mercurial/i18n.py

rust-discovery: using from Python code

As previously done in other topics, the Rust version is used if it's been
built.

The version fully in Rust of the partialdiscovery class has the performance
advantage over the Python version (actually using the Rust MissingAncestor) if
the undecided set is big enough. Otherwise no sampling occurs, and the
discovery is reasonably fast anyway.

Note: it's hard to predict the size of the initial undecided set, it can
depend on the kind of topological changes between the local and remote graphs.
The point of the Rust version is to make the bad cases acceptable.

More specifically, the performance advantages are:

- faster sampling, especially takefullsample()
- much faster addmissings() in almost all cases (see commit message in
  grandparent of the present changeset)
- no conversion cost of the undecided set at the interface between Rust and
  Python

== Measurements with big undecided sets

For an extreme example, discovery between mozilla-try and mozilla-unified
(over one million undecided revisions, same case as in dbd0fcca6dfc), we
get roughly a x2.5/x3 better performance:

Growing sample size (5% starting with 200): time goes down from
210 to 72 seconds.
Constant sample size of 200: time down from 1853 to 659 seconds.

With a sample size computed from number of roots and heads of the
undecided set (`respectsize` is `False`), here are perfdiscovery results:

Before ! wall 9.358729 comb 9.360000 user 9.310000 sys 0.050000 (median of 50)
After  ! wall 3.793819 comb 3.790000 user 3.750000 sys 0.040000 (median of 50)

In that later case, the sample sizes are routinely in the hundreds of
thousands of revisions. While still faster, the Rust iteration in
addmissings has less of an advantage than with smaller sample sizes, but
one sees addcommons becoming faster, probably a consequence of not having
to copy big sets back and forth.

This example is not a goal in itself, but it showcases several different
areas in which the process can become slow, due to different factors, and
how this full Rust version can help.

== Measurements with small undecided sets

In cases the undecided set is small enough than no sampling occurs,
the Rust version has a disadvantage at init if `targetheads` is really big
(some time is lost in the translation to Rust data structures),
and that is compensated by the faster `addmissings()`.

On a private repository with over one million commits, we still get a minor
improvement, of 6.8%:

  Before ! wall 0.593585 comb 0.590000 user 0.550000 sys 0.040000 (median of 50)
  After  ! wall 0.553035 comb 0.550000 user 0.520000 sys 0.030000 (median of 50)

What's interesting in that case is the first addinfo() at 180ms for Rust and
233ms for Python+C, mostly due to add_missings and the children cache
computation being done in less than 0.2ms on the Rust side vs over 40ms on the
Python side.

The worst case we have on hand is with mozilla-try, prepared with
discovery-helper.sh for 10 heads and depth 10, time goes up 2.2% on the median.
In this case `targetheads` is really huge with 165842 server heads.

  Before ! wall 0.823884 comb 0.810000 user 0.790000 sys 0.020000 (median of 50)
  After  ! wall 0.842607 comb 0.840000 user 0.800000 sys 0.040000 (median of 50)

If that would be considered a problem, more adjustments can be made, which are
prematurate at this stage: cooking special variants of methods of the inner
MissingAncestors object, retrieving local heads directly from Rust to avoid
the cost of conversion. Effort would probably be better spent at this point
improving the surroundings if needed.

Here's another data point with a smaller repository, pypy, where performance
is almost identical

 Before ! wall 0.015121 comb 0.030000 user 0.020000 sys 0.010000 (median of 186)
 After  ! wall 0.015009 comb 0.010000 user 0.010000 sys 0.000000 (median of 184)

Differential Revision: https://phab.mercurial-scm.org/D6430

# i18n.py - internationalization support for mercurial
#
# Copyright 2005, 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 gettext as gettextmod
import locale
import os
import sys

from . import (
    encoding,
    pycompat,
)

# modelled after templater.templatepath:
if getattr(sys, 'frozen', None) is not None:
    module = pycompat.sysexecutable
else:
    module = pycompat.fsencode(__file__)

_languages = None
if (pycompat.iswindows
    and 'LANGUAGE' not in encoding.environ
    and 'LC_ALL' not in encoding.environ
    and 'LC_MESSAGES' not in encoding.environ
    and 'LANG' not in encoding.environ):
    # Try to detect UI language by "User Interface Language Management" API
    # if no locale variables are set. Note that locale.getdefaultlocale()
    # uses GetLocaleInfo(), which may be different from UI language.
    # (See http://msdn.microsoft.com/en-us/library/dd374098(v=VS.85).aspx )
    try:
        import ctypes
        langid = ctypes.windll.kernel32.GetUserDefaultUILanguage()
        _languages = [locale.windows_locale[langid]]
    except (ImportError, AttributeError, KeyError):
        # ctypes not found or unknown langid
        pass

_ugettext = None

def setdatapath(datapath):
    datapath = pycompat.fsdecode(datapath)
    localedir = os.path.join(datapath, r'locale')
    t = gettextmod.translation(r'hg', localedir, _languages, fallback=True)
    global _ugettext
    try:
        _ugettext = t.ugettext
    except AttributeError:
        _ugettext = t.gettext

_msgcache = {}  # encoding: {message: translation}

def gettext(message):
    """Translate message.

    The message is looked up in the catalog to get a Unicode string,
    which is encoded in the local encoding before being returned.

    Important: message is restricted to characters in the encoding
    given by sys.getdefaultencoding() which is most likely 'ascii'.
    """
    # If message is None, t.ugettext will return u'None' as the
    # translation whereas our callers expect us to return None.
    if message is None or not _ugettext:
        return message

    cache = _msgcache.setdefault(encoding.encoding, {})
    if message not in cache:
        if type(message) is pycompat.unicode:
            # goofy unicode docstrings in test
            paragraphs = message.split(u'\n\n')
        else:
            # should be ascii, but we have unicode docstrings in test, which
            # are converted to utf-8 bytes on Python 3.
            paragraphs = [p.decode("utf-8") for p in message.split('\n\n')]
        # Be careful not to translate the empty string -- it holds the
        # meta data of the .po file.
        u = u'\n\n'.join([p and _ugettext(p) or u'' for p in paragraphs])
        try:
            # encoding.tolocal cannot be used since it will first try to
            # decode the Unicode string. Calling u.decode(enc) really
            # means u.encode(sys.getdefaultencoding()).decode(enc). Since
            # the Python encoding defaults to 'ascii', this fails if the
            # translated string use non-ASCII characters.
            encodingstr = pycompat.sysstr(encoding.encoding)
            cache[message] = u.encode(encodingstr, "replace")
        except LookupError:
            # An unknown encoding results in a LookupError.
            cache[message] = message
    return cache[message]

def _plain():
    if ('HGPLAIN' not in encoding.environ
        and 'HGPLAINEXCEPT' not in encoding.environ):
        return False
    exceptions = encoding.environ.get('HGPLAINEXCEPT', '').strip().split(',')
    return 'i18n' not in exceptions

if _plain():
    _ = lambda message: message
else:
    _ = gettext