annotate hgext/lfs/wrapper.py @ 46582:b0a3ca02d17a

copies-rust: implement PartialEqual manually

Now that we know that each (dest, rev) pair has at most a unique CopySource, we
can simplify comparison a lot.

This "simple" step buy a good share of the previous slowdown back in some case:

Repo            Case                                  Source-Rev   Dest-Rev      # of revisions  old time    new time      Difference    Factor    time per rev
---------------------------------------------------------------------------------------------------------------------------------------------------------------
mozilla-try     x00000_revs_x00000_added_x000_copies  9b2a99adc05e 8e29777b48e6 : 382065 revs,  43.304637 s,  34.443661 s,  -8.860976 s, × 0.7954,    90 µs/rev


Full benchmark:

Repo            Case                                  Source-Rev   Dest-Rev      # of revisions  old time    new time      Difference    Factor    time per rev
---------------------------------------------------------------------------------------------------------------------------------------------------------------
mercurial       x_revs_x_added_0_copies               ad6b123de1c7 39cfcef4f463 :      1 revs,   0.000043 s,   0.000043 s,  +0.000000 s, × 1.0000,    43 µs/rev
mercurial       x_revs_x_added_x_copies               2b1c78674230 0c1d10351869 :      6 revs,   0.000114 s,   0.000117 s,  +0.000003 s, × 1.0263,    19 µs/rev
mercurial       x000_revs_x000_added_x_copies         81f8ff2a9bf2 dd3267698d84 :   1032 revs,   0.004937 s,   0.004892 s,  -0.000045 s, × 0.9909,     4 µs/rev
pypy            x_revs_x_added_0_copies               aed021ee8ae8 099ed31b181b :      9 revs,   0.000339 s,   0.000196 s,  -0.000143 s, × 0.5782,    21 µs/rev
pypy            x_revs_x000_added_0_copies            4aa4e1f8e19a 359343b9ac0e :      1 revs,   0.000049 s,   0.000050 s,  +0.000001 s, × 1.0204,    50 µs/rev
pypy            x_revs_x_added_x_copies               ac52eb7bbbb0 72e022663155 :      7 revs,   0.000202 s,   0.000117 s,  -0.000085 s, × 0.5792,    16 µs/rev
pypy            x_revs_x00_added_x_copies             c3b14617fbd7 ace7255d9a26 :      1 revs,   0.000409 s,   0.6f1f4a s,  -0.000087 s, × 0.7873,   322 µs/rev
pypy            x_revs_x000_added_x000_copies         df6f7a526b60 a83dc6a2d56f :      6 revs,   0.011984 s,   0.011949 s,  -0.000035 s, × 0.9971,  1991 µs/rev
pypy            x000_revs_xx00_added_0_copies         89a76aede314 2f22446ff07e :   4785 revs,   0.050820 s,   0.050802 s,  -0.000018 s, × 0.9996,    10 µs/rev
pypy            x000_revs_x000_added_x_copies         8a3b5bfd266e 2c68e87c3efe :   6780 revs,   0.087953 s,   0.088090 s,  +0.000137 s, × 1.0016,    12 µs/rev
pypy            x000_revs_x000_added_x000_copies      89a76aede314 7b3dda341c84 :   5441 revs,   0.062902 s,   0.062079 s,  -0.000823 s, × 0.9869,    11 µs/rev
pypy            x0000_revs_x_added_0_copies           d1defd0dc478 c9cb1334cc78 :  43645 revs,   0.679234 s,   0.635337 s,  -0.043897 s, × 0.9354,    14 µs/rev
pypy            x0000_revs_xx000_added_0_copies       bf2c629d0071 4ffed77c095c :      2 revs,   0.013095 s,   0.013262 s,  +0.000167 s, × 1.0128,  6631 µs/rev
pypy            x0000_revs_xx000_added_x000_copies    08ea3258278e d9fa043f30c0 :  11316 revs,   0.120910 s,   0.120085 s,  -0.000825 s, × 0.9932,    10 µs/rev
netbeans        x_revs_x_added_0_copies               fb0955ffcbcd a01e9239f9e7 :      2 revs,   0.000087 s,   0.000085 s,  -0.000002 s, × 0.9770,    42 µs/rev
