view mercurial/metadata.py @ 48687:f8f2ecdde4b5

branchmap: skip obsolete revisions while computing heads

It's time to make this part of core Mercurial obsolescence-aware.

Not considering obsolete revisions when computing heads is clearly what
Mercurial should do. But there are a couple of small issues:

- Let's say tip of the repo is obsolete. There are two ways of finding tiprev
  for branchcache (both are in use): looking at input data for update() and
  looking at computed heads after update(). Previously, repo tip would be
  tiprev of the branchcache. With this patch, an obsolete revision can no
  longer be tiprev. And depending on what way we use for finding tiprev (input
  data vs computed heads) we'll get a different result. This is relevant when
  recomputing cache key from cache contents, and may lead to updating cache for
  obsolete revisions multiple times (not from scratch, because it still would
  be considered valid for a subset of revisions in the repo).

- If all commits on a branch are obsolete, the branchcache will include that
  branch, but the list of heads will be empty (that's why there's now `if not
  heads` when recomputing tiprev/tipnode from cache contents). Having an entry
  for every branch is currently required for notify extension (and
  test-notify.t to pass), because notify doesn't handle revsets in its
  subscription config very well and will throw an error if e.g. a branch
  doesn't exist.

- Cloning static HTTP repos may try to stat() a non-existent obsstore file. The
  issue is that we now care about obsolescence during clone, but statichttpvfs
  doesn't implement a stat method, so a regular vfs.stat() is used, and it
  assumes that file is local and calls os.stat().  During a clone, we're trying
  to stat() .hg/store/obsstore, but in static HTTP case we provide a literal
  URL to the obsstore file on the remote as if it were a local file path. On
  windows it actually results in a failure in test-static-http.t.

The first issue is going to be addressed in a series dedicated to making sure
branchcache is properly and timely written on disk (it wasn't perfect even
before this patch, but there aren't enough tests to demonstrate that). The
second issue will be addressed in a future patch for notify extension that will
make it not raise an exception if a branch doesn't exist. And the third one was
partially addressed in the previous patch in this series and will be properly
fixed in a future patch when this series is accepted.

filteredhash() grows a keyword argument to make sure that branchcache is also
invalidated when there are new obsolete revisions in its repo view. This way
the on-disk cache format is unchanged and compatible between versions (although
it will obviously be recomputed when switching versions before/after this patch
and the repo has obsolete revisions).

There's one test that uses plain `hg up` without arguments while updated to a
pruned commit. To make this test pass, simply return current working directory
parent. Later in this series this code will be replaced by what prune command
does: updating to the closest non-obsolete ancestor.

Test changes:

test-branch-change.t: update branch head and cache update message. The head of
default listed in hg heads is changed because revision 2 was rewritten as 7,
and 1 is the closest ancestor on the same branch, so it's the head of default
now.

The cache invalidation message appears now because of the cache hash change,
since we're now accounting for obsolete revisions. Here's some context:

"served.hidden" repo filter means everything is visible (no filtered
revisions), so before this series branch2-served.hidden file would not contain
any cache hash, only revnum and node. Now it also has a hash when there are
obsolete changesets in the repo. The command that the message appears for is
changing branch of 5 and 6, which are now obsolete, so the cache hash changes.

In general, when cache is simply out-of-date, it can be updated using the old
version as a base. But if cache hash differs, then the cache for that
particular repo filter is recomputed (at least with the current
implementation). This is what happens here.

test-obsmarker-template.t: the pull reports 2 heads changed, but after that the
repo correctly sees only 1. The new message could be better, but it's still an
improvement over the previous one where hg pull suggested merging with an
obsolete revision.

test-obsolete.t: we can see these revisions in hg log --hidden, but they
shouldn't be considered heads even with --hidden.

test-rebase-obsolete{,2}.t: there were new heads created previously after
making new orphan changesets, but they weren't detected. Now we are properly
detecting and reporting them.

test-rebase-obsolete4.t: there's only one head now because the other head is
pruned and was falsely reported before.

test-static-http.t: add obsstore to the list of requested files. This file
doesn't exist on the remotes, but clients want it anyway (they get 404). This
is fine, because there are other nonexistent files that clients request, like
.hg/bookmarks or .hg/cache/tags2-served.

