mercurial/help/patterns.txt

revlog: use an LRU cache for delta chain bases

Profiling using statprof revealed a hotspot during changegroup
application calculating delta chain bases on generaldelta repos.
Essentially, revlog._addrevision() was performing a lot of redundant
work tracing the delta chain as part of determining when the chain
distance was acceptable. This was most pronounced when adding
revisions to manifests, which can have delta chains thousands of
revisions long.

There was a delta chain base cache on revlogs before, but it only
captured a single revision. This was acceptable before generaldelta,
when _addrevision would build deltas from the previous revision and
thus we'd pretty much guarantee a cache hit when resolving the delta
chain base on a subsequent _addrevision call. However, it isn't
suitable for generaldelta because parent revisions aren't necessarily
the last processed revision.

This patch converts the delta chain base cache to an LRU dict cache.
The cache can hold multiple entries, so generaldelta repos have a
higher chance of getting a cache hit.

The impact of this change when processing changegroup additions is
significant. On a generaldelta conversion of the "mozilla-unified"
repo (which contains heads of the main Firefox repositories in
chronological order - this means there are lots of transitions between
heads in revlog order), this change has the following impact when
performing an `hg unbundle` of an uncompressed bundle of the repo:

before: 5:42 CPU time
after:  4:34 CPU time

Most of this time is saved when applying the changelog and manifest
revlogs:

before: 2:30 CPU time
after:  1:17 CPU time

That nearly a 50% reduction in CPU time applying changesets and
manifests!

Applying a gzipped bundle of the same repo (effectively simulating a
`hg clone` over HTTP) showed a similar speedup:

before: 5:53 CPU time
after:  4:46 CPU time

Wall time improvements were basically the same as CPU time.

I didn't measure explicitly, but it feels like most of the time
is saved when processing manifests. This makes sense, as large
manifests tend to have very long delta chains and thus benefit the
most from this cache.

So, this change effectively makes changegroup application (which is
used by `hg unbundle`, `hg clone`, `hg pull`, `hg unshelve`, and
various other commands) significantly faster when delta chains are
long (which can happen on repos with large numbers of files and thus
large manifests).

In theory, this change can result in more memory utilization. However,
we're caching a dict of ints. At most we have 200 ints + Python object
overhead per revlog. And, the cache is really only populated when
performing read-heavy operations, such as adding changegroups or
scanning an individual revlog. For memory bloat to be an issue, we'd
need to scan/read several revisions from several revlogs all while
having active references to several revlogs. I don't think there are
many operations that do this, so I don't think memory bloat from the
cache will be an issue.

Mercurial accepts several notations for identifying one or more files
at a time.

By default, Mercurial treats filenames as shell-style extended glob
patterns.

Alternate pattern notations must be specified explicitly.

.. note::

  Patterns specified in ``.hgignore`` are not rooted.
  Please see :hg:`help hgignore` for details.

To use a plain path name without any pattern matching, start it with
``path:``. These path names must completely match starting at the
current repository root.

To use an extended glob, start a name with ``glob:``. Globs are rooted
at the current directory; a glob such as ``*.c`` will only match files
in the current directory ending with ``.c``.

The supported glob syntax extensions are ``**`` to match any string
across path separators and ``{a,b}`` to mean "a or b".

To use a Perl/Python regular expression, start a name with ``re:``.
Regexp pattern matching is anchored at the root of the repository.

To read name patterns from a file, use ``listfile:`` or ``listfile0:``.
The latter expects null delimited patterns while the former expects line
feeds. Each string read from the file is itself treated as a file
pattern.

To read a set of patterns from a file, use ``include:`` or ``subinclude:``.
``include:`` will use all the patterns from the given file and treat them as if
they had been passed in manually.  ``subinclude:`` will only apply the patterns
against files that are under the subinclude file's directory. See :hg:`help
hgignore` for details on the format of these files.

All patterns, except for ``glob:`` specified in command line (not for
``-I`` or ``-X`` options), can match also against directories: files
under matched directories are treated as matched.

Plain examples::

  path:foo/bar   a name bar in a directory named foo in the root
                 of the repository
  path:path:name a file or directory named "path:name"

Glob examples::

  glob:*.c       any name ending in ".c" in the current directory
  *.c            any name ending in ".c" in the current directory
  **.c           any name ending in ".c" in any subdirectory of the
                 current directory including itself.
  foo/*.c        any name ending in ".c" in the directory foo
  foo/**.c       any name ending in ".c" in any subdirectory of foo
                 including itself.

Regexp examples::

  re:.*\.c$      any name ending in ".c", anywhere in the repository

File examples::

  listfile:list.txt  read list from list.txt with one file pattern per line
  listfile0:list.txt read list from list.txt with null byte delimiters

See also :hg:`help filesets`.

Include examples::

  include:path/to/mypatternfile    reads patterns to be applied to all paths
  subinclude:path/to/subignorefile reads patterns specifically for paths in the
                                   subdirectory