netbeans        x_revs_x000_added_0_copies            6f360122949f 20eb231cc7d0 :      2 revs,   0.000107 s,   0.000110 s,  +0.000003 s, × 1.0280,    55 µs/rev
netbeans        x_revs_x_added_x_copies               1ada3faf6fb6 5a39d12eecf4 :      3 revs,   0.000186 s,   0.000177 s,  -0.000009 s, × 0.9516,    59 µs/rev
netbeans        x_revs_x00_added_x_copies             35be93ba1e2c 9eec5e90c05f :      9 revs,   0.000754 s,   0.000743 s,  -0.000011 s, × 0.9854,    82 µs/rev
netbeans        x000_revs_xx00_added_0_copies         eac3045b4fdd 51d4ae7f1290 :   1421 revs,   0.010443 s,   0.010168 s,  -0.000275 s, × 0.9737,     7 µs/rev
netbeans        x000_revs_x000_added_x_copies         e2063d266acd 6081d72689dc :   1533 revs,   0.015697 s,   0.015946 s,  +0.000249 s, × 1.0159,    10 µs/rev
netbeans        x000_revs_x000_added_x000_copies      ff453e9fee32 411350406ec2 :   5750 revs,   0.063528 s,   0.062712 s,  -0.000816 s, × 0.9872,    10 µs/rev
netbeans        x0000_revs_xx000_added_x000_copies    588c2d1ced70 1aad62e59ddd :  66949 revs,   0.545515 s,   0.523832 s,  -0.021683 s, × 0.9603,     7 µs/rev
mozilla-central x_revs_x_added_0_copies               3697f962bb7b 7015fcdd43a2 :      2 revs,   0.000089 s,   0.000090 s,  +0.000001 s, × 1.0112,    45 µs/rev
mozilla-central x_revs_x000_added_0_copies            dd390860c6c9 40d0c5bed75d :      8 revs,   0.000265 s,   0.000264 s,  -0.000001 s, × 0.9962,    33 µs/rev
mozilla-central x_revs_x_added_x_copies               8d198483ae3b 14207ffc2b2f :      9 revs,   0.000381 s,   0.000187 s,  -0.000194 s, × 0.4908,    20 µs/rev
mozilla-central x_revs_x00_added_x_copies             98cbc58cc6bc 446a150332c3 :      7 revs,   0.000672 s,   0.000665 s,  -0.000007 s, × 0.9896,    95 µs/rev
mozilla-central x_revs_x000_added_x000_copies         3c684b4b8f68 0a5e72d1b479 :      3 revs,   0.003497 s,   0.003556 s,  +0.000059 s, × 1.0169,  1185 µs/rev
mozilla-central x_revs_x0000_added_x0000_copies       effb563bb7e5 c07a39dc4e80 :      6 revs,   0.073204 s,   0.071345 s,  -0.001859 s, × 0.9746, 11890 µs/rev
mozilla-central x000_revs_xx00_added_0_copies         6100d773079a 04a55431795e :   1593 revs,   0.006482 s,   0.006551 s,  +0.000069 s, × 1.0106,     4 µs/rev
mozilla-central x000_revs_x000_added_x_copies         9f17a6fc04f9 2d37b966abed :     41 revs,   0.005066 s,   0.005078 s,  +0.000012 s, × 1.0024,   123 µs/rev
mozilla-central x000_revs_x000_added_x000_copies      7c97034feb78 4407bd0c6330 :   7839 revs,   0.065707 s,   0.065823 s,  +0.000116 s, × 1.0018,     8 µs/rev
mozilla-central x0000_revs_xx000_added_0_copies       9eec5917337d 67118cc6dcad :    615 revs,   0.026800 s,   0.027050 s,  +0.000250 s, × 1.0093,    43 µs/rev
mozilla-central x0000_revs_xx000_added_x000_copies    f78c615a656c 96a38b690156 :  30263 revs,   0.203856 s,   0.202443 s,  -0.001413 s, × 0.9931,     6 µs/rev
mozilla-central x00000_revs_x0000_added_x0000_copies  6832ae71433c 4c222a1d9a00 : 153721 revs,   1.293394 s,   1.261583 s,  -0.031811 s, × 0.9754,     8 µs/rev
mozilla-central x00000_revs_x00000_added_x000_copies  76caed42cf7c 1daa622bbe42 : 204976 revs,   1.698239 s,   1.643869 s,  -0.054370 s, × 0.9680,     8 µs/rev
mozilla-try     x_revs_x_added_0_copies               aaf6dde0deb8 9790f499805a :      2 revs,   0.000875 s,   0.000868 s,  -0.000007 s, × 0.9920,   434 µs/rev
