, filename, vfs=None):
     """Write chunks to a file and return its filename.
 
     The stream is assumed to be a bundle file.
     Existing files will not be overwritten.
     If no filename is specified, a temporary file is created.
     """
     fh = None
     cleanup = None
     try:
         if filename:
             if vfs:
                 fh = vfs.open(filename, b"wb")
             else:
                 # Increase default buffer size because default is usually
                 # small (4k is common on Linux).
                 fh = open(filename, b"wb", 131072)
         else:
             fd, filename = pycompat.mkstemp(prefix=b"hg-bundle-", suffix=b".hg")
             fh = os.fdopen(fd, "wb")
         cleanup = filename
         for c in chunks:
             fh.write(c)
         cleanup = None
         return filename
     finally:
         if fh is not None:
             fh.close()
         if cleanup is not None:
             if filename and vfs:
                 vfs.unlink(cleanup)
             else:
                 os.unlink(cleanup)
 
 
 class cg1unpacker(object):
     """Unpacker for cg1 changegroup streams.
 
     A changegroup unpacker handles the framing of the revision data in
     the wire format. Most consumers will want to use the apply()
     method to add the changes from the changegroup to a repository.
 
     If you're forwarding a changegroup unmodified to another consumer,
     use getchunks(), which returns an iterator of changegroup
     chunks. This is mostly useful for cases where you need to know the
     data stream has ended by observing the end of the changegroup.
 
     deltachunk() is useful only if you're applying delta data. Most
     consumers should prefer apply() instead.
 
     A few other public methods exist. Those are used only for
     bundlerepo and some debug commands - their use is discouraged.
     """
 
     deltaheader = _CHANGEGROUPV1_DELTA_HEADER
     deltaheadersize = deltaheader.size
     version = b'01'
     _grouplistcount = 1  # One list of files after the manifests
 
     def __init__(self, fh, alg, extras=None):
         if alg is None:
             alg = b'UN'
         if alg not in util.compengines.supportedbundletypes:
             raise error.Abort(_(b'unknown stream compression type: %s') % alg)
         if alg == b'BZ':
             alg = b'_truncatedBZ'
 
         compengine = util.compengines.forbundletype(alg)
         self._stream = compengine.decompressorreader(fh)
         self._type = alg
         self.extras = extras or {}
         self.callback = None
 
     # These methods (compressed, read, seek, tell) all appear to only
     # be used by bundlerepo, but it's a little hard to tell.
     def compressed(self):
         return self._type is not None and self._type != b'UN'
 
     def read(self, l):
         return self._stream.read(l)
 
     def seek(self, pos):
         return self._stream.seek(pos)
 
     def tell(self):
         return self._stream.tell()
 
     def close(self):
         return self._stream.close()
 
     def _chunklength(self):
         d = readexactly(self._stream, 4)
         l = struct.unpack(b">l", d)[0]
         if l <= 4:
             if l:
                 raise error.Abort(_(b"invalid chunk length %d") % l)
             return 0
         if self.callback:
             self.callback()
         return l - 4
 
     def changelogheader(self):
         """v10 does not have a changelog header chunk"""
         return {}
 
     def manifestheader(self):
         """v10 does not have a manifest header chunk"""
         return {}
 
     def filelogheader(self):
         """return the header of the filelogs chunk, v10 only has the filename"""
         l = self._chunklength()
         if not l:
             return {}
         fname = readexactly(self._stream, l)
         return {b'filename': fname}
 
     def _deltaheader(self, headertuple, prevnode):
         node, p1, p2, cs = headertuple
         if prevnode is None:
             deltabase = p1
         else:
             deltabase = prevnode
         flags = 0
         return node, p1, p2, deltabase, cs, flags
 
     def deltachunk(self, prevnode):
         l = self._chunklength()
         if not l:
             return {}
         headerdata = readexactly(self._stream, self.deltaheadersize)
         header = self.deltaheader.unpack(headerdata)
         delta = readexactly(self._stream, l - self.deltaheadersize)
         node, p1, p2, deltabase, cs, flags = self._deltaheader(header, prevnode)
         return (node, p1, p2, cs, deltabase, delta, flags)
 
     def getchunks(self):
         """returns all the chunks contains in the bundle
 
         Used when you need to forward the binary stream to a file or another
         network API. To do so, it parse the changegroup data, otherwise it will
         block in case of sshrepo because it don't know the end of the stream.
         """
         # For changegroup 1 and 2, we expect 3 parts: changelog, manifestlog,
         # and a list of filelogs. For changegroup 3, we expect 4 parts:
         # changelog, manifestlog, a list of tree manifestlogs, and a list of
         # filelogs.
         #
         # Changelog and manifestlog parts are terminated with empty chunks. The
         # tree and file parts are a list of entry sections. Each entry section
         # is a series of chunks terminating in an empty chunk. The list of these
         # entry sections is terminated in yet another empty chunk, so we know
         # we've reached the end of the tree/file list when we reach an empty
         # chunk that was proceeded by no non-empty chunks.
 
         parts = 0
         while parts < 2 + self._grouplistcount:
             noentries = True
             while True:
                 chunk = getchunk(self)
                 if not chunk:
                     # The first two empty chunks represent the end of the
                     # changelog and the manifestlog portions. The remaining
                     # empty chunks represent either A) the end of individual
                     # tree or file entries in the file list, or B) the end of
                     # the entire list. It's the end of the entire list if there
                     # were no entries (i.e. noentries is True).
                     if parts < 2:
                         parts += 1
                     elif noentries:
                         parts += 1
                     break
                 noentries = False
                 yield chunkheader(len(chunk))
                 pos = 0
                 while pos < len(chunk):
                     next = pos + 2 ** 20
                     yield chunk[pos:next]
                     pos = next
             yield closechunk()
 
     def _unpackmanifests(self, repo, revmap, trp, prog):
         self.callback = prog.increment
         # no need to check for empty manifest group here:
         # if the result of the merge of 1 and 2 is the same in 3 and 4,
         # no new manifest will be created and the manifest group will
         # be empty during the pull
         self.manifestheader()
         deltas = self.deltaiter()
         repo.manifestlog.getstorage(b'').addgroup(deltas, revmap, trp)
         prog.complete()
         self.callback = None
 
     def apply(
         self,
         repo,
         tr,
         srctype,
         url,
         targetphase=phases.draft,
         expectedtotal=None,
     ):
         """Add the changegroup returned by source.read() to this repo.
         srctype is a string like 'push', 'pull', or 'unbundle'.  url is
         the URL of the repo where this changegroup is coming from.
 
         Return an integer summarizing the change to this repo:
         - nothing changed or no source: 0
         - more heads than before: 1+added heads (2..n)
         - fewer heads than before: -1-removed heads (-2..-n)
         - number of heads stays the same: 1
         """
         repo = repo.unfiltered()
 
         def csmap(x):
             repo.ui.debug(b"add changeset %s\n" % short(x))
             return len(cl)
 
         def revmap(x):
             return cl.rev(x)
 
         try:
             # The transaction may already carry source information. In this
             # case we use the top level data. We overwrite the argument
             # because we need to use the top level value (if they exist)
             # in this function.
             srctype = tr.hookargs.setdefault(b'source', srctype)
             tr.hookargs.setdefault(b'url', url)
             repo.hook(
                 b'prechangegroup', throw=True, **pycompat.strkwargs(tr.hookargs)
             )
 
             # write changelog data to temp files so concurrent readers
             # will not see an inconsistent view
             cl = repo.changelog
             cl.delayupdate(tr)
             oldheads = set(cl.heads())
 
             trp = weakref.proxy(tr)
             # pull off the changeset group
             repo.ui.status(_(b"adding changesets\n"))
             clstart = len(cl)
             progress = repo.ui.makeprogress(
                 _(b'changesets'), unit=_(b'chunks'), total=expectedtotal
             )
             self.callback = progress.increment
 
             efilesset = set()
             cgnodes = []
 
             def ondupchangelog(cl, node):
                 if cl.rev(node) < clstart:
                     cgnodes.append(node)
 
             def onchangelog(cl, node):
                 rev = cl.rev(node)
                 ctx = cl.changelogrevision(rev)
                 efilesset.update(ctx.files)
                 repo.register_changeset(rev, ctx)
 
             self.changelogheader()
             deltas = self.deltaiter()
             if not cl.addgroup(
                 deltas,
                 csmap,
                 trp,
+                alwayscache=True,
                 addrevisioncb=onchangelog,
                 duplicaterevisioncb=ondupchangelog,
             ):
                 repo.ui.develwarn(
                     b'applied empty changelog from changegroup',
                     config=b'warn-empty-changegroup',
                 )
             efiles = len(efilesset)
             clend = len(cl)
             changesets = clend - clstart
             progress.complete()
             del deltas
             # TODO Python 2.7 removal
             # del efilesset
             efilesset = None
             self.callback = None
 
             # pull off the manifest group
             repo.ui.status(_(b"adding manifests\n"))
             # We know that we'll never have more manifests than we had
             # changesets.
             progress = repo.ui.makeprogress(
                 _(b'manifests'), unit=_(b'chunks'), total=changesets
             )
             self._unpackmanifests(repo, revmap, trp, progress)
 
             needfiles = {}
             if repo.ui.configbool(b'server', b'validate'):
                 cl = repo.changelog
                 ml = repo.manifestlog
                 # validate incoming csets have their manifests
                 for cset in pycompat.xrange(clstart, clend):
                     mfnode = cl.changelogrevision(cset).manifest
                     mfest = ml[mfnode].readdelta()
                     # store file nodes we must see
                     for f, n in pycompat.iteritems(mfest):
                         needfiles.setdefault(f, set()).add(n)
 
             # process the files
             repo.ui.status(_(b"adding file changes\n"))
             newrevs, newfiles = _addchangegroupfiles(
                 repo, self, revmap, trp, efiles, needfiles
             )
 
             # making sure the value exists
             tr.changes.setdefault(b'changegroup-count-changesets', 0)
             tr.changes.setdefault(b'changegroup-count-revisions', 0)
             tr.changes.setdefault(b'changegroup-count-files', 0)
             tr.changes.setdefault(b'changegroup-count-heads', 0)
 
             # some code use bundle operation for internal purpose. They usually
             # set `ui.quiet` to do this outside of user sight. Size the report
             # of such operation now happens at the end of the transaction, that
             # ui.quiet has not direct effect on the output.
             #
             # To preserve this intend use an inelegant hack, we fail to report
             # the change if `quiet` is set. We should probably move to
             # something better, but this is a good first step to allow the "end
             # of transaction report" to pass tests.
             if not repo.ui.quiet:
                 tr.changes[b'changegroup-count-changesets'] += changesets
                 tr.changes[b'changegroup-count-revisions'] += newrevs
                 tr.changes[b'changegroup-count-files'] += newfiles
 
             deltaheads = 0
             if oldheads:
                 heads = cl.heads()
                 deltaheads += len(heads) - len(oldheads)
                 for h in heads:
                     if h not in oldheads and repo[h].closesbranch():
                         deltaheads -= 1
 
             # see previous comment about checking ui.quiet
             if not repo.ui.quiet:
                 tr.changes[b'changegroup-count-heads'] += deltaheads
             repo.invalidatevolatilesets()
 
             if changesets > 0:
                 if b'node' not in tr.hookargs:
                     tr.hookargs[b'node'] = hex(cl.node(clstart))
                     tr.hookargs[b'node_last'] = hex(cl.node(clend - 1))
                     hookargs = dict(tr.hookargs)
                 else:
                     hookargs = dict(tr.hookargs)
                     hookargs[b'node'] = hex(cl.node(clstart))
                     hookargs[b'node_last'] = hex(cl.node(clend - 1))
                 repo.hook(
                     b'pretxnchangegroup',
                     throw=True,
                     **pycompat.strkwargs(hookargs)
                 )
 
             added = pycompat.xrange(clstart, clend)
             phaseall = None
             if srctype in (b'push', b'serve'):
                 # Old servers can not push the boundary themselves.
                 # New servers won't push the boundary if changeset already
                 # exists locally as secret
                 #
                 # We should not use added here but the list of all change in
                 # the bundle
                 if repo.publishing():
                     targetphase = phaseall = phases.public
                 else:
                     # closer target phase computation
 
                     # Those changesets have been pushed from the
                     # outside, their phases are going to be pushed
                     # alongside. Therefor `targetphase` is
                     # ignored.
                     targetphase = phaseall = phases.draft
             if added:
                 phases.registernew(repo, tr, targetphase, added)
             if phaseall is not None:
                 phases.advanceboundary(repo, tr, phaseall, cgnodes, revs=added)
                 cgnodes = []
 
             if changesets > 0:
 
                 def runhooks(unused_success):
                     # These hooks run when the lock releases, not when the
                     # transaction closes. So it's possible for the changelog
                     # to have changed since we last saw it.
                     if clstart >= len(repo):
                         return
 
                     repo.hook(b"changegroup", **pycompat.strkwargs(hookargs))
 
                     for rev in added:
                         args = hookargs.copy()
                         args[b'node'] = hex(cl.node(rev))
                         del args[b'node_last']
                         repo.hook(b"incoming", **pycompat.strkwargs(args))
 
                     newheads = [h for h in repo.heads() if h not in oldheads]
                     repo.ui.log(
                         b"incoming",
                         b"%d incoming changes - new heads: %s\n",
                         len(added),
                         b', '.join([hex(c[:6]) for c in newheads]),
                     )
 
                 tr.addpostclose(
                     b'changegroup-runhooks-%020i' % clstart,
                     lambda tr: repo._afterlock(runhooks),
                 )
         finally:
             repo.ui.flush()
         # never return 0 here:
         if deltaheads < 0:
             ret = deltaheads - 1
         else:
             ret = deltaheads + 1
         return ret
 
     def deltaiter(self):
         """
         returns an iterator of the deltas in this changegroup
 
         Useful for passing to the underlying storage system to be stored.
         """
         chain = None
         for chunkdata in iter(lambda: self.deltachunk(chain), {}):
             # Chunkdata: (node, p1, p2, cs, deltabase, delta, flags)
             yield chunkdata
             chain = chunkdata[0]
 
 
 class cg2unpacker(cg1unpacker):
     """Unpacker for cg2 streams.
 
     cg2 streams add support for generaldelta, so the delta header
     format is slightly different. All other features about the data
     remain the same.
     """
 
     deltaheader = _CHANGEGROUPV2_DELTA_HEADER
     deltaheadersize = deltaheader.size
     version = b'02'
 
     def _deltaheader(self, headertuple, prevnode):
         node, p1, p2, deltabase, cs = headertuple
         flags = 0
         return node, p1, p2, deltabase, cs, flags
 
 
 class cg3unpacker(cg2unpacker):
     """Unpacker for cg3 streams.
 
     cg3 streams add support for exchanging treemanifests and revlog
     flags. It adds the revlog flags to the delta header and an empty chunk
     separating manifests and files.
     """
 
     deltaheader = _CHANGEGROUPV3_DELTA_HEADER
     deltaheadersize = deltaheader.size
     version = b'03'
     _grouplistcount = 2  # One list of manifests and one list of files
 
     def _deltaheader(self, headertuple, prevnode):
         node, p1, p2, deltabase, cs, flags = headertuple
         return node, p1, p2, deltabase, cs, flags
 
     def _unpackmanifests(self, repo, revmap, trp, prog):
         super(cg3unpacker, self)._unpackmanifests(repo, revmap, trp, prog)
         for chunkdata in iter(self.filelogheader, {}):
             # If we get here, there are directory manifests in the changegroup
             d = chunkdata[b"filename"]
             repo.ui.debug(b"adding %s revisions\n" % d)
             deltas = self.deltaiter()
             if not repo.manifestlog.getstorage(d).addgroup(deltas, revmap, trp):
                 raise error.Abort(_(b"received dir revlog group is empty"))
 
 
 class headerlessfixup(object):
     def __init__(self, fh, h):
         self._h = h
         self._fh = fh
 
     def read(self, n):
         if self._h:
             d, self._h = self._h[:n], self._h[n:]
             if len(d) < n:
                 d += readexactly(self._fh, n - len(d))
             return d
         return readexactly(self._fh, n)
 
 
 def _revisiondeltatochunks(delta, headerfn):
     """Serialize a revisiondelta to changegroup chunks."""
 
     # The captured revision delta may be encoded as a delta against
     # a base revision or as a full revision. The changegroup format
     # requires that everything on the wire be deltas. So for full
     # revisions, we need to invent a header that says to rewrite
     # data.
 
     if delta.delta is not None:
         prefix, data = b'', delta.delta
     elif delta.basenode == nullid:
         data = delta.revision
         prefix = mdiff.trivialdiffheader(len(data))
     else:
         data = delta.revision
         prefix = mdiff.replacediffheader(delta.baserevisionsize, len(data))
 
     meta = headerfn(delta)
 
     yield chunkheader(len(meta) + len(prefix) + len(data))
     yield meta
     if prefix:
         yield prefix
     yield data
 
 
 def _sortnodesellipsis(store, nodes, cl, lookup):
     """Sort nodes for changegroup generation."""
     # Ellipses serving mode.
     #
     # In a perfect world, we'd generate better ellipsis-ified graphs
     # for non-changelog revlogs. In practice, we haven't started doing
     # that yet, so the resulting DAGs for the manifestlog and filelogs
     # are actually full of bogus parentage on all the ellipsis
     # nodes. This has the side effect that, while the contents are
     # correct, the individual DAGs might be completely out of whack in
     # a case like 882681bc3166 and its ancestors (back about 10
     # revisions or so) in the main hg repo.
     #
     # The one invariant we *know* holds is that the new (potentially
     # bogus) DAG shape will be valid if we order the nodes in the
     # order that they're introduced in dramatis personae by the
     # changelog, so what we do is we sort the non-changelog histories
     # by the order in which they are used by the changelog.
     key = lambda n: cl.rev(lookup(n))
     return sorted(nodes, key=key)
 
 
 def _resolvenarrowrevisioninfo(
     cl,
     store,
     ischangelog,
     rev,
     linkrev,
     linknode,
     clrevtolocalrev,
     fullclnodes,
     precomputedellipsis,
 ):
     linkparents = precomputedellipsis[linkrev]
 
     def local(clrev):
         """Turn a changelog revnum into a local revnum.
 
         The ellipsis dag is stored as revnums on the changelog,
         but when we're producing ellipsis entries for
         non-changelog revlogs, we need to turn those numbers into
         something local. This does that for us, and during the
         changelog sending phase will also expand the stored
         mappings as needed.
         """
         if clrev == nullrev:
             return nullrev
 
         if ischangelog:
             return clrev
 
         # Walk the ellipsis-ized changelog breadth-first looking for a
         # change that has been linked from the current revlog.
         #
         # For a flat manifest revlog only a single step should be necessary
         # as all relevant changelog entries are relevant to the flat
         # manifest.
         #
         # For a filelog or tree manifest dirlog however not every changelog
         # entry will have been relevant, so we need to skip some changelog
         # nodes even after ellipsis-izing.
         walk = [clrev]
         while walk:
             p = walk[0]
             walk = walk[1:]
             if p in clrevtolocalrev:
                 return clrevtolocalrev[p]
             elif p in fullclnodes:
                 walk.extend([pp for pp in cl.parentrevs(p) if pp != nullrev])
             elif p in precomputedellipsis:
                 walk.extend(
                     [pp for pp in precomputedellipsis[p] if pp != nullrev]
                 )
             else:
                 # In this case, we've got an ellipsis with parents
                 # outside the current bundle (likely an
                 # incremental pull). We "know" that we can use the
                 # value of this same revlog at whatever revision
                 # is pointed to by linknode. "Know" is in scare
                 # quotes because I haven't done enough examination
                 # of edge cases to convince myself this is really
                 # a fact - it works for all the (admittedly
                 # thorough) cases in our testsuite, but I would be
                 # somewhat unsurprised to find a case in the wild
                 # where this breaks down a bit. That said, I don't
                 # know if it would hurt anything.
                 for i in pycompat.xrange(rev, 0, -1):
                     if store.linkrev(i) == clrev:
                         return i
                 # We failed to resolve a parent for this node, so
                 # we crash the changegroup construction.
                 raise error.Abort(
                     b'unable to resolve parent while packing %r %r'
                     b' for changeset %r' % (store.indexfile, rev, clrev)
                 )
 
         return nullrev
 
     if not linkparents or (store.parentrevs(rev) == (nullrev, nullrev)):
         p1, p2 = nullrev, nullrev
     elif len(linkparents) == 1:
         (p1,) = sorted(local(p) for p in linkparents)
         p2 = nullrev
     else:
         p1, p2 = sorted(local(p) for p in linkparents)
 
     p1node, p2node = store.node(p1), store.node(p2)
 
     return p1node, p2node, linknode
 
 
 def deltagroup(
     repo,
     store,
     nodes,
     ischangelog,
     lookup,
     forcedeltaparentprev,
     topic=None,
     ellipses=False,
     clrevtolocalrev=None,
     fullclnodes=None,
     precomputedellipsis=None,
 ):
     """Calculate deltas for a set of revisions.
 
     Is a generator of ``revisiondelta`` instances.
 
     If topic is not None, progress detail will be generated using this
     topic name (e.g. changesets, manifests, etc).
     """
     if not nodes:
         return
 
     cl = repo.changelog
 
     if ischangelog:
         # `hg log` shows changesets in storage order. To preserve order
         # across clones, send out changesets in storage order.
         nodesorder = b'storage'
     elif ellipses:
         nodes = _sortnodesellipsis(store, nodes, cl, lookup)
         nodesorder = b'nodes'
     else:
         nodesorder = None
 
     # Perform ellipses filtering and revision massaging. We do this before
     # emitrevisions() because a) filtering out revisions creates less work
     # for emitrevisions() b) dropping revisions would break emitrevisions()'s
     # assumptions about delta choices and we would possibly send a delta
     # referencing a missing base revision.
     #
     # Also, calling lookup() has side-effects with regards to populating
     # data structures. If we don't call lookup() for each node or if we call
     # lookup() after the first pass through each node, things can break -
     # possibly intermittently depending on the python hash seed! For that
     # reason, we store a mapping of all linknodes during the initial node
     # pass rather than use lookup() on the output side.
     if ellipses:
         filtered = []
         adjustedparents = {}
         linknodes = {}
 
         for node in nodes:
             rev = store.rev(node)
             linknode = lookup(node)
             linkrev = cl.rev(linknode)
             clrevtolocalrev[linkrev] = rev
 
             # If linknode is in fullclnodes, it means the corresponding
             # changeset was a full changeset and is being sent unaltered.
             if linknode in fullclnodes:
                 linknodes[node] = linknode
 
             # If the corresponding changeset wasn't in the set computed
             # as relevant to us, it should be dropped outright.
             elif linkrev not in precomputedellipsis:
                 continue
 
             else:
                 # We could probably do this later and avoid the dict
                 # holding state. But it likely doesn't matter.
                 p1node, p2node, linknode = _resolvenarrowrevisioninfo(
                     cl,
                     store,
                     ischangelog,
                     rev,
                     linkrev,
                     linknode,
                     clrevtolocalrev,
                     fullclnodes,
                     precomputedellipsis,
                 )
 
                 adjustedparents[node] = (p1node, p2node)
                 linknodes[node] = linknode
 
             filtered.append(node)
 
         nodes = filtered
 
     # We expect the first pass to be fast, so we only engage the progress
     # meter for constructing the revision deltas.
     progress = None
     if topic is not None:
         progress = repo.ui.makeprogress(
             topic, unit=_(b'chunks'), total=len(nodes)
         )
 
     configtarget = repo.ui.config(b'devel', b'bundle.delta')
     if configtarget not in (b'', b'p1', b'full'):
         msg = _("""config "devel.bundle.delta" as unknown value: %s""")
         repo.ui.warn(msg % configtarget)
 
     deltamode = repository.CG_DELTAMODE_STD
     if forcedeltaparentprev:
         deltamode = repository.CG_DELTAMODE_PREV
     elif configtarget == b'p1':
         deltamode = repository.CG_DELTAMODE_P1
     elif configtarget == b'full':
         deltamode = repository.CG_DELTAMODE_FULL
 
     revisions = store.emitrevisions(
         nodes,
         nodesorder=nodesorder,
         revisiondata=True,
         assumehaveparentrevisions=not ellipses,
         deltamode=deltamode,
     )
 
     for i, revision in enumerate(revisions):
         if progress:
             progress.update(i + 1)
 
         if ellipses:
             linknode = linknodes[revision.node]
 
             if revision.node in adjustedparents:
                 p1node, p2node = adjustedparents[revision.node]
                 revision.p1node = p1node
                 revision.p2node = p2node
                 revision.flags |= repository.REVISION_FLAG_ELLIPSIS
 
         else:
             linknode = lookup(revision.node)
 
         revision.linknode = linknode
         yield revision
 
     if progress:
         progress.complete()
 
 
 class cgpacker(object):
     def __init__(
         self,
         repo,
         oldmatcher,
         matcher,
         version,
         builddeltaheader,
         manifestsend,
         forcedeltaparentprev=False,
         bundlecaps=None,
         ellipses=False,
         shallow=False,
         ellipsisroots=None,
         fullnodes=None,
     ):
         """Given a source repo, construct a bundler.
 
         oldmatcher is a matcher that matches on files the client already has.
         These will not be included in the changegroup.
 
         matcher is a matcher that matches on files to include in the
         changegroup. Used to facilitate sparse changegroups.
 
         forcedeltaparentprev indicates whether delta parents must be against
         the previous revision in a delta group. This should only be used for
         compatibility with changegroup version 1.
 
         builddeltaheader is a callable that constructs the header for a group
         delta.
 
         manifestsend is a chunk to send after manifests have been fully emitted.
 
         ellipses indicates whether ellipsis serving mode is enabled.
 
         bundlecaps is optional and can be used to specify the set of
         capabilities which can be used to build the bundle. While bundlecaps is
         unused in core Mercurial, extensions rely on this feature to communicate
         capabilities to customize the changegroup packer.
 
         shallow indicates whether shallow data might be sent. The packer may
         need to pack file contents not introduced by the changes being packed.
 
         fullnodes is the set of changelog nodes which should not be ellipsis
         nodes. We store this rather than the set of nodes that should be
         ellipsis because for very large histories we expect this to be
         significantly smaller.
         """
         assert oldmatcher
         assert matcher
         self._oldmatcher = oldmatcher
         self._matcher = matcher
 
         self.version = version
         self._forcedeltaparentprev = forcedeltaparentprev
         self._builddeltaheader = builddeltaheader
         self._manifestsend = manifestsend
         self._ellipses = ellipses
 
         # Set of capabilities we can use to build the bundle.
         if bundlecaps is None:
             bundlecaps = set()
         self._bundlecaps = bundlecaps
         self._isshallow = shallow
         self._fullclnodes = fullnodes
 
         # Maps ellipsis revs to their roots at the changelog level.
         self._precomputedellipsis = ellipsisroots
 
         self._repo = repo
 
         if self._repo.ui.verbose and not self._repo.ui.debugflag:
             self._verbosenote = self._repo.ui.note
         else:
             self._verbosenote = lambda s: None
 
     def generate(
         self, commonrevs, clnodes, fastpathlinkrev, source, changelog=True
     ):
         """Yield a sequence of changegroup byte chunks.
         If changelog is False, changelog data won't be added to changegroup
         """
 
         repo = self._repo
         cl = repo.changelog
 
         self._verbosenote(_(b'uncompressed size of bundle content:\n'))
         size = 0
 
         clstate, deltas = self._generatechangelog(
             cl, clnodes, generate=changelog
         )
         for delta in deltas:
             for chunk in _revisiondeltatochunks(delta, self._builddeltaheader):
                 size += len(chunk)
                 yield chunk
 
         close = closechunk()
         size += len(close)
         yield closechunk()
 
         self._verbosenote(_(b'%8.i (changelog)\n') % size)
 
         clrevorder = clstate[b'clrevorder']
         manifests = clstate[b'manifests']
         changedfiles = clstate[b'changedfiles']
 
         # We need to make sure that the linkrev in the changegroup refers to
         # the first changeset that introduced the manifest or file revision.
         # The fastpath is usually safer than the slowpath, because the filelogs
         # are walked in revlog order.
         #
         # When taking the slowpath when the manifest revlog uses generaldelta,
         # the manifest may be walked in the "wrong" order. Without 'clrevorder',
         # we would get an incorrect linkrev (see fix in cc0ff93d0c0c).
         #
         # When taking the fastpath, we are only vulnerable to reordering
         # of the changelog itself. The changelog never uses generaldelta and is
         # never reordered. To handle this case, we simply take the slowpath,
         # which already has the 'clrevorder' logic. This was also fixed in
         # cc0ff93d0c0c.
 