Differential Revision: https://phab.mercurial-scm.org/D12097

# coding: utf-8
# metadata.py -- code related to various metadata computation and access.
#
# Copyright 2019 Google, Inc <[e-mail]>
# Copyright 2020 Pierre-Yves David <[e-mail]>
#
# This software may be used and distributed according to the terms of the
# GNU General Public License version 2 or any later version.
from __future__ import absolute_import, print_function

import multiprocessing
import struct

from .node import nullrev
from . import (
    error,
    util,
)

from .revlogutils import (
    flagutil as sidedataflag,
    sidedata as sidedatamod,
)


class ChangingFiles(object):
    """A class recording the changes made to files by a changeset

    Actions performed on files are gathered into 3 sets:

    - added:   files actively added in the changeset.
    - merged:  files whose history got merged
    - removed: files removed in the revision
    - salvaged: files that might have been deleted by a merge but were not
    - touched: files affected by the merge

    and copies information is held by 2 mappings

    - copied_from_p1: {"<new-name>": "<source-name-in-p1>"} mapping for copies
    - copied_from_p2: {"<new-name>": "<source-name-in-p2>"} mapping for copies

    See their inline help for details.
    """

    def __init__(
        self,
        touched=None,
        added=None,
        removed=None,
        merged=None,
        salvaged=None,
        p1_copies=None,
        p2_copies=None,
    ):
        self._added = set(() if added is None else added)
        self._merged = set(() if merged is None else merged)
        self._removed = set(() if removed is None else removed)
        self._touched = set(() if touched is None else touched)
        self._salvaged = set(() if salvaged is None else salvaged)
        self._touched.update(self._added)
        self._touched.update(self._merged)
        self._touched.update(self._removed)
        self._p1_copies = dict(() if p1_copies is None else p1_copies)
        self._p2_copies = dict(() if p2_copies is None else p2_copies)

    def __eq__(self, other):
        return (
            self.added == other.added
            and self.merged == other.merged
            and self.removed == other.removed
            and self.salvaged == other.salvaged
            and self.touched == other.touched
            and self.copied_from_p1 == other.copied_from_p1
            and self.copied_from_p2 == other.copied_from_p2
        )

    @property
    def has_copies_info(self):
        return bool(
            self.removed
            or self.merged
            or self.salvaged
            or self.copied_from_p1
            or self.copied_from_p2
        )

    @util.propertycache
    def added(self):
        """files actively added in the changeset

        Any file present in that revision that was absent in all the changeset's
        parents.

        In case of merge, this means a file absent in one of the parents but
        existing in the other will *not* be contained in this set. (They were
        added by an ancestor)
        """
        return frozenset(self._added)

    def mark_added(self, filename):
        if 'added' in vars(self):
            del self.added
        self._added.add(filename)
        self.mark_touched(filename)

    def update_added(self, filenames):
        for f in filenames:
            self.mark_added(f)

    @util.propertycache
    def merged(self):
        """files actively merged during a merge

        Any modified files which had modification on both size that needed merging.

        In this case a new filenode was created and it has two parents.
        """
        return frozenset(self._merged)

    def mark_merged(self, filename):
        if 'merged' in vars(self):
            del self.merged
        self._merged.add(filename)
        self.mark_touched(filename)

    def update_merged(self, filenames):
        for f in filenames:
            self.mark_merged(f)

    @util.propertycache
    def removed(self):
        """files actively removed by the changeset

        In case of merge this will only contain the set of files removing "new"
        content. For any file absent in the current changeset:

        a) If the file exists in both parents, it is clearly "actively" removed
        by this changeset.

        b) If a file exists in only one parent and in none of the common
        ancestors, then the file was newly added in one of the merged branches
        and then got "actively" removed.

        c) If a file exists in only one parent and at least one of the common
        ancestors using the same filenode, then the file was unchanged on one
        side and deleted on the other side. The merge "passively" propagated
        that deletion, but didn't "actively" remove the file. In this case the
        file is *not* included in the `removed` set.

        d) If a file exists in only one parent and at least one of the common
        ancestors using a different filenode, then the file was changed on one
        side and removed on the other side. The merge process "actively"
        decided to drop the new change and delete the file. Unlike in the
        previous case, (c), the file included in the `removed` set.