mozilla-try     x_revs_x000_added_0_copies            d8d0222927b4 5bb8ce8c7450 :      2 revs,   0.000891 s,   0.000887 s,  -0.000004 s, × 0.9955,   443 µs/rev
mozilla-try     x_revs_x_added_x_copies               092fcca11bdb 936255a0384a :      4 revs,   0.000292 s,   0.000168 s,  -0.000124 s, × 0.5753,    42 µs/rev
mozilla-try     x_revs_x00_added_x_copies             b53d2fadbdb5 017afae788ec :      2 revs,   0.003939 s,   0.001160 s,  -0.002779 s, × 0.2945,   580 µs/rev
mozilla-try     x_revs_x000_added_x000_copies         20408ad61ce5 6f0ee96e21ad :      1 revs,   0.033027 s,   0.033016 s,  -0.000011 s, × 0.9997, 33016 µs/rev
mozilla-try     x_revs_x0000_added_x0000_copies       effb563bb7e5 c07a39dc4e80 :      6 revs,   0.073703 s,   0.073312 s,  -0.39ae31 s, × 0.9947, 12218 µs/rev
mozilla-try     x000_revs_xx00_added_0_copies         6100d773079a 04a55431795e :   1593 revs,   0.006469 s,   0.006485 s,  +0.000016 s, × 1.0025,     4 µs/rev
mozilla-try     x000_revs_x000_added_x_copies         9f17a6fc04f9 2d37b966abed :     41 revs,   0.005278 s,   0.005494 s,  +0.000216 s, × 1.0409,   134 µs/rev
mozilla-try     x000_revs_x000_added_x000_copies      1346fd0130e4 4c65cbdabc1f :   6657 revs,   0.064995 s,   0.064879 s,  -0.000116 s, × 0.9982,     9 µs/rev
mozilla-try     x0000_revs_x_added_0_copies           63519bfd42ee a36a2a865d92 :  40314 revs,   0.301041 s,   0.301469 s,  +0.000428 s, × 1.0014,     7 µs/rev
mozilla-try     x0000_revs_x_added_x_copies           9fe69ff0762d bcabf2a78927 :  38690 revs,   0.285575 s,   0.297113 s,  +0.011538 s, × 1.0404,     7 µs/rev
mozilla-try     x0000_revs_xx000_added_x_copies       156f6e2674f2 4d0f2c178e66 :   8598 revs,   0.085597 s,   0.085890 s,  +0.000293 s, × 1.0034,     9 µs/rev
mozilla-try     x0000_revs_xx000_added_0_copies       9eec5917337d 67118cc6dcad :    615 revs,   0.027118 s,   0.027718 s,  +0.000600 s, × 1.0221,    45 µs/rev
mozilla-try     x0000_revs_xx000_added_x000_copies    89294cd501d9 7ccb2fc7ccb5 :  97052 revs,   2.119204 s,   2.048949 s,  -0.070255 s, × 0.9668,    21 µs/rev
mozilla-try     x0000_revs_x0000_added_x0000_copies   e928c65095ed e951f4ad123a :  52031 revs,   0.701479 s,   0.685924 s,  -0.015555 s, × 0.9778,    13 µs/rev
mozilla-try     x00000_revs_x_added_0_copies          6a320851d377 1ebb79acd503 : 363753 revs,   4.482399 s,   4.482891 s,  +0.000492 s, × 1.0001,    12 µs/rev
mozilla-try     x00000_revs_x00000_added_0_copies     dc8a3ca7010e d16fde900c9c :  34414 revs,   0.574082 s,   0.577633 s,  +0.003551 s, × 1.0062,    16 µs/rev
mozilla-try     x00000_revs_x_added_x_copies          5173c4b6f97c 95d83ee7242d : 362229 revs,   4.480366 s,   4.397816 s,  -0.082550 s, × 0.9816,    12 µs/rev
mozilla-try     x00000_revs_x000_added_x_copies       9126823d0e9c ca82787bb23c : 359344 revs,   4.369070 s,   4.370538 s,  +0.001468 s, × 1.0003,    12 µs/rev
mozilla-try     x00000_revs_x0000_added_x0000_copies  8d3fafa80d4b eb884023b810 : 192665 revs,   1.592506 s,   1.570439 s,  -0.022067 s, × 0.9861,     8 µs/rev
mozilla-try     x00000_revs_x00000_added_x0000_copies 1b661134e2ca 1ae03d022d6d : 228985 revs,  87.824489 s,  88.388512 s,  +0.564023 s, × 1.0064,   386 µs/rev
mozilla-try     x00000_revs_x00000_added_x000_copies  9b2a99adc05e 8e29777b48e6 : 382065 revs,  43.304637 s,  34.443661 s,  -8.860976 s, × 0.7954,    90 µs/rev
private                                                                         : 459513 revs,  33.853687 s,  27.370148 s,  -6.483539 s, × 0.8085,    59 µs/rev