         # Treemanifests don't work correctly with fastpathlinkrev
         # either, because we don't discover which directory nodes to
         # send along with files. This could probably be fixed.
         fastpathlinkrev = fastpathlinkrev and not scmutil.istreemanifest(repo)
 
         fnodes = {}  # needed file nodes
 
         size = 0
         it = self.generatemanifests(
             commonrevs,
             clrevorder,
             fastpathlinkrev,
             manifests,
             fnodes,
             source,
             clstate[b'clrevtomanifestrev'],
         )
 
         for tree, deltas in it:
             if tree:
                 assert self.version == b'03'
                 chunk = _fileheader(tree)
                 size += len(chunk)
                 yield chunk
 
             for delta in deltas:
                 chunks = _revisiondeltatochunks(delta, self._builddeltaheader)
                 for chunk in chunks:
                     size += len(chunk)
                     yield chunk
 
             close = closechunk()
             size += len(close)
             yield close
 
         self._verbosenote(_(b'%8.i (manifests)\n') % size)
         yield self._manifestsend
 
         mfdicts = None
         if self._ellipses and self._isshallow:
             mfdicts = [
                 (self._repo.manifestlog[n].read(), lr)
                 for (n, lr) in pycompat.iteritems(manifests)
             ]
 
         manifests.clear()
         clrevs = {cl.rev(x) for x in clnodes}
 
         it = self.generatefiles(
             changedfiles,
             commonrevs,
             source,
             mfdicts,
             fastpathlinkrev,
             fnodes,
             clrevs,
         )
 
         for path, deltas in it:
             h = _fileheader(path)
             size = len(h)
             yield h
 
             for delta in deltas:
                 chunks = _revisiondeltatochunks(delta, self._builddeltaheader)
                 for chunk in chunks:
                     size += len(chunk)
                     yield chunk
 
             close = closechunk()
             size += len(close)
             yield close
 
             self._verbosenote(_(b'%8.i  %s\n') % (size, path))
 
         yield closechunk()
 
         if clnodes:
             repo.hook(b'outgoing', node=hex(clnodes[0]), source=source)
 
     def _generatechangelog(self, cl, nodes, generate=True):
         """Generate data for changelog chunks.
 
         Returns a 2-tuple of a dict containing state and an iterable of
         byte chunks. The state will not be fully populated until the
         chunk stream has been fully consumed.
 
         if generate is False, the state will be fully populated and no chunk
         stream will be yielded
         """
         clrevorder = {}
         manifests = {}
         mfl = self._repo.manifestlog
         changedfiles = set()
         clrevtomanifestrev = {}
 
         state = {
             b'clrevorder': clrevorder,
             b'manifests': manifests,
             b'changedfiles': changedfiles,
             b'clrevtomanifestrev': clrevtomanifestrev,
         }
 
         if not (generate or self._ellipses):
             # sort the nodes in storage order
             nodes = sorted(nodes, key=cl.rev)
             for node in nodes:
                 c = cl.changelogrevision(node)
                 clrevorder[node] = len(clrevorder)
                 # record the first changeset introducing this manifest version
                 manifests.setdefault(c.manifest, node)
                 # Record a complete list of potentially-changed files in
                 # this manifest.
                 changedfiles.update(c.files)
 
             return state, ()
 
         # Callback for the changelog, used to collect changed files and
         # manifest nodes.
         # Returns the linkrev node (identity in the changelog case).
         def lookupcl(x):
             c = cl.changelogrevision(x)
             clrevorder[x] = len(clrevorder)
 
             if self._ellipses:
                 # Only update manifests if x is going to be sent. Otherwise we
                 # end up with bogus linkrevs specified for manifests and
                 # we skip some manifest nodes that we should otherwise
                 # have sent.
                 if (
                     x in self._fullclnodes
                     or cl.rev(x) in self._precomputedellipsis
                 ):
 
                     manifestnode = c.manifest
                     # Record the first changeset introducing this manifest
                     # version.
                     manifests.setdefault(manifestnode, x)
                     # Set this narrow-specific dict so we have the lowest
                     # manifest revnum to look up for this cl revnum. (Part of
                     # mapping changelog ellipsis parents to manifest ellipsis
                     # parents)
                     clrevtomanifestrev.setdefault(
                         cl.rev(x), mfl.rev(manifestnode)
                     )
                 # We can't trust the changed files list in the changeset if the
                 # client requested a shallow clone.
                 if self._isshallow:
                     changedfiles.update(mfl[c.manifest].read().keys())
                 else:
                     changedfiles.update(c.files)
             else:
                 # record the first changeset introducing this manifest version
                 manifests.setdefault(c.manifest, x)
                 # Record a complete list of potentially-changed files in
                 # this manifest.
                 changedfiles.update(c.files)
 
             return x
 
         gen = deltagroup(
             self._repo,
             cl,
             nodes,
             True,
             lookupcl,
             self._forcedeltaparentprev,
             ellipses=self._ellipses,
             topic=_(b'changesets'),
             clrevtolocalrev={},
             fullclnodes=self._fullclnodes,
             precomputedellipsis=self._precomputedellipsis,
         )
 
         return state, gen
 
     def generatemanifests(
         self,
         commonrevs,
         clrevorder,
         fastpathlinkrev,
         manifests,
         fnodes,
         source,
         clrevtolocalrev,
     ):
         """Returns an iterator of changegroup chunks containing manifests.
 
         `source` is unused here, but is used by extensions like remotefilelog to
         change what is sent based in pulls vs pushes, etc.
         """
         repo = self._repo
         mfl = repo.manifestlog
         tmfnodes = {b'': manifests}
 
         # Callback for the manifest, used to collect linkrevs for filelog
         # revisions.
         # Returns the linkrev node (collected in lookupcl).
         def makelookupmflinknode(tree, nodes):
             if fastpathlinkrev:
                 assert not tree
                 return (
                     manifests.__getitem__
                 )  # pytype: disable=unsupported-operands
 
             def lookupmflinknode(x):
                 """Callback for looking up the linknode for manifests.
 
                 Returns the linkrev node for the specified manifest.
 
                 SIDE EFFECT:
 
                 1) fclnodes gets populated with the list of relevant
                    file nodes if we're not using fastpathlinkrev
                 2) When treemanifests are in use, collects treemanifest nodes
                    to send
 
                 Note that this means manifests must be completely sent to
                 the client before you can trust the list of files and
                 treemanifests to send.
                 """
                 clnode = nodes[x]
                 mdata = mfl.get(tree, x).readfast(shallow=True)
                 for p, n, fl in mdata.iterentries():
                     if fl == b't':  # subdirectory manifest
                         subtree = tree + p + b'/'
                         tmfclnodes = tmfnodes.setdefault(subtree, {})
                         tmfclnode = tmfclnodes.setdefault(n, clnode)
                         if clrevorder[clnode] < clrevorder[tmfclnode]:
                             tmfclnodes[n] = clnode
                     else:
                         f = tree + p
                         fclnodes = fnodes.setdefault(f, {})
                         fclnode = fclnodes.setdefault(n, clnode)
                         if clrevorder[clnode] < clrevorder[fclnode]:
                             fclnodes[n] = clnode
                 return clnode
 
             return lookupmflinknode
 
         while tmfnodes:
             tree, nodes = tmfnodes.popitem()
 
             should_visit = self._matcher.visitdir(tree[:-1])
             if tree and not should_visit:
                 continue
 
             store = mfl.getstorage(tree)
 
             if not should_visit:
                 # No nodes to send because this directory is out of
                 # the client's view of the repository (probably
                 # because of narrow clones). Do this even for the root
                 # directory (tree=='')
                 prunednodes = []
             else:
                 # Avoid sending any manifest nodes we can prove the
                 # client already has by checking linkrevs. See the
                 # related comment in generatefiles().
                 prunednodes = self._prunemanifests(store, nodes, commonrevs)
 
             if tree and not prunednodes:
                 continue
 
             lookupfn = makelookupmflinknode(tree, nodes)
 
             deltas = deltagroup(
                 self._repo,
                 store,
                 prunednodes,
                 False,
                 lookupfn,
                 self._forcedeltaparentprev,
                 ellipses=self._ellipses,
                 topic=_(b'manifests'),
                 clrevtolocalrev=clrevtolocalrev,
                 fullclnodes=self._fullclnodes,
                 precomputedellipsis=self._precomputedellipsis,
             )
 
             if not self._oldmatcher.visitdir(store.tree[:-1]):
                 yield tree, deltas
             else:
                 # 'deltas' is a generator and we need to consume it even if
                 # we are not going to send it because a side-effect is that
                 # it updates tmdnodes (via lookupfn)
                 for d in deltas:
                     pass
                 if not tree:
                     yield tree, []
 
     def _prunemanifests(self, store, nodes, commonrevs):
         if not self._ellipses:
             # In non-ellipses case and large repositories, it is better to
             # prevent calling of store.rev and store.linkrev on a lot of
             # nodes as compared to sending some extra data
             return nodes.copy()
         # This is split out as a separate method to allow filtering
         # commonrevs in extension code.
         #
         # TODO(augie): this shouldn't be required, instead we should
         # make filtering of revisions to send delegated to the store
         # layer.
         frev, flr = store.rev, store.linkrev
         return [n for n in nodes if flr(frev(n)) not in commonrevs]
 
     # The 'source' parameter is useful for extensions
     def generatefiles(
         self,
         changedfiles,
         commonrevs,
         source,
         mfdicts,
         fastpathlinkrev,
         fnodes,
         clrevs,
     ):
         changedfiles = [
             f
             for f in changedfiles
             if self._matcher(f) and not self._oldmatcher(f)
         ]
 
         if not fastpathlinkrev:
 
             def normallinknodes(unused, fname):
                 return fnodes.get(fname, {})
 
         else:
             cln = self._repo.changelog.node
 
             def normallinknodes(store, fname):
                 flinkrev = store.linkrev
                 fnode = store.node
                 revs = ((r, flinkrev(r)) for r in store)
                 return {fnode(r): cln(lr) for r, lr in revs if lr in clrevs}
 
         clrevtolocalrev = {}
 
         if self._isshallow:
             # In a shallow clone, the linknodes callback needs to also include
             # those file nodes that are in the manifests we sent but weren't
             # introduced by those manifests.
             commonctxs = [self._repo[c] for c in commonrevs]
             clrev = self._repo.changelog.rev
 
             def linknodes(flog, fname):
                 for c in commonctxs:
                     try:
                         fnode = c.filenode(fname)
                         clrevtolocalrev[c.rev()] = flog.rev(fnode)
                     except error.ManifestLookupError:
                         pass
                 links = normallinknodes(flog, fname)
                 if len(links) != len(mfdicts):
                     for mf, lr in mfdicts:
                         fnode = mf.get(fname, None)
                         if fnode in links:
                             links[fnode] = min(links[fnode], lr, key=clrev)
                         elif fnode:
                             links[fnode] = lr
                 return links
 
         else:
             linknodes = normallinknodes
 
         repo = self._repo
         progress = repo.ui.makeprogress(
             _(b'files'), unit=_(b'files'), total=len(changedfiles)
         )
         for i, fname in enumerate(sorted(changedfiles)):
             filerevlog = repo.file(fname)
             if not filerevlog:
                 raise error.Abort(
                     _(b"empty or missing file data for %s") % fname
                 )
 
             clrevtolocalrev.clear()
 
             linkrevnodes = linknodes(filerevlog, fname)
             # Lookup for filenodes, we collected the linkrev nodes above in the
             # fastpath case and with lookupmf in the slowpath case.
             def lookupfilelog(x):
                 return linkrevnodes[x]
 
             frev, flr = filerevlog.rev, filerevlog.linkrev
             # Skip sending any filenode we know the client already
             # has. This avoids over-sending files relatively
             # inexpensively, so it's not a problem if we under-filter
             # here.
             filenodes = [
                 n for n in linkrevnodes if flr(frev(n)) not in commonrevs
             ]
 
             if not filenodes:
                 continue
 
             progress.update(i + 1, item=fname)
 
             deltas = deltagroup(
                 self._repo,
                 filerevlog,
                 filenodes,
                 False,
                 lookupfilelog,
                 self._forcedeltaparentprev,
                 ellipses=self._ellipses,
                 clrevtolocalrev=clrevtolocalrev,
                 fullclnodes=self._fullclnodes,
                 precomputedellipsis=self._precomputedellipsis,
             )
 
             yield fname, deltas
 
         progress.complete()
 
 
 def _makecg1packer(
     repo,
     oldmatcher,
     matcher,
     bundlecaps,
     ellipses=False,
     shallow=False,
     ellipsisroots=None,
     fullnodes=None,
 ):
     builddeltaheader = lambda d: _CHANGEGROUPV1_DELTA_HEADER.pack(
         d.node, d.p1node, d.p2node, d.linknode
     )
 
     return cgpacker(
         repo,
         oldmatcher,
         matcher,
         b'01',
         builddeltaheader=builddeltaheader,
         manifestsend=b'',
         forcedeltaparentprev=True,
         bundlecaps=bundlecaps,
         ellipses=ellipses,
         shallow=shallow,
         ellipsisroots=ellipsisroots,
         fullnodes=fullnodes,
     )
 
 
 def _makecg2packer(
     repo,
     oldmatcher,
     matcher,
     bundlecaps,
     ellipses=False,
     shallow=False,
     ellipsisroots=None,
     fullnodes=None,
 ):
     builddeltaheader = lambda d: _CHANGEGROUPV2_DELTA_HEADER.pack(
         d.node, d.p1node, d.p2node, d.basenode, d.linknode
     )
 
     return cgpacker(
         repo,
         oldmatcher,
         matcher,
         b'02',
         builddeltaheader=builddeltaheader,
         manifestsend=b'',
         bundlecaps=bundlecaps,
         ellipses=ellipses,
         shallow=shallow,
         ellipsisroots=ellipsisroots,
         fullnodes=fullnodes,
     )
 
 
 def _makecg3packer(
     repo,
     oldmatcher,
     matcher,
     bundlecaps,
     ellipses=False,
     shallow=False,
     ellipsisroots=None,
     fullnodes=None,
 ):
     builddeltaheader = lambda d: _CHANGEGROUPV3_DELTA_HEADER.pack(
         d.node, d.p1node, d.p2node, d.basenode, d.linknode, d.flags
     )
 
     return cgpacker(
         repo,
         oldmatcher,
         matcher,
         b'03',
         builddeltaheader=builddeltaheader,
         manifestsend=closechunk(),
         bundlecaps=bundlecaps,
         ellipses=ellipses,
         shallow=shallow,
         ellipsisroots=ellipsisroots,
         fullnodes=fullnodes,
     )
 
 
 _packermap = {
     b'01': (_makecg1packer, cg1unpacker),
     # cg2 adds support for exchanging generaldelta
     b'02': (_makecg2packer, cg2unpacker),
     # cg3 adds support for exchanging revlog flags and treemanifests
     b'03': (_makecg3packer, cg3unpacker),
 }
 
 
 def allsupportedversions(repo):
     versions = set(_packermap.keys())
     needv03 = False
     if (
         repo.ui.configbool(b'experimental', b'changegroup3')
         or repo.ui.configbool(b'experimental', b'treemanifest')
         or scmutil.istreemanifest(repo)
     ):
         # we keep version 03 because we need to to exchange treemanifest data
         #
         # we also keep vresion 01 and 02, because it is possible for repo to
         # contains both normal and tree manifest at the same time. so using
         # older version to pull data is viable
         #
         # (or even to push subset of history)
         needv03 = True
     if b'exp-sidedata-flag' in repo.requirements:
         needv03 = True
         # don't attempt to use 01/02 until we do sidedata cleaning
         versions.discard(b'01')
         versions.discard(b'02')
     if not needv03:
         versions.discard(b'03')
     return versions
 
 
 # Changegroup versions that can be applied to the repo
 def supportedincomingversions(repo):
     return allsupportedversions(repo)
 
 
 # Changegroup versions that can be created from the repo
 def supportedoutgoingversions(repo):
     versions = allsupportedversions(repo)
     if scmutil.istreemanifest(repo):
         # Versions 01 and 02 support only flat manifests and it's just too
         # expensive to convert between the flat manifest and tree manifest on
         # the fly. Since tree manifests are hashed differently, all of history
         # would have to be converted. Instead, we simply don't even pretend to
         # support versions 01 and 02.
         versions.discard(b'01')
         versions.discard(b'02')
     if requirements.NARROW_REQUIREMENT in repo.requirements:
         # Versions 01 and 02 don't support revlog flags, and we need to
         # support that for stripping and unbundling to work.
         versions.discard(b'01')
         versions.discard(b'02')
     if LFS_REQUIREMENT in repo.requirements:
         # Versions 01 and 02 don't support revlog flags, and we need to
         # mark LFS entries with REVIDX_EXTSTORED.
         versions.discard(b'01')
         versions.discard(b'02')
 
     return versions
 
 
 def localversion(repo):
     # Finds the best version to use for bundles that are meant to be used
     # locally, such as those from strip and shelve, and temporary bundles.
     return max(supportedoutgoingversions(repo))
 
 
 def safeversion(repo):
     # Finds the smallest version that it's safe to assume clients of the repo
     # will support. For example, all hg versions that support generaldelta also
     # support changegroup 02.
     versions = supportedoutgoingversions(repo)
     if b'generaldelta' in repo.requirements:
         versions.discard(b'01')
     assert versions
     return min(versions)
 
 
 def getbundler(
     version,
     repo,
     bundlecaps=None,
     oldmatcher=None,
     matcher=None,
     ellipses=False,
     shallow=False,
     ellipsisroots=None,
     fullnodes=None,
 ):
     assert version in supportedoutgoingversions(repo)
 
     if matcher is None:
         matcher = matchmod.always()
     if oldmatcher is None:
         oldmatcher = matchmod.never()
 
     if version == b'01' and not matcher.always():
         raise error.ProgrammingError(
             b'version 01 changegroups do not support sparse file matchers'
         )
 
     if ellipses and version in (b'01', b'02'):
         raise error.Abort(
             _(
                 b'ellipsis nodes require at least cg3 on client and server, '
                 b'but negotiated version %s'
             )
             % version
         )
 
     # Requested files could include files not in the local store. So
     # filter those out.
     matcher = repo.narrowmatch(matcher)
 
     fn = _packermap[version][0]
     return fn(
         repo,
         oldmatcher,
         matcher,
         bundlecaps,
         ellipses=ellipses,
         shallow=shallow,
         ellipsisroots=ellipsisroots,
         fullnodes=fullnodes,
     )
 
 
 def getunbundler(version, fh, alg, extras=None):
     return _packermap[version][1](fh, alg, extras=extras)
 
 
 def _changegroupinfo(repo, nodes, source):
     if repo.ui.verbose or source == b'bundle':
         repo.ui.status(_(b"%d changesets found\n") % len(nodes))
     if repo.ui.debugflag:
         repo.ui.debug(b"list of changesets:\n")
         for node in nodes:
             repo.ui.debug(b"%s\n" % hex(node))
 
 
 def makechangegroup(
     repo, outgoing, version, source, fastpath=False, bundlecaps=None
 ):
     cgstream = makestream(
         repo,
         outgoing,
         version,
         source,
         fastpath=fastpath,
         bundlecaps=bundlecaps,
     )
     return getunbundler(
         version,
         util.chunkbuffer(cgstream),
         None,
         {b'clcount': len(outgoing.missing)},
     )
 
 
 def makestream(
     repo,
     outgoing,
     version,
     source,
     fastpath=False,
     bundlecaps=None,
     matcher=None,
 ):
     bundler = getbundler(version, repo, bundlecaps=bundlecaps, matcher=matcher)
 
     repo = repo.unfiltered()
     commonrevs = outgoing.common
     csets = outgoing.missing
     heads = outgoing.ancestorsof
     # We go through the fast path if we get told to, or if all (unfiltered
     # heads have been requested (since we then know there all linkrevs will
     # be pulled by the client).
     heads.sort()
     fastpathlinkrev = fastpath or (
         repo.filtername is None and heads == sorted(repo.heads())
     )
 
     repo.hook(b'preoutgoing', throw=True, source=source)
     _changegroupinfo(repo, csets, source)
     return bundler.generate(commonrevs, csets, fastpathlinkrev, source)
 
 
 def _addchangegroupfiles(repo, source, revmap, trp, expectedfiles, needfiles):
     revisions = 0
     files = 0
     progress = repo.ui.makeprogress(
         _(b'files'), unit=_(b'files'), total=expectedfiles
     )
     for chunkdata in iter(source.filelogheader, {}):
         files += 1
         f = chunkdata[b"filename"]
         repo.ui.debug(b"adding %s revisions\n" % f)
         progress.increment()
         fl = repo.file(f)
         o = len(fl)
         try:
             deltas = source.deltaiter()
             if not fl.addgroup(deltas, revmap, trp):
                 raise error.Abort(_(b"received file revlog group is empty"))
         except error.CensoredBaseError as e:
             raise error.Abort(_(b"received delta base is censored: %s") % e)
         revisions += len(fl) - o
         if f in needfiles:
             needs = needfiles[f]
             for new in pycompat.xrange(o, len(fl)):
                 n = fl.node(new)
                 if n in needs:
                     needs.remove(n)
                 else:
                     raise error.Abort(_(b"received spurious file revlog entry"))
             if not needs:
                 del needfiles[f]
     progress.complete()
 
     for f, needs in pycompat.iteritems(needfiles):
         fl = repo.file(f)
         for n in needs:
             try:
                 fl.rev(n)
             except error.LookupError:
                 raise error.Abort(
                     _(b'missing file data for %s:%s - run hg verify')
                     % (f, hex(n))
                 )
 
     return revisions, files

mercurial/exchangev2.py

@@ -1,798 +1,799 b''
 # exchangev2.py - repository exchange for wire protocol version 2
 #
 # Copyright 2018 Gregory Szorc <[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 collections
 import weakref
 
 from .i18n import _
 from .node import (
     nullid,
     short,
 )
 from . import (
     bookmarks,
     error,
     mdiff,
     narrowspec,
     phases,
     pycompat,
     setdiscovery,
 )
 from .interfaces import repository
 
 
 def pull(pullop):
     """Pull using wire protocol version 2."""
     repo = pullop.repo
     remote = pullop.remote
 
     usingrawchangelogandmanifest = _checkuserawstorefiledata(pullop)
 
     # If this is a clone and it was requested to perform a "stream clone",
     # we obtain the raw files data from the remote then fall back to an
     # incremental pull. This is somewhat hacky and is not nearly robust enough
     # for long-term usage.
     if usingrawchangelogandmanifest:
         with repo.transaction(b'clone'):
             _fetchrawstorefiles(repo, remote)
             repo.invalidate(clearfilecache=True)
 
     tr = pullop.trmanager.transaction()
 
     # We don't use the repo's narrow matcher here because the patterns passed
     # to exchange.pull() could be different.
     narrowmatcher = narrowspec.match(
         repo.root,
         # Empty maps to nevermatcher. So always
         # set includes if missing.
         pullop.includepats or {b'path:.'},
         pullop.excludepats,
     )
 
     if pullop.includepats or pullop.excludepats:
         pathfilter = {}
         if pullop.includepats:
             pathfilter[b'include'] = sorted(pullop.includepats)
         if pullop.excludepats:
             pathfilter[b'exclude'] = sorted(pullop.excludepats)
     else:
         pathfilter = None
 
     # Figure out what needs to be fetched.
     common, fetch, remoteheads = _pullchangesetdiscovery(
         repo, remote, pullop.heads, abortwhenunrelated=pullop.force
     )
 
     # And fetch the data.
     pullheads = pullop.heads or remoteheads
     csetres = _fetchchangesets(repo, tr, remote, common, fetch, pullheads)
 
     # New revisions are written to the changelog. But all other updates
     # are deferred. Do those now.
 
     # Ensure all new changesets are draft by default. If the repo is
     # publishing, the phase will be adjusted by the loop below.
     if csetres[b'added']:
         phases.registernew(
             repo, tr, phases.draft, [repo[n].rev() for n in csetres[b'added']]
         )
 
     # And adjust the phase of all changesets accordingly.
     for phasenumber, phase in phases.phasenames.items():
         if phase == b'secret' or not csetres[b'nodesbyphase'][phase]:
             continue
 
         phases.advanceboundary(
             repo,
             tr,
             phasenumber,
             csetres[b'nodesbyphase'][phase],
         )
 
     # Write bookmark updates.
     bookmarks.updatefromremote(
         repo.ui,
         repo,
         csetres[b'bookmarks'],
         remote.url(),
         pullop.gettransaction,
         explicit=pullop.explicitbookmarks,
     )
 
     manres = _fetchmanifests(repo, tr, remote, csetres[b'manifestnodes'])
 
     # We don't properly support shallow changeset and manifest yet. So we apply
     # depth limiting locally.
     if pullop.depth:
         relevantcsetnodes = set()
         clnode = repo.changelog.node
 
         for rev in repo.revs(
             b'ancestors(%ln, %s)', pullheads, pullop.depth - 1
         ):
             relevantcsetnodes.add(clnode(rev))
 
         csetrelevantfilter = lambda n: n in relevantcsetnodes
 
     else:
         csetrelevantfilter = lambda n: True
 
     # If obtaining the raw store files, we need to scan the full repo to
     # derive all the changesets, manifests, and linkrevs.
     if usingrawchangelogandmanifest:
         csetsforfiles = []
         mnodesforfiles = []
         manifestlinkrevs = {}
 
         for rev in repo:
             ctx = repo[rev]
             node = ctx.node()
 
             if not csetrelevantfilter(node):
                 continue
 
             mnode = ctx.manifestnode()
 
             csetsforfiles.append(node)
             mnodesforfiles.append(mnode)
             manifestlinkrevs[mnode] = rev
 
     else:
         csetsforfiles = [n for n in csetres[b'added'] if csetrelevantfilter(n)]
         mnodesforfiles = manres[b'added']
         manifestlinkrevs = manres[b'linkrevs']
 
     # Find all file nodes referenced by added manifests and fetch those
     # revisions.
     fnodes = _derivefilesfrommanifests(repo, narrowmatcher, mnodesforfiles)
     _fetchfilesfromcsets(
         repo,
         tr,
         remote,
         pathfilter,
         fnodes,
         csetsforfiles,
         manifestlinkrevs,
         shallow=bool(pullop.depth),
     )
 
 
 def _checkuserawstorefiledata(pullop):
     """Check whether we should use rawstorefiledata command to retrieve data."""
 
     repo = pullop.repo
     remote = pullop.remote
 
     # Command to obtain raw store data isn't available.
     if b'rawstorefiledata' not in remote.apidescriptor[b'commands']:
         return False
 
     # Only honor if user requested stream clone operation.
     if not pullop.streamclonerequested:
         return False
 
     # Only works on empty repos.
     if len(repo):
         return False
 
     # TODO This is super hacky. There needs to be a storage API for this. We
     # also need to check for compatibility with the remote.
     if b'revlogv1' not in repo.requirements:
         return False
 
     return True
 
 
 def _fetchrawstorefiles(repo, remote):
     with remote.commandexecutor() as e:
         objs = e.callcommand(
             b'rawstorefiledata',
             {
                 b'files': [b'changelog', b'manifestlog'],
             },
         ).result()
 
         # First object is a summary of files data that follows.
         overall = next(objs)
 
         progress = repo.ui.makeprogress(
             _(b'clone'), total=overall[b'totalsize'], unit=_(b'bytes')
         )
         with progress:
             progress.update(0)
 
             # Next are pairs of file metadata, data.
             while True:
                 try:
                     filemeta = next(objs)
                 except StopIteration:
                     break
 
                 for k in (b'location', b'path', b'size'):
                     if k not in filemeta:
                         raise error.Abort(
                             _(b'remote file data missing key: %s') % k
                         )
 
                 if filemeta[b'location'] == b'store':
                     vfs = repo.svfs
                 else:
                     raise error.Abort(
                         _(b'invalid location for raw file data: %s')
                         % filemeta[b'location']
                     )
 
                 bytesremaining = filemeta[b'size']
 
                 with vfs.open(filemeta[b'path'], b'wb') as fh:
                     while True:
                         try:
                             chunk = next(objs)
                         except StopIteration:
                             break
 
                         bytesremaining -= len(chunk)
 
                         if bytesremaining < 0:
                             raise error.Abort(
                                 _(
                                     b'received invalid number of bytes for file '
                                     b'data; expected %d, got extra'
                                 )
                                 % filemeta[b'size']
                             )
 
                         progress.increment(step=len(chunk))
                         fh.write(chunk)
 
                         try:
                             if chunk.islast:
                                 break
                         except AttributeError:
                             raise error.Abort(
                                 _(
                                     b'did not receive indefinite length bytestring '
                                     b'for file data'
                                 )
                             )
 
                 if bytesremaining:
                     raise error.Abort(
                         _(
                             b'received invalid number of bytes for'
                             b'file data; expected %d got %d'
                         )
                         % (
                             filemeta[b'size'],
                             filemeta[b'size'] - bytesremaining,
                         )
                     )
 
 
 def _pullchangesetdiscovery(repo, remote, heads, abortwhenunrelated=True):
     """Determine which changesets need to be pulled."""
 
     if heads:
         knownnode = repo.changelog.hasnode
         if all(knownnode(head) for head in heads):
             return heads, False, heads
 
     # TODO wire protocol version 2 is capable of more efficient discovery
     # than setdiscovery. Consider implementing something better.
     common, fetch, remoteheads = setdiscovery.findcommonheads(
         repo.ui, repo, remote, abortwhenunrelated=abortwhenunrelated
     )
 
     common = set(common)
     remoteheads = set(remoteheads)
 
     # If a remote head is filtered locally, put it back in the common set.
     # See the comment in exchange._pulldiscoverychangegroup() for more.
 