        Summary table for merge:

        case | exists in parents | exists in gca || removed
         (a) |       both        |     *         ||   yes
         (b) |       one         |     none      ||   yes
         (c) |       one         | same filenode ||   no
         (d) |       one         |  new filenode ||   yes
        """
        return frozenset(self._removed)

    def mark_removed(self, filename):
        if 'removed' in vars(self):
            del self.removed
        self._removed.add(filename)
        self.mark_touched(filename)

    def update_removed(self, filenames):
        for f in filenames:
            self.mark_removed(f)

    @util.propertycache
    def salvaged(self):
        """files that might have been deleted by a merge, but still exists.

        During a merge, the manifest merging might select some files for
        removal, or for a removed/changed conflict. If at commit time the file
        still exists, its removal was "reverted" and the file is "salvaged"
        """
        return frozenset(self._salvaged)

    def mark_salvaged(self, filename):
        if "salvaged" in vars(self):
            del self.salvaged
        self._salvaged.add(filename)
        self.mark_touched(filename)

    def update_salvaged(self, filenames):
        for f in filenames:
            self.mark_salvaged(f)

    @util.propertycache
    def touched(self):
        """files either actively modified, added or removed"""
        return frozenset(self._touched)

    def mark_touched(self, filename):
        if 'touched' in vars(self):
            del self.touched
        self._touched.add(filename)

    def update_touched(self, filenames):
        for f in filenames:
            self.mark_touched(f)

    @util.propertycache
    def copied_from_p1(self):
        return self._p1_copies.copy()

    def mark_copied_from_p1(self, source, dest):
        if 'copied_from_p1' in vars(self):
            del self.copied_from_p1
        self._p1_copies[dest] = source

    def update_copies_from_p1(self, copies):
        for dest, source in copies.items():
            self.mark_copied_from_p1(source, dest)

    @util.propertycache
    def copied_from_p2(self):
        return self._p2_copies.copy()

    def mark_copied_from_p2(self, source, dest):
        if 'copied_from_p2' in vars(self):
            del self.copied_from_p2
        self._p2_copies[dest] = source

    def update_copies_from_p2(self, copies):
        for dest, source in copies.items():
            self.mark_copied_from_p2(source, dest)


def compute_all_files_changes(ctx):
    """compute the files changed by a revision"""
    p1 = ctx.p1()
    p2 = ctx.p2()
    if p1.rev() == nullrev and p2.rev() == nullrev:
        return _process_root(ctx)
    elif p1.rev() != nullrev and p2.rev() == nullrev:
        return _process_linear(p1, ctx)
    elif p1.rev() == nullrev and p2.rev() != nullrev:
        # In the wild, one can encounter changeset where p1 is null but p2 is not
        return _process_linear(p1, ctx, parent=2)
    elif p1.rev() == p2.rev():
        # In the wild, one can encounter such "non-merge"
        return _process_linear(p1, ctx)
    else:
        return _process_merge(p1, p2, ctx)


def _process_root(ctx):
    """compute the appropriate changed files for a changeset with no parents"""
    # Simple, there was nothing before it, so everything is added.
    md = ChangingFiles()
    manifest = ctx.manifest()
    for filename in manifest:
        md.mark_added(filename)
    return md


def _process_linear(parent_ctx, children_ctx, parent=1):
    """compute the appropriate changed files for a changeset with a single parent"""
    md = ChangingFiles()
    parent_manifest = parent_ctx.manifest()
    children_manifest = children_ctx.manifest()

    copies_candidate = []

    for filename, d in parent_manifest.diff(children_manifest).items():
        if d[1][0] is None:
            # no filenode for the "new" value, file is absent
            md.mark_removed(filename)
        else:
            copies_candidate.append(filename)
            if d[0][0] is None:
                # not filenode for the "old" value file was absent
                md.mark_added(filename)
            else:
                # filenode for both "old" and "new"
                md.mark_touched(filename)

    if parent == 1:
        copied = md.mark_copied_from_p1
    elif parent == 2:
        copied = md.mark_copied_from_p2
    else:
        assert False, "bad parent value %d" % parent

    for filename in copies_candidate:
        copy_info = children_ctx[filename].renamed()
        if copy_info:
            source, srcnode = copy_info
            copied(source, filename)

    return md


def _process_merge(p1_ctx, p2_ctx, ctx):
    """compute the appropriate changed files for a changeset with two parents