Differential Revision: https://phab.mercurial-scm.org/D9653

# wrapper.py - methods wrapping core mercurial logic
#
# Copyright 2017 Facebook, Inc.
#
# This software may be used and distributed according to the terms of the
# GNU General Public License version 2 or any later version.

from __future__ import absolute_import

import hashlib

from mercurial.i18n import _
from mercurial.node import bin, hex, nullid, short
from mercurial.pycompat import (
    getattr,
    setattr,
)

from mercurial import (
    bundle2,
    changegroup,
    cmdutil,
    context,
    error,
    exchange,
    exthelper,
    localrepo,
    pycompat,
    revlog,
    scmutil,
    util,
    vfs as vfsmod,
    wireprotov1server,
)

from mercurial.upgrade_utils import (
    actions as upgrade_actions,
    engine as upgrade_engine,
)

from mercurial.interfaces import repository

from mercurial.utils import (
    storageutil,
    stringutil,
)

from ..largefiles import lfutil

from . import (
    blobstore,
    pointer,
)

eh = exthelper.exthelper()


@eh.wrapfunction(localrepo, b'makefilestorage')
def localrepomakefilestorage(orig, requirements, features, **kwargs):
    if b'lfs' in requirements:
        features.add(repository.REPO_FEATURE_LFS)

    return orig(requirements=requirements, features=features, **kwargs)


@eh.wrapfunction(changegroup, b'allsupportedversions')
def allsupportedversions(orig, ui):
    versions = orig(ui)
    versions.add(b'03')
    return versions


@eh.wrapfunction(wireprotov1server, b'_capabilities')
def _capabilities(orig, repo, proto):
    '''Wrap server command to announce lfs server capability'''
    caps = orig(repo, proto)
    if util.safehasattr(repo.svfs, b'lfslocalblobstore'):
        # Advertise a slightly different capability when lfs is *required*, so
        # that the client knows it MUST load the extension.  If lfs is not
        # required on the server, there's no reason to autoload the extension
        # on the client.
        if b'lfs' in repo.requirements:
            caps.append(b'lfs-serve')

        caps.append(b'lfs')
    return caps


def bypasscheckhash(self, text):
    return False


def readfromstore(self, text):
    """Read filelog content from local blobstore transform for flagprocessor.

