i18n/hggettext

changelog: lazily parse date/extra field

This is probably the most complicated patch in the parsing
refactor.

Because the date and extras are encoded in the same field, we
stuff the entire field into a dedicated variable and add a
property for accessing the sub-components of each. There is
some duplicated code here. But the code is relatively simple,
so it shouldn't be a big deal.

We see revset performance wins across the board:

author(mpm)
0.896565
0.876713
0.822961

desc(bug)
0.887169
0.895514
0.847054

date(2015)
0.878797
0.820987
0.811613

extra(rebase_source)
0.865446
0.823811
0.797756

author(mpm) or author(greg)
1.801832
1.784160
1.668172

author(mpm) or desc(bug)
1.812438
1.822756
1.677608

date(2015) or branch(default)
0.968276
0.910981
0.896032

author(mpm) or desc(bug) or date(2015) or extra(rebase_source)
3.656193
3.516788
3.265024

We see a speed-up on revsets accessing date and extras because the new
parsing code only parses what you access. Even though they are stored
the same text field, we avoid parsing dates when accessing extras and
vice-versa.

But strangely revsets accessing both date and extras appeared to speed
up as well! I'm not sure if this is due to refactoring the parsing
code or due to an optimization in revsets. You can't argue with the
results!

#!/usr/bin/env python
#
# hggettext - carefully extract docstrings for Mercurial
#
# Copyright 2009 Matt Mackall <[e-mail]> and others
#
# This software may be used and distributed according to the terms of the
# GNU General Public License version 2 or any later version.

# The normalize function is taken from pygettext which is distributed
# with Python under the Python License, which is GPL compatible.

"""Extract docstrings from Mercurial commands.

Compared to pygettext, this script knows about the cmdtable and table
dictionaries used by Mercurial, and will only extract docstrings from
functions mentioned therein.

Use xgettext like normal to extract strings marked as translatable and
join the message cataloges to get the final catalog.
"""

import os, sys, inspect


def escape(s):
    # The order is important, the backslash must be escaped first
    # since the other replacements introduce new backslashes
    # themselves.
    s = s.replace('\\', '\\\\')
    s = s.replace('\n', '\\n')
    s = s.replace('\r', '\\r')
    s = s.replace('\t', '\\t')
    s = s.replace('"', '\\"')
    return s


def normalize(s):
    # This converts the various Python string types into a format that
    # is appropriate for .po files, namely much closer to C style.
    lines = s.split('\n')
    if len(lines) == 1:
        s = '"' + escape(s) + '"'
    else:
        if not lines[-1]:
            del lines[-1]
            lines[-1] = lines[-1] + '\n'
        lines = map(escape, lines)
        lineterm = '\\n"\n"'
        s = '""\n"' + lineterm.join(lines) + '"'
    return s


def poentry(path, lineno, s):
    return ('#: %s:%d\n' % (path, lineno) +
            'msgid %s\n' % normalize(s) +
            'msgstr ""\n')


def offset(src, doc, name, default):
    """Compute offset or issue a warning on stdout."""
    # Backslashes in doc appear doubled in src.
    end = src.find(doc.replace('\\', '\\\\'))
    if end == -1:
        # This can happen if the docstring contains unnecessary escape
        # sequences such as \" in a triple-quoted string. The problem
        # is that \" is turned into " and so doc wont appear in src.
        sys.stderr.write("warning: unknown offset in %s, assuming %d lines\n"
                         % (name, default))
        return default
    else:
        return src.count('\n', 0, end)


def importpath(path):
    """Import a path like foo/bar/baz.py and return the baz module."""
    if path.endswith('.py'):
        path = path[:-3]
    if path.endswith('/__init__'):
        path = path[:-9]
    path = path.replace('/', '.')
    mod = __import__(path)
    for comp in path.split('.')[1:]:
        mod = getattr(mod, comp)
    return mod


def docstrings(path):
    """Extract docstrings from path.

    This respects the Mercurial cmdtable/table convention and will
    only extract docstrings from functions mentioned in these tables.
    """
    mod = importpath(path)
    if mod.__doc__:
        src = open(path).read()
        lineno = 1 + offset(src, mod.__doc__, path, 7)
        print poentry(path, lineno, mod.__doc__)

    functions = list(getattr(mod, 'i18nfunctions', []))
    functions = [(f, True) for f in functions]

    cmdtable = getattr(mod, 'cmdtable', {})
    if not cmdtable:
        # Maybe we are processing mercurial.commands?
        cmdtable = getattr(mod, 'table', {})
    functions.extend((c[0], False) for c in cmdtable.itervalues())

    for func, rstrip in functions:
        if func.__doc__:
            src = inspect.getsource(func)
            name = "%s.%s" % (path, func.__name__)
            lineno = func.func_code.co_firstlineno
            doc = func.__doc__
            if rstrip:
                doc = doc.rstrip()
            lineno += offset(src, doc, name, 1)
            print poentry(path, lineno, doc)


def rawtext(path):
    src = open(path).read()
    print poentry(path, 1, src)


if __name__ == "__main__":
    # It is very important that we import the Mercurial modules from
    # the source tree where hggettext is executed. Otherwise we might
    # accidentally import and extract strings from a Mercurial
    # installation mentioned in PYTHONPATH.
    sys.path.insert(0, os.getcwd())
    from mercurial import demandimport; demandimport.enable()
    for path in sys.argv[1:]:
        if path.endswith('.txt'):
            rawtext(path)
        else:
            docstrings(path)