     if fetch and remoteheads:
         has_node = repo.unfiltered().changelog.index.has_node
 
         common |= {head for head in remoteheads if has_node(head)}
 
         if set(remoteheads).issubset(common):
             fetch = []
 
     common.discard(nullid)
 
     return common, fetch, remoteheads
 
 
 def _fetchchangesets(repo, tr, remote, common, fetch, remoteheads):
     # TODO consider adding a step here where we obtain the DAG shape first
     # (or ask the server to slice changesets into chunks for us) so that
     # we can perform multiple fetches in batches. This will facilitate
     # resuming interrupted clones, higher server-side cache hit rates due
     # to smaller segments, etc.
     with remote.commandexecutor() as e:
         objs = e.callcommand(
             b'changesetdata',
             {
                 b'revisions': [
                     {
                         b'type': b'changesetdagrange',
                         b'roots': sorted(common),
                         b'heads': sorted(remoteheads),
                     }
                 ],
                 b'fields': {b'bookmarks', b'parents', b'phase', b'revision'},
             },
         ).result()
 
         # The context manager waits on all response data when exiting. So
         # we need to remain in the context manager in order to stream data.
         return _processchangesetdata(repo, tr, objs)
 
 
 def _processchangesetdata(repo, tr, objs):
     repo.hook(b'prechangegroup', throw=True, **pycompat.strkwargs(tr.hookargs))
 
     urepo = repo.unfiltered()
     cl = urepo.changelog
 
     cl.delayupdate(tr)
 
     # The first emitted object is a header describing the data that
     # follows.
     meta = next(objs)
 
     progress = repo.ui.makeprogress(
         _(b'changesets'), unit=_(b'chunks'), total=meta.get(b'totalitems')
     )
 
     manifestnodes = {}
     added = []
 
     def linkrev(node):
         repo.ui.debug(b'add changeset %s\n' % short(node))
         # Linkrev for changelog is always self.
         return len(cl)
 
     def ondupchangeset(cl, node):
         added.append(node)
 
     def onchangeset(cl, node):
         progress.increment()
 
         rev = cl.rev(node)
         revision = cl.changelogrevision(rev)
         added.append(node)
 
         # We need to preserve the mapping of changelog revision to node
         # so we can set the linkrev accordingly when manifests are added.
         manifestnodes[rev] = revision.manifest
 
         repo.register_changeset(rev, revision)
 
     nodesbyphase = {phase: set() for phase in phases.phasenames.values()}
     remotebookmarks = {}
 
     # addgroup() expects a 7-tuple describing revisions. This normalizes
     # the wire data to that format.
     #
     # This loop also aggregates non-revision metadata, such as phase
     # data.
     def iterrevisions():
         for cset in objs:
             node = cset[b'node']
 
             if b'phase' in cset:
                 nodesbyphase[cset[b'phase']].add(node)
 
             for mark in cset.get(b'bookmarks', []):
                 remotebookmarks[mark] = node
 
             # TODO add mechanism for extensions to examine records so they
             # can siphon off custom data fields.
 
             extrafields = {}
 
             for field, size in cset.get(b'fieldsfollowing', []):
                 extrafields[field] = next(objs)
 
             # Some entries might only be metadata only updates.
             if b'revision' not in extrafields:
                 continue
 
             data = extrafields[b'revision']
 
             yield (
                 node,
                 cset[b'parents'][0],
                 cset[b'parents'][1],
                 # Linknode is always itself for changesets.
                 cset[b'node'],
                 # We always send full revisions. So delta base is not set.
                 nullid,
                 mdiff.trivialdiffheader(len(data)) + data,
                 # Flags not yet supported.
                 0,
             )
 
     cl.addgroup(
         iterrevisions(),
         linkrev,
         weakref.proxy(tr),
+        alwayscache=True,
         addrevisioncb=onchangeset,
         duplicaterevisioncb=ondupchangeset,
     )
 
     progress.complete()
 
     return {
         b'added': added,
         b'nodesbyphase': nodesbyphase,
         b'bookmarks': remotebookmarks,
         b'manifestnodes': manifestnodes,
     }
 
 
 def _fetchmanifests(repo, tr, remote, manifestnodes):
     rootmanifest = repo.manifestlog.getstorage(b'')
 
     # Some manifests can be shared between changesets. Filter out revisions
     # we already know about.
     fetchnodes = []
     linkrevs = {}
     seen = set()
 
     for clrev, node in sorted(pycompat.iteritems(manifestnodes)):
         if node in seen:
             continue
 
         try:
             rootmanifest.rev(node)
         except error.LookupError:
             fetchnodes.append(node)
             linkrevs[node] = clrev
 
         seen.add(node)
 
     # TODO handle tree manifests
 
     # addgroup() expects 7-tuple describing revisions. This normalizes
     # the wire data to that format.
     def iterrevisions(objs, progress):
         for manifest in objs:
             node = manifest[b'node']
 
             extrafields = {}
 
             for field, size in manifest.get(b'fieldsfollowing', []):
                 extrafields[field] = next(objs)
 
             if b'delta' in extrafields:
                 basenode = manifest[b'deltabasenode']
                 delta = extrafields[b'delta']
             elif b'revision' in extrafields:
                 basenode = nullid
                 revision = extrafields[b'revision']
                 delta = mdiff.trivialdiffheader(len(revision)) + revision
             else:
                 continue
 
             yield (
                 node,
                 manifest[b'parents'][0],
                 manifest[b'parents'][1],
                 # The value passed in is passed to the lookup function passed
                 # to addgroup(). We already have a map of manifest node to
                 # changelog revision number. So we just pass in the
                 # manifest node here and use linkrevs.__getitem__ as the
                 # resolution function.
                 node,
                 basenode,
                 delta,
                 # Flags not yet supported.
                 0,
             )
 
             progress.increment()
 
     progress = repo.ui.makeprogress(
         _(b'manifests'), unit=_(b'chunks'), total=len(fetchnodes)
     )
 
     commandmeta = remote.apidescriptor[b'commands'][b'manifestdata']
     batchsize = commandmeta.get(b'recommendedbatchsize', 10000)
     # TODO make size configurable on client?
 
     # We send commands 1 at a time to the remote. This is not the most
     # efficient because we incur a round trip at the end of each batch.
     # However, the existing frame-based reactor keeps consuming server
     # data in the background. And this results in response data buffering
     # in memory. This can consume gigabytes of memory.
     # TODO send multiple commands in a request once background buffering
     # issues are resolved.
 
     added = []
 
     for i in pycompat.xrange(0, len(fetchnodes), batchsize):
         batch = [node for node in fetchnodes[i : i + batchsize]]
         if not batch:
             continue
 
         with remote.commandexecutor() as e:
             objs = e.callcommand(
                 b'manifestdata',
                 {
                     b'tree': b'',
                     b'nodes': batch,
                     b'fields': {b'parents', b'revision'},
                     b'haveparents': True,
                 },
             ).result()
 
             # Chomp off header object.
             next(objs)
 
             def onchangeset(cl, node):
                 added.append(node)
 
             rootmanifest.addgroup(
                 iterrevisions(objs, progress),
                 linkrevs.__getitem__,
                 weakref.proxy(tr),
                 addrevisioncb=onchangeset,
                 duplicaterevisioncb=onchangeset,
             )
 
     progress.complete()
 
     return {
         b'added': added,
         b'linkrevs': linkrevs,
     }
 
 
 def _derivefilesfrommanifests(repo, matcher, manifestnodes):
     """Determine what file nodes are relevant given a set of manifest nodes.
 
     Returns a dict mapping file paths to dicts of file node to first manifest
     node.
     """
     ml = repo.manifestlog
     fnodes = collections.defaultdict(dict)
 
     progress = repo.ui.makeprogress(
         _(b'scanning manifests'), total=len(manifestnodes)
     )
 
     with progress:
         for manifestnode in manifestnodes:
             m = ml.get(b'', manifestnode)
 
             # TODO this will pull in unwanted nodes because it takes the storage
             # delta into consideration. What we really want is something that
             # takes the delta between the manifest's parents. And ideally we
             # would ignore file nodes that are known locally. For now, ignore
             # both these limitations. This will result in incremental fetches
             # requesting data we already have. So this is far from ideal.
             md = m.readfast()
 
             for path, fnode in md.items():
                 if matcher(path):
                     fnodes[path].setdefault(fnode, manifestnode)
 
             progress.increment()
 
     return fnodes
 
 
 def _fetchfiles(repo, tr, remote, fnodes, linkrevs):
     """Fetch file data from explicit file revisions."""
 
     def iterrevisions(objs, progress):
         for filerevision in objs:
             node = filerevision[b'node']
 
             extrafields = {}
 
             for field, size in filerevision.get(b'fieldsfollowing', []):
                 extrafields[field] = next(objs)
 
             if b'delta' in extrafields:
                 basenode = filerevision[b'deltabasenode']
                 delta = extrafields[b'delta']
             elif b'revision' in extrafields:
                 basenode = nullid
                 revision = extrafields[b'revision']
                 delta = mdiff.trivialdiffheader(len(revision)) + revision
             else:
                 continue
 
             yield (
                 node,
                 filerevision[b'parents'][0],
                 filerevision[b'parents'][1],
                 node,
                 basenode,
                 delta,
                 # Flags not yet supported.
                 0,
             )
 
             progress.increment()
 
     progress = repo.ui.makeprogress(
         _(b'files'),
         unit=_(b'chunks'),
         total=sum(len(v) for v in pycompat.itervalues(fnodes)),
     )
 
     # TODO make batch size configurable
     batchsize = 10000
     fnodeslist = [x for x in sorted(fnodes.items())]
 
     for i in pycompat.xrange(0, len(fnodeslist), batchsize):
         batch = [x for x in fnodeslist[i : i + batchsize]]
         if not batch:
             continue
 
         with remote.commandexecutor() as e:
             fs = []
             locallinkrevs = {}
 
             for path, nodes in batch:
                 fs.append(
                     (
                         path,
                         e.callcommand(
                             b'filedata',
                             {
                                 b'path': path,
                                 b'nodes': sorted(nodes),
                                 b'fields': {b'parents', b'revision'},
                                 b'haveparents': True,
                             },
                         ),
                     )
                 )
 
                 locallinkrevs[path] = {
                     node: linkrevs[manifestnode]
                     for node, manifestnode in pycompat.iteritems(nodes)
                 }
 
             for path, f in fs:
                 objs = f.result()
 
                 # Chomp off header objects.
                 next(objs)
 
                 store = repo.file(path)
                 store.addgroup(
                     iterrevisions(objs, progress),
                     locallinkrevs[path].__getitem__,
                     weakref.proxy(tr),
                 )
 
 
 def _fetchfilesfromcsets(
     repo, tr, remote, pathfilter, fnodes, csets, manlinkrevs, shallow=False
 ):
     """Fetch file data from explicit changeset revisions."""
 
     def iterrevisions(objs, remaining, progress):
         while remaining:
             filerevision = next(objs)
 
             node = filerevision[b'node']
 
             extrafields = {}
 
             for field, size in filerevision.get(b'fieldsfollowing', []):
                 extrafields[field] = next(objs)
 
             if b'delta' in extrafields:
                 basenode = filerevision[b'deltabasenode']
                 delta = extrafields[b'delta']
             elif b'revision' in extrafields:
                 basenode = nullid
                 revision = extrafields[b'revision']
                 delta = mdiff.trivialdiffheader(len(revision)) + revision
             else:
                 continue
 
             if b'linknode' in filerevision:
                 linknode = filerevision[b'linknode']
             else:
                 linknode = node
 
             yield (
                 node,
                 filerevision[b'parents'][0],
                 filerevision[b'parents'][1],
                 linknode,
                 basenode,
                 delta,
                 # Flags not yet supported.
                 0,
             )
 
             progress.increment()
             remaining -= 1
 
     progress = repo.ui.makeprogress(
         _(b'files'),
         unit=_(b'chunks'),
         total=sum(len(v) for v in pycompat.itervalues(fnodes)),
     )
 
     commandmeta = remote.apidescriptor[b'commands'][b'filesdata']
     batchsize = commandmeta.get(b'recommendedbatchsize', 50000)
 
     shallowfiles = repository.REPO_FEATURE_SHALLOW_FILE_STORAGE in repo.features
     fields = {b'parents', b'revision'}
     clrev = repo.changelog.rev
 
     # There are no guarantees that we'll have ancestor revisions if
     # a) this repo has shallow file storage b) shallow data fetching is enabled.
     # Force remote to not delta against possibly unknown revisions when these
     # conditions hold.
     haveparents = not (shallowfiles or shallow)
 
     # Similarly, we may not have calculated linkrevs for all incoming file
     # revisions. Ask the remote to do work for us in this case.
     if not haveparents:
         fields.add(b'linknode')
 
     for i in pycompat.xrange(0, len(csets), batchsize):
         batch = [x for x in csets[i : i + batchsize]]
         if not batch:
             continue
 
         with remote.commandexecutor() as e:
             args = {
                 b'revisions': [
                     {
                         b'type': b'changesetexplicit',
                         b'nodes': batch,
                     }
                 ],
                 b'fields': fields,
                 b'haveparents': haveparents,
             }
 
             if pathfilter:
                 args[b'pathfilter'] = pathfilter
 
             objs = e.callcommand(b'filesdata', args).result()
 
             # First object is an overall header.
             overall = next(objs)
 
             # We have overall['totalpaths'] segments.
             for i in pycompat.xrange(overall[b'totalpaths']):
                 header = next(objs)
 
                 path = header[b'path']
                 store = repo.file(path)
 
                 linkrevs = {
                     fnode: manlinkrevs[mnode]
                     for fnode, mnode in pycompat.iteritems(fnodes[path])
                 }
 
                 def getlinkrev(node):
                     if node in linkrevs:
                         return linkrevs[node]
                     else:
                         return clrev(node)
 
                 store.addgroup(
                     iterrevisions(objs, header[b'totalitems'], progress),
                     getlinkrev,
                     weakref.proxy(tr),
                     maybemissingparents=shallow,
                 )

mercurial/interfaces/repository.py

@@ -1,1987 +1,1993 b''
 # repository.py - Interfaces and base classes for repositories and peers.
 #
 # Copyright 2017 Gregory Szorc <[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
 
 from ..i18n import _
 from .. import error
 from . import util as interfaceutil
 
 # Local repository feature string.
 
 # Revlogs are being used for file storage.
 REPO_FEATURE_REVLOG_FILE_STORAGE = b'revlogfilestorage'
 # The storage part of the repository is shared from an external source.
 REPO_FEATURE_SHARED_STORAGE = b'sharedstore'
 # LFS supported for backing file storage.
 REPO_FEATURE_LFS = b'lfs'
 # Repository supports being stream cloned.
 REPO_FEATURE_STREAM_CLONE = b'streamclone'
 # Files storage may lack data for all ancestors.
 REPO_FEATURE_SHALLOW_FILE_STORAGE = b'shallowfilestorage'
 
 REVISION_FLAG_CENSORED = 1 << 15
 REVISION_FLAG_ELLIPSIS = 1 << 14
 REVISION_FLAG_EXTSTORED = 1 << 13
 REVISION_FLAG_SIDEDATA = 1 << 12
 REVISION_FLAG_HASCOPIESINFO = 1 << 11
 
 REVISION_FLAGS_KNOWN = (
     REVISION_FLAG_CENSORED
     | REVISION_FLAG_ELLIPSIS
     | REVISION_FLAG_EXTSTORED
     | REVISION_FLAG_SIDEDATA
     | REVISION_FLAG_HASCOPIESINFO
 )
 
 CG_DELTAMODE_STD = b'default'
 CG_DELTAMODE_PREV = b'previous'
 CG_DELTAMODE_FULL = b'fulltext'
 CG_DELTAMODE_P1 = b'p1'
 
 
 class ipeerconnection(interfaceutil.Interface):
     """Represents a "connection" to a repository.
 
     This is the base interface for representing a connection to a repository.
     It holds basic properties and methods applicable to all peer types.
 
     This is not a complete interface definition and should not be used
     outside of this module.
     """
 
     ui = interfaceutil.Attribute("""ui.ui instance""")
 
     def url():
         """Returns a URL string representing this peer.
 
         Currently, implementations expose the raw URL used to construct the
         instance. It may contain credentials as part of the URL. The
         expectations of the value aren't well-defined and this could lead to
         data leakage.
 
         TODO audit/clean consumers and more clearly define the contents of this
         value.
         """
 
     def local():
         """Returns a local repository instance.
 
         If the peer represents a local repository, returns an object that
         can be used to interface with it. Otherwise returns ``None``.
         """
 
     def peer():
         """Returns an object conforming to this interface.
 
         Most implementations will ``return self``.
         """
 
     def canpush():
         """Returns a boolean indicating if this peer can be pushed to."""
 
     def close():
         """Close the connection to this peer.
 
         This is called when the peer will no longer be used. Resources
         associated with the peer should be cleaned up.
         """
 
 
 class ipeercapabilities(interfaceutil.Interface):
     """Peer sub-interface related to capabilities."""
 
     def capable(name):
         """Determine support for a named capability.
 
         Returns ``False`` if capability not supported.
 
         Returns ``True`` if boolean capability is supported. Returns a string
         if capability support is non-boolean.
 
         Capability strings may or may not map to wire protocol capabilities.
         """
 
     def requirecap(name, purpose):
         """Require a capability to be present.
 
         Raises a ``CapabilityError`` if the capability isn't present.
         """
 
 
 class ipeercommands(interfaceutil.Interface):
     """Client-side interface for communicating over the wire protocol.
 
     This interface is used as a gateway to the Mercurial wire protocol.
     methods commonly call wire protocol commands of the same name.
     """
 
     def branchmap():
         """Obtain heads in named branches.
 
         Returns a dict mapping branch name to an iterable of nodes that are
         heads on that branch.
         """
 
     def capabilities():
         """Obtain capabilities of the peer.
 
         Returns a set of string capabilities.
         """
 
     def clonebundles():
         """Obtains the clone bundles manifest for the repo.
 
         Returns the manifest as unparsed bytes.
         """
 
     def debugwireargs(one, two, three=None, four=None, five=None):
         """Used to facilitate debugging of arguments passed over the wire."""
 
     def getbundle(source, **kwargs):
         """Obtain remote repository data as a bundle.
 
         This command is how the bulk of repository data is transferred from
         the peer to the local repository
 
         Returns a generator of bundle data.
         """
 
     def heads():
         """Determine all known head revisions in the peer.
 
         Returns an iterable of binary nodes.
         """
 
     def known(nodes):
         """Determine whether multiple nodes are known.
 
         Accepts an iterable of nodes whose presence to check for.
 
         Returns an iterable of booleans indicating of the corresponding node
         at that index is known to the peer.
         """
 
     def listkeys(namespace):
         """Obtain all keys in a pushkey namespace.
 
         Returns an iterable of key names.
         """
 
     def lookup(key):
         """Resolve a value to a known revision.
 
         Returns a binary node of the resolved revision on success.
         """
 
     def pushkey(namespace, key, old, new):
         """Set a value using the ``pushkey`` protocol.
 
         Arguments correspond to the pushkey namespace and key to operate on and
         the old and new values for that key.
 
         Returns a string with the peer result. The value inside varies by the
         namespace.
         """
 
     def stream_out():
         """Obtain streaming clone data.
 
         Successful result should be a generator of data chunks.
         """
 
     def unbundle(bundle, heads, url):
         """Transfer repository data to the peer.
 
         This is how the bulk of data during a push is transferred.
 
         Returns the integer number of heads added to the peer.
         """
 
 
 class ipeerlegacycommands(interfaceutil.Interface):
     """Interface for implementing support for legacy wire protocol commands.
 
     Wire protocol commands transition to legacy status when they are no longer
     used by modern clients. To facilitate identifying which commands are
     legacy, the interfaces are split.
     """
 
     def between(pairs):
         """Obtain nodes between pairs of nodes.
 
         ``pairs`` is an iterable of node pairs.
 
         Returns an iterable of iterables of nodes corresponding to each
         requested pair.
         """
 
     def branches(nodes):
         """Obtain ancestor changesets of specific nodes back to a branch point.
 
         For each requested node, the peer finds the first ancestor node that is
         a DAG root or is a merge.
 
         Returns an iterable of iterables with the resolved values for each node.
         """
 
     def changegroup(nodes, source):
         """Obtain a changegroup with data for descendants of specified nodes."""
 
     def changegroupsubset(bases, heads, source):
         pass
 
 
 class ipeercommandexecutor(interfaceutil.Interface):
     """Represents a mechanism to execute remote commands.
 
     This is the primary interface for requesting that wire protocol commands
     be executed. Instances of this interface are active in a context manager
     and have a well-defined lifetime. When the context manager exits, all
     outstanding requests are waited on.
     """
 
     def callcommand(name, args):
         """Request that a named command be executed.
 
         Receives the command name and a dictionary of command arguments.
 
         Returns a ``concurrent.futures.Future`` that will resolve to the
         result of that command request. That exact value is left up to
         the implementation and possibly varies by command.
 
         Not all commands can coexist with other commands in an executor
         instance: it depends on the underlying wire protocol transport being
         used and the command itself.
 
         Implementations MAY call ``sendcommands()`` automatically if the
         requested command can not coexist with other commands in this executor.
 
         Implementations MAY call ``sendcommands()`` automatically when the
         future's ``result()`` is called. So, consumers using multiple
         commands with an executor MUST ensure that ``result()`` is not called
         until all command requests have been issued.
         """
 
     def sendcommands():
         """Trigger submission of queued command requests.
 
         Not all transports submit commands as soon as they are requested to
         run. When called, this method forces queued command requests to be
         issued. It will no-op if all commands have already been sent.
 
         When called, no more new commands may be issued with this executor.
         """
 
     def close():
         """Signal that this command request is finished.
 
         When called, no more new commands may be issued. All outstanding
         commands that have previously been issued are waited on before
         returning. This not only includes waiting for the futures to resolve,
         but also waiting for all response data to arrive. In other words,
         calling this waits for all on-wire state for issued command requests
         to finish.
 
         When used as a context manager, this method is called when exiting the
         context manager.
 
         This method may call ``sendcommands()`` if there are buffered commands.
         """
 
 
 class ipeerrequests(interfaceutil.Interface):
     """Interface for executing commands on a peer."""
 
     limitedarguments = interfaceutil.Attribute(
         """True if the peer cannot receive large argument value for commands."""
     )
 
     def commandexecutor():
         """A context manager that resolves to an ipeercommandexecutor.
 
         The object this resolves to can be used to issue command requests
         to the peer.
 
         Callers should call its ``callcommand`` method to issue command
         requests.
 
         A new executor should be obtained for each distinct set of commands
         (possibly just a single command) that the consumer wants to execute
         as part of a single operation or round trip. This is because some
         peers are half-duplex and/or don't support persistent connections.
         e.g. in the case of HTTP peers, commands sent to an executor represent
         a single HTTP request. While some peers may support multiple command
         sends over the wire per executor, consumers need to code to the least
         capable peer. So it should be assumed that command executors buffer
         called commands until they are told to send them and that each
         command executor could result in a new connection or wire-level request
         being issued.
         """
 
 
 class ipeerbase(ipeerconnection, ipeercapabilities, ipeerrequests):
     """Unified interface for peer repositories.
 
     All peer instances must conform to this interface.
     """
 
 
 class ipeerv2(ipeerconnection, ipeercapabilities, ipeerrequests):
     """Unified peer interface for wire protocol version 2 peers."""
 
     apidescriptor = interfaceutil.Attribute(
         """Data structure holding description of server API."""
     )
 
 
 @interfaceutil.implementer(ipeerbase)
 class peer(object):
     """Base class for peer repositories."""
 
     limitedarguments = False
 
     def capable(self, name):
         caps = self.capabilities()
         if name in caps:
             return True
 
         name = b'%s=' % name
         for cap in caps:
             if cap.startswith(name):
                 return cap[len(name) :]
 
         return False
 
     def requirecap(self, name, purpose):
         if self.capable(name):
             return
 
         raise error.CapabilityError(
             _(
                 b'cannot %s; remote repository does not support the '
                 b'\'%s\' capability'
             )
             % (purpose, name)
         )
 
 
 class iverifyproblem(interfaceutil.Interface):
     """Represents a problem with the integrity of the repository.
 
     Instances of this interface are emitted to describe an integrity issue
     with a repository (e.g. corrupt storage, missing data, etc).
 
     Instances are essentially messages associated with severity.
     """
 
     warning = interfaceutil.Attribute(
         """Message indicating a non-fatal problem."""
     )
 
     error = interfaceutil.Attribute("""Message indicating a fatal problem.""")
 
     node = interfaceutil.Attribute(
         """Revision encountering the problem.
 
         ``None`` means the problem doesn't apply to a single revision.
         """
     )
 
 
 class irevisiondelta(interfaceutil.Interface):
     """Represents a delta between one revision and another.
 
     Instances convey enough information to allow a revision to be exchanged
     with another repository.
 
     Instances represent the fulltext revision data or a delta against
     another revision. Therefore the ``revision`` and ``delta`` attributes
     are mutually exclusive.
 
     Typically used for changegroup generation.
     """
 
     node = interfaceutil.Attribute("""20 byte node of this revision.""")
 
     p1node = interfaceutil.Attribute(
         """20 byte node of 1st parent of this revision."""
     )
 
     p2node = interfaceutil.Attribute(
         """20 byte node of 2nd parent of this revision."""
     )
 
     linknode = interfaceutil.Attribute(
         """20 byte node of the changelog revision this node is linked to."""
     )
 
     flags = interfaceutil.Attribute(
         """2 bytes of integer flags that apply to this revision.
 
         This is a bitwise composition of the ``REVISION_FLAG_*`` constants.
         """
     )
 
     basenode = interfaceutil.Attribute(
         """20 byte node of the revision this data is a delta against.
 
         ``nullid`` indicates that the revision is a full revision and not
         a delta.
         """
     )
 
     baserevisionsize = interfaceutil.Attribute(
         """Size of base revision this delta is against.
 
         May be ``None`` if ``basenode`` is ``nullid``.
         """
     )
 
     revision = interfaceutil.Attribute(
         """Raw fulltext of revision data for this node."""
     )
 
     delta = interfaceutil.Attribute(
         """Delta between ``basenode`` and ``node``.
 
         Stored in the bdiff delta format.
         """
     )
 
 
 class ifilerevisionssequence(interfaceutil.Interface):
     """Contains index data for all revisions of a file.
 
     Types implementing this behave like lists of tuples. The index
     in the list corresponds to the revision number. The values contain
     index metadata.
 
     The *null* revision (revision number -1) is always the last item
     in the index.
     """
 
     def __len__():
         """The total number of revisions."""
 
     def __getitem__(rev):
         """Returns the object having a specific revision number.
 
         Returns an 8-tuple with the following fields:
 
         offset+flags
            Contains the offset and flags for the revision. 64-bit unsigned
            integer where first 6 bytes are the offset and the next 2 bytes
            are flags. The offset can be 0 if it is not used by the store.
         compressed size
             Size of the revision data in the store. It can be 0 if it isn't
             needed by the store.
         uncompressed size
             Fulltext size. It can be 0 if it isn't needed by the store.
         base revision
             Revision number of revision the delta for storage is encoded
             against. -1 indicates not encoded against a base revision.
         link revision
             Revision number of changelog revision this entry is related to.
         p1 revision
             Revision number of 1st parent. -1 if no 1st parent.
         p2 revision
             Revision number of 2nd parent. -1 if no 1st parent.
         node
             Binary node value for this revision number.
 
         Negative values should index off the end of the sequence. ``-1``
         should return the null revision. ``-2`` should return the most
         recent revision.
         """
 
     def __contains__(rev):
         """Whether a revision number exists."""
 
     def insert(self, i, entry):
         """Add an item to the index at specific revision."""
 
 
 class ifileindex(interfaceutil.Interface):
     """Storage interface for index data of a single file.
 
     File storage data is divided into index metadata and data storage.
     This interface defines the index portion of the interface.
 
     The index logically consists of:
 
     * A mapping between revision numbers and nodes.
     * DAG data (storing and querying the relationship between nodes).
     * Metadata to facilitate storage.
     """
 
     def __len__():
         """Obtain the number of revisions stored for this file."""
 
     def __iter__():
         """Iterate over revision numbers for this file."""
 
     def hasnode(node):
         """Returns a bool indicating if a node is known to this store.
 
         Implementations must only return True for full, binary node values:
         hex nodes, revision numbers, and partial node matches must be
         rejected.
 
         The null node is never present.
         """
 
     def revs(start=0, stop=None):
         """Iterate over revision numbers for this file, with control."""
 
     def parents(node):
         """Returns a 2-tuple of parent nodes for a revision.
 
         Values will be ``nullid`` if the parent is empty.
         """
 
     def parentrevs(rev):
         """Like parents() but operates on revision numbers."""
 
     def rev(node):
         """Obtain the revision number given a node.
 
         Raises ``error.LookupError`` if the node is not known.
         """
 
     def node(rev):
         """Obtain the node value given a revision number.
 
         Raises ``IndexError`` if the node is not known.
         """
 
     def lookup(node):
         """Attempt to resolve a value to a node.
 
         Value can be a binary node, hex node, revision number, or a string
         that can be converted to an integer.
 
         Raises ``error.LookupError`` if a node could not be resolved.
         """
 
     def linkrev(rev):
         """Obtain the changeset revision number a revision is linked to."""
 
     def iscensored(rev):
         """Return whether a revision's content has been censored."""
 
     def commonancestorsheads(node1, node2):
         """Obtain an iterable of nodes containing heads of common ancestors.
 