    Default tranform for flagprocessor, returning contents from blobstore.
    Returns a 2-typle (text, validatehash) where validatehash is True as the
    contents of the blobstore should be checked using checkhash.
    """
    p = pointer.deserialize(text)
    oid = p.oid()
    store = self.opener.lfslocalblobstore
    if not store.has(oid):
        p.filename = self.filename
        self.opener.lfsremoteblobstore.readbatch([p], store)

    # The caller will validate the content
    text = store.read(oid, verify=False)

    # pack hg filelog metadata
    hgmeta = {}
    for k in p.keys():
        if k.startswith(b'x-hg-'):
            name = k[len(b'x-hg-') :]
            hgmeta[name] = p[k]
    if hgmeta or text.startswith(b'\1\n'):
        text = storageutil.packmeta(hgmeta, text)

    return (text, True, {})


def writetostore(self, text, sidedata):
    # hg filelog metadata (includes rename, etc)
    hgmeta, offset = storageutil.parsemeta(text)
    if offset and offset > 0:
        # lfs blob does not contain hg filelog metadata
        text = text[offset:]

    # git-lfs only supports sha256
    oid = hex(hashlib.sha256(text).digest())
    self.opener.lfslocalblobstore.write(oid, text)

    # replace contents with metadata
    longoid = b'sha256:%s' % oid
    metadata = pointer.gitlfspointer(oid=longoid, size=b'%d' % len(text))

    # by default, we expect the content to be binary. however, LFS could also
    # be used for non-binary content. add a special entry for non-binary data.
    # this will be used by filectx.isbinary().
    if not stringutil.binary(text):
        # not hg filelog metadata (affecting commit hash), no "x-hg-" prefix
        metadata[b'x-is-binary'] = b'0'

    # translate hg filelog metadata to lfs metadata with "x-hg-" prefix
    if hgmeta is not None:
        for k, v in pycompat.iteritems(hgmeta):
            metadata[b'x-hg-%s' % k] = v

    rawtext = metadata.serialize()
    return (rawtext, False)


def _islfs(rlog, node=None, rev=None):
    if rev is None:
        if node is None:
            # both None - likely working copy content where node is not ready
            return False
        rev = rlog.rev(node)
    else:
        node = rlog.node(rev)
    if node == nullid:
        return False
    flags = rlog.flags(rev)
    return bool(flags & revlog.REVIDX_EXTSTORED)


# Wrapping may also be applied by remotefilelog
def filelogaddrevision(
    orig,
    self,
    text,
    transaction,
    link,
    p1,
    p2,
    cachedelta=None,
    node=None,
    flags=revlog.REVIDX_DEFAULT_FLAGS,
    **kwds
):
    # The matcher isn't available if reposetup() wasn't called.
    lfstrack = self._revlog.opener.options.get(b'lfstrack')

    if lfstrack:
        textlen = len(text)
        # exclude hg rename meta from file size
        meta, offset = storageutil.parsemeta(text)
        if offset:
            textlen -= offset

        if lfstrack(self._revlog.filename, textlen):
            flags |= revlog.REVIDX_EXTSTORED

    return orig(
        self,
        text,
        transaction,
        link,
        p1,
        p2,
        cachedelta=cachedelta,
        node=node,
        flags=flags,
        **kwds
    )