    This is a more advance case. The information we need to record is summarise
    in the following table:

    ┌──────────────┬──────────────┬──────────────┬──────────────┬──────────────┐
    │ diff ╲  diff │       ø      │ (Some, None) │ (None, Some) │ (Some, Some) │
    │  p2   ╲  p1  │              │              │              │              │
    ├──────────────┼──────────────┼──────────────┼──────────────┼──────────────┤
    │              │              │🄱  No Changes │🄳  No Changes │              │
    │  ø           │🄰  No Changes │      OR      │     OR       │🄵  No Changes │
    │              │              │🄲  Deleted[1] │🄴  Salvaged[2]│     [3]      │
    ├──────────────┼──────────────┼──────────────┼──────────────┼──────────────┤
    │              │🄶  No Changes │              │              │              │
    │ (Some, None) │      OR      │🄻  Deleted    │       ø      │      ø       │
    │              │🄷  Deleted[1] │              │              │              │
    ├──────────────┼──────────────┼──────────────┼──────────────┼──────────────┤
    │              │🄸  No Changes │              │              │   🄽 Touched  │
    │ (None, Some) │     OR       │      ø       │🄼   Added     │OR 🅀 Salvaged │
    │              │🄹  Salvaged[2]│              │   (copied?)  │   (copied?)  │
    ├──────────────┼──────────────┼──────────────┼──────────────┼──────────────┤
    │              │              │              │   🄾 Touched  │   🄿 Merged   │
    │ (Some, Some) │🄺  No Changes │      ø       │OR 🅁 Salvaged │OR 🅂 Touched  │
    │              │     [3]      │              │   (copied?)  │   (copied?)  │
    └──────────────┴──────────────┴──────────────┴──────────────┴──────────────┘

    Special case [1]:

      The situation is:
        - parent-A:     file exists,
        - parent-B:     no file,
        - working-copy: no file.

      Detecting a "deletion" will depend on the presence of actual change on
      the "parent-A" branch:

      Subcase 🄱 or 🄶 : if the state of the file in "parent-A" is unchanged
      compared to the merge ancestors, then parent-A branch left the file
      untouched while parent-B deleted it. We simply apply the change from
      "parent-B" branch the file was automatically dropped.
      The result is:
          - file is not recorded as touched by the merge.

      Subcase 🄲 or 🄷 : otherwise, the change from parent-A branch were explicitly dropped and
      the file was "deleted again". From a user perspective, the message
      about "locally changed" while "remotely deleted" (or the other way
      around) was issued and the user chose to deleted the file.
      The result:
          - file is recorded as touched by the merge.


    Special case [2]:

      The situation is:
        - parent-A:     no file,
        - parent-B:     file,
        - working-copy: file (same content as parent-B).

      There are three subcases depending on the ancestors contents:

      - A) the file is missing in all ancestors,
      - B) at least one ancestor has the file with filenode ≠ from parent-B,
      - C) all ancestors use the same filenode as parent-B,

      Subcase (A) is the simpler, nothing happend on parent-A side while
      parent-B added it.

        The result:
            - the file is not marked as touched by the merge.

      Subcase (B) is the counter part of "Special case [1]", the file was
        modified on parent-B side, while parent-A side deleted it. However this
        time, the conflict was solved by keeping the file (and its
        modification). We consider the file as "salvaged".

        The result:
            - the file is marked as "salvaged" by the merge.

      Subcase (C) is subtle variation of the case above. In this case, the
        file in unchanged on the parent-B side and actively removed on the
        parent-A side. So the merge machinery correctly decide it should be
        removed. However, the file was explicitly restored to its parent-B
        content before the merge was commited. The file is be marked
        as salvaged too. From the merge result perspective, this is similar to
        Subcase (B), however from the merge resolution perspective they differ
        since in (C), there was some conflict not obvious solution to the
        merge (That got reversed)

    Special case [3]:

      The situation is:
        - parent-A:     file,
        - parent-B:     file (different filenode as parent-A),
        - working-copy: file (same filenode as parent-B).