         See ``ancestor.commonancestorsheads()``.
         """
 
     def descendants(revs):
         """Obtain descendant revision numbers for a set of revision numbers.
 
         If ``nullrev`` is in the set, this is equivalent to ``revs()``.
         """
 
     def heads(start=None, stop=None):
         """Obtain a list of nodes that are DAG heads, with control.
 
         The set of revisions examined can be limited by specifying
         ``start`` and ``stop``. ``start`` is a node. ``stop`` is an
         iterable of nodes. DAG traversal starts at earlier revision
         ``start`` and iterates forward until any node in ``stop`` is
         encountered.
         """
 
     def children(node):
         """Obtain nodes that are children of a node.
 
         Returns a list of nodes.
         """
 
 
 class ifiledata(interfaceutil.Interface):
     """Storage interface for data storage of a specific file.
 
     This complements ``ifileindex`` and provides an interface for accessing
     data for a tracked file.
     """
 
     def size(rev):
         """Obtain the fulltext size of file data.
 
         Any metadata is excluded from size measurements.
         """
 
     def revision(node, raw=False):
         """Obtain fulltext data for a node.
 
         By default, any storage transformations are applied before the data
         is returned. If ``raw`` is True, non-raw storage transformations
         are not applied.
 
         The fulltext data may contain a header containing metadata. Most
         consumers should use ``read()`` to obtain the actual file data.
         """
 
     def rawdata(node):
         """Obtain raw data for a node."""
 
     def read(node):
         """Resolve file fulltext data.
 
         This is similar to ``revision()`` except any metadata in the data
         headers is stripped.
         """
 
     def renamed(node):
         """Obtain copy metadata for a node.
 
         Returns ``False`` if no copy metadata is stored or a 2-tuple of
         (path, node) from which this revision was copied.
         """
 
     def cmp(node, fulltext):
         """Compare fulltext to another revision.
 
         Returns True if the fulltext is different from what is stored.
 
         This takes copy metadata into account.
 
         TODO better document the copy metadata and censoring logic.
         """
 
     def emitrevisions(
         nodes,
         nodesorder=None,
         revisiondata=False,
         assumehaveparentrevisions=False,
         deltamode=CG_DELTAMODE_STD,
     ):
         """Produce ``irevisiondelta`` for revisions.
 
         Given an iterable of nodes, emits objects conforming to the
         ``irevisiondelta`` interface that describe revisions in storage.
 
         This method is a generator.
 
         The input nodes may be unordered. Implementations must ensure that a
         node's parents are emitted before the node itself. Transitively, this
         means that a node may only be emitted once all its ancestors in
         ``nodes`` have also been emitted.
 
         By default, emits "index" data (the ``node``, ``p1node``, and
         ``p2node`` attributes). If ``revisiondata`` is set, revision data
         will also be present on the emitted objects.
 
         With default argument values, implementations can choose to emit
         either fulltext revision data or a delta. When emitting deltas,
         implementations must consider whether the delta's base revision
         fulltext is available to the receiver.
 
         The base revision fulltext is guaranteed to be available if any of
         the following are met:
 
         * Its fulltext revision was emitted by this method call.
         * A delta for that revision was emitted by this method call.
         * ``assumehaveparentrevisions`` is True and the base revision is a
           parent of the node.
 
         ``nodesorder`` can be used to control the order that revisions are
         emitted. By default, revisions can be reordered as long as they are
         in DAG topological order (see above). If the value is ``nodes``,
         the iteration order from ``nodes`` should be used. If the value is
         ``storage``, then the native order from the backing storage layer
         is used. (Not all storage layers will have strong ordering and behavior
         of this mode is storage-dependent.) ``nodes`` ordering can force
         revisions to be emitted before their ancestors, so consumers should
         use it with care.
 
         The ``linknode`` attribute on the returned ``irevisiondelta`` may not
         be set and it is the caller's responsibility to resolve it, if needed.
 
         If ``deltamode`` is CG_DELTAMODE_PREV and revision data is requested,
         all revision data should be emitted as deltas against the revision
         emitted just prior. The initial revision should be a delta against its
         1st parent.
         """
 
 
 class ifilemutation(interfaceutil.Interface):
     """Storage interface for mutation events of a tracked file."""
 
     def add(filedata, meta, transaction, linkrev, p1, p2):
         """Add a new revision to the store.
 
         Takes file data, dictionary of metadata, a transaction, linkrev,
         and parent nodes.
 
         Returns the node that was added.
 
         May no-op if a revision matching the supplied data is already stored.
         """
 
     def addrevision(
         revisiondata,
         transaction,
         linkrev,
         p1,
         p2,
         node=None,
         flags=0,
         cachedelta=None,
     ):
         """Add a new revision to the store.
 
         This is similar to ``add()`` except it operates at a lower level.
 
         The data passed in already contains a metadata header, if any.
 
         ``node`` and ``flags`` can be used to define the expected node and
         the flags to use with storage. ``flags`` is a bitwise value composed
         of the various ``REVISION_FLAG_*`` constants.
 
         ``add()`` is usually called when adding files from e.g. the working
         directory. ``addrevision()`` is often called by ``add()`` and for
         scenarios where revision data has already been computed, such as when
         applying raw data from a peer repo.
         """
 
     def addgroup(
         deltas,
         linkmapper,
         transaction,
         addrevisioncb=None,
         duplicaterevisioncb=None,
         maybemissingparents=False,
     ):
         """Process a series of deltas for storage.
 
         ``deltas`` is an iterable of 7-tuples of
         (node, p1, p2, linknode, deltabase, delta, flags) defining revisions
         to add.
 
         The ``delta`` field contains ``mpatch`` data to apply to a base
         revision, identified by ``deltabase``. The base node can be
         ``nullid``, in which case the header from the delta can be ignored
         and the delta used as the fulltext.
 
+        ``alwayscache`` instructs the lower layers to cache the content of the
+        newly added revision, even if it needs to be explicitly computed.
+        This used to be the default when ``addrevisioncb`` was provided up to
+        Mercurial 5.8.
+
         ``addrevisioncb`` should be called for each node as it is committed.
+        ``duplicaterevisioncb`` should be called for each pre-existing node.
 
         ``maybemissingparents`` is a bool indicating whether the incoming
         data may reference parents/ancestor revisions that aren't present.
         This flag is set when receiving data into a "shallow" store that
         doesn't hold all history.
 
         Returns a list of nodes that were processed. A node will be in the list
         even if it existed in the store previously.
         """
 
     def censorrevision(tr, node, tombstone=b''):
         """Remove the content of a single revision.
 
         The specified ``node`` will have its content purged from storage.
         Future attempts to access the revision data for this node will
         result in failure.
 
         A ``tombstone`` message can optionally be stored. This message may be
         displayed to users when they attempt to access the missing revision
         data.
 
         Storage backends may have stored deltas against the previous content
         in this revision. As part of censoring a revision, these storage
         backends are expected to rewrite any internally stored deltas such
         that they no longer reference the deleted content.
         """
 
     def getstrippoint(minlink):
         """Find the minimum revision that must be stripped to strip a linkrev.
 
         Returns a 2-tuple containing the minimum revision number and a set
         of all revisions numbers that would be broken by this strip.
 
         TODO this is highly revlog centric and should be abstracted into
         a higher-level deletion API. ``repair.strip()`` relies on this.
         """
 
     def strip(minlink, transaction):
         """Remove storage of items starting at a linkrev.
 
         This uses ``getstrippoint()`` to determine the first node to remove.
         Then it effectively truncates storage for all revisions after that.
 
         TODO this is highly revlog centric and should be abstracted into a
         higher-level deletion API.
         """
 
 
 class ifilestorage(ifileindex, ifiledata, ifilemutation):
     """Complete storage interface for a single tracked file."""
 
     def files():
         """Obtain paths that are backing storage for this file.
 
         TODO this is used heavily by verify code and there should probably
         be a better API for that.
         """
 
     def storageinfo(
         exclusivefiles=False,
         sharedfiles=False,
         revisionscount=False,
         trackedsize=False,
         storedsize=False,
     ):
         """Obtain information about storage for this file's data.
 
         Returns a dict describing storage for this tracked path. The keys
         in the dict map to arguments of the same. The arguments are bools
         indicating whether to calculate and obtain that data.
 
         exclusivefiles
            Iterable of (vfs, path) describing files that are exclusively
            used to back storage for this tracked path.
 
         sharedfiles
            Iterable of (vfs, path) describing files that are used to back
            storage for this tracked path. Those files may also provide storage
            for other stored entities.
 
         revisionscount
            Number of revisions available for retrieval.
 
         trackedsize
            Total size in bytes of all tracked revisions. This is a sum of the
            length of the fulltext of all revisions.
 
         storedsize
            Total size in bytes used to store data for all tracked revisions.
            This is commonly less than ``trackedsize`` due to internal usage
            of deltas rather than fulltext revisions.
 
         Not all storage backends may support all queries are have a reasonable
         value to use. In that case, the value should be set to ``None`` and
         callers are expected to handle this special value.
         """
 
     def verifyintegrity(state):
         """Verifies the integrity of file storage.
 
         ``state`` is a dict holding state of the verifier process. It can be
         used to communicate data between invocations of multiple storage
         primitives.
 
         If individual revisions cannot have their revision content resolved,
         the method is expected to set the ``skipread`` key to a set of nodes
         that encountered problems.  If set, the method can also add the node(s)
         to ``safe_renamed`` in order to indicate nodes that may perform the
         rename checks with currently accessible data.
 
         The method yields objects conforming to the ``iverifyproblem``
         interface.
         """
 
 
 class idirs(interfaceutil.Interface):
     """Interface representing a collection of directories from paths.
 
     This interface is essentially a derived data structure representing
     directories from a collection of paths.
     """
 
     def addpath(path):
         """Add a path to the collection.
 
         All directories in the path will be added to the collection.
         """
 
     def delpath(path):
         """Remove a path from the collection.
 
         If the removal was the last path in a particular directory, the
         directory is removed from the collection.
         """
 
     def __iter__():
         """Iterate over the directories in this collection of paths."""
 
     def __contains__(path):
         """Whether a specific directory is in this collection."""
 
 
 class imanifestdict(interfaceutil.Interface):
     """Interface representing a manifest data structure.
 
     A manifest is effectively a dict mapping paths to entries. Each entry
     consists of a binary node and extra flags affecting that entry.
     """
 
     def __getitem__(path):
         """Returns the binary node value for a path in the manifest.
 
         Raises ``KeyError`` if the path does not exist in the manifest.
 
         Equivalent to ``self.find(path)[0]``.
         """
 
     def find(path):
         """Returns the entry for a path in the manifest.
 
         Returns a 2-tuple of (node, flags).
 
         Raises ``KeyError`` if the path does not exist in the manifest.
         """
 
     def __len__():
         """Return the number of entries in the manifest."""
 
     def __nonzero__():
         """Returns True if the manifest has entries, False otherwise."""
 
     __bool__ = __nonzero__
 
     def __setitem__(path, node):
         """Define the node value for a path in the manifest.
 
         If the path is already in the manifest, its flags will be copied to
         the new entry.
         """
 
     def __contains__(path):
         """Whether a path exists in the manifest."""
 
     def __delitem__(path):
         """Remove a path from the manifest.
 
         Raises ``KeyError`` if the path is not in the manifest.
         """
 
     def __iter__():
         """Iterate over paths in the manifest."""
 
     def iterkeys():
         """Iterate over paths in the manifest."""
 
     def keys():
         """Obtain a list of paths in the manifest."""
 
     def filesnotin(other, match=None):
         """Obtain the set of paths in this manifest but not in another.
 
         ``match`` is an optional matcher function to be applied to both
         manifests.
 
         Returns a set of paths.
         """
 
     def dirs():
         """Returns an object implementing the ``idirs`` interface."""
 
     def hasdir(dir):
         """Returns a bool indicating if a directory is in this manifest."""
 
     def walk(match):
         """Generator of paths in manifest satisfying a matcher.
 
         If the matcher has explicit files listed and they don't exist in
         the manifest, ``match.bad()`` is called for each missing file.
         """
 
     def diff(other, match=None, clean=False):
         """Find differences between this manifest and another.
 
         This manifest is compared to ``other``.
 
         If ``match`` is provided, the two manifests are filtered against this
         matcher and only entries satisfying the matcher are compared.
 
         If ``clean`` is True, unchanged files are included in the returned
         object.
 
         Returns a dict with paths as keys and values of 2-tuples of 2-tuples of
         the form ``((node1, flag1), (node2, flag2))`` where ``(node1, flag1)``
         represents the node and flags for this manifest and ``(node2, flag2)``
         are the same for the other manifest.
         """
 
     def setflag(path, flag):
         """Set the flag value for a given path.
 
         Raises ``KeyError`` if the path is not already in the manifest.
         """
 
     def get(path, default=None):
         """Obtain the node value for a path or a default value if missing."""
 
     def flags(path):
         """Return the flags value for a path (default: empty bytestring)."""
 
     def copy():
         """Return a copy of this manifest."""
 
     def items():
         """Returns an iterable of (path, node) for items in this manifest."""
 
     def iteritems():
         """Identical to items()."""
 
     def iterentries():
         """Returns an iterable of (path, node, flags) for this manifest.
 
         Similar to ``iteritems()`` except items are a 3-tuple and include
         flags.
         """
 
     def text():
         """Obtain the raw data representation for this manifest.
 
         Result is used to create a manifest revision.
         """
 
     def fastdelta(base, changes):
         """Obtain a delta between this manifest and another given changes.
 
         ``base`` in the raw data representation for another manifest.
 
         ``changes`` is an iterable of ``(path, to_delete)``.
 
         Returns a 2-tuple containing ``bytearray(self.text())`` and the
         delta between ``base`` and this manifest.
 
         If this manifest implementation can't support ``fastdelta()``,
         raise ``mercurial.manifest.FastdeltaUnavailable``.
         """
 
 
 class imanifestrevisionbase(interfaceutil.Interface):
     """Base interface representing a single revision of a manifest.
 
     Should not be used as a primary interface: should always be inherited
     as part of a larger interface.
     """
 
     def copy():
         """Obtain a copy of this manifest instance.
 
         Returns an object conforming to the ``imanifestrevisionwritable``
         interface. The instance will be associated with the same
         ``imanifestlog`` collection as this instance.
         """
 
     def read():
         """Obtain the parsed manifest data structure.
 
         The returned object conforms to the ``imanifestdict`` interface.
         """
 
 
 class imanifestrevisionstored(imanifestrevisionbase):
     """Interface representing a manifest revision committed to storage."""
 
     def node():
         """The binary node for this manifest."""
 
     parents = interfaceutil.Attribute(
         """List of binary nodes that are parents for this manifest revision."""
     )
 
     def readdelta(shallow=False):
         """Obtain the manifest data structure representing changes from parent.
 
         This manifest is compared to its 1st parent. A new manifest representing
         those differences is constructed.
 
         The returned object conforms to the ``imanifestdict`` interface.
         """
 
     def readfast(shallow=False):
         """Calls either ``read()`` or ``readdelta()``.
 
         The faster of the two options is called.
         """
 
     def find(key):
         """Calls self.read().find(key)``.
 
         Returns a 2-tuple of ``(node, flags)`` or raises ``KeyError``.
         """
 
 
 class imanifestrevisionwritable(imanifestrevisionbase):
     """Interface representing a manifest revision that can be committed."""
 
     def write(transaction, linkrev, p1node, p2node, added, removed, match=None):
         """Add this revision to storage.
 
         Takes a transaction object, the changeset revision number it will
         be associated with, its parent nodes, and lists of added and
         removed paths.
 
         If match is provided, storage can choose not to inspect or write out
         items that do not match. Storage is still required to be able to provide
         the full manifest in the future for any directories written (these
         manifests should not be "narrowed on disk").
 
         Returns the binary node of the created revision.
         """
 
 
 class imanifeststorage(interfaceutil.Interface):
     """Storage interface for manifest data."""
 
     tree = interfaceutil.Attribute(
         """The path to the directory this manifest tracks.
 
         The empty bytestring represents the root manifest.
         """
     )
 
     index = interfaceutil.Attribute(
         """An ``ifilerevisionssequence`` instance."""
     )
 
     indexfile = interfaceutil.Attribute(
         """Path of revlog index file.
 
         TODO this is revlog specific and should not be exposed.
         """
     )
 
     opener = interfaceutil.Attribute(
         """VFS opener to use to access underlying files used for storage.
 
         TODO this is revlog specific and should not be exposed.
         """
     )
 
     version = interfaceutil.Attribute(
         """Revlog version number.
 
         TODO this is revlog specific and should not be exposed.
         """
     )
 
     _generaldelta = interfaceutil.Attribute(
         """Whether generaldelta storage is being used.
 
         TODO this is revlog specific and should not be exposed.
         """
     )
 
     fulltextcache = interfaceutil.Attribute(
         """Dict with cache of fulltexts.
 
         TODO this doesn't feel appropriate for the storage interface.
         """
     )
 
     def __len__():
         """Obtain the number of revisions stored for this manifest."""
 
     def __iter__():
         """Iterate over revision numbers for this manifest."""
 
     def rev(node):
         """Obtain the revision number given a binary node.
 
         Raises ``error.LookupError`` if the node is not known.
         """
 
     def node(rev):
         """Obtain the node value given a revision number.
 
         Raises ``error.LookupError`` if the revision is not known.
         """
 
     def lookup(value):
         """Attempt to resolve a value to a node.
 
         Value can be a binary node, hex node, revision number, or a bytes
         that can be converted to an integer.
 
         Raises ``error.LookupError`` if a ndoe could not be resolved.
         """
 
     def parents(node):
         """Returns a 2-tuple of parent nodes for a node.
 
         Values will be ``nullid`` if the parent is empty.
         """
 
     def parentrevs(rev):
         """Like parents() but operates on revision numbers."""
 
     def linkrev(rev):
         """Obtain the changeset revision number a revision is linked to."""
 
     def revision(node, _df=None, raw=False):
         """Obtain fulltext data for a node."""
 
     def rawdata(node, _df=None):
         """Obtain raw data for a node."""
 
     def revdiff(rev1, rev2):
         """Obtain a delta between two revision numbers.
 
         The returned data is the result of ``bdiff.bdiff()`` on the raw
         revision data.
         """
 
     def cmp(node, fulltext):
         """Compare fulltext to another revision.
 
         Returns True if the fulltext is different from what is stored.
         """
 
     def emitrevisions(
         nodes,
         nodesorder=None,
         revisiondata=False,
         assumehaveparentrevisions=False,
     ):
         """Produce ``irevisiondelta`` describing revisions.
 
         See the documentation for ``ifiledata`` for more.
         """
 
     def addgroup(
         deltas,
         linkmapper,
         transaction,
         addrevisioncb=None,
         duplicaterevisioncb=None,
     ):
         """Process a series of deltas for storage.
 
         See the documentation in ``ifilemutation`` for more.
         """
 
     def rawsize(rev):
         """Obtain the size of tracked data.
 
         Is equivalent to ``len(m.rawdata(node))``.
 
         TODO this method is only used by upgrade code and may be removed.
         """
 
     def getstrippoint(minlink):
         """Find minimum revision that must be stripped to strip a linkrev.
 
         See the documentation in ``ifilemutation`` for more.
         """
 
     def strip(minlink, transaction):
         """Remove storage of items starting at a linkrev.
 
         See the documentation in ``ifilemutation`` for more.
         """
 
     def checksize():
         """Obtain the expected sizes of backing files.
 
         TODO this is used by verify and it should not be part of the interface.
         """
 
     def files():
         """Obtain paths that are backing storage for this manifest.
 
         TODO this is used by verify and there should probably be a better API
         for this functionality.
         """
 
     def deltaparent(rev):
         """Obtain the revision that a revision is delta'd against.
 
         TODO delta encoding is an implementation detail of storage and should
         not be exposed to the storage interface.
         """
 
     def clone(tr, dest, **kwargs):
         """Clone this instance to another."""
 
     def clearcaches(clear_persisted_data=False):
         """Clear any caches associated with this instance."""
 
     def dirlog(d):
         """Obtain a manifest storage instance for a tree."""
 
     def add(
         m, transaction, link, p1, p2, added, removed, readtree=None, match=None
     ):
         """Add a revision to storage.
 
         ``m`` is an object conforming to ``imanifestdict``.
 
         ``link`` is the linkrev revision number.
 
         ``p1`` and ``p2`` are the parent revision numbers.
 
         ``added`` and ``removed`` are iterables of added and removed paths,
         respectively.
 
         ``readtree`` is a function that can be used to read the child tree(s)
         when recursively writing the full tree structure when using
         treemanifets.
 
         ``match`` is a matcher that can be used to hint to storage that not all
         paths must be inspected; this is an optimization and can be safely
         ignored. Note that the storage must still be able to reproduce a full
         manifest including files that did not match.
         """
 
     def storageinfo(
         exclusivefiles=False,
         sharedfiles=False,
         revisionscount=False,
         trackedsize=False,
         storedsize=False,
     ):
         """Obtain information about storage for this manifest's data.
 
         See ``ifilestorage.storageinfo()`` for a description of this method.
         This one behaves the same way, except for manifest data.
         """
 
 
 class imanifestlog(interfaceutil.Interface):
     """Interface representing a collection of manifest snapshots.
 
     Represents the root manifest in a repository.
 
     Also serves as a means to access nested tree manifests and to cache
     tree manifests.
     """
 
     def __getitem__(node):
         """Obtain a manifest instance for a given binary node.
 
         Equivalent to calling ``self.get('', node)``.
 
         The returned object conforms to the ``imanifestrevisionstored``
         interface.
         """
 
     def get(tree, node, verify=True):
         """Retrieve the manifest instance for a given directory and binary node.
 
         ``node`` always refers to the node of the root manifest (which will be
         the only manifest if flat manifests are being used).
 
         If ``tree`` is the empty string, the root manifest is returned.
         Otherwise the manifest for the specified directory will be returned
         (requires tree manifests).
 
         If ``verify`` is True, ``LookupError`` is raised if the node is not
         known.
 
         The returned object conforms to the ``imanifestrevisionstored``
         interface.
         """
 
     def getstorage(tree):
         """Retrieve an interface to storage for a particular tree.
 
         If ``tree`` is the empty bytestring, storage for the root manifest will
         be returned. Otherwise storage for a tree manifest is returned.
 
         TODO formalize interface for returned object.
         """
 
     def clearcaches():
         """Clear caches associated with this collection."""
 
     def rev(node):
         """Obtain the revision number for a binary node.
 
         Raises ``error.LookupError`` if the node is not known.
         """
 
     def update_caches(transaction):
         """update whatever cache are relevant for the used storage."""
 
 
 class ilocalrepositoryfilestorage(interfaceutil.Interface):
     """Local repository sub-interface providing access to tracked file storage.
 
     This interface defines how a repository accesses storage for a single
     tracked file path.
     """
 
     def file(f):
         """Obtain a filelog for a tracked path.
 
         The returned type conforms to the ``ifilestorage`` interface.
         """
 
 
 class ilocalrepositorymain(interfaceutil.Interface):
     """Main interface for local repositories.
 
     This currently captures the reality of things - not how things should be.
     """
 
     supportedformats = interfaceutil.Attribute(
         """Set of requirements that apply to stream clone.
 
         This is actually a class attribute and is shared among all instances.
         """
     )
 
     supported = interfaceutil.Attribute(
         """Set of requirements that this repo is capable of opening."""
     )
 
     requirements = interfaceutil.Attribute(
         """Set of requirements this repo uses."""
     )
 
     features = interfaceutil.Attribute(
         """Set of "features" this repository supports.
 
         A "feature" is a loosely-defined term. It can refer to a feature
         in the classical sense or can describe an implementation detail
         of the repository. For example, a ``readonly`` feature may denote
         the repository as read-only. Or a ``revlogfilestore`` feature may
         denote that the repository is using revlogs for file storage.
 
         The intent of features is to provide a machine-queryable mechanism
         for repo consumers to test for various repository characteristics.
 
         Features are similar to ``requirements``. The main difference is that
         requirements are stored on-disk and represent requirements to open the
         repository. Features are more run-time capabilities of the repository
         and more granular capabilities (which may be derived from requirements).
         """
     )
 
     filtername = interfaceutil.Attribute(
         """Name of the repoview that is active on this repo."""
     )
 
     wvfs = interfaceutil.Attribute(
         """VFS used to access the working directory."""
     )
 
     vfs = interfaceutil.Attribute(
         """VFS rooted at the .hg directory.
 
         Used to access repository data not in the store.
         """
     )
 
     svfs = interfaceutil.Attribute(
         """VFS rooted at the store.
 
         Used to access repository data in the store. Typically .hg/store.
         But can point elsewhere if the store is shared.
         """
     )
 
     root = interfaceutil.Attribute(
         """Path to the root of the working directory."""
     )
 
     path = interfaceutil.Attribute("""Path to the .hg directory.""")
 
     origroot = interfaceutil.Attribute(
         """The filesystem path that was used to construct the repo."""
     )
 
     auditor = interfaceutil.Attribute(
         """A pathauditor for the working directory.
 
         This checks if a path refers to a nested repository.
 
         Operates on the filesystem.
         """
     )
 
     nofsauditor = interfaceutil.Attribute(
         """A pathauditor for the working directory.
 
         This is like ``auditor`` except it doesn't do filesystem checks.
         """
     )
 
     baseui = interfaceutil.Attribute(
         """Original ui instance passed into constructor."""
     )
 
     ui = interfaceutil.Attribute("""Main ui instance for this instance.""")
 
     sharedpath = interfaceutil.Attribute(
         """Path to the .hg directory of the repo this repo was shared from."""
     )
 
     store = interfaceutil.Attribute("""A store instance.""")
 
     spath = interfaceutil.Attribute("""Path to the store.""")
 
     sjoin = interfaceutil.Attribute("""Alias to self.store.join.""")
 
     cachevfs = interfaceutil.Attribute(
         """A VFS used to access the cache directory.
 
         Typically .hg/cache.
         """
     )
 
     wcachevfs = interfaceutil.Attribute(
         """A VFS used to access the cache directory dedicated to working copy
 
         Typically .hg/wcache.
         """
     )
 
     filteredrevcache = interfaceutil.Attribute(
         """Holds sets of revisions to be filtered."""
     )
 
     names = interfaceutil.Attribute("""A ``namespaces`` instance.""")
 
     filecopiesmode = interfaceutil.Attribute(
         """The way files copies should be dealt with in this repo."""
     )
 
     def close():
         """Close the handle on this repository."""
 
     def peer():
         """Obtain an object conforming to the ``peer`` interface."""
 
     def unfiltered():
         """Obtain an unfiltered/raw view of this repo."""
 
     def filtered(name, visibilityexceptions=None):
         """Obtain a named view of this repository."""
 
     obsstore = interfaceutil.Attribute("""A store of obsolescence data.""")
 
     changelog = interfaceutil.Attribute("""A handle on the changelog revlog.""")
 
     manifestlog = interfaceutil.Attribute(
         """An instance conforming to the ``imanifestlog`` interface.
 
         Provides access to manifests for the repository.
         """
     )
 
     dirstate = interfaceutil.Attribute("""Working directory state.""")
 
     narrowpats = interfaceutil.Attribute(
         """Matcher patterns for this repository's narrowspec."""
     )
 
     def narrowmatch(match=None, includeexact=False):
         """Obtain a matcher for the narrowspec."""
 
     def setnarrowpats(newincludes, newexcludes):
         """Define the narrowspec for this repository."""
 
     def __getitem__(changeid):
         """Try to resolve a changectx."""
 
     def __contains__(changeid):
         """Whether a changeset exists."""
 
     def __nonzero__():
         """Always returns True."""
         return True
 
     __bool__ = __nonzero__
 
     def __len__():
         """Returns the number of changesets in the repo."""
 
     def __iter__():
         """Iterate over revisions in the changelog."""
 
     def revs(expr, *args):
         """Evaluate a revset.
 
         Emits revisions.
         """
 
     def set(expr, *args):
         """Evaluate a revset.
 
         Emits changectx instances.
         """
 
     def anyrevs(specs, user=False, localalias=None):
         """Find revisions matching one of the given revsets."""
 
     def url():
         """Returns a string representing the location of this repo."""
 
     def hook(name, throw=False, **args):
         """Call a hook."""
 
     def tags():
         """Return a mapping of tag to node."""
 
     def tagtype(tagname):
         """Return the type of a given tag."""
 
     def tagslist():
         """Return a list of tags ordered by revision."""
 
     def nodetags(node):
         """Return the tags associated with a node."""
 
     def nodebookmarks(node):
         """Return the list of bookmarks pointing to the specified node."""
 
     def branchmap():
         """Return a mapping of branch to heads in that branch."""
 
     def revbranchcache():
         pass
 
     def register_changeset(rev, changelogrevision):
         """Extension point for caches for new nodes.
 
         Multiple consumers are expected to need parts of the changelogrevision,
         so it is provided as optimization to avoid duplicate lookups. A simple
         cache would be fragile when other revisions are accessed, too."""
         pass
 
     def branchtip(branchtip, ignoremissing=False):
         """Return the tip node for a given branch."""
 
     def lookup(key):
         """Resolve the node for a revision."""
 
     def lookupbranch(key):
         """Look up the branch name of the given revision or branch name."""
 
     def known(nodes):
         """Determine whether a series of nodes is known.
 