# Wrapping may also be applied by remotefilelog
def filelogrenamed(orig, self, node):
    if _islfs(self._revlog, node):
        rawtext = self._revlog.rawdata(node)
        if not rawtext:
            return False
        metadata = pointer.deserialize(rawtext)
        if b'x-hg-copy' in metadata and b'x-hg-copyrev' in metadata:
            return metadata[b'x-hg-copy'], bin(metadata[b'x-hg-copyrev'])
        else:
            return False
    return orig(self, node)


# Wrapping may also be applied by remotefilelog
def filelogsize(orig, self, rev):
    if _islfs(self._revlog, rev=rev):
        # fast path: use lfs metadata to answer size
        rawtext = self._revlog.rawdata(rev)
        metadata = pointer.deserialize(rawtext)
        return int(metadata[b'size'])
    return orig(self, rev)


@eh.wrapfunction(revlog, b'_verify_revision')
def _verify_revision(orig, rl, skipflags, state, node):
    if _islfs(rl, node=node):
        rawtext = rl.rawdata(node)
        metadata = pointer.deserialize(rawtext)

        # Don't skip blobs that are stored locally, as local verification is
        # relatively cheap and there's no other way to verify the raw data in
        # the revlog.
        if rl.opener.lfslocalblobstore.has(metadata.oid()):
            skipflags &= ~revlog.REVIDX_EXTSTORED
        elif skipflags & revlog.REVIDX_EXTSTORED:
            # The wrapped method will set `skipread`, but there's enough local
            # info to check renames.
            state[b'safe_renamed'].add(node)

    orig(rl, skipflags, state, node)


@eh.wrapfunction(context.basefilectx, b'cmp')
def filectxcmp(orig, self, fctx):
    """returns True if text is different than fctx"""
    # some fctx (ex. hg-git) is not based on basefilectx and do not have islfs
    if self.islfs() and getattr(fctx, 'islfs', lambda: False)():
        # fast path: check LFS oid
        p1 = pointer.deserialize(self.rawdata())
        p2 = pointer.deserialize(fctx.rawdata())
        return p1.oid() != p2.oid()
    return orig(self, fctx)


@eh.wrapfunction(context.basefilectx, b'isbinary')
def filectxisbinary(orig, self):
    if self.islfs():
        # fast path: use lfs metadata to answer isbinary
        metadata = pointer.deserialize(self.rawdata())
        # if lfs metadata says nothing, assume it's binary by default
        return bool(int(metadata.get(b'x-is-binary', 1)))
    return orig(self)


def filectxislfs(self):
    return _islfs(self.filelog()._revlog, self.filenode())


@eh.wrapfunction(cmdutil, b'_updatecatformatter')
def _updatecatformatter(orig, fm, ctx, matcher, path, decode):
    orig(fm, ctx, matcher, path, decode)
    fm.data(rawdata=ctx[path].rawdata())


@eh.wrapfunction(scmutil, b'wrapconvertsink')
def convertsink(orig, sink):
    sink = orig(sink)
    if sink.repotype == b'hg':

        class lfssink(sink.__class__):
            def putcommit(
                self,
                files,
                copies,
                parents,
                commit,
                source,
                revmap,
                full,
                cleanp2,
            ):
                pc = super(lfssink, self).putcommit
                node = pc(
                    files,
                    copies,
                    parents,
                    commit,
                    source,
                    revmap,
                    full,
                    cleanp2,
                )

                if b'lfs' not in self.repo.requirements:
                    ctx = self.repo[node]