      This case is in theory much simple, for this to happens, this mean the
      filenode in parent-A is purely replacing the one in parent-B (either a
      descendant, or a full new file history, see changeset). So the merge
      introduce no changes, and the file is not affected by the merge...

      However, in the wild it is possible to find commit with the above is not
      True. For example repository have some commit where the *new* node is an
      ancestor of the node in parent-A, or where parent-A and parent-B are two
      branches of the same file history, yet not merge-filenode were created
      (while the "merge" should have led to a "modification").

      Detecting such cases (and not recording the file as modified) would be a
      nice bonus. However do not any of this yet.
    """

    repo = ctx.repo()
    md = ChangingFiles()

    m = ctx.manifest()
    p1m = p1_ctx.manifest()
    p2m = p2_ctx.manifest()
    diff_p1 = p1m.diff(m)
    diff_p2 = p2m.diff(m)

    cahs = ctx.repo().changelog.commonancestorsheads(
        p1_ctx.node(), p2_ctx.node()
    )
    if not cahs:
        cahs = [nullrev]
    mas = [ctx.repo()[r].manifest() for r in cahs]

    copy_candidates = []

    # Dealing with case 🄰 happens automatically.  Since there are no entry in
    # d1 nor d2, we won't iterate on it ever.

    # Iteration over d1 content will deal with all cases, but the one in the
    # first column of the table.
    for filename, d1 in diff_p1.items():

        d2 = diff_p2.pop(filename, None)

        if d2 is None:
            # this deal with the first line of the table.
            _process_other_unchanged(md, mas, filename, d1)
        else:

            if d1[0][0] is None and d2[0][0] is None:
                # case 🄼 — both deleted the file.
                md.mark_added(filename)
                copy_candidates.append(filename)
            elif d1[1][0] is None and d2[1][0] is None:
                # case 🄻 — both deleted the file.
                md.mark_removed(filename)
            elif d1[1][0] is not None and d2[1][0] is not None:
                if d1[0][0] is None or d2[0][0] is None:
                    if any(_find(ma, filename) is not None for ma in mas):
                        # case 🅀 or 🅁
                        md.mark_salvaged(filename)
                    else:
                        # case 🄽 🄾 : touched
                        md.mark_touched(filename)
                else:
                    fctx = repo.filectx(filename, fileid=d1[1][0])
                    if fctx.p2().rev() == nullrev:
                        # case 🅂
                        # lets assume we can trust the file history. If the
                        # filenode is not a merge, the file was not merged.
                        md.mark_touched(filename)
                    else:
                        # case 🄿
                        md.mark_merged(filename)
                copy_candidates.append(filename)
            else:
                # Impossible case, the post-merge file status cannot be None on
                # one side and Something on the other side.
                assert False, "unreachable"

    # Iteration over remaining d2 content deal with the first column of the
    # table.
    for filename, d2 in diff_p2.items():
        _process_other_unchanged(md, mas, filename, d2)

    for filename in copy_candidates:
        copy_info = ctx[filename].renamed()
        if copy_info:
            source, srcnode = copy_info
            if source in p1_ctx and p1_ctx[source].filenode() == srcnode:
                md.mark_copied_from_p1(source, filename)
            elif source in p2_ctx and p2_ctx[source].filenode() == srcnode:
                md.mark_copied_from_p2(source, filename)
    return md


def _find(manifest, filename):
    """return the associate filenode or None"""
    if filename not in manifest:
        return None
    return manifest.find(filename)[0]


def _process_other_unchanged(md, mas, filename, diff):
    source_node = diff[0][0]
    target_node = diff[1][0]