         Returns a list of bools.
         """
 
     def local():
         """Whether the repository is local."""
         return True
 
     def publishing():
         """Whether the repository is a publishing repository."""
 
     def cancopy():
         pass
 
     def shared():
         """The type of shared repository or None."""
 
     def wjoin(f, *insidef):
         """Calls self.vfs.reljoin(self.root, f, *insidef)"""
 
     def setparents(p1, p2):
         """Set the parent nodes of the working directory."""
 
     def filectx(path, changeid=None, fileid=None):
         """Obtain a filectx for the given file revision."""
 
     def getcwd():
         """Obtain the current working directory from the dirstate."""
 
     def pathto(f, cwd=None):
         """Obtain the relative path to a file."""
 
     def adddatafilter(name, fltr):
         pass
 
     def wread(filename):
         """Read a file from wvfs, using data filters."""
 
     def wwrite(filename, data, flags, backgroundclose=False, **kwargs):
         """Write data to a file in the wvfs, using data filters."""
 
     def wwritedata(filename, data):
         """Resolve data for writing to the wvfs, using data filters."""
 
     def currenttransaction():
         """Obtain the current transaction instance or None."""
 
     def transaction(desc, report=None):
         """Open a new transaction to write to the repository."""
 
     def undofiles():
         """Returns a list of (vfs, path) for files to undo transactions."""
 
     def recover():
         """Roll back an interrupted transaction."""
 
     def rollback(dryrun=False, force=False):
         """Undo the last transaction.
 
         DANGEROUS.
         """
 
     def updatecaches(tr=None, full=False):
         """Warm repo caches."""
 
     def invalidatecaches():
         """Invalidate cached data due to the repository mutating."""
 
     def invalidatevolatilesets():
         pass
 
     def invalidatedirstate():
         """Invalidate the dirstate."""
 
     def invalidate(clearfilecache=False):
         pass
 
     def invalidateall():
         pass
 
     def lock(wait=True):
         """Lock the repository store and return a lock instance."""
 
     def wlock(wait=True):
         """Lock the non-store parts of the repository."""
 
     def currentwlock():
         """Return the wlock if it's held or None."""
 
     def checkcommitpatterns(wctx, match, status, fail):
         pass
 
     def commit(
         text=b'',
         user=None,
         date=None,
         match=None,
         force=False,
         editor=False,
         extra=None,
     ):
         """Add a new revision to the repository."""
 
     def commitctx(ctx, error=False, origctx=None):
         """Commit a commitctx instance to the repository."""
 
     def destroying():
         """Inform the repository that nodes are about to be destroyed."""
 
     def destroyed():
         """Inform the repository that nodes have been destroyed."""
 
     def status(
         node1=b'.',
         node2=None,
         match=None,
         ignored=False,
         clean=False,
         unknown=False,
         listsubrepos=False,
     ):
         """Convenience method to call repo[x].status()."""
 
     def addpostdsstatus(ps):
         pass
 
     def postdsstatus():
         pass
 
     def clearpostdsstatus():
         pass
 
     def heads(start=None):
         """Obtain list of nodes that are DAG heads."""
 
     def branchheads(branch=None, start=None, closed=False):
         pass
 
     def branches(nodes):
         pass
 
     def between(pairs):
         pass
 
     def checkpush(pushop):
         pass
 
     prepushoutgoinghooks = interfaceutil.Attribute("""util.hooks instance.""")
 
     def pushkey(namespace, key, old, new):
         pass
 
     def listkeys(namespace):
         pass
 
     def debugwireargs(one, two, three=None, four=None, five=None):
         pass
 
     def savecommitmessage(text):
         pass
 
 
 class completelocalrepository(
     ilocalrepositorymain, ilocalrepositoryfilestorage
 ):
     """Complete interface for a local repository."""
 
 
 class iwireprotocolcommandcacher(interfaceutil.Interface):
     """Represents a caching backend for wire protocol commands.
 
     Wire protocol version 2 supports transparent caching of many commands.
     To leverage this caching, servers can activate objects that cache
     command responses. Objects handle both cache writing and reading.
     This interface defines how that response caching mechanism works.
 
     Wire protocol version 2 commands emit a series of objects that are
     serialized and sent to the client. The caching layer exists between
     the invocation of the command function and the sending of its output
     objects to an output layer.
 
     Instances of this interface represent a binding to a cache that
     can serve a response (in place of calling a command function) and/or
     write responses to a cache for subsequent use.
 
     When a command request arrives, the following happens with regards
     to this interface:
 
     1. The server determines whether the command request is cacheable.
     2. If it is, an instance of this interface is spawned.
     3. The cacher is activated in a context manager (``__enter__`` is called).
     4. A cache *key* for that request is derived. This will call the
        instance's ``adjustcachekeystate()`` method so the derivation
        can be influenced.
     5. The cacher is informed of the derived cache key via a call to
        ``setcachekey()``.
     6. The cacher's ``lookup()`` method is called to test for presence of
        the derived key in the cache.
     7. If ``lookup()`` returns a hit, that cached result is used in place
        of invoking the command function. ``__exit__`` is called and the instance
        is discarded.
     8. The command function is invoked.
     9. ``onobject()`` is called for each object emitted by the command
        function.
     10. After the final object is seen, ``onfinished()`` is called.
     11. ``__exit__`` is called to signal the end of use of the instance.
 
     Cache *key* derivation can be influenced by the instance.
 
     Cache keys are initially derived by a deterministic representation of
     the command request. This includes the command name, arguments, protocol
     version, etc. This initial key derivation is performed by CBOR-encoding a
     data structure and feeding that output into a hasher.
 
     Instances of this interface can influence this initial key derivation
     via ``adjustcachekeystate()``.
 
     The instance is informed of the derived cache key via a call to
     ``setcachekey()``. The instance must store the key locally so it can
     be consulted on subsequent operations that may require it.
 
     When constructed, the instance has access to a callable that can be used
     for encoding response objects. This callable receives as its single
     argument an object emitted by a command function. It returns an iterable
     of bytes chunks representing the encoded object. Unless the cacher is
     caching native Python objects in memory or has a way of reconstructing
     the original Python objects, implementations typically call this function
     to produce bytes from the output objects and then store those bytes in
     the cache. When it comes time to re-emit those bytes, they are wrapped
     in a ``wireprototypes.encodedresponse`` instance to tell the output
     layer that they are pre-encoded.
 
     When receiving the objects emitted by the command function, instances
     can choose what to do with those objects. The simplest thing to do is
     re-emit the original objects. They will be forwarded to the output
     layer and will be processed as if the cacher did not exist.
 
     Implementations could also choose to not emit objects - instead locally
     buffering objects or their encoded representation. They could then emit
     a single "coalesced" object when ``onfinished()`` is called. In
     this way, the implementation would function as a filtering layer of
     sorts.
 
     When caching objects, typically the encoded form of the object will
     be stored. Keep in mind that if the original object is forwarded to
     the output layer, it will need to be encoded there as well. For large
     output, this redundant encoding could add overhead. Implementations
     could wrap the encoded object data in ``wireprototypes.encodedresponse``
     instances to avoid this overhead.
     """
 
     def __enter__():
         """Marks the instance as active.
 
         Should return self.
         """
 
     def __exit__(exctype, excvalue, exctb):
         """Called when cacher is no longer used.
 
         This can be used by implementations to perform cleanup actions (e.g.
         disconnecting network sockets, aborting a partially cached response.
         """
 
     def adjustcachekeystate(state):
         """Influences cache key derivation by adjusting state to derive key.
 
         A dict defining the state used to derive the cache key is passed.
 
         Implementations can modify this dict to record additional state that
         is wanted to influence key derivation.
 
         Implementations are *highly* encouraged to not modify or delete
         existing keys.
         """
 
     def setcachekey(key):
         """Record the derived cache key for this request.
 
         Instances may mutate the key for internal usage, as desired. e.g.
         instances may wish to prepend the repo name, introduce path
         components for filesystem or URL addressing, etc. Behavior is up to
         the cache.
 
         Returns a bool indicating if the request is cacheable by this
         instance.
         """
 
     def lookup():
         """Attempt to resolve an entry in the cache.
 
         The instance is instructed to look for the cache key that it was
         informed about via the call to ``setcachekey()``.
 
         If there's no cache hit or the cacher doesn't wish to use the cached
         entry, ``None`` should be returned.
 
         Else, a dict defining the cached result should be returned. The
         dict may have the following keys:
 
         objs
            An iterable of objects that should be sent to the client. That
            iterable of objects is expected to be what the command function
            would return if invoked or an equivalent representation thereof.
         """
 
     def onobject(obj):
         """Called when a new object is emitted from the command function.
 
         Receives as its argument the object that was emitted from the
         command function.
 
         This method returns an iterator of objects to forward to the output
         layer. The easiest implementation is a generator that just
         ``yield obj``.
         """
 
     def onfinished():
         """Called after all objects have been emitted from the command function.
 
         Implementations should return an iterator of objects to forward to
         the output layer.
 
         This method can be a generator.
         """

mercurial/manifest.py

@@ -1,2349 +1,2351 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,
     match as matchmod,
     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')
         nl = len(n)
         flags = n[-1:]
         if flags in _manifestflags:
             n = n[:-1]
             nl -= 1
         else:
             flags = b''
         if nl not in (40, 64):
             raise ValueError(b'Invalid manifest line')
 
         yield f, bin(n), flags
 
 
 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)
         nlpos = data.find(b'\n', pos)
         if zeropos == -1 or nlpos == -1 or nlpos < zeropos:
             raise error.StorageError(b'Invalid manifest line')
         flags = data[nlpos - 1 : nlpos]
         if flags in _manifestflags:
             hlen = nlpos - zeropos - 2
         else:
             hlen = nlpos - zeropos - 1
             flags = b''
         if hlen not in (40, 64):
             raise error.StorageError(b'Invalid manifest line')
         hashval = unhexlify(
             data, self.lm.extrainfo[self.pos], zeropos + 1, hlen
         )
         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)
 
 
 _manifestflags = {b'', b'l', b't', b'x'}
 
 
 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 __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)
         nlpos = data.find(b'\n', zeropos)
         assert 0 <= needle <= len(self.positions)
         assert len(self.extrainfo) == len(self.positions)
         if zeropos == -1 or nlpos == -1 or nlpos < zeropos:
             raise error.StorageError(b'Invalid manifest line')
         hlen = nlpos - zeropos - 1
         flags = data[nlpos - 1 : nlpos]
         if flags in _manifestflags:
             hlen -= 1
         else:
             flags = b''
         if hlen not in (40, 64):
             raise error.StorageError(b'Invalid manifest line')
         hashval = unhexlify(data, self.extrainfo[needle], zeropos + 1, hlen)
         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 len(hashval) not in (20, 32):
             raise TypeError(b"node must be a 20-byte or 32-byte byte string")
         flags = value[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
                     # Hashes are either 20 bytes (old sha1s) or 32
                     # bytes (new non-sha1).
                     hlen = 20
                     if len(t[1]) > 25:
                         hlen = 32
                     if len(t[1]) > hlen:
                         self.extrainfo[i] = ord(t[1][hlen + 1])
                     offset += len(l[-1])
                     i += 1
         self.data = b''.join(l)
         self.hasremovals = False
         self.extradata = []
 
     def _pack(self, d):
         n = d[1]
         assert len(n) in (20, 32)
         return d[0] + b'\x00' + hex(n) + 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)
 
     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 is not None:
             match = matchmod.badmatch(match, lambda path, msg: None)
             sm2 = set(m2.walk(match))
             return {f for f in self.walk(match) if f not in sm2}
         return {f for f in self if f not in m2}
 
     @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):
                 if fn in self:
                     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):
         if flag not in _manifestflags:
             raise TypeError(b"Invalid manifest flag set.")
         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):
         try:
             return self._lm[key][1]
         except KeyError:
             return b''
 
     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
 
 
 @interfaceutil.implementer(repository.imanifestdict)
 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
         subpath = self._subpath
         for d, (node, readsubtree, docopy) in pycompat.iteritems(
             self._lazydirs
         ):
             if docopy:
                 selfdirs[d] = readsubtree(subpath(d), node).copy()
             else:
                 selfdirs[d] = readsubtree(subpath(d), node)
         self._lazydirs = {}
 
     def _loadlazy(self, d):
         v = self._lazydirs.get(d)
         if v:
             node, readsubtree, docopy = v
             if docopy:
                 self._dirs[d] = readsubtree(self._subpath(d), node).copy()
             else:
                 self._dirs[d] = readsubtree(self._subpath(d), 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[0] != v1[0]:
                 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=%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:
             # manifest nodes are either 20 bytes or 32 bytes,
             # depending on the hash in use. Assert this as historically
             # sometimes extra bytes were added.
             assert len(n) in (20, 32)
             self._files[f] = n
         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."""
         if flags not in _manifestflags:
             raise TypeError(b"Invalid manifest flag set.")
         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: (n, r, True)
                     for d, (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.
 
         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):
         """recursively generate a new manifest filtered by the match argument."""
         if match.always():
             return self.copy()
         return self._matches_inner(match)
 
     def _matches_inner(self, match):
         if match.always():
             return self.copy()
 
         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_inner(match)
             if not m._isempty():
                 ret._dirs[dir] = m
 
         if not ret._isempty():
             ret._dirty = True
         return ret
 
     def fastdelta(self, base, changes):
         raise FastdeltaUnavailable()
 
     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
         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] = (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[0], 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[0]
             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:
                     # TODO do we need to do work here for sha1 portability?
                     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
         try:
             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
         except IOError:
             # We could not write the cache (eg: permission error)
             # the content can be missing.
             #
             # We could try harder and see if we could recreate a wcache
             # directory were we coudl write too.
             #
             # XXX the error pass silently, having some way to issue an error
             # log `ui.log` would be nice.
             pass
 
     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
 
 
 class FastdeltaUnavailable(Exception):
     """Exception raised when fastdelta isn't usable on a manifest."""
 
 
 @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,
             persistentnodemap=opener.options.get(b'persistent-nodemap', False),
         )
 
         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,
     ):
         """add some manifest entry in to the manifest log
 
         input:
 
           m:           the manifest dict we want to store
           transaction: the open transaction
           p1:          manifest-node of p1
           p2:          manifest-node of p2
           added:       file added/changed compared to parent
           removed:     file removed compared to parent
 
         tree manifest input:
 
           readtree:    a function to read a subtree
           match:       a filematcher for the subpart of the tree manifest
         """
         try:
             if p1 not in self.fulltextcache:
                 raise FastdeltaUnavailable()
             # 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
             )
         except FastdeltaUnavailable:
             # The first parent manifest isn't already loaded or the
             # manifest implementation doesn't support fastdelta, 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,
+        alwayscache=False,
         addrevisioncb=None,
         duplicaterevisioncb=None,
     ):
         return self._revlog.addgroup(
             deltas,
             linkmapper,
             transaction,
+            alwayscache=alwayscache,
             addrevisioncb=addrevisioncb,
             duplicaterevisioncb=duplicaterevisioncb,
         )
 
     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)
 
     def update_caches(self, transaction):
         return self._rootstore._revlog.update_caches(transaction=transaction)
 
 
 @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 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 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 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 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 readfast(self, shallow=False):
         # special version of readfast since we don't have underlying storage
         return self.read()
 
     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/revlog.py

@@ -1,3089 +1,3090 b''
 # revlog.py - storage back-end 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.
 
 """Storage back-end for Mercurial.
 
 This provides efficient delta storage with O(1) retrieve and append
 and O(changes) merge between branches.
 """
 
 from __future__ import absolute_import
 
 import collections
 import contextlib
 import errno
 import io
 import os
 import struct
 import zlib
 
 # import stuff from node for others to import from revlog
 from .node import (
     bin,
     hex,
     nullhex,
     nullid,
     nullrev,
     short,
     wdirfilenodeids,
     wdirhex,
     wdirid,
     wdirrev,
 )
 from .i18n import _
 from .pycompat import getattr
 from .revlogutils.constants import (
     FLAG_GENERALDELTA,
     FLAG_INLINE_DATA,
     REVLOGV0,
     REVLOGV1,
     REVLOGV1_FLAGS,
     REVLOGV2,
     REVLOGV2_FLAGS,
     REVLOG_DEFAULT_FLAGS,
     REVLOG_DEFAULT_FORMAT,
     REVLOG_DEFAULT_VERSION,
 )
 from .revlogutils.flagutil import (
     REVIDX_DEFAULT_FLAGS,
     REVIDX_ELLIPSIS,
     REVIDX_EXTSTORED,
     REVIDX_FLAGS_ORDER,
     REVIDX_HASCOPIESINFO,
     REVIDX_ISCENSORED,
     REVIDX_RAWTEXT_CHANGING_FLAGS,
     REVIDX_SIDEDATA,
 )
 from .thirdparty import attr
 from . import (
     ancestor,
     dagop,
     error,
     mdiff,
     policy,
     pycompat,
     templatefilters,
     util,
 )
 from .interfaces import (
     repository,
     util as interfaceutil,
 )
 from .revlogutils import (
     deltas as deltautil,
     flagutil,
     nodemap as nodemaputil,
     sidedata as sidedatautil,
 )
 from .utils import (
     storageutil,
     stringutil,
 )
 
 # blanked usage of all the name to prevent pyflakes constraints
 # We need these name available in the module for extensions.
 REVLOGV0
 REVLOGV1
 REVLOGV2
 FLAG_INLINE_DATA
 FLAG_GENERALDELTA
 REVLOG_DEFAULT_FLAGS
 REVLOG_DEFAULT_FORMAT
 REVLOG_DEFAULT_VERSION
 REVLOGV1_FLAGS
 REVLOGV2_FLAGS
 REVIDX_ISCENSORED
 REVIDX_ELLIPSIS
 REVIDX_SIDEDATA
 REVIDX_HASCOPIESINFO
 REVIDX_EXTSTORED
 REVIDX_DEFAULT_FLAGS
 REVIDX_FLAGS_ORDER
 REVIDX_RAWTEXT_CHANGING_FLAGS
 
 parsers = policy.importmod('parsers')
 rustancestor = policy.importrust('ancestor')
 rustdagop = policy.importrust('dagop')
 rustrevlog = policy.importrust('revlog')
 
 # Aliased for performance.
 _zlibdecompress = zlib.decompress
 
 # max size of revlog with inline data
 _maxinline = 131072
 _chunksize = 1048576
 
 # Flag processors for REVIDX_ELLIPSIS.
 def ellipsisreadprocessor(rl, text):
     return text, False, {}
 
 
 def ellipsiswriteprocessor(rl, text, sidedata):
     return text, False
 
 
 def ellipsisrawprocessor(rl, text):
     return False
 
 
 ellipsisprocessor = (
     ellipsisreadprocessor,
     ellipsiswriteprocessor,
     ellipsisrawprocessor,
 )
 
 
 def getoffset(q):
     return int(q >> 16)
 
 
 def gettype(q):
     return int(q & 0xFFFF)
 
 
 def offset_type(offset, type):
     if (type & ~flagutil.REVIDX_KNOWN_FLAGS) != 0:
         raise ValueError(b'unknown revlog index flags')
     return int(int(offset) << 16 | type)
 
 
 def _verify_revision(rl, skipflags, state, node):
     """Verify the integrity of the given revlog ``node`` while providing a hook
     point for extensions to influence the operation."""
     if skipflags:
         state[b'skipread'].add(node)
     else:
         # Side-effect: read content and verify hash.
         rl.revision(node)
 
 
 # True if a fast implementation for persistent-nodemap is available
 #
 # We also consider we have a "fast" implementation in "pure" python because
 # people using pure don't really have performance consideration (and a
 # wheelbarrow of other slowness source)
 HAS_FAST_PERSISTENT_NODEMAP = rustrevlog is not None or util.safehasattr(
     parsers, 'BaseIndexObject'
 )
 
 
 @attr.s(slots=True, frozen=True)
 class _revisioninfo(object):
     """Information about a revision that allows building its fulltext
     node:       expected hash of the revision
     p1, p2:     parent revs of the revision
     btext:      built text cache consisting of a one-element list
     cachedelta: (baserev, uncompressed_delta) or None
     flags:      flags associated to the revision storage
 
     One of btext[0] or cachedelta must be set.
     """
 
     node = attr.ib()
     p1 = attr.ib()
     p2 = attr.ib()
     btext = attr.ib()
     textlen = attr.ib()
     cachedelta = attr.ib()
     flags = attr.ib()
 
 
 @interfaceutil.implementer(repository.irevisiondelta)
 @attr.s(slots=True)
 class revlogrevisiondelta(object):
     node = attr.ib()
     p1node = attr.ib()
     p2node = attr.ib()
     basenode = attr.ib()
     flags = attr.ib()
     baserevisionsize = attr.ib()
     revision = attr.ib()
     delta = attr.ib()
     linknode = attr.ib(default=None)
 
 
 @interfaceutil.implementer(repository.iverifyproblem)
 @attr.s(frozen=True)
 class revlogproblem(object):
     warning = attr.ib(default=None)
     error = attr.ib(default=None)
     node = attr.ib(default=None)
 
 
 # index v0:
 #  4 bytes: offset
 #  4 bytes: compressed length
 #  4 bytes: base rev
 #  4 bytes: link rev
 # 20 bytes: parent 1 nodeid
 # 20 bytes: parent 2 nodeid
 # 20 bytes: nodeid
 indexformatv0 = struct.Struct(b">4l20s20s20s")
 indexformatv0_pack = indexformatv0.pack
 indexformatv0_unpack = indexformatv0.unpack
 
 
 class revlogoldindex(list):
     @property
     def nodemap(self):
         msg = b"index.nodemap is deprecated, use index.[has_node|rev|get_rev]"
         util.nouideprecwarn(msg, b'5.3', stacklevel=2)
         return self._nodemap
 
     @util.propertycache
     def _nodemap(self):
         nodemap = nodemaputil.NodeMap({nullid: nullrev})
         for r in range(0, len(self)):
             n = self[r][7]
             nodemap[n] = r
         return nodemap
 
     def has_node(self, node):
         """return True if the node exist in the index"""
         return node in self._nodemap
 
     def rev(self, node):
         """return a revision for a node
 
         If the node is unknown, raise a RevlogError"""
         return self._nodemap[node]
 
     def get_rev(self, node):
         """return a revision for a node
 
         If the node is unknown, return None"""
         return self._nodemap.get(node)
 
     def append(self, tup):
         self._nodemap[tup[7]] = len(self)
         super(revlogoldindex, self).append(tup)
 
     def __delitem__(self, i):
         if not isinstance(i, slice) or not i.stop == -1 or i.step is not None:
             raise ValueError(b"deleting slices only supports a:-1 with step 1")
         for r in pycompat.xrange(i.start, len(self)):
             del self._nodemap[self[r][7]]
         super(revlogoldindex, self).__delitem__(i)
 
     def clearcaches(self):
         self.__dict__.pop('_nodemap', None)
 
     def __getitem__(self, i):
         if i == -1:
             return (0, 0, 0, -1, -1, -1, -1, nullid)
         return list.__getitem__(self, i)
 
 
 class revlogoldio(object):
     def __init__(self):
         self.size = indexformatv0.size
 
     def parseindex(self, data, inline):
         s = self.size
         index = []
         nodemap = nodemaputil.NodeMap({nullid: nullrev})
         n = off = 0
         l = len(data)
         while off + s <= l:
             cur = data[off : off + s]
             off += s
             e = indexformatv0_unpack(cur)
             # transform to revlogv1 format
             e2 = (
                 offset_type(e[0], 0),
                 e[1],
                 -1,
                 e[2],
                 e[3],
                 nodemap.get(e[4], nullrev),
                 nodemap.get(e[5], nullrev),
                 e[6],
             )
             index.append(e2)
             nodemap[e[6]] = n
             n += 1
 
         index = revlogoldindex(index)
         return index, None
 
     def packentry(self, entry, node, version, rev):
         if gettype(entry[0]):
             raise error.RevlogError(
                 _(b'index entry flags need revlog version 1')
             )
         e2 = (
             getoffset(entry[0]),
             entry[1],
             entry[3],
             entry[4],
             node(entry[5]),
             node(entry[6]),
             entry[7],
         )
         return indexformatv0_pack(*e2)
 
 
 # index ng:
 #  6 bytes: offset
 #  2 bytes: flags
 #  4 bytes: compressed length
 #  4 bytes: uncompressed length
 #  4 bytes: base rev
 #  4 bytes: link rev
 #  4 bytes: parent 1 rev
 #  4 bytes: parent 2 rev
 # 32 bytes: nodeid
 indexformatng = struct.Struct(b">Qiiiiii20s12x")
 indexformatng_pack = indexformatng.pack
 versionformat = struct.Struct(b">I")
 versionformat_pack = versionformat.pack
 versionformat_unpack = versionformat.unpack
 
 # corresponds to uncompressed length of indexformatng (2 gigs, 4-byte
 # signed integer)
 _maxentrysize = 0x7FFFFFFF
 
 
 class revlogio(object):
     def __init__(self):
         self.size = indexformatng.size
 
     def parseindex(self, data, inline):
         # call the C implementation to parse the index data
         index, cache = parsers.parse_index2(data, inline)
         return index, cache
 
     def packentry(self, entry, node, version, rev):
         p = indexformatng_pack(*entry)
         if rev == 0:
             p = versionformat_pack(version) + p[4:]
         return p
 
 
 NodemapRevlogIO = None
 
 if util.safehasattr(parsers, 'parse_index_devel_nodemap'):
 
     class NodemapRevlogIO(revlogio):
         """A debug oriented IO class that return a PersistentNodeMapIndexObject
 
         The PersistentNodeMapIndexObject object is meant to test the persistent nodemap feature.
         """
 
         def parseindex(self, data, inline):
             index, cache = parsers.parse_index_devel_nodemap(data, inline)
             return index, cache
 
 
 class rustrevlogio(revlogio):
     def parseindex(self, data, inline):
         index, cache = super(rustrevlogio, self).parseindex(data, inline)
         return rustrevlog.MixedIndex(index), cache
 
 
 class revlog(object):
     """
     the underlying revision storage object
 
     A revlog consists of two parts, an index and the revision data.
 
     The index is a file with a fixed record size containing
     information on each revision, including its nodeid (hash), the
     nodeids of its parents, the position and offset of its data within
     the data file, and the revision it's based on. Finally, each entry
     contains a linkrev entry that can serve as a pointer to external
     data.
 
     The revision data itself is a linear collection of data chunks.
     Each chunk represents a revision and is usually represented as a
     delta against the previous chunk. To bound lookup time, runs of
     deltas are limited to about 2 times the length of the original
     version data. This makes retrieval of a version proportional to
     its size, or O(1) relative to the number of revisions.
 
     Both pieces of the revlog are written to in an append-only
     fashion, which means we never need to rewrite a file to insert or
     remove data, and can use some simple techniques to avoid the need
     for locking while reading.
 
     If checkambig, indexfile is opened with checkambig=True at
     writing, to avoid file stat ambiguity.
 
     If mmaplargeindex is True, and an mmapindexthreshold is set, the
     index will be mmapped rather than read if it is larger than the
     configured threshold.
 
     If censorable is True, the revlog can have censored revisions.
 
     If `upperboundcomp` is not None, this is the expected maximal gain from
     compression for the data content.
     """
 
     _flagserrorclass = error.RevlogError
 
     def __init__(
         self,
         opener,
         indexfile,
         datafile=None,
         checkambig=False,
         mmaplargeindex=False,
         censorable=False,
         upperboundcomp=None,
         persistentnodemap=False,
     ):
         """
         create a revlog object
 
         opener is a function that abstracts the file opening operation
         and can be used to implement COW semantics or the like.
 