                    # The file list may contain removed files, so check for
                    # membership before assuming it is in the context.
                    if any(f in ctx and ctx[f].islfs() for f, n in files):
                        self.repo.requirements.add(b'lfs')
                        scmutil.writereporequirements(self.repo)

                return node

        sink.__class__ = lfssink

    return sink


# bundlerepo uses "vfsmod.readonlyvfs(othervfs)", we need to make sure lfs
# options and blob stores are passed from othervfs to the new readonlyvfs.
@eh.wrapfunction(vfsmod.readonlyvfs, b'__init__')
def vfsinit(orig, self, othervfs):
    orig(self, othervfs)
    # copy lfs related options
    for k, v in othervfs.options.items():
        if k.startswith(b'lfs'):
            self.options[k] = v
    # also copy lfs blobstores. note: this can run before reposetup, so lfs
    # blobstore attributes are not always ready at this time.
    for name in [b'lfslocalblobstore', b'lfsremoteblobstore']:
        if util.safehasattr(othervfs, name):
            setattr(self, name, getattr(othervfs, name))


def _prefetchfiles(repo, revmatches):
    """Ensure that required LFS blobs are present, fetching them as a group if
    needed."""
    if not util.safehasattr(repo.svfs, b'lfslocalblobstore'):
        return

    pointers = []
    oids = set()
    localstore = repo.svfs.lfslocalblobstore

    for rev, match in revmatches:
        ctx = repo[rev]
        for f in ctx.walk(match):
            p = pointerfromctx(ctx, f)
            if p and p.oid() not in oids and not localstore.has(p.oid()):
                p.filename = f
                pointers.append(p)
                oids.add(p.oid())

    if pointers:
        # Recalculating the repo store here allows 'paths.default' that is set
        # on the repo by a clone command to be used for the update.
        blobstore.remote(repo).readbatch(pointers, localstore)


def _canskipupload(repo):
    # Skip if this hasn't been passed to reposetup()
    if not util.safehasattr(repo.svfs, b'lfsremoteblobstore'):
        return True

    # if remotestore is a null store, upload is a no-op and can be skipped
    return isinstance(repo.svfs.lfsremoteblobstore, blobstore._nullremote)


def candownload(repo):
    # Skip if this hasn't been passed to reposetup()
    if not util.safehasattr(repo.svfs, b'lfsremoteblobstore'):
        return False

    # if remotestore is a null store, downloads will lead to nothing
    return not isinstance(repo.svfs.lfsremoteblobstore, blobstore._nullremote)


def uploadblobsfromrevs(repo, revs):
    """upload lfs blobs introduced by revs

    Note: also used by other extensions e. g. infinitepush. avoid renaming.
    """
    if _canskipupload(repo):
        return
    pointers = extractpointers(repo, revs)
    uploadblobs(repo, pointers)


def prepush(pushop):
    """Prepush hook.

    Read through the revisions to push, looking for filelog entries that can be
    deserialized into metadata so that we can block the push on their upload to
    the remote blobstore.
    """
    return uploadblobsfromrevs(pushop.repo, pushop.outgoing.missing)


@eh.wrapfunction(exchange, b'push')
def push(orig, repo, remote, *args, **kwargs):
    """bail on push if the extension isn't enabled on remote when needed, and
    update the remote store based on the destination path."""
    if b'lfs' in repo.requirements:
        # If the remote peer is for a local repo, the requirement tests in the
        # base class method enforce lfs support.  Otherwise, some revisions in
        # this repo use lfs, and the remote repo needs the extension loaded.
        if not remote.local() and not remote.capable(b'lfs'):
            # This is a copy of the message in exchange.push() when requirements
            # are missing between local repos.
            m = _(b"required features are not supported in the destination: %s")
            raise error.Abort(
                m % b'lfs', hint=_(b'enable the lfs extension on the server')
            )

        # Repositories where this extension is disabled won't have the field.
        # But if there's a requirement, then the extension must be loaded AND
        # there may be blobs to push.
        remotestore = repo.svfs.lfsremoteblobstore
        try:
            repo.svfs.lfsremoteblobstore = blobstore.remote(repo, remote.url())
            return orig(repo, remote, *args, **kwargs)
        finally:
            repo.svfs.lfsremoteblobstore = remotestore
    else:
        return orig(repo, remote, *args, **kwargs)