    if source_node is not None and target_node is None:
        if any(not _find(ma, filename) == source_node for ma in mas):
            # case 🄲 of 🄷
            md.mark_removed(filename)
        # else, we have case 🄱 or 🄶 : no change need to be recorded
    elif source_node is None and target_node is not None:
        if any(_find(ma, filename) is not None for ma in mas):
            # case 🄴 or 🄹
            md.mark_salvaged(filename)
        # else, we have case 🄳 or 🄸 : simple merge without intervention
    elif source_node is not None and target_node is not None:
        # case 🄵  or 🄺 : simple merge without intervention
        #
        # In buggy case where source_node is not an ancestors of target_node.
        # There should have a been a new filenode created, recording this as
        # "modified". We do not deal with them yet.
        pass
    else:
        # An impossible case, the diff algorithm should not return entry if the
        # file is missing on both side.
        assert False, "unreachable"


def _missing_from_all_ancestors(mas, filename):
    return all(_find(ma, filename) is None for ma in mas)


def computechangesetfilesadded(ctx):
    """return the list of files added in a changeset"""
    added = []
    for f in ctx.files():
        if not any(f in p for p in ctx.parents()):
            added.append(f)
    return added


def get_removal_filter(ctx, x=None):
    """return a function to detect files "wrongly" detected as `removed`

    When a file is removed relative to p1 in a merge, this
    function determines whether the absence is due to a
    deletion from a parent, or whether the merge commit
    itself deletes the file. We decide this by doing a
    simplified three way merge of the manifest entry for
    the file. There are two ways we decide the merge
    itself didn't delete a file:
    - neither parent (nor the merge) contain the file
    - exactly one parent contains the file, and that
      parent has the same filelog entry as the merge
      ancestor (or all of them if there two). In other
      words, that parent left the file unchanged while the
      other one deleted it.
    One way to think about this is that deleting a file is
    similar to emptying it, so the list of changed files
    should be similar either way. The computation
    described above is not done directly in _filecommit
    when creating the list of changed files, however
    it does something very similar by comparing filelog
    nodes.
    """

    if x is not None:
        p1, p2, m1, m2 = x
    else:
        p1 = ctx.p1()
        p2 = ctx.p2()
        m1 = p1.manifest()
        m2 = p2.manifest()

    @util.cachefunc
    def mas():
        p1n = p1.node()
        p2n = p2.node()
        cahs = ctx.repo().changelog.commonancestorsheads(p1n, p2n)
        if not cahs:
            cahs = [nullrev]
        return [ctx.repo()[r].manifest() for r in cahs]

    def deletionfromparent(f):
        if f in m1:
            return f not in m2 and all(
                f in ma and ma.find(f) == m1.find(f) for ma in mas()
            )
        elif f in m2:
            return all(f in ma and ma.find(f) == m2.find(f) for ma in mas())
        else:
            return True

    return deletionfromparent


def computechangesetfilesremoved(ctx):
    """return the list of files removed in a changeset"""
    removed = []
    for f in ctx.files():
        if f not in ctx:
            removed.append(f)
    if removed:
        rf = get_removal_filter(ctx)
        removed = [r for r in removed if not rf(r)]
    return removed


def computechangesetfilesmerged(ctx):
    """return the list of files merged in a changeset"""
    merged = []
    if len(ctx.parents()) < 2:
        return merged
    for f in ctx.files():
        if f in ctx:
            fctx = ctx[f]
            parents = fctx._filelog.parents(fctx._filenode)
            if parents[1] != ctx.repo().nullid:
                merged.append(f)
    return merged


def computechangesetcopies(ctx):
    """return the copies data for a changeset

    The copies data are returned as a pair of dictionnary (p1copies, p2copies).

    Each dictionnary are in the form: `{newname: oldname}`
    """
    p1copies = {}
    p2copies = {}
    p1 = ctx.p1()
    p2 = ctx.p2()
    narrowmatch = ctx._repo.narrowmatch()
    for dst in ctx.files():
        if not narrowmatch(dst) or dst not in ctx:
            continue
        copied = ctx[dst].renamed()
        if not copied:
            continue
        src, srcnode = copied
        if src in p1 and p1[src].filenode() == srcnode:
            p1copies[dst] = src
        elif src in p2 and p2[src].filenode() == srcnode:
            p2copies[dst] = src
    return p1copies, p2copies


def encodecopies(files, copies):
    items = []
    for i, dst in enumerate(files):
        if dst in copies:
            items.append(b'%d\0%s' % (i, copies[dst]))
    if len(items) != len(copies):
        raise error.ProgrammingError(
            b'some copy targets missing from file list'
        )
    return b"\n".join(items)