         """
         self.upperboundcomp = upperboundcomp
         self.indexfile = indexfile
         self.datafile = datafile or (indexfile[:-2] + b".d")
         self.nodemap_file = None
         if persistentnodemap:
             if indexfile.endswith(b'.a'):
                 pending_path = indexfile[:-4] + b".n.a"
                 if opener.exists(pending_path):
                     self.nodemap_file = pending_path
                 else:
                     self.nodemap_file = indexfile[:-4] + b".n"
             else:
                 self.nodemap_file = indexfile[:-2] + b".n"
 
         self.opener = opener
         #  When True, indexfile is opened with checkambig=True at writing, to
         #  avoid file stat ambiguity.
         self._checkambig = checkambig
         self._mmaplargeindex = mmaplargeindex
         self._censorable = censorable
         # 3-tuple of (node, rev, text) for a raw revision.
         self._revisioncache = None
         # Maps rev to chain base rev.
         self._chainbasecache = util.lrucachedict(100)
         # 2-tuple of (offset, data) of raw data from the revlog at an offset.
         self._chunkcache = (0, b'')
         # How much data to read and cache into the raw revlog data cache.
         self._chunkcachesize = 65536
         self._maxchainlen = None
         self._deltabothparents = True
         self.index = None
         self._nodemap_docket = None
         # Mapping of partial identifiers to full nodes.
         self._pcache = {}
         # Mapping of revision integer to full node.
         self._compengine = b'zlib'
         self._compengineopts = {}
         self._maxdeltachainspan = -1
         self._withsparseread = False
         self._sparserevlog = False
         self._srdensitythreshold = 0.50
         self._srmingapsize = 262144
 
         # Make copy of flag processors so each revlog instance can support
         # custom flags.
         self._flagprocessors = dict(flagutil.flagprocessors)
 
         # 2-tuple of file handles being used for active writing.
         self._writinghandles = None
 
         self._loadindex()
 
     def _loadindex(self):
         mmapindexthreshold = None
         opts = self.opener.options
 
         if b'revlogv2' in opts:
             newversionflags = REVLOGV2 | FLAG_INLINE_DATA
         elif b'revlogv1' in opts:
             newversionflags = REVLOGV1 | FLAG_INLINE_DATA
             if b'generaldelta' in opts:
                 newversionflags |= FLAG_GENERALDELTA
         elif b'revlogv0' in self.opener.options:
             newversionflags = REVLOGV0
         else:
             newversionflags = REVLOG_DEFAULT_VERSION
 
         if b'chunkcachesize' in opts:
             self._chunkcachesize = opts[b'chunkcachesize']
         if b'maxchainlen' in opts:
             self._maxchainlen = opts[b'maxchainlen']
         if b'deltabothparents' in opts:
             self._deltabothparents = opts[b'deltabothparents']
         self._lazydelta = bool(opts.get(b'lazydelta', True))
         self._lazydeltabase = False
         if self._lazydelta:
             self._lazydeltabase = bool(opts.get(b'lazydeltabase', False))
         if b'compengine' in opts:
             self._compengine = opts[b'compengine']
         if b'zlib.level' in opts:
             self._compengineopts[b'zlib.level'] = opts[b'zlib.level']
         if b'zstd.level' in opts:
             self._compengineopts[b'zstd.level'] = opts[b'zstd.level']
         if b'maxdeltachainspan' in opts:
             self._maxdeltachainspan = opts[b'maxdeltachainspan']
         if self._mmaplargeindex and b'mmapindexthreshold' in opts:
             mmapindexthreshold = opts[b'mmapindexthreshold']
         self.hassidedata = bool(opts.get(b'side-data', False))
         if self.hassidedata:
             self._flagprocessors[REVIDX_SIDEDATA] = sidedatautil.processors
         self._sparserevlog = bool(opts.get(b'sparse-revlog', False))
         withsparseread = bool(opts.get(b'with-sparse-read', False))
         # sparse-revlog forces sparse-read
         self._withsparseread = self._sparserevlog or withsparseread
         if b'sparse-read-density-threshold' in opts:
             self._srdensitythreshold = opts[b'sparse-read-density-threshold']
         if b'sparse-read-min-gap-size' in opts:
             self._srmingapsize = opts[b'sparse-read-min-gap-size']
         if opts.get(b'enableellipsis'):
             self._flagprocessors[REVIDX_ELLIPSIS] = ellipsisprocessor
 
         # revlog v0 doesn't have flag processors
         for flag, processor in pycompat.iteritems(
             opts.get(b'flagprocessors', {})
         ):
             flagutil.insertflagprocessor(flag, processor, self._flagprocessors)
 
         if self._chunkcachesize <= 0:
             raise error.RevlogError(
                 _(b'revlog chunk cache size %r is not greater than 0')
                 % self._chunkcachesize
             )
         elif self._chunkcachesize & (self._chunkcachesize - 1):
             raise error.RevlogError(
                 _(b'revlog chunk cache size %r is not a power of 2')
                 % self._chunkcachesize
             )
 
         indexdata = b''
         self._initempty = True
         try:
             with self._indexfp() as f:
                 if (
                     mmapindexthreshold is not None
                     and self.opener.fstat(f).st_size >= mmapindexthreshold
                 ):
                     # TODO: should .close() to release resources without
                     # relying on Python GC
                     indexdata = util.buffer(util.mmapread(f))
                 else:
                     indexdata = f.read()
             if len(indexdata) > 0:
                 versionflags = versionformat_unpack(indexdata[:4])[0]
                 self._initempty = False
             else:
                 versionflags = newversionflags
         except IOError as inst:
             if inst.errno != errno.ENOENT:
                 raise
 
             versionflags = newversionflags
 
         self.version = versionflags
 
         flags = versionflags & ~0xFFFF
         fmt = versionflags & 0xFFFF
 
         if fmt == REVLOGV0:
             if flags:
                 raise error.RevlogError(
                     _(b'unknown flags (%#04x) in version %d revlog %s')
                     % (flags >> 16, fmt, self.indexfile)
                 )
 
             self._inline = False
             self._generaldelta = False
 
         elif fmt == REVLOGV1:
             if flags & ~REVLOGV1_FLAGS:
                 raise error.RevlogError(
                     _(b'unknown flags (%#04x) in version %d revlog %s')
                     % (flags >> 16, fmt, self.indexfile)
                 )
 
             self._inline = versionflags & FLAG_INLINE_DATA
             self._generaldelta = versionflags & FLAG_GENERALDELTA
 
         elif fmt == REVLOGV2:
             if flags & ~REVLOGV2_FLAGS:
                 raise error.RevlogError(
                     _(b'unknown flags (%#04x) in version %d revlog %s')
                     % (flags >> 16, fmt, self.indexfile)
                 )
 
             self._inline = versionflags & FLAG_INLINE_DATA
             # generaldelta implied by version 2 revlogs.
             self._generaldelta = True
 
         else:
             raise error.RevlogError(
                 _(b'unknown version (%d) in revlog %s') % (fmt, self.indexfile)
             )
         # sparse-revlog can't be on without general-delta (issue6056)
         if not self._generaldelta:
             self._sparserevlog = False
 
         self._storedeltachains = True
 
         devel_nodemap = (
             self.nodemap_file
             and opts.get(b'devel-force-nodemap', False)
             and NodemapRevlogIO is not None
         )
 
         use_rust_index = False
         if rustrevlog is not None:
             if self.nodemap_file is not None:
                 use_rust_index = True
             else:
                 use_rust_index = self.opener.options.get(b'rust.index')
 
         self._io = revlogio()
         if self.version == REVLOGV0:
             self._io = revlogoldio()
         elif devel_nodemap:
             self._io = NodemapRevlogIO()
         elif use_rust_index:
             self._io = rustrevlogio()
         try:
             d = self._io.parseindex(indexdata, self._inline)
             index, _chunkcache = d
             use_nodemap = (
                 not self._inline
                 and self.nodemap_file is not None
                 and util.safehasattr(index, 'update_nodemap_data')
             )
             if use_nodemap:
                 nodemap_data = nodemaputil.persisted_data(self)
                 if nodemap_data is not None:
                     docket = nodemap_data[0]
                     if (
                         len(d[0]) > docket.tip_rev
                         and d[0][docket.tip_rev][7] == docket.tip_node
                     ):
                         # no changelog tampering
                         self._nodemap_docket = docket
                         index.update_nodemap_data(*nodemap_data)
         except (ValueError, IndexError):
             raise error.RevlogError(
                 _(b"index %s is corrupted") % self.indexfile
             )
         self.index, self._chunkcache = d
         if not self._chunkcache:
             self._chunkclear()
         # revnum -> (chain-length, sum-delta-length)
         self._chaininfocache = util.lrucachedict(500)
         # revlog header -> revlog compressor
         self._decompressors = {}
 
     @util.propertycache
     def _compressor(self):
         engine = util.compengines[self._compengine]
         return engine.revlogcompressor(self._compengineopts)
 
     def _indexfp(self, mode=b'r'):
         """file object for the revlog's index file"""
         args = {'mode': mode}
         if mode != b'r':
             args['checkambig'] = self._checkambig
         if mode == b'w':
             args['atomictemp'] = True
         return self.opener(self.indexfile, **args)
 
     def _datafp(self, mode=b'r'):
         """file object for the revlog's data file"""
         return self.opener(self.datafile, mode=mode)
 
     @contextlib.contextmanager
     def _datareadfp(self, existingfp=None):
         """file object suitable to read data"""
         # Use explicit file handle, if given.
         if existingfp is not None:
             yield existingfp
 
         # Use a file handle being actively used for writes, if available.
         # There is some danger to doing this because reads will seek the
         # file. However, _writeentry() performs a SEEK_END before all writes,
         # so we should be safe.
         elif self._writinghandles:
             if self._inline:
                 yield self._writinghandles[0]
             else:
                 yield self._writinghandles[1]
 
         # Otherwise open a new file handle.
         else:
             if self._inline:
                 func = self._indexfp
             else:
                 func = self._datafp
             with func() as fp:
                 yield fp
 
     def tiprev(self):
         return len(self.index) - 1
 
     def tip(self):
         return self.node(self.tiprev())
 
     def __contains__(self, rev):
         return 0 <= rev < len(self)
 
     def __len__(self):
         return len(self.index)
 
     def __iter__(self):
         return iter(pycompat.xrange(len(self)))
 
     def revs(self, start=0, stop=None):
         """iterate over all rev in this revlog (from start to stop)"""
         return storageutil.iterrevs(len(self), start=start, stop=stop)
 
     @property
     def nodemap(self):
         msg = (
             b"revlog.nodemap is deprecated, "
             b"use revlog.index.[has_node|rev|get_rev]"
         )
         util.nouideprecwarn(msg, b'5.3', stacklevel=2)
         return self.index.nodemap
 
     @property
     def _nodecache(self):
         msg = b"revlog._nodecache is deprecated, use revlog.index.nodemap"
         util.nouideprecwarn(msg, b'5.3', stacklevel=2)
         return self.index.nodemap
 
     def hasnode(self, node):
         try:
             self.rev(node)
             return True
         except KeyError:
             return False
 
     def candelta(self, baserev, rev):
         """whether two revisions (baserev, rev) can be delta-ed or not"""
         # Disable delta if either rev requires a content-changing flag
         # processor (ex. LFS). This is because such flag processor can alter
         # the rawtext content that the delta will be based on, and two clients
         # could have a same revlog node with different flags (i.e. different
         # rawtext contents) and the delta could be incompatible.
         if (self.flags(baserev) & REVIDX_RAWTEXT_CHANGING_FLAGS) or (
             self.flags(rev) & REVIDX_RAWTEXT_CHANGING_FLAGS
         ):
             return False
         return True
 
     def update_caches(self, transaction):
         if self.nodemap_file is not None:
             if transaction is None:
                 nodemaputil.update_persistent_nodemap(self)
             else:
                 nodemaputil.setup_persistent_nodemap(transaction, self)
 
     def clearcaches(self):
         self._revisioncache = None
         self._chainbasecache.clear()
         self._chunkcache = (0, b'')
         self._pcache = {}
         self._nodemap_docket = None
         self.index.clearcaches()
         # The python code is the one responsible for validating the docket, we
         # end up having to refresh it here.
         use_nodemap = (
             not self._inline
             and self.nodemap_file is not None
             and util.safehasattr(self.index, 'update_nodemap_data')
         )
         if use_nodemap:
             nodemap_data = nodemaputil.persisted_data(self)
             if nodemap_data is not None:
                 self._nodemap_docket = nodemap_data[0]
                 self.index.update_nodemap_data(*nodemap_data)
 
     def rev(self, node):
         try:
             return self.index.rev(node)
         except TypeError:
             raise
         except error.RevlogError:
             # parsers.c radix tree lookup failed
             if node == wdirid or node in wdirfilenodeids:
                 raise error.WdirUnsupported
             raise error.LookupError(node, self.indexfile, _(b'no node'))
 
     # Accessors for index entries.
 
     # First tuple entry is 8 bytes. First 6 bytes are offset. Last 2 bytes
     # are flags.
     def start(self, rev):
         return int(self.index[rev][0] >> 16)
 
     def flags(self, rev):
         return self.index[rev][0] & 0xFFFF
 
     def length(self, rev):
         return self.index[rev][1]
 
     def rawsize(self, rev):
         """return the length of the uncompressed text for a given revision"""
         l = self.index[rev][2]
         if l >= 0:
             return l
 
         t = self.rawdata(rev)
         return len(t)
 
     def size(self, rev):
         """length of non-raw text (processed by a "read" flag processor)"""
         # fast path: if no "read" flag processor could change the content,
         # size is rawsize. note: ELLIPSIS is known to not change the content.
         flags = self.flags(rev)
         if flags & (flagutil.REVIDX_KNOWN_FLAGS ^ REVIDX_ELLIPSIS) == 0:
             return self.rawsize(rev)
 
         return len(self.revision(rev, raw=False))
 
     def chainbase(self, rev):
         base = self._chainbasecache.get(rev)
         if base is not None:
             return base
 
         index = self.index
         iterrev = rev
         base = index[iterrev][3]
         while base != iterrev:
             iterrev = base
             base = index[iterrev][3]
 
         self._chainbasecache[rev] = base
         return base
 
     def linkrev(self, rev):
         return self.index[rev][4]
 
     def parentrevs(self, rev):
         try:
             entry = self.index[rev]
         except IndexError:
             if rev == wdirrev:
                 raise error.WdirUnsupported
             raise
 
         return entry[5], entry[6]
 
     # fast parentrevs(rev) where rev isn't filtered
     _uncheckedparentrevs = parentrevs
 
     def node(self, rev):
         try:
             return self.index[rev][7]
         except IndexError:
             if rev == wdirrev:
                 raise error.WdirUnsupported
             raise
 
     # Derived from index values.
 
     def end(self, rev):
         return self.start(rev) + self.length(rev)
 
     def parents(self, node):
         i = self.index
         d = i[self.rev(node)]
         return i[d[5]][7], i[d[6]][7]  # map revisions to nodes inline
 
     def chainlen(self, rev):
         return self._chaininfo(rev)[0]
 
     def _chaininfo(self, rev):
         chaininfocache = self._chaininfocache
         if rev in chaininfocache:
             return chaininfocache[rev]
         index = self.index
         generaldelta = self._generaldelta
         iterrev = rev
         e = index[iterrev]
         clen = 0
         compresseddeltalen = 0
         while iterrev != e[3]:
             clen += 1
             compresseddeltalen += e[1]
             if generaldelta:
                 iterrev = e[3]
             else:
                 iterrev -= 1
             if iterrev in chaininfocache:
                 t = chaininfocache[iterrev]
                 clen += t[0]
                 compresseddeltalen += t[1]
                 break
             e = index[iterrev]
         else:
             # Add text length of base since decompressing that also takes
             # work. For cache hits the length is already included.
             compresseddeltalen += e[1]
         r = (clen, compresseddeltalen)
         chaininfocache[rev] = r
         return r
 
     def _deltachain(self, rev, stoprev=None):
         """Obtain the delta chain for a revision.
 
         ``stoprev`` specifies a revision to stop at. If not specified, we
         stop at the base of the chain.
 
         Returns a 2-tuple of (chain, stopped) where ``chain`` is a list of
         revs in ascending order and ``stopped`` is a bool indicating whether
         ``stoprev`` was hit.
         """
         # Try C implementation.
         try:
             return self.index.deltachain(rev, stoprev, self._generaldelta)
         except AttributeError:
             pass
 
         chain = []
 
         # Alias to prevent attribute lookup in tight loop.
         index = self.index
         generaldelta = self._generaldelta
 
         iterrev = rev
         e = index[iterrev]
         while iterrev != e[3] and iterrev != stoprev:
             chain.append(iterrev)
             if generaldelta:
                 iterrev = e[3]
             else:
                 iterrev -= 1
             e = index[iterrev]
 
         if iterrev == stoprev:
             stopped = True
         else:
             chain.append(iterrev)
             stopped = False
 
         chain.reverse()
         return chain, stopped
 
     def ancestors(self, revs, stoprev=0, inclusive=False):
         """Generate the ancestors of 'revs' in reverse revision order.
         Does not generate revs lower than stoprev.
 
         See the documentation for ancestor.lazyancestors for more details."""
 
         # first, make sure start revisions aren't filtered
         revs = list(revs)
         checkrev = self.node
         for r in revs:
             checkrev(r)
         # and we're sure ancestors aren't filtered as well
 
         if rustancestor is not None:
             lazyancestors = rustancestor.LazyAncestors
             arg = self.index
         else:
             lazyancestors = ancestor.lazyancestors
             arg = self._uncheckedparentrevs
         return lazyancestors(arg, revs, stoprev=stoprev, inclusive=inclusive)
 
     def descendants(self, revs):
         return dagop.descendantrevs(revs, self.revs, self.parentrevs)
 
     def findcommonmissing(self, common=None, heads=None):
         """Return a tuple of the ancestors of common and the ancestors of heads
         that are not ancestors of common. In revset terminology, we return the
         tuple:
 
           ::common, (::heads) - (::common)
 
         The list is sorted by revision number, meaning it is
         topologically sorted.
 
         'heads' and 'common' are both lists of node IDs.  If heads is
         not supplied, uses all of the revlog's heads.  If common is not
         supplied, uses nullid."""
         if common is None:
             common = [nullid]
         if heads is None:
             heads = self.heads()
 
         common = [self.rev(n) for n in common]
         heads = [self.rev(n) for n in heads]
 
         # we want the ancestors, but inclusive
         class lazyset(object):
             def __init__(self, lazyvalues):
                 self.addedvalues = set()
                 self.lazyvalues = lazyvalues
 
             def __contains__(self, value):
                 return value in self.addedvalues or value in self.lazyvalues
 
             def __iter__(self):
                 added = self.addedvalues
                 for r in added:
                     yield r
                 for r in self.lazyvalues:
                     if not r in added:
                         yield r
 
             def add(self, value):
                 self.addedvalues.add(value)
 
             def update(self, values):
                 self.addedvalues.update(values)
 
         has = lazyset(self.ancestors(common))
         has.add(nullrev)
         has.update(common)
 
         # take all ancestors from heads that aren't in has
         missing = set()
         visit = collections.deque(r for r in heads if r not in has)
         while visit:
             r = visit.popleft()
             if r in missing:
                 continue
             else:
                 missing.add(r)
                 for p in self.parentrevs(r):
                     if p not in has:
                         visit.append(p)
         missing = list(missing)
         missing.sort()
         return has, [self.node(miss) for miss in missing]
 
     def incrementalmissingrevs(self, common=None):
         """Return an object that can be used to incrementally compute the
         revision numbers of the ancestors of arbitrary sets that are not
         ancestors of common. This is an ancestor.incrementalmissingancestors
         object.
 
         'common' is a list of revision numbers. If common is not supplied, uses
         nullrev.
         """
         if common is None:
             common = [nullrev]
 
         if rustancestor is not None:
             return rustancestor.MissingAncestors(self.index, common)
         return ancestor.incrementalmissingancestors(self.parentrevs, common)
 
     def findmissingrevs(self, common=None, heads=None):
         """Return the revision numbers of the ancestors of heads that
         are not ancestors of common.
 
         More specifically, return a list of revision numbers corresponding to
         nodes N such that every N satisfies the following constraints:
 
           1. N is an ancestor of some node in 'heads'
           2. N is not an ancestor of any node in 'common'
 
         The list is sorted by revision number, meaning it is
         topologically sorted.
 
         'heads' and 'common' are both lists of revision numbers.  If heads is
         not supplied, uses all of the revlog's heads.  If common is not
         supplied, uses nullid."""
         if common is None:
             common = [nullrev]
         if heads is None:
             heads = self.headrevs()
 
         inc = self.incrementalmissingrevs(common=common)
         return inc.missingancestors(heads)
 
     def findmissing(self, common=None, heads=None):
         """Return the ancestors of heads that are not ancestors of common.
 
         More specifically, return a list of nodes N such that every N
         satisfies the following constraints:
 
           1. N is an ancestor of some node in 'heads'
           2. N is not an ancestor of any node in 'common'
 
         The list is sorted by revision number, meaning it is
         topologically sorted.
 
         'heads' and 'common' are both lists of node IDs.  If heads is
         not supplied, uses all of the revlog's heads.  If common is not
         supplied, uses nullid."""
         if common is None:
             common = [nullid]
         if heads is None:
             heads = self.heads()
 
         common = [self.rev(n) for n in common]
         heads = [self.rev(n) for n in heads]
 
         inc = self.incrementalmissingrevs(common=common)
         return [self.node(r) for r in inc.missingancestors(heads)]
 
     def nodesbetween(self, roots=None, heads=None):
         """Return a topological path from 'roots' to 'heads'.
 
         Return a tuple (nodes, outroots, outheads) where 'nodes' is a
         topologically sorted list of all nodes N that satisfy both of
         these constraints:
 
           1. N is a descendant of some node in 'roots'
           2. N is an ancestor of some node in 'heads'
 
         Every node is considered to be both a descendant and an ancestor
         of itself, so every reachable node in 'roots' and 'heads' will be
         included in 'nodes'.
 
         'outroots' is the list of reachable nodes in 'roots', i.e., the
         subset of 'roots' that is returned in 'nodes'.  Likewise,
         'outheads' is the subset of 'heads' that is also in 'nodes'.
 
         'roots' and 'heads' are both lists of node IDs.  If 'roots' is
         unspecified, uses nullid as the only root.  If 'heads' is
         unspecified, uses list of all of the revlog's heads."""
         nonodes = ([], [], [])
         if roots is not None:
             roots = list(roots)
             if not roots:
                 return nonodes
             lowestrev = min([self.rev(n) for n in roots])
         else:
             roots = [nullid]  # Everybody's a descendant of nullid
             lowestrev = nullrev
         if (lowestrev == nullrev) and (heads is None):
             # We want _all_ the nodes!
             return ([self.node(r) for r in self], [nullid], list(self.heads()))
         if heads is None:
             # All nodes are ancestors, so the latest ancestor is the last
             # node.
             highestrev = len(self) - 1
             # Set ancestors to None to signal that every node is an ancestor.
             ancestors = None
             # Set heads to an empty dictionary for later discovery of heads
             heads = {}
         else:
             heads = list(heads)
             if not heads:
                 return nonodes
             ancestors = set()
             # Turn heads into a dictionary so we can remove 'fake' heads.
             # Also, later we will be using it to filter out the heads we can't
             # find from roots.
             heads = dict.fromkeys(heads, False)
             # Start at the top and keep marking parents until we're done.
             nodestotag = set(heads)
             # Remember where the top was so we can use it as a limit later.
             highestrev = max([self.rev(n) for n in nodestotag])
             while nodestotag:
                 # grab a node to tag
                 n = nodestotag.pop()
                 # Never tag nullid
                 if n == nullid:
                     continue
                 # A node's revision number represents its place in a
                 # topologically sorted list of nodes.
                 r = self.rev(n)
                 if r >= lowestrev:
                     if n not in ancestors:
                         # If we are possibly a descendant of one of the roots
                         # and we haven't already been marked as an ancestor
                         ancestors.add(n)  # Mark as ancestor
                         # Add non-nullid parents to list of nodes to tag.
                         nodestotag.update(
                             [p for p in self.parents(n) if p != nullid]
                         )
                     elif n in heads:  # We've seen it before, is it a fake head?
                         # So it is, real heads should not be the ancestors of
                         # any other heads.
                         heads.pop(n)
             if not ancestors:
                 return nonodes
             # Now that we have our set of ancestors, we want to remove any
             # roots that are not ancestors.
 
             # If one of the roots was nullid, everything is included anyway.
             if lowestrev > nullrev:
                 # But, since we weren't, let's recompute the lowest rev to not
                 # include roots that aren't ancestors.
 
                 # Filter out roots that aren't ancestors of heads
                 roots = [root for root in roots if root in ancestors]
                 # Recompute the lowest revision
                 if roots:
                     lowestrev = min([self.rev(root) for root in roots])
                 else:
                     # No more roots?  Return empty list
                     return nonodes
             else:
                 # We are descending from nullid, and don't need to care about
                 # any other roots.
                 lowestrev = nullrev
                 roots = [nullid]
         # Transform our roots list into a set.
         descendants = set(roots)
         # Also, keep the original roots so we can filter out roots that aren't
         # 'real' roots (i.e. are descended from other roots).
         roots = descendants.copy()
         # Our topologically sorted list of output nodes.
         orderedout = []
         # Don't start at nullid since we don't want nullid in our output list,
         # and if nullid shows up in descendants, empty parents will look like
         # they're descendants.
         for r in self.revs(start=max(lowestrev, 0), stop=highestrev + 1):
             n = self.node(r)
             isdescendant = False
             if lowestrev == nullrev:  # Everybody is a descendant of nullid
                 isdescendant = True
             elif n in descendants:
                 # n is already a descendant
                 isdescendant = True
                 # This check only needs to be done here because all the roots
                 # will start being marked is descendants before the loop.
                 if n in roots:
                     # If n was a root, check if it's a 'real' root.
                     p = tuple(self.parents(n))
                     # If any of its parents are descendants, it's not a root.
                     if (p[0] in descendants) or (p[1] in descendants):
                         roots.remove(n)
             else:
                 p = tuple(self.parents(n))
                 # A node is a descendant if either of its parents are
                 # descendants.  (We seeded the dependents list with the roots
                 # up there, remember?)
                 if (p[0] in descendants) or (p[1] in descendants):
                     descendants.add(n)
                     isdescendant = True
             if isdescendant and ((ancestors is None) or (n in ancestors)):
                 # Only include nodes that are both descendants and ancestors.
                 orderedout.append(n)
                 if (ancestors is not None) and (n in heads):
                     # We're trying to figure out which heads are reachable
                     # from roots.
                     # Mark this head as having been reached
                     heads[n] = True
                 elif ancestors is None:
                     # Otherwise, we're trying to discover the heads.
                     # Assume this is a head because if it isn't, the next step
                     # will eventually remove it.
                     heads[n] = True
                     # But, obviously its parents aren't.
                     for p in self.parents(n):
                         heads.pop(p, None)
         heads = [head for head, flag in pycompat.iteritems(heads) if flag]
         roots = list(roots)
         assert orderedout
         assert roots
         assert heads
         return (orderedout, roots, heads)
 
     def headrevs(self, revs=None):
         if revs is None:
             try:
                 return self.index.headrevs()
             except AttributeError:
                 return self._headrevs()
         if rustdagop is not None:
             return rustdagop.headrevs(self.index, revs)
         return dagop.headrevs(revs, self._uncheckedparentrevs)
 
     def computephases(self, roots):
         return self.index.computephasesmapsets(roots)
 
     def _headrevs(self):
         count = len(self)
         if not count:
             return [nullrev]
         # we won't iter over filtered rev so nobody is a head at start
         ishead = [0] * (count + 1)
         index = self.index
         for r in self:
             ishead[r] = 1  # I may be an head
             e = index[r]
             ishead[e[5]] = ishead[e[6]] = 0  # my parent are not
         return [r for r, val in enumerate(ishead) if val]
 
     def heads(self, start=None, stop=None):
         """return the list of all nodes that have no children
 
         if start is specified, only heads that are descendants of
         start will be returned
         if stop is specified, it will consider all the revs from stop
         as if they had no children
         """
         if start is None and stop is None:
             if not len(self):
                 return [nullid]
             return [self.node(r) for r in self.headrevs()]
 
         if start is None:
             start = nullrev
         else:
             start = self.rev(start)
 
         stoprevs = {self.rev(n) for n in stop or []}
 
         revs = dagop.headrevssubset(
             self.revs, self.parentrevs, startrev=start, stoprevs=stoprevs
         )
 
         return [self.node(rev) for rev in revs]
 
     def children(self, node):
         """find the children of a given node"""
         c = []
         p = self.rev(node)
         for r in self.revs(start=p + 1):
             prevs = [pr for pr in self.parentrevs(r) if pr != nullrev]
             if prevs:
                 for pr in prevs:
                     if pr == p:
                         c.append(self.node(r))
             elif p == nullrev:
                 c.append(self.node(r))
         return c
 
     def commonancestorsheads(self, a, b):
         """calculate all the heads of the common ancestors of nodes a and b"""
         a, b = self.rev(a), self.rev(b)
         ancs = self._commonancestorsheads(a, b)
         return pycompat.maplist(self.node, ancs)
 
     def _commonancestorsheads(self, *revs):
         """calculate all the heads of the common ancestors of revs"""
         try:
             ancs = self.index.commonancestorsheads(*revs)
         except (AttributeError, OverflowError):  # C implementation failed
             ancs = ancestor.commonancestorsheads(self.parentrevs, *revs)
         return ancs
 
     def isancestor(self, a, b):
         """return True if node a is an ancestor of node b
 
         A revision is considered an ancestor of itself."""
         a, b = self.rev(a), self.rev(b)
         return self.isancestorrev(a, b)
 
     def isancestorrev(self, a, b):
         """return True if revision a is an ancestor of revision b
 
         A revision is considered an ancestor of itself.
 