# when writing a bundle via "hg bundle" command, upload related LFS blobs
@eh.wrapfunction(bundle2, b'writenewbundle')
def writenewbundle(
    orig, ui, repo, source, filename, bundletype, outgoing, *args, **kwargs
):
    """upload LFS blobs added by outgoing revisions on 'hg bundle'"""
    uploadblobsfromrevs(repo, outgoing.missing)
    return orig(
        ui, repo, source, filename, bundletype, outgoing, *args, **kwargs
    )


def extractpointers(repo, revs):
    """return a list of lfs pointers added by given revs"""
    repo.ui.debug(b'lfs: computing set of blobs to upload\n')
    pointers = {}

    makeprogress = repo.ui.makeprogress
    with makeprogress(
        _(b'lfs search'), _(b'changesets'), len(revs)
    ) as progress:
        for r in revs:
            ctx = repo[r]
            for p in pointersfromctx(ctx).values():
                pointers[p.oid()] = p
            progress.increment()
        return sorted(pointers.values(), key=lambda p: p.oid())


def pointerfromctx(ctx, f, removed=False):
    """return a pointer for the named file from the given changectx, or None if
    the file isn't LFS.

    Optionally, the pointer for a file deleted from the context can be returned.
    Since no such pointer is actually stored, and to distinguish from a non LFS
    file, this pointer is represented by an empty dict.
    """
    _ctx = ctx
    if f not in ctx:
        if not removed:
            return None
        if f in ctx.p1():
            _ctx = ctx.p1()
        elif f in ctx.p2():
            _ctx = ctx.p2()
        else:
            return None
    fctx = _ctx[f]
    if not _islfs(fctx.filelog()._revlog, fctx.filenode()):
        return None
    try:
        p = pointer.deserialize(fctx.rawdata())
        if ctx == _ctx:
            return p
        return {}
    except pointer.InvalidPointer as ex:
        raise error.Abort(
            _(b'lfs: corrupted pointer (%s@%s): %s\n')
            % (f, short(_ctx.node()), ex)
        )


def pointersfromctx(ctx, removed=False):
    """return a dict {path: pointer} for given single changectx.

    If ``removed`` == True and the LFS file was removed from ``ctx``, the value
    stored for the path is an empty dict.
    """
    result = {}
    m = ctx.repo().narrowmatch()

    # TODO: consider manifest.fastread() instead
    for f in ctx.files():
        if not m(f):
            continue
        p = pointerfromctx(ctx, f, removed=removed)
        if p is not None:
            result[f] = p
    return result


def uploadblobs(repo, pointers):
    """upload given pointers from local blobstore"""
    if not pointers:
        return

    remoteblob = repo.svfs.lfsremoteblobstore
    remoteblob.writebatch(pointers, repo.svfs.lfslocalblobstore)


@eh.wrapfunction(upgrade_engine, b'finishdatamigration')
def upgradefinishdatamigration(orig, ui, srcrepo, dstrepo, requirements):
    orig(ui, srcrepo, dstrepo, requirements)

    # Skip if this hasn't been passed to reposetup()
    if util.safehasattr(
        srcrepo.svfs, b'lfslocalblobstore'
    ) and util.safehasattr(dstrepo.svfs, b'lfslocalblobstore'):
        srclfsvfs = srcrepo.svfs.lfslocalblobstore.vfs
        dstlfsvfs = dstrepo.svfs.lfslocalblobstore.vfs

        for dirpath, dirs, files in srclfsvfs.walk():
            for oid in files:
                ui.write(_(b'copying lfs blob %s\n') % oid)
                lfutil.link(srclfsvfs.join(oid), dstlfsvfs.join(oid))


@eh.wrapfunction(upgrade_actions, b'preservedrequirements')
@eh.wrapfunction(upgrade_actions, b'supporteddestrequirements')
def upgraderequirements(orig, repo):
    reqs = orig(repo)
    if b'lfs' in repo.requirements:
        reqs.add(b'lfs')
    return reqs