def decodecopies(files, data):
    try:
        copies = {}
        if not data:
            return copies
        for l in data.split(b'\n'):
            strindex, src = l.split(b'\0')
            i = int(strindex)
            dst = files[i]
            copies[dst] = src
        return copies
    except (ValueError, IndexError):
        # Perhaps someone had chosen the same key name (e.g. "p1copies") and
        # used different syntax for the value.
        return None


def encodefileindices(files, subset):
    subset = set(subset)
    indices = []
    for i, f in enumerate(files):
        if f in subset:
            indices.append(b'%d' % i)
    return b'\n'.join(indices)


def decodefileindices(files, data):
    try:
        subset = []
        if not data:
            return subset
        for strindex in data.split(b'\n'):
            i = int(strindex)
            if i < 0 or i >= len(files):
                return None
            subset.append(files[i])
        return subset
    except (ValueError, IndexError):
        # Perhaps someone had chosen the same key name (e.g. "added") and
        # used different syntax for the value.
        return None


# see mercurial/helptext/internals/revlogs.txt for details about the format

ACTION_MASK = int("111" "00", 2)
# note: untouched file used as copy source will as `000` for this mask.
ADDED_FLAG = int("001" "00", 2)
MERGED_FLAG = int("010" "00", 2)
REMOVED_FLAG = int("011" "00", 2)
SALVAGED_FLAG = int("100" "00", 2)
TOUCHED_FLAG = int("101" "00", 2)

COPIED_MASK = int("11", 2)
COPIED_FROM_P1_FLAG = int("10", 2)
COPIED_FROM_P2_FLAG = int("11", 2)

# structure is <flag><filename-end><copy-source>
INDEX_HEADER = struct.Struct(">L")
INDEX_ENTRY = struct.Struct(">bLL")


def encode_files_sidedata(files):
    all_files = set(files.touched)
    all_files.update(files.copied_from_p1.values())
    all_files.update(files.copied_from_p2.values())
    all_files = sorted(all_files)
    file_idx = {f: i for (i, f) in enumerate(all_files)}
    file_idx[None] = 0

    chunks = [INDEX_HEADER.pack(len(all_files))]

    filename_length = 0
    for f in all_files:
        filename_size = len(f)
        filename_length += filename_size
        flag = 0
        if f in files.added:
            flag |= ADDED_FLAG
        elif f in files.merged:
            flag |= MERGED_FLAG
        elif f in files.removed:
            flag |= REMOVED_FLAG
        elif f in files.salvaged:
            flag |= SALVAGED_FLAG
        elif f in files.touched:
            flag |= TOUCHED_FLAG

        copy = None
        if f in files.copied_from_p1:
            flag |= COPIED_FROM_P1_FLAG
            copy = files.copied_from_p1.get(f)
        elif f in files.copied_from_p2:
            copy = files.copied_from_p2.get(f)
            flag |= COPIED_FROM_P2_FLAG
        copy_idx = file_idx[copy]
        chunks.append(INDEX_ENTRY.pack(flag, filename_length, copy_idx))
    chunks.extend(all_files)
    return {sidedatamod.SD_FILES: b''.join(chunks)}


def decode_files_sidedata(sidedata):
    md = ChangingFiles()
    raw = sidedata.get(sidedatamod.SD_FILES)

    if raw is None:
        return md

    copies = []
    all_files = []

    assert len(raw) >= INDEX_HEADER.size
    total_files = INDEX_HEADER.unpack_from(raw, 0)[0]

    offset = INDEX_HEADER.size
    file_offset_base = offset + (INDEX_ENTRY.size * total_files)
    file_offset_last = file_offset_base

    assert len(raw) >= file_offset_base

    for idx in range(total_files):
        flag, file_end, copy_idx = INDEX_ENTRY.unpack_from(raw, offset)
        file_end += file_offset_base
        filename = raw[file_offset_last:file_end]
        filesize = file_end - file_offset_last
        assert len(filename) == filesize
        offset += INDEX_ENTRY.size
        file_offset_last = file_end
        all_files.append(filename)
        if flag & ACTION_MASK == ADDED_FLAG:
            md.mark_added(filename)
        elif flag & ACTION_MASK == MERGED_FLAG:
            md.mark_merged(filename)
        elif flag & ACTION_MASK == REMOVED_FLAG:
            md.mark_removed(filename)
        elif flag & ACTION_MASK == SALVAGED_FLAG:
            md.mark_salvaged(filename)
        elif flag & ACTION_MASK == TOUCHED_FLAG:
            md.mark_touched(filename)