         The implementation of this is trivial but the use of
         reachableroots is not."""
         if a == nullrev:
             return True
         elif a == b:
             return True
         elif a > b:
             return False
         return bool(self.reachableroots(a, [b], [a], includepath=False))
 
     def reachableroots(self, minroot, heads, roots, includepath=False):
         """return (heads(::(<roots> and <roots>::<heads>)))
 
         If includepath is True, return (<roots>::<heads>)."""
         try:
             return self.index.reachableroots2(
                 minroot, heads, roots, includepath
             )
         except AttributeError:
             return dagop._reachablerootspure(
                 self.parentrevs, minroot, roots, heads, includepath
             )
 
     def ancestor(self, a, b):
         """calculate the "best" common ancestor of nodes a and b"""
 
         a, b = self.rev(a), self.rev(b)
         try:
             ancs = self.index.ancestors(a, b)
         except (AttributeError, OverflowError):
             ancs = ancestor.ancestors(self.parentrevs, a, b)
         if ancs:
             # choose a consistent winner when there's a tie
             return min(map(self.node, ancs))
         return nullid
 
     def _match(self, id):
         if isinstance(id, int):
             # rev
             return self.node(id)
         if len(id) == 20:
             # possibly a binary node
             # odds of a binary node being all hex in ASCII are 1 in 10**25
             try:
                 node = id
                 self.rev(node)  # quick search the index
                 return node
             except error.LookupError:
                 pass  # may be partial hex id
         try:
             # str(rev)
             rev = int(id)
             if b"%d" % rev != id:
                 raise ValueError
             if rev < 0:
                 rev = len(self) + rev
             if rev < 0 or rev >= len(self):
                 raise ValueError
             return self.node(rev)
         except (ValueError, OverflowError):
             pass
         if len(id) == 40:
             try:
                 # a full hex nodeid?
                 node = bin(id)
                 self.rev(node)
                 return node
             except (TypeError, error.LookupError):
                 pass
 
     def _partialmatch(self, id):
         # we don't care wdirfilenodeids as they should be always full hash
         maybewdir = wdirhex.startswith(id)
         try:
             partial = self.index.partialmatch(id)
             if partial and self.hasnode(partial):
                 if maybewdir:
                     # single 'ff...' match in radix tree, ambiguous with wdir
                     raise error.RevlogError
                 return partial
             if maybewdir:
                 # no 'ff...' match in radix tree, wdir identified
                 raise error.WdirUnsupported
             return None
         except error.RevlogError:
             # parsers.c radix tree lookup gave multiple matches
             # fast path: for unfiltered changelog, radix tree is accurate
             if not getattr(self, 'filteredrevs', None):
                 raise error.AmbiguousPrefixLookupError(
                     id, self.indexfile, _(b'ambiguous identifier')
                 )
             # fall through to slow path that filters hidden revisions
         except (AttributeError, ValueError):
             # we are pure python, or key was too short to search radix tree
             pass
 
         if id in self._pcache:
             return self._pcache[id]
 
         if len(id) <= 40:
             try:
                 # hex(node)[:...]
                 l = len(id) // 2  # grab an even number of digits
                 prefix = bin(id[: l * 2])
                 nl = [e[7] for e in self.index if e[7].startswith(prefix)]
                 nl = [
                     n for n in nl if hex(n).startswith(id) and self.hasnode(n)
                 ]
                 if nullhex.startswith(id):
                     nl.append(nullid)
                 if len(nl) > 0:
                     if len(nl) == 1 and not maybewdir:
                         self._pcache[id] = nl[0]
                         return nl[0]
                     raise error.AmbiguousPrefixLookupError(
                         id, self.indexfile, _(b'ambiguous identifier')
                     )
                 if maybewdir:
                     raise error.WdirUnsupported
                 return None
             except TypeError:
                 pass
 
     def lookup(self, id):
         """locate a node based on:
         - revision number or str(revision number)
         - nodeid or subset of hex nodeid
         """
         n = self._match(id)
         if n is not None:
             return n
         n = self._partialmatch(id)
         if n:
             return n
 
         raise error.LookupError(id, self.indexfile, _(b'no match found'))
 
     def shortest(self, node, minlength=1):
         """Find the shortest unambiguous prefix that matches node."""
 
         def isvalid(prefix):
             try:
                 matchednode = self._partialmatch(prefix)
             except error.AmbiguousPrefixLookupError:
                 return False
             except error.WdirUnsupported:
                 # single 'ff...' match
                 return True
             if matchednode is None:
                 raise error.LookupError(node, self.indexfile, _(b'no node'))
             return True
 
         def maybewdir(prefix):
             return all(c == b'f' for c in pycompat.iterbytestr(prefix))
 
         hexnode = hex(node)
 
         def disambiguate(hexnode, minlength):
             """Disambiguate against wdirid."""
             for length in range(minlength, len(hexnode) + 1):
                 prefix = hexnode[:length]
                 if not maybewdir(prefix):
                     return prefix
 
         if not getattr(self, 'filteredrevs', None):
             try:
                 length = max(self.index.shortest(node), minlength)
                 return disambiguate(hexnode, length)
             except error.RevlogError:
                 if node != wdirid:
                     raise error.LookupError(node, self.indexfile, _(b'no node'))
             except AttributeError:
                 # Fall through to pure code
                 pass
 
         if node == wdirid:
             for length in range(minlength, len(hexnode) + 1):
                 prefix = hexnode[:length]
                 if isvalid(prefix):
                     return prefix
 
         for length in range(minlength, len(hexnode) + 1):
             prefix = hexnode[:length]
             if isvalid(prefix):
                 return disambiguate(hexnode, length)
 
     def cmp(self, node, text):
         """compare text with a given file revision
 
         returns True if text is different than what is stored.
         """
         p1, p2 = self.parents(node)
         return storageutil.hashrevisionsha1(text, p1, p2) != node
 
     def _cachesegment(self, offset, data):
         """Add a segment to the revlog cache.
 
         Accepts an absolute offset and the data that is at that location.
         """
         o, d = self._chunkcache
         # try to add to existing cache
         if o + len(d) == offset and len(d) + len(data) < _chunksize:
             self._chunkcache = o, d + data
         else:
             self._chunkcache = offset, data
 
     def _readsegment(self, offset, length, df=None):
         """Load a segment of raw data from the revlog.
 
         Accepts an absolute offset, length to read, and an optional existing
         file handle to read from.
 
         If an existing file handle is passed, it will be seeked and the
         original seek position will NOT be restored.
 
         Returns a str or buffer of raw byte data.
 
         Raises if the requested number of bytes could not be read.
         """
         # Cache data both forward and backward around the requested
         # data, in a fixed size window. This helps speed up operations
         # involving reading the revlog backwards.
         cachesize = self._chunkcachesize
         realoffset = offset & ~(cachesize - 1)
         reallength = (
             (offset + length + cachesize) & ~(cachesize - 1)
         ) - realoffset
         with self._datareadfp(df) as df:
             df.seek(realoffset)
             d = df.read(reallength)
 
         self._cachesegment(realoffset, d)
         if offset != realoffset or reallength != length:
             startoffset = offset - realoffset
             if len(d) - startoffset < length:
                 raise error.RevlogError(
                     _(
                         b'partial read of revlog %s; expected %d bytes from '
                         b'offset %d, got %d'
                     )
                     % (
                         self.indexfile if self._inline else self.datafile,
                         length,
                         realoffset,
                         len(d) - startoffset,
                     )
                 )
 
             return util.buffer(d, startoffset, length)
 
         if len(d) < length:
             raise error.RevlogError(
                 _(
                     b'partial read of revlog %s; expected %d bytes from offset '
                     b'%d, got %d'
                 )
                 % (
                     self.indexfile if self._inline else self.datafile,
                     length,
                     offset,
                     len(d),
                 )
             )
 
         return d
 
     def _getsegment(self, offset, length, df=None):
         """Obtain a segment of raw data from the revlog.
 
         Accepts an absolute offset, length of bytes to obtain, and an
         optional file handle to the already-opened revlog. If the file
         handle is used, it's original seek position will not be preserved.
 
         Requests for data may be returned from a cache.
 
         Returns a str or a buffer instance of raw byte data.
         """
         o, d = self._chunkcache
         l = len(d)
 
         # is it in the cache?
         cachestart = offset - o
         cacheend = cachestart + length
         if cachestart >= 0 and cacheend <= l:
             if cachestart == 0 and cacheend == l:
                 return d  # avoid a copy
             return util.buffer(d, cachestart, cacheend - cachestart)
 
         return self._readsegment(offset, length, df=df)
 
     def _getsegmentforrevs(self, startrev, endrev, df=None):
         """Obtain a segment of raw data corresponding to a range of revisions.
 
         Accepts the start and end revisions and an optional already-open
         file handle to be used for reading. If the file handle is read, its
         seek position will not be preserved.
 
         Requests for data may be satisfied by a cache.
 
         Returns a 2-tuple of (offset, data) for the requested range of
         revisions. Offset is the integer offset from the beginning of the
         revlog and data is a str or buffer of the raw byte data.
 
         Callers will need to call ``self.start(rev)`` and ``self.length(rev)``
         to determine where each revision's data begins and ends.
         """
         # Inlined self.start(startrev) & self.end(endrev) for perf reasons
         # (functions are expensive).
         index = self.index
         istart = index[startrev]
         start = int(istart[0] >> 16)
         if startrev == endrev:
             end = start + istart[1]
         else:
             iend = index[endrev]
             end = int(iend[0] >> 16) + iend[1]
 
         if self._inline:
             start += (startrev + 1) * self._io.size
             end += (endrev + 1) * self._io.size
         length = end - start
 
         return start, self._getsegment(start, length, df=df)
 
     def _chunk(self, rev, df=None):
         """Obtain a single decompressed chunk for a revision.
 
         Accepts an integer revision and an optional already-open file handle
         to be used for reading. If used, the seek position of the file will not
         be preserved.
 
         Returns a str holding uncompressed data for the requested revision.
         """
         return self.decompress(self._getsegmentforrevs(rev, rev, df=df)[1])
 
     def _chunks(self, revs, df=None, targetsize=None):
         """Obtain decompressed chunks for the specified revisions.
 
         Accepts an iterable of numeric revisions that are assumed to be in
         ascending order. Also accepts an optional already-open file handle
         to be used for reading. If used, the seek position of the file will
         not be preserved.
 
         This function is similar to calling ``self._chunk()`` multiple times,
         but is faster.
 
         Returns a list with decompressed data for each requested revision.
         """
         if not revs:
             return []
         start = self.start
         length = self.length
         inline = self._inline
         iosize = self._io.size
         buffer = util.buffer
 
         l = []
         ladd = l.append
 
         if not self._withsparseread:
             slicedchunks = (revs,)
         else:
             slicedchunks = deltautil.slicechunk(
                 self, revs, targetsize=targetsize
             )
 
         for revschunk in slicedchunks:
             firstrev = revschunk[0]
             # Skip trailing revisions with empty diff
             for lastrev in revschunk[::-1]:
                 if length(lastrev) != 0:
                     break
 
             try:
                 offset, data = self._getsegmentforrevs(firstrev, lastrev, df=df)
             except OverflowError:
                 # issue4215 - we can't cache a run of chunks greater than
                 # 2G on Windows
                 return [self._chunk(rev, df=df) for rev in revschunk]
 
             decomp = self.decompress
             for rev in revschunk:
                 chunkstart = start(rev)
                 if inline:
                     chunkstart += (rev + 1) * iosize
                 chunklength = length(rev)
                 ladd(decomp(buffer(data, chunkstart - offset, chunklength)))
 
         return l
 
     def _chunkclear(self):
         """Clear the raw chunk cache."""
         self._chunkcache = (0, b'')
 
     def deltaparent(self, rev):
         """return deltaparent of the given revision"""
         base = self.index[rev][3]
         if base == rev:
             return nullrev
         elif self._generaldelta:
             return base
         else:
             return rev - 1
 
     def issnapshot(self, rev):
         """tells whether rev is a snapshot"""
         if not self._sparserevlog:
             return self.deltaparent(rev) == nullrev
         elif util.safehasattr(self.index, b'issnapshot'):
             # directly assign the method to cache the testing and access
             self.issnapshot = self.index.issnapshot
             return self.issnapshot(rev)
         if rev == nullrev:
             return True
         entry = self.index[rev]
         base = entry[3]
         if base == rev:
             return True
         if base == nullrev:
             return True
         p1 = entry[5]
         p2 = entry[6]
         if base == p1 or base == p2:
             return False
         return self.issnapshot(base)
 
     def snapshotdepth(self, rev):
         """number of snapshot in the chain before this one"""
         if not self.issnapshot(rev):
             raise error.ProgrammingError(b'revision %d not a snapshot')
         return len(self._deltachain(rev)[0]) - 1
 
     def revdiff(self, rev1, rev2):
         """return or calculate a delta between two revisions
 
         The delta calculated is in binary form and is intended to be written to
         revlog data directly. So this function needs raw revision data.
         """
         if rev1 != nullrev and self.deltaparent(rev2) == rev1:
             return bytes(self._chunk(rev2))
 
         return mdiff.textdiff(self.rawdata(rev1), self.rawdata(rev2))
 
     def _processflags(self, text, flags, operation, raw=False):
         """deprecated entry point to access flag processors"""
         msg = b'_processflag(...) use the specialized variant'
         util.nouideprecwarn(msg, b'5.2', stacklevel=2)
         if raw:
             return text, flagutil.processflagsraw(self, text, flags)
         elif operation == b'read':
             return flagutil.processflagsread(self, text, flags)
         else:  # write operation
             return flagutil.processflagswrite(self, text, flags, None)
 
     def revision(self, nodeorrev, _df=None, raw=False):
         """return an uncompressed revision of a given node or revision
         number.
 
         _df - an existing file handle to read from. (internal-only)
         raw - an optional argument specifying if the revision data is to be
         treated as raw data when applying flag transforms. 'raw' should be set
         to True when generating changegroups or in debug commands.
         """
         if raw:
             msg = (
                 b'revlog.revision(..., raw=True) is deprecated, '
                 b'use revlog.rawdata(...)'
             )
             util.nouideprecwarn(msg, b'5.2', stacklevel=2)
         return self._revisiondata(nodeorrev, _df, raw=raw)[0]
 
     def sidedata(self, nodeorrev, _df=None):
         """a map of extra data related to the changeset but not part of the hash
 
         This function currently return a dictionary. However, more advanced
         mapping object will likely be used in the future for a more
         efficient/lazy code.
         """
         return self._revisiondata(nodeorrev, _df)[1]
 
     def _revisiondata(self, nodeorrev, _df=None, raw=False):
         # deal with <nodeorrev> argument type
         if isinstance(nodeorrev, int):
             rev = nodeorrev
             node = self.node(rev)
         else:
             node = nodeorrev
             rev = None
 
         # fast path the special `nullid` rev
         if node == nullid:
             return b"", {}
 
         # ``rawtext`` is the text as stored inside the revlog. Might be the
         # revision or might need to be processed to retrieve the revision.
         rev, rawtext, validated = self._rawtext(node, rev, _df=_df)
 
         if raw and validated:
             # if we don't want to process the raw text and that raw
             # text is cached, we can exit early.
             return rawtext, {}
         if rev is None:
             rev = self.rev(node)
         # the revlog's flag for this revision
         # (usually alter its state or content)
         flags = self.flags(rev)
 
         if validated and flags == REVIDX_DEFAULT_FLAGS:
             # no extra flags set, no flag processor runs, text = rawtext
             return rawtext, {}
 
         sidedata = {}
         if raw:
             validatehash = flagutil.processflagsraw(self, rawtext, flags)
             text = rawtext
         else:
             try:
                 r = flagutil.processflagsread(self, rawtext, flags)
             except error.SidedataHashError as exc:
                 msg = _(b"integrity check failed on %s:%s sidedata key %d")
                 msg %= (self.indexfile, pycompat.bytestr(rev), exc.sidedatakey)
                 raise error.RevlogError(msg)
             text, validatehash, sidedata = r
         if validatehash:
             self.checkhash(text, node, rev=rev)
         if not validated:
             self._revisioncache = (node, rev, rawtext)
 
         return text, sidedata
 
     def _rawtext(self, node, rev, _df=None):
         """return the possibly unvalidated rawtext for a revision
 
         returns (rev, rawtext, validated)
         """
 
         # revision in the cache (could be useful to apply delta)
         cachedrev = None
         # An intermediate text to apply deltas to
         basetext = None
 
         # Check if we have the entry in cache
         # The cache entry looks like (node, rev, rawtext)
         if self._revisioncache:
             if self._revisioncache[0] == node:
                 return (rev, self._revisioncache[2], True)
             cachedrev = self._revisioncache[1]
 
         if rev is None:
             rev = self.rev(node)
 
         chain, stopped = self._deltachain(rev, stoprev=cachedrev)
         if stopped:
             basetext = self._revisioncache[2]
 
         # drop cache to save memory, the caller is expected to
         # update self._revisioncache after validating the text
         self._revisioncache = None
 
         targetsize = None
         rawsize = self.index[rev][2]
         if 0 <= rawsize:
             targetsize = 4 * rawsize
 
         bins = self._chunks(chain, df=_df, targetsize=targetsize)
         if basetext is None:
             basetext = bytes(bins[0])
             bins = bins[1:]
 
         rawtext = mdiff.patches(basetext, bins)
         del basetext  # let us have a chance to free memory early
         return (rev, rawtext, False)
 
     def rawdata(self, nodeorrev, _df=None):
         """return an uncompressed raw data of a given node or revision number.
 
         _df - an existing file handle to read from. (internal-only)
         """
         return self._revisiondata(nodeorrev, _df, raw=True)[0]
 
     def hash(self, text, p1, p2):
         """Compute a node hash.
 
         Available as a function so that subclasses can replace the hash
         as needed.
         """
         return storageutil.hashrevisionsha1(text, p1, p2)
 
     def checkhash(self, text, node, p1=None, p2=None, rev=None):
         """Check node hash integrity.
 
         Available as a function so that subclasses can extend hash mismatch
         behaviors as needed.
         """
         try:
             if p1 is None and p2 is None:
                 p1, p2 = self.parents(node)
             if node != self.hash(text, p1, p2):
                 # Clear the revision cache on hash failure. The revision cache
                 # only stores the raw revision and clearing the cache does have
                 # the side-effect that we won't have a cache hit when the raw
                 # revision data is accessed. But this case should be rare and
                 # it is extra work to teach the cache about the hash
                 # verification state.
                 if self._revisioncache and self._revisioncache[0] == node:
                     self._revisioncache = None
 
                 revornode = rev
                 if revornode is None:
                     revornode = templatefilters.short(hex(node))
                 raise error.RevlogError(
                     _(b"integrity check failed on %s:%s")
                     % (self.indexfile, pycompat.bytestr(revornode))
                 )
         except error.RevlogError:
             if self._censorable and storageutil.iscensoredtext(text):
                 raise error.CensoredNodeError(self.indexfile, node, text)
             raise
 
     def _enforceinlinesize(self, tr, fp=None):
         """Check if the revlog is too big for inline and convert if so.
 
         This should be called after revisions are added to the revlog. If the
         revlog has grown too large to be an inline revlog, it will convert it
         to use multiple index and data files.
         """
         tiprev = len(self) - 1
         if (
             not self._inline
             or (self.start(tiprev) + self.length(tiprev)) < _maxinline
         ):
             return
 
         troffset = tr.findoffset(self.indexfile)
         if troffset is None:
             raise error.RevlogError(
                 _(b"%s not found in the transaction") % self.indexfile
             )
         trindex = 0
         tr.add(self.datafile, 0)
 
         if fp:
             fp.flush()
             fp.close()
             # We can't use the cached file handle after close(). So prevent
             # its usage.
             self._writinghandles = None
 
         with self._indexfp(b'r') as ifh, self._datafp(b'w') as dfh:
             for r in self:
                 dfh.write(self._getsegmentforrevs(r, r, df=ifh)[1])
                 if troffset <= self.start(r):
                     trindex = r
 
         with self._indexfp(b'w') as fp:
             self.version &= ~FLAG_INLINE_DATA
             self._inline = False
             io = self._io
             for i in self:
                 e = io.packentry(self.index[i], self.node, self.version, i)
                 fp.write(e)
 
             # the temp file replace the real index when we exit the context
             # manager
 
         tr.replace(self.indexfile, trindex * self._io.size)
         nodemaputil.setup_persistent_nodemap(tr, self)
         self._chunkclear()
 
     def _nodeduplicatecallback(self, transaction, node):
         """called when trying to add a node already stored."""
 
     def addrevision(
         self,
         text,
         transaction,
         link,
         p1,
         p2,
         cachedelta=None,
         node=None,
         flags=REVIDX_DEFAULT_FLAGS,
         deltacomputer=None,
         sidedata=None,
     ):
         """add a revision to the log
 
         text - the revision data to add
         transaction - the transaction object used for rollback
         link - the linkrev data to add
         p1, p2 - the parent nodeids of the revision
         cachedelta - an optional precomputed delta
         node - nodeid of revision; typically node is not specified, and it is
             computed by default as hash(text, p1, p2), however subclasses might
             use different hashing method (and override checkhash() in such case)
         flags - the known flags to set on the revision
         deltacomputer - an optional deltacomputer instance shared between
             multiple calls
         """
         if link == nullrev:
             raise error.RevlogError(
                 _(b"attempted to add linkrev -1 to %s") % self.indexfile
             )
 
         if sidedata is None:
             sidedata = {}
             flags = flags & ~REVIDX_SIDEDATA
         elif not self.hassidedata:
             raise error.ProgrammingError(
                 _(b"trying to add sidedata to a revlog who don't support them")
             )
         else:
             flags |= REVIDX_SIDEDATA
 
         if flags:
             node = node or self.hash(text, p1, p2)
 
         rawtext, validatehash = flagutil.processflagswrite(
             self, text, flags, sidedata=sidedata
         )
 
         # If the flag processor modifies the revision data, ignore any provided
         # cachedelta.
         if rawtext != text:
             cachedelta = None
 
         if len(rawtext) > _maxentrysize:
             raise error.RevlogError(
                 _(
                     b"%s: size of %d bytes exceeds maximum revlog storage of 2GiB"
                 )
                 % (self.indexfile, len(rawtext))
             )
 
         node = node or self.hash(rawtext, p1, p2)
         if self.index.has_node(node):
             return node
 
         if validatehash:
             self.checkhash(rawtext, node, p1=p1, p2=p2)
 
         return self.addrawrevision(
             rawtext,
             transaction,
             link,
             p1,
             p2,
             node,
             flags,
             cachedelta=cachedelta,
             deltacomputer=deltacomputer,
         )
 
     def addrawrevision(
         self,
         rawtext,
         transaction,
         link,
         p1,
         p2,
         node,
         flags,
         cachedelta=None,
         deltacomputer=None,
     ):
         """add a raw revision with known flags, node and parents
         useful when reusing a revision not stored in this revlog (ex: received
         over wire, or read from an external bundle).
         """
         dfh = None
         if not self._inline:
             dfh = self._datafp(b"a+")
         ifh = self._indexfp(b"a+")
         try:
             return self._addrevision(
                 node,
                 rawtext,
                 transaction,
                 link,
                 p1,
                 p2,
                 flags,
                 cachedelta,
                 ifh,
                 dfh,
                 deltacomputer=deltacomputer,
             )
         finally:
             if dfh:
                 dfh.close()
             ifh.close()
 
     def compress(self, data):
         """Generate a possibly-compressed representation of data."""
         if not data:
             return b'', data
 
         compressed = self._compressor.compress(data)
 
         if compressed:
             # The revlog compressor added the header in the returned data.
             return b'', compressed
 
         if data[0:1] == b'\0':
             return b'', data
         return b'u', data
 
     def decompress(self, data):
         """Decompress a revlog chunk.
 
         The chunk is expected to begin with a header identifying the
         format type so it can be routed to an appropriate decompressor.
         """
         if not data:
             return data
 
         # Revlogs are read much more frequently than they are written and many
         # chunks only take microseconds to decompress, so performance is
         # important here.
         #
         # We can make a few assumptions about revlogs:
         #
         # 1) the majority of chunks will be compressed (as opposed to inline
         #    raw data).
         # 2) decompressing *any* data will likely by at least 10x slower than
         #    returning raw inline data.
         # 3) we want to prioritize common and officially supported compression
         #    engines
         #
         # It follows that we want to optimize for "decompress compressed data
         # when encoded with common and officially supported compression engines"
         # case over "raw data" and "data encoded by less common or non-official
         # compression engines." That is why we have the inline lookup first
         # followed by the compengines lookup.
         #
         # According to `hg perfrevlogchunks`, this is ~0.5% faster for zlib
         # compressed chunks. And this matters for changelog and manifest reads.
         t = data[0:1]
 
         if t == b'x':
             try:
                 return _zlibdecompress(data)
             except zlib.error as e:
                 raise error.RevlogError(
                     _(b'revlog decompress error: %s')
                     % stringutil.forcebytestr(e)
                 )
         # '\0' is more common than 'u' so it goes first.
         elif t == b'\0':
             return data
         elif t == b'u':
             return util.buffer(data, 1)
 
         try:
             compressor = self._decompressors[t]
         except KeyError:
             try:
                 engine = util.compengines.forrevlogheader(t)
                 compressor = engine.revlogcompressor(self._compengineopts)
                 self._decompressors[t] = compressor
             except KeyError:
                 raise error.RevlogError(_(b'unknown compression type %r') % t)
 
         return compressor.decompress(data)
 
     def _addrevision(
         self,
         node,
         rawtext,
         transaction,
         link,
         p1,
         p2,
         flags,
         cachedelta,
         ifh,
         dfh,
         alwayscache=False,
         deltacomputer=None,
     ):
         """internal function to add revisions to the log
 
         see addrevision for argument descriptions.
 
         note: "addrevision" takes non-raw text, "_addrevision" takes raw text.
 
         if "deltacomputer" is not provided or None, a defaultdeltacomputer will
         be used.
 
         invariants:
         - rawtext is optional (can be None); if not set, cachedelta must be set.
           if both are set, they must correspond to each other.
         """
         if node == nullid:
             raise error.RevlogError(
                 _(b"%s: attempt to add null revision") % self.indexfile
             )
         if node == wdirid or node in wdirfilenodeids:
             raise error.RevlogError(
                 _(b"%s: attempt to add wdir revision") % self.indexfile
             )
 
         if self._inline:
             fh = ifh
         else:
             fh = dfh
 
         btext = [rawtext]
 
         curr = len(self)
         prev = curr - 1
         offset = self.end(prev)
         p1r, p2r = self.rev(p1), self.rev(p2)
 
         # full versions are inserted when the needed deltas
         # become comparable to the uncompressed text
         if rawtext is None:
             # need rawtext size, before changed by flag processors, which is
             # the non-raw size. use revlog explicitly to avoid filelog's extra
             # logic that might remove metadata size.
             textlen = mdiff.patchedsize(
                 revlog.size(self, cachedelta[0]), cachedelta[1]
             )
         else:
             textlen = len(rawtext)
 
         if deltacomputer is None:
             deltacomputer = deltautil.deltacomputer(self)
 
         revinfo = _revisioninfo(node, p1, p2, btext, textlen, cachedelta, flags)
 
         deltainfo = deltacomputer.finddeltainfo(revinfo, fh)
 
         e = (
             offset_type(offset, flags),
             deltainfo.deltalen,
             textlen,
             deltainfo.base,
             link,
             p1r,
             p2r,
             node,
         )
         self.index.append(e)
 
         entry = self._io.packentry(e, self.node, self.version, curr)
         self._writeentry(
             transaction, ifh, dfh, entry, deltainfo.data, link, offset
         )
 
         rawtext = btext[0]
 
         if alwayscache and rawtext is None:
             rawtext = deltacomputer.buildtext(revinfo, fh)
 
         if type(rawtext) == bytes:  # only accept immutable objects
             self._revisioncache = (node, curr, rawtext)
         self._chainbasecache[curr] = deltainfo.chainbase
         return node
 
     def _writeentry(self, transaction, ifh, dfh, entry, data, link, offset):
         # Files opened in a+ mode have inconsistent behavior on various
         # platforms. Windows requires that a file positioning call be made
         # when the file handle transitions between reads and writes. See
         # 3686fa2b8eee and the mixedfilemodewrapper in windows.py. On other
         # platforms, Python or the platform itself can be buggy. Some versions
         # of Solaris have been observed to not append at the end of the file
         # if the file was seeked to before the end. See issue4943 for more.
         #
         # We work around this issue by inserting a seek() before writing.
         # Note: This is likely not necessary on Python 3. However, because
         # the file handle is reused for reads and may be seeked there, we need
         # to be careful before changing this.
         ifh.seek(0, os.SEEK_END)
         if dfh:
             dfh.seek(0, os.SEEK_END)
 
         curr = len(self) - 1
         if not self._inline:
             transaction.add(self.datafile, offset)
             transaction.add(self.indexfile, curr * len(entry))
             if data[0]:
                 dfh.write(data[0])
             dfh.write(data[1])
             ifh.write(entry)
         else:
             offset += curr * self._io.size
             transaction.add(self.indexfile, offset)
             ifh.write(entry)
             ifh.write(data[0])
             ifh.write(data[1])
             self._enforceinlinesize(transaction, ifh)
         nodemaputil.setup_persistent_nodemap(transaction, self)
 
     def addgroup(
         self,
         deltas,
         linkmapper,
         transaction,
+        alwayscache=False,
         addrevisioncb=None,
         duplicaterevisioncb=None,
     ):
         """
         add a delta group
 
         given a set of deltas, add them to the revision log. the
         first delta is against its parent, which should be in our
         log, the rest are against the previous delta.
 