        copied = None
        if flag & COPIED_MASK == COPIED_FROM_P1_FLAG:
            copied = md.mark_copied_from_p1
        elif flag & COPIED_MASK == COPIED_FROM_P2_FLAG:
            copied = md.mark_copied_from_p2

        if copied is not None:
            copies.append((copied, filename, copy_idx))

    for copied, filename, copy_idx in copies:
        copied(all_files[copy_idx], filename)

    return md


def _getsidedata(srcrepo, rev):
    ctx = srcrepo[rev]
    files = compute_all_files_changes(ctx)
    return encode_files_sidedata(files), files.has_copies_info


def copies_sidedata_computer(repo, revlog, rev, existing_sidedata):
    sidedata, has_copies_info = _getsidedata(repo, rev)
    flags_to_add = sidedataflag.REVIDX_HASCOPIESINFO if has_copies_info else 0
    return sidedata, (flags_to_add, 0)


def _sidedata_worker(srcrepo, revs_queue, sidedata_queue, tokens):
    """The function used by worker precomputing sidedata

    It read an input queue containing revision numbers
    It write in an output queue containing (rev, <sidedata-map>)

    The `None` input value is used as a stop signal.

    The `tokens` semaphore is user to avoid having too many unprocessed
    entries. The workers needs to acquire one token before fetching a task.
    They will be released by the consumer of the produced data.
    """
    tokens.acquire()
    rev = revs_queue.get()
    while rev is not None:
        data = _getsidedata(srcrepo, rev)
        sidedata_queue.put((rev, data))
        tokens.acquire()
        rev = revs_queue.get()
    # processing of `None` is completed, release the token.
    tokens.release()


BUFF_PER_WORKER = 50


def _get_worker_sidedata_adder(srcrepo, destrepo):
    """The parallel version of the sidedata computation

    This code spawn a pool of worker that precompute a buffer of sidedata
    before we actually need them"""
    # avoid circular import copies -> scmutil -> worker -> copies
    from . import worker

    nbworkers = worker._numworkers(srcrepo.ui)

    tokens = multiprocessing.BoundedSemaphore(nbworkers * BUFF_PER_WORKER)
    revsq = multiprocessing.Queue()
    sidedataq = multiprocessing.Queue()

    assert srcrepo.filtername is None
    # queue all tasks beforehand, revision numbers are small and it make
    # synchronisation simpler
    #
    # Since the computation for each node can be quite expensive, the overhead
    # of using a single queue is not revelant. In practice, most computation
    # are fast but some are very expensive and dominate all the other smaller
    # cost.
    for r in srcrepo.changelog.revs():
        revsq.put(r)
    # queue the "no more tasks" markers
    for i in range(nbworkers):
        revsq.put(None)

    allworkers = []
    for i in range(nbworkers):
        args = (srcrepo, revsq, sidedataq, tokens)
        w = multiprocessing.Process(target=_sidedata_worker, args=args)
        allworkers.append(w)
        w.start()

    # dictionnary to store results for revision higher than we one we are
    # looking for. For example, if we need the sidedatamap for 42, and 43 is
    # received, when shelve 43 for later use.
    staging = {}

    def sidedata_companion(repo, revlog, rev, old_sidedata):
        # Is the data previously shelved ?
        data = staging.pop(rev, None)
        if data is None:
            # look at the queued result until we find the one we are lookig
            # for (shelve the other ones)
            r, data = sidedataq.get()
            while r != rev:
                staging[r] = data
                r, data = sidedataq.get()
        tokens.release()
        sidedata, has_copies_info = data
        new_flag = 0
        if has_copies_info:
            new_flag = sidedataflag.REVIDX_HASCOPIESINFO
        return sidedata, (new_flag, 0)

    return sidedata_companion