         If ``addrevisioncb`` is defined, it will be called with arguments of
         this revlog and the node that was added.
         """
 
         if self._writinghandles:
             raise error.ProgrammingError(b'cannot nest addgroup() calls')
 
         r = len(self)
         end = 0
         if r:
             end = self.end(r - 1)
         ifh = self._indexfp(b"a+")
         isize = r * self._io.size
         if self._inline:
             transaction.add(self.indexfile, end + isize)
             dfh = None
         else:
             transaction.add(self.indexfile, isize)
             transaction.add(self.datafile, end)
             dfh = self._datafp(b"a+")
 
         def flush():
             if dfh:
                 dfh.flush()
             ifh.flush()
 
         self._writinghandles = (ifh, dfh)
         empty = True
 
         try:
             deltacomputer = deltautil.deltacomputer(self)
             # loop through our set of deltas
             for data in deltas:
                 node, p1, p2, linknode, deltabase, delta, flags = data
                 link = linkmapper(linknode)
                 flags = flags or REVIDX_DEFAULT_FLAGS
 
                 if self.index.has_node(node):
                     # this can happen if two branches make the same change
                     self._nodeduplicatecallback(transaction, node)
                     if duplicaterevisioncb:
                         duplicaterevisioncb(self, node)
                     empty = False
                     continue
 
                 for p in (p1, p2):
                     if not self.index.has_node(p):
                         raise error.LookupError(
                             p, self.indexfile, _(b'unknown parent')
                         )
 
                 if not self.index.has_node(deltabase):
                     raise error.LookupError(
                         deltabase, self.indexfile, _(b'unknown delta base')
                     )
 
                 baserev = self.rev(deltabase)
 
                 if baserev != nullrev and self.iscensored(baserev):
                     # if base is censored, delta must be full replacement in a
                     # single patch operation
                     hlen = struct.calcsize(b">lll")
                     oldlen = self.rawsize(baserev)
                     newlen = len(delta) - hlen
                     if delta[:hlen] != mdiff.replacediffheader(oldlen, newlen):
                         raise error.CensoredBaseError(
                             self.indexfile, self.node(baserev)
                         )
 
                 if not flags and self._peek_iscensored(baserev, delta, flush):
                     flags |= REVIDX_ISCENSORED
 
                 # We assume consumers of addrevisioncb will want to retrieve
                 # the added revision, which will require a call to
                 # revision(). revision() will fast path if there is a cache
                 # hit. So, we tell _addrevision() to always cache in this case.
                 # We're only using addgroup() in the context of changegroup
                 # generation so the revision data can always be handled as raw
                 # by the flagprocessor.
                 self._addrevision(
                     node,
                     None,
                     transaction,
                     link,
                     p1,
                     p2,
                     flags,
                     (baserev, delta),
                     ifh,
                     dfh,
-                    alwayscache=bool(addrevisioncb),
+                    alwayscache=alwayscache,
                     deltacomputer=deltacomputer,
                 )
 
                 if addrevisioncb:
                     addrevisioncb(self, node)
                 empty = False
 
                 if not dfh and not self._inline:
                     # addrevision switched from inline to conventional
                     # reopen the index
                     ifh.close()
                     dfh = self._datafp(b"a+")
                     ifh = self._indexfp(b"a+")
                     self._writinghandles = (ifh, dfh)
         finally:
             self._writinghandles = None
 
             if dfh:
                 dfh.close()
             ifh.close()
         return not empty
 
     def iscensored(self, rev):
         """Check if a file revision is censored."""
         if not self._censorable:
             return False
 
         return self.flags(rev) & REVIDX_ISCENSORED
 
     def _peek_iscensored(self, baserev, delta, flush):
         """Quickly check if a delta produces a censored revision."""
         if not self._censorable:
             return False
 
         return storageutil.deltaiscensored(delta, baserev, self.rawsize)
 
     def getstrippoint(self, minlink):
         """find the minimum rev that must be stripped to strip the linkrev
 
         Returns a tuple containing the minimum rev and a set of all revs that
         have linkrevs that will be broken by this strip.
         """
         return storageutil.resolvestripinfo(
             minlink,
             len(self) - 1,
             self.headrevs(),
             self.linkrev,
             self.parentrevs,
         )
 
     def strip(self, minlink, transaction):
         """truncate the revlog on the first revision with a linkrev >= minlink
 
         This function is called when we're stripping revision minlink and
         its descendants from the repository.
 
         We have to remove all revisions with linkrev >= minlink, because
         the equivalent changelog revisions will be renumbered after the
         strip.
 
         So we truncate the revlog on the first of these revisions, and
         trust that the caller has saved the revisions that shouldn't be
         removed and that it'll re-add them after this truncation.
         """
         if len(self) == 0:
             return
 
         rev, _ = self.getstrippoint(minlink)
         if rev == len(self):
             return
 
         # first truncate the files on disk
         end = self.start(rev)
         if not self._inline:
             transaction.add(self.datafile, end)
             end = rev * self._io.size
         else:
             end += rev * self._io.size
 
         transaction.add(self.indexfile, end)
 
         # then reset internal state in memory to forget those revisions
         self._revisioncache = None
         self._chaininfocache = util.lrucachedict(500)
         self._chunkclear()
 
         del self.index[rev:-1]
 
     def checksize(self):
         """Check size of index and data files
 
         return a (dd, di) tuple.
         - dd: extra bytes for the "data" file
         - di: extra bytes for the "index" file
 
         A healthy revlog will return (0, 0).
         """
         expected = 0
         if len(self):
             expected = max(0, self.end(len(self) - 1))
 
         try:
             with self._datafp() as f:
                 f.seek(0, io.SEEK_END)
                 actual = f.tell()
             dd = actual - expected
         except IOError as inst:
             if inst.errno != errno.ENOENT:
                 raise
             dd = 0
 
         try:
             f = self.opener(self.indexfile)
             f.seek(0, io.SEEK_END)
             actual = f.tell()
             f.close()
             s = self._io.size
             i = max(0, actual // s)
             di = actual - (i * s)
             if self._inline:
                 databytes = 0
                 for r in self:
                     databytes += max(0, self.length(r))
                 dd = 0
                 di = actual - len(self) * s - databytes
         except IOError as inst:
             if inst.errno != errno.ENOENT:
                 raise
             di = 0
 
         return (dd, di)
 
     def files(self):
         res = [self.indexfile]
         if not self._inline:
             res.append(self.datafile)
         return res
 
     def emitrevisions(
         self,
         nodes,
         nodesorder=None,
         revisiondata=False,
         assumehaveparentrevisions=False,
         deltamode=repository.CG_DELTAMODE_STD,
     ):
         if nodesorder not in (b'nodes', b'storage', b'linear', None):
             raise error.ProgrammingError(
                 b'unhandled value for nodesorder: %s' % nodesorder
             )
 
         if nodesorder is None and not self._generaldelta:
             nodesorder = b'storage'
 
         if (
             not self._storedeltachains
             and deltamode != repository.CG_DELTAMODE_PREV
         ):
             deltamode = repository.CG_DELTAMODE_FULL
 
         return storageutil.emitrevisions(
             self,
             nodes,
             nodesorder,
             revlogrevisiondelta,
             deltaparentfn=self.deltaparent,
             candeltafn=self.candelta,
             rawsizefn=self.rawsize,
             revdifffn=self.revdiff,
             flagsfn=self.flags,
             deltamode=deltamode,
             revisiondata=revisiondata,
             assumehaveparentrevisions=assumehaveparentrevisions,
         )
 
     DELTAREUSEALWAYS = b'always'
     DELTAREUSESAMEREVS = b'samerevs'
     DELTAREUSENEVER = b'never'
 
     DELTAREUSEFULLADD = b'fulladd'
 
     DELTAREUSEALL = {b'always', b'samerevs', b'never', b'fulladd'}
 
     def clone(
         self,
         tr,
         destrevlog,
         addrevisioncb=None,
         deltareuse=DELTAREUSESAMEREVS,
         forcedeltabothparents=None,
         sidedatacompanion=None,
     ):
         """Copy this revlog to another, possibly with format changes.
 
         The destination revlog will contain the same revisions and nodes.
         However, it may not be bit-for-bit identical due to e.g. delta encoding
         differences.
 
         The ``deltareuse`` argument control how deltas from the existing revlog
         are preserved in the destination revlog. The argument can have the
         following values:
 
         DELTAREUSEALWAYS
            Deltas will always be reused (if possible), even if the destination
            revlog would not select the same revisions for the delta. This is the
            fastest mode of operation.
         DELTAREUSESAMEREVS
            Deltas will be reused if the destination revlog would pick the same
            revisions for the delta. This mode strikes a balance between speed
            and optimization.
         DELTAREUSENEVER
            Deltas will never be reused. This is the slowest mode of execution.
            This mode can be used to recompute deltas (e.g. if the diff/delta
            algorithm changes).
         DELTAREUSEFULLADD
            Revision will be re-added as if their were new content. This is
            slower than DELTAREUSEALWAYS but allow more mechanism to kicks in.
            eg: large file detection and handling.
 
         Delta computation can be slow, so the choice of delta reuse policy can
         significantly affect run time.
 
         The default policy (``DELTAREUSESAMEREVS``) strikes a balance between
         two extremes. Deltas will be reused if they are appropriate. But if the
         delta could choose a better revision, it will do so. This means if you
         are converting a non-generaldelta revlog to a generaldelta revlog,
         deltas will be recomputed if the delta's parent isn't a parent of the
         revision.
 
         In addition to the delta policy, the ``forcedeltabothparents``
         argument controls whether to force compute deltas against both parents
         for merges. By default, the current default is used.
 
         If not None, the `sidedatacompanion` is callable that accept two
         arguments:
 
             (srcrevlog, rev)
 
         and return a quintet that control changes to sidedata content from the
         old revision to the new clone result:
 
             (dropall, filterout, update, new_flags, dropped_flags)
 
         * if `dropall` is True, all sidedata should be dropped
         * `filterout` is a set of sidedata keys that should be dropped
         * `update` is a mapping of additionnal/new key -> value
         * new_flags is a bitfields of new flags that the revision should get
         * dropped_flags is a bitfields of new flags that the revision shoudl not longer have
         """
         if deltareuse not in self.DELTAREUSEALL:
             raise ValueError(
                 _(b'value for deltareuse invalid: %s') % deltareuse
             )
 
         if len(destrevlog):
             raise ValueError(_(b'destination revlog is not empty'))
 
         if getattr(self, 'filteredrevs', None):
             raise ValueError(_(b'source revlog has filtered revisions'))
         if getattr(destrevlog, 'filteredrevs', None):
             raise ValueError(_(b'destination revlog has filtered revisions'))
 
         # lazydelta and lazydeltabase controls whether to reuse a cached delta,
         # if possible.
         oldlazydelta = destrevlog._lazydelta
         oldlazydeltabase = destrevlog._lazydeltabase
         oldamd = destrevlog._deltabothparents
 
         try:
             if deltareuse == self.DELTAREUSEALWAYS:
                 destrevlog._lazydeltabase = True
                 destrevlog._lazydelta = True
             elif deltareuse == self.DELTAREUSESAMEREVS:
                 destrevlog._lazydeltabase = False
                 destrevlog._lazydelta = True
             elif deltareuse == self.DELTAREUSENEVER:
                 destrevlog._lazydeltabase = False
                 destrevlog._lazydelta = False
 
             destrevlog._deltabothparents = forcedeltabothparents or oldamd
 
             self._clone(
                 tr,
                 destrevlog,
                 addrevisioncb,
                 deltareuse,
                 forcedeltabothparents,
                 sidedatacompanion,
             )
 
         finally:
             destrevlog._lazydelta = oldlazydelta
             destrevlog._lazydeltabase = oldlazydeltabase
             destrevlog._deltabothparents = oldamd
 
     def _clone(
         self,
         tr,
         destrevlog,
         addrevisioncb,
         deltareuse,
         forcedeltabothparents,
         sidedatacompanion,
     ):
         """perform the core duty of `revlog.clone` after parameter processing"""
         deltacomputer = deltautil.deltacomputer(destrevlog)
         index = self.index
         for rev in self:
             entry = index[rev]
 
             # Some classes override linkrev to take filtered revs into
             # account. Use raw entry from index.
             flags = entry[0] & 0xFFFF
             linkrev = entry[4]
             p1 = index[entry[5]][7]
             p2 = index[entry[6]][7]
             node = entry[7]
 
             sidedataactions = (False, [], {}, 0, 0)
             if sidedatacompanion is not None:
                 sidedataactions = sidedatacompanion(self, rev)
 
             # (Possibly) reuse the delta from the revlog if allowed and
             # the revlog chunk is a delta.
             cachedelta = None
             rawtext = None
             if any(sidedataactions) or deltareuse == self.DELTAREUSEFULLADD:
                 dropall = sidedataactions[0]
                 filterout = sidedataactions[1]
                 update = sidedataactions[2]
                 new_flags = sidedataactions[3]
                 dropped_flags = sidedataactions[4]
                 text, sidedata = self._revisiondata(rev)
                 if dropall:
                     sidedata = {}
                 for key in filterout:
                     sidedata.pop(key, None)
                 sidedata.update(update)
                 if not sidedata:
                     sidedata = None
 
                 flags |= new_flags
                 flags &= ~dropped_flags
 
                 destrevlog.addrevision(
                     text,
                     tr,
                     linkrev,
                     p1,
                     p2,
                     cachedelta=cachedelta,
                     node=node,
                     flags=flags,
                     deltacomputer=deltacomputer,
                     sidedata=sidedata,
                 )
             else:
                 if destrevlog._lazydelta:
                     dp = self.deltaparent(rev)
                     if dp != nullrev:
                         cachedelta = (dp, bytes(self._chunk(rev)))
 
                 if not cachedelta:
                     rawtext = self.rawdata(rev)
 
                 ifh = destrevlog.opener(
                     destrevlog.indexfile, b'a+', checkambig=False
                 )
                 dfh = None
                 if not destrevlog._inline:
                     dfh = destrevlog.opener(destrevlog.datafile, b'a+')
                 try:
                     destrevlog._addrevision(
                         node,
                         rawtext,
                         tr,
                         linkrev,
                         p1,
                         p2,
                         flags,
                         cachedelta,
                         ifh,
                         dfh,
                         deltacomputer=deltacomputer,
                     )
                 finally:
                     if dfh:
                         dfh.close()
                     ifh.close()
 
             if addrevisioncb:
                 addrevisioncb(self, rev, node)
 
     def censorrevision(self, tr, censornode, tombstone=b''):
         if (self.version & 0xFFFF) == REVLOGV0:
             raise error.RevlogError(
                 _(b'cannot censor with version %d revlogs') % self.version
             )
 
         censorrev = self.rev(censornode)
         tombstone = storageutil.packmeta({b'censored': tombstone}, b'')
 
         if len(tombstone) > self.rawsize(censorrev):
             raise error.Abort(
                 _(b'censor tombstone must be no longer than censored data')
             )
 
         # Rewriting the revlog in place is hard. Our strategy for censoring is
         # to create a new revlog, copy all revisions to it, then replace the
         # revlogs on transaction close.
 
         newindexfile = self.indexfile + b'.tmpcensored'
         newdatafile = self.datafile + b'.tmpcensored'
 
         # This is a bit dangerous. We could easily have a mismatch of state.
         newrl = revlog(self.opener, newindexfile, newdatafile, censorable=True)
         newrl.version = self.version
         newrl._generaldelta = self._generaldelta
         newrl._io = self._io
 
         for rev in self.revs():
             node = self.node(rev)
             p1, p2 = self.parents(node)
 
             if rev == censorrev:
                 newrl.addrawrevision(
                     tombstone,
                     tr,
                     self.linkrev(censorrev),
                     p1,
                     p2,
                     censornode,
                     REVIDX_ISCENSORED,
                 )
 
                 if newrl.deltaparent(rev) != nullrev:
                     raise error.Abort(
                         _(
                             b'censored revision stored as delta; '
                             b'cannot censor'
                         ),
                         hint=_(
                             b'censoring of revlogs is not '
                             b'fully implemented; please report '
                             b'this bug'
                         ),
                     )
                 continue
 
             if self.iscensored(rev):
                 if self.deltaparent(rev) != nullrev:
                     raise error.Abort(
                         _(
                             b'cannot censor due to censored '
                             b'revision having delta stored'
                         )
                     )
                 rawtext = self._chunk(rev)
             else:
                 rawtext = self.rawdata(rev)
 
             newrl.addrawrevision(
                 rawtext, tr, self.linkrev(rev), p1, p2, node, self.flags(rev)
             )
 
         tr.addbackup(self.indexfile, location=b'store')
         if not self._inline:
             tr.addbackup(self.datafile, location=b'store')
 
         self.opener.rename(newrl.indexfile, self.indexfile)
         if not self._inline:
             self.opener.rename(newrl.datafile, self.datafile)
 
         self.clearcaches()
         self._loadindex()
 
     def verifyintegrity(self, state):
         """Verifies the integrity of the revlog.
 
         Yields ``revlogproblem`` instances describing problems that are
         found.
         """
         dd, di = self.checksize()
         if dd:
             yield revlogproblem(error=_(b'data length off by %d bytes') % dd)
         if di:
             yield revlogproblem(error=_(b'index contains %d extra bytes') % di)
 
         version = self.version & 0xFFFF
 
         # The verifier tells us what version revlog we should be.
         if version != state[b'expectedversion']:
             yield revlogproblem(
                 warning=_(b"warning: '%s' uses revlog format %d; expected %d")
                 % (self.indexfile, version, state[b'expectedversion'])
             )
 
         state[b'skipread'] = set()
         state[b'safe_renamed'] = set()
 
         for rev in self:
             node = self.node(rev)
 
             # Verify contents. 4 cases to care about:
             #
             #   common: the most common case
             #   rename: with a rename
             #   meta: file content starts with b'\1\n', the metadata
             #         header defined in filelog.py, but without a rename
             #   ext: content stored externally
             #
             # More formally, their differences are shown below:
             #
             #                       | common | rename | meta  | ext
             #  -------------------------------------------------------
             #   flags()             | 0      | 0      | 0     | not 0
             #   renamed()           | False  | True   | False | ?
             #   rawtext[0:2]=='\1\n'| False  | True   | True  | ?
             #
             # "rawtext" means the raw text stored in revlog data, which
             # could be retrieved by "rawdata(rev)". "text"
             # mentioned below is "revision(rev)".
             #
             # There are 3 different lengths stored physically:
             #  1. L1: rawsize, stored in revlog index
             #  2. L2: len(rawtext), stored in revlog data
             #  3. L3: len(text), stored in revlog data if flags==0, or
             #     possibly somewhere else if flags!=0
             #
             # L1 should be equal to L2. L3 could be different from them.
             # "text" may or may not affect commit hash depending on flag
             # processors (see flagutil.addflagprocessor).
             #
             #              | common  | rename | meta  | ext
             # -------------------------------------------------
             #    rawsize() | L1      | L1     | L1    | L1
             #       size() | L1      | L2-LM  | L1(*) | L1 (?)
             # len(rawtext) | L2      | L2     | L2    | L2
             #    len(text) | L2      | L2     | L2    | L3
             #  len(read()) | L2      | L2-LM  | L2-LM | L3 (?)
             #
             # LM:  length of metadata, depending on rawtext
             # (*): not ideal, see comment in filelog.size
             # (?): could be "- len(meta)" if the resolved content has
             #      rename metadata
             #
             # Checks needed to be done:
             #  1. length check: L1 == L2, in all cases.
             #  2. hash check: depending on flag processor, we may need to
             #     use either "text" (external), or "rawtext" (in revlog).
 
             try:
                 skipflags = state.get(b'skipflags', 0)
                 if skipflags:
                     skipflags &= self.flags(rev)
 
                 _verify_revision(self, skipflags, state, node)
 
                 l1 = self.rawsize(rev)
                 l2 = len(self.rawdata(node))
 
                 if l1 != l2:
                     yield revlogproblem(
                         error=_(b'unpacked size is %d, %d expected') % (l2, l1),
                         node=node,
                     )
 
             except error.CensoredNodeError:
                 if state[b'erroroncensored']:
                     yield revlogproblem(
                         error=_(b'censored file data'), node=node
                     )
                     state[b'skipread'].add(node)
             except Exception as e:
                 yield revlogproblem(
                     error=_(b'unpacking %s: %s')
                     % (short(node), stringutil.forcebytestr(e)),
                     node=node,
                 )
                 state[b'skipread'].add(node)
 
     def storageinfo(
         self,
         exclusivefiles=False,
         sharedfiles=False,
         revisionscount=False,
         trackedsize=False,
         storedsize=False,
     ):
         d = {}
 
         if exclusivefiles:
             d[b'exclusivefiles'] = [(self.opener, self.indexfile)]
             if not self._inline:
                 d[b'exclusivefiles'].append((self.opener, self.datafile))
 
         if sharedfiles:
             d[b'sharedfiles'] = []
 
         if revisionscount:
             d[b'revisionscount'] = len(self)
 
         if trackedsize:
             d[b'trackedsize'] = sum(map(self.rawsize, iter(self)))
 
         if storedsize:
             d[b'storedsize'] = sum(
                 self.opener.stat(path).st_size for path in self.files()
             )
 
         return d

mercurial/unionrepo.py

@@ -1,287 +1,288 b''
 # unionrepo.py - repository class for viewing union of repository changesets
 #
 # Derived from bundlerepo.py
 # Copyright 2006, 2007 Benoit Boissinot <[e-mail]>
 # Copyright 2013 Unity Technologies, Mads Kiilerich <[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.
 
 """Repository class for "in-memory pull" of one local repository to another,
 allowing operations like diff and log with revsets.
 """
 
 from __future__ import absolute_import
 
 from .i18n import _
 from .pycompat import getattr
 
 from . import (
     changelog,
     cmdutil,
     encoding,
     error,
     filelog,
     localrepo,
     manifest,
     mdiff,
     pathutil,
     revlog,
     util,
     vfs as vfsmod,
 )
 
 
 class unionrevlog(revlog.revlog):
     def __init__(self, opener, indexfile, revlog2, linkmapper):
         # How it works:
         # To retrieve a revision, we just need to know the node id so we can
         # look it up in revlog2.
         #
         # To differentiate a rev in the second revlog from a rev in the revlog,
         # we check revision against repotiprev.
         opener = vfsmod.readonlyvfs(opener)
         revlog.revlog.__init__(self, opener, indexfile)
         self.revlog2 = revlog2
 
         n = len(self)
         self.repotiprev = n - 1
         self.bundlerevs = set()  # used by 'bundle()' revset expression
         for rev2 in self.revlog2:
             rev = self.revlog2.index[rev2]
             # rev numbers - in revlog2, very different from self.rev
             _start, _csize, rsize, base, linkrev, p1rev, p2rev, node = rev
             flags = _start & 0xFFFF
 
             if linkmapper is None:  # link is to same revlog
                 assert linkrev == rev2  # we never link back
                 link = n
             else:  # rev must be mapped from repo2 cl to unified cl by linkmapper
                 link = linkmapper(linkrev)
 
             if linkmapper is not None:  # link is to same revlog
                 base = linkmapper(base)
 
             this_rev = self.index.get_rev(node)
             if this_rev is not None:
                 # this happens for the common revlog revisions
                 self.bundlerevs.add(this_rev)
                 continue
 
             p1node = self.revlog2.node(p1rev)
             p2node = self.revlog2.node(p2rev)
 
             # TODO: it's probably wrong to set compressed length to -1, but
             # I have no idea if csize is valid in the base revlog context.
             e = (
                 flags,
                 -1,
                 rsize,
                 base,
                 link,
                 self.rev(p1node),
                 self.rev(p2node),
                 node,
             )
             self.index.append(e)
             self.bundlerevs.add(n)
             n += 1
 
     def _chunk(self, rev):
         if rev <= self.repotiprev:
             return revlog.revlog._chunk(self, rev)
         return self.revlog2._chunk(self.node(rev))
 
     def revdiff(self, rev1, rev2):
         """return or calculate a delta between two revisions"""
         if rev1 > self.repotiprev and rev2 > self.repotiprev:
             return self.revlog2.revdiff(
                 self.revlog2.rev(self.node(rev1)),
                 self.revlog2.rev(self.node(rev2)),
             )
         elif rev1 <= self.repotiprev and rev2 <= self.repotiprev:
             return super(unionrevlog, self).revdiff(rev1, rev2)
 
         return mdiff.textdiff(self.rawdata(rev1), self.rawdata(rev2))
 
     def _revisiondata(self, nodeorrev, _df=None, raw=False):
         if isinstance(nodeorrev, int):
             rev = nodeorrev
             node = self.node(rev)
         else:
             node = nodeorrev
             rev = self.rev(node)
 
         if rev > self.repotiprev:
             # work around manifestrevlog NOT being a revlog
             revlog2 = getattr(self.revlog2, '_revlog', self.revlog2)
             func = revlog2._revisiondata
         else:
             func = super(unionrevlog, self)._revisiondata
         return func(node, _df=_df, raw=raw)
 
     def addrevision(self, text, transaction, link, p1=None, p2=None, d=None):
         raise NotImplementedError
 
     def addgroup(
         self,
         deltas,
         linkmapper,
         transaction,
+        alwayscache=False,
         addrevisioncb=None,
         duplicaterevisioncb=None,
         maybemissingparents=False,
     ):
         raise NotImplementedError
 
     def strip(self, minlink, transaction):
         raise NotImplementedError
 
     def checksize(self):
         raise NotImplementedError
 
 
 class unionchangelog(unionrevlog, changelog.changelog):
     def __init__(self, opener, opener2):
         changelog.changelog.__init__(self, opener)
         linkmapper = None
         changelog2 = changelog.changelog(opener2)
         unionrevlog.__init__(
             self, opener, self.indexfile, changelog2, linkmapper
         )
 
 
 class unionmanifest(unionrevlog, manifest.manifestrevlog):
     def __init__(self, opener, opener2, linkmapper):
         manifest.manifestrevlog.__init__(self, opener)
         manifest2 = manifest.manifestrevlog(opener2)
         unionrevlog.__init__(
             self, opener, self.indexfile, manifest2, linkmapper
         )
 
 
 class unionfilelog(filelog.filelog):
     def __init__(self, opener, path, opener2, linkmapper, repo):
         filelog.filelog.__init__(self, opener, path)
         filelog2 = filelog.filelog(opener2, path)
         self._revlog = unionrevlog(
             opener, self.indexfile, filelog2._revlog, linkmapper
         )
         self._repo = repo
         self.repotiprev = self._revlog.repotiprev
         self.revlog2 = self._revlog.revlog2
 
     def iscensored(self, rev):
         """Check if a revision is censored."""
         if rev <= self.repotiprev:
             return filelog.filelog.iscensored(self, rev)
         node = self.node(rev)
         return self.revlog2.iscensored(self.revlog2.rev(node))
 
 
 class unionpeer(localrepo.localpeer):
     def canpush(self):
         return False
 
 
 class unionrepository(object):
     """Represents the union of data in 2 repositories.
 
     Instances are not usable if constructed directly. Use ``instance()``
     or ``makeunionrepository()`` to create a usable instance.
     """
 
     def __init__(self, repo2, url):
         self.repo2 = repo2
         self._url = url
 
         self.ui.setconfig(b'phases', b'publish', False, b'unionrepo')
 
     @localrepo.unfilteredpropertycache
     def changelog(self):
         return unionchangelog(self.svfs, self.repo2.svfs)
 
     @localrepo.unfilteredpropertycache
     def manifestlog(self):
         rootstore = unionmanifest(
             self.svfs, self.repo2.svfs, self.unfiltered()._clrev
         )
         return manifest.manifestlog(
             self.svfs, self, rootstore, self.narrowmatch()
         )
 
     def _clrev(self, rev2):
         """map from repo2 changelog rev to temporary rev in self.changelog"""
         node = self.repo2.changelog.node(rev2)
         return self.changelog.rev(node)
 
     def url(self):
         return self._url
 
     def file(self, f):
         return unionfilelog(
             self.svfs, f, self.repo2.svfs, self.unfiltered()._clrev, self
         )
 
     def close(self):
         self.repo2.close()
 
     def cancopy(self):
         return False
 
     def peer(self):
         return unionpeer(self)
 
     def getcwd(self):
         return encoding.getcwd()  # always outside the repo
 
 
 def instance(ui, path, create, intents=None, createopts=None):
     if create:
         raise error.Abort(_(b'cannot create new union repository'))
     parentpath = ui.config(b"bundle", b"mainreporoot")
     if not parentpath:
         # try to find the correct path to the working directory repo
         parentpath = cmdutil.findrepo(encoding.getcwd())
         if parentpath is None:
             parentpath = b''
     if parentpath:
         # Try to make the full path relative so we get a nice, short URL.
         # In particular, we don't want temp dir names in test outputs.
         cwd = encoding.getcwd()
         if parentpath == cwd:
             parentpath = b''
         else:
             cwd = pathutil.normasprefix(cwd)
             if parentpath.startswith(cwd):
                 parentpath = parentpath[len(cwd) :]
     if path.startswith(b'union:'):
         s = path.split(b":", 1)[1].split(b"+", 1)
         if len(s) == 1:
             repopath, repopath2 = parentpath, s[0]
         else:
             repopath, repopath2 = s
     else:
         repopath, repopath2 = parentpath, path
 
     return makeunionrepository(ui, repopath, repopath2)
 
 
 def makeunionrepository(ui, repopath1, repopath2):
     """Make a union repository object from 2 local repo paths."""
     repo1 = localrepo.instance(ui, repopath1, create=False)
     repo2 = localrepo.instance(ui, repopath2, create=False)
 
     url = b'union:%s+%s' % (
         util.expandpath(repopath1),
         util.expandpath(repopath2),
     )
 
     class derivedunionrepository(unionrepository, repo1.__class__):
         pass
 
     repo = repo1
     repo.__class__ = derivedunionrepository
     unionrepository.__init__(repo1, repo2, url)
 
     return repo