# keyword.py - keyword expansion for mercurial
#
# Copyright 2006 Christian Ebert <blacktrash@gmx.net>
#
# This software may be used and distributed according to the terms
# of the GNU General Public License, incorporated herein by reference.

'''keyword expansion hack against the grain of a DSCM

This extension lets you expand RCS/CVS-like keywords in a Mercurial
repository.

There are many good reasons why this is not needed in a distributed
SCM, still it may be useful in very small projects based on single
files (like LaTeX packages), that are mostly addressed to an audience
not running a version control system.

Supported keywords are (changeset 000000000000):
    $Revision: 000000000000 $
    $Author: Your Name <address@example.com> $
    $Date: %a %b %d %H:%M:%S %Y %z $
    $RCSFile: basename,v $
    $Source: /path/to/basename,v $
    $Id: basename,v 000000000000 %Y-%m-%d %H:%M:%S %z shortname $
    $Header: /path/to/basename,v 000000000000 %Y-%m-%d %H:%M:%S %z shortname $

The extension, according to its hackish nature, is a hybrid and consists
actually in 2 parts:

    1. pure extension code (reposetup) that is triggered on checkout and
       logging of changes.
    2. a pretxncommit hook (hgrc (5)) that expands keywords immediately
       at commit time in the working directory.

Simple setup in hgrc:

    # enable extension
    hgext.keyword =
    
    # filename patterns for expansion are configured in this section
    [keyword]
    *.sty = expand
    ...

    # set up pretxncommit hook
    [hooks]
    pretxncommit =
    pretxncommit.keyword = python:hgext.keyword.pretxnkw
'''

from mercurial.i18n import _
from mercurial import context, util
import os.path, re


re_kw = re.compile(
        r'\$(Id|Header|Author|Date|Revision|RCSFile|Source)[^$]*?\$')


def kwexpand(matchobj, repo, path, changeid=None, fileid=None, filelog=None):
    '''Called by kwfilelog.read and pretxnkw.
    Sets supported keywords as local variables and evaluates them to
    their expansion if matchobj is equal to string representation.'''

    c = context.filectx(repo, path,
            changeid=changeid, fileid=fileid, filelog=filelog)
    date = c.date()

    Revision = c.changectx()
    Author = c.user()
    RCSFile = os.path.basename(path)+',v'
    Source = repo.wjoin(path)+',v'
    Date = util.datestr(date=date)
    revdateauth = '%s %s %s' % (Revision,
            util.datestr(date=date, format=util.defaultdateformats[0]),
            util.shortuser(Author))
    Header = '%s %s' % (Source, revdateauth)
    Id = '%s %s' % (RCSFile, revdateauth)

    return '$%s: %s $' % (matchobj.group(1), eval(matchobj.group(1)))

def kwfmatchers(ui, repo):
    '''Returns filename matchers from ui keyword section.'''
    return [util.matcher(repo.root, '', [pat], [], [])[1]
            for pat, opt in ui.configitems('keyword')
            if opt == 'expand']


def reposetup(ui, repo):
    from mercurial import filelog, revlog

    if not repo.local():
        return

    class kwrepo(repo.__class__):
        def file(self, f):
            if f[0] == '/':
                f = f[1:]
            return filelog.filelog(self.sopener, f, self, self.revlogversion)

    class kwfilelog(filelog.filelog):
        def __init__(self, opener, path, repo,
                     defversion=revlog.REVLOG_DEFAULT_VERSION):
            super(kwfilelog, self).__init__(opener, path, defversion)
            self._repo = repo
            self._path = path

        def read(self, node):
            data = super(kwfilelog, self).read(node)
            if not self._path.startswith('.hg') and not util.binary(data):
                for mf in kwfmatchers(ui, self._repo):
                    if mf(self._path):
                        ui.debug(_('expanding keywords in %s\n' % self._path))
                        return re_kw.sub(lambda m:
                                kwexpand(m, self._repo, self._path,
                                    fileid=node, filelog=self),
                                data)
            return data

        def add(self, text, meta, tr, link, p1=None, p2=None):
            if not util.binary(text):
                ui.debug(_('removing keyword substitutions in %s\n')
                        % self._path)
                text = re_kw.sub(r'$\1$', text)
            return super(kwfilelog, self).add(text, meta, tr, link, p1, p2)

        def size(self, rev):
            '''Overrides filelog's size() to use kwfilelog.read().'''
            node = revlog.node(self, rev)
            if super(kwfilelog, self).renamed(node):
                return len(self.read(node))
            return revlog.size(self, rev)

        def cmp(self, node, text):
            '''Overrides filelog's cmp() to use kwfilelog.read().'''
            if super(kwfilelog, self).renamed(node):
                t2 = self.read(node)
                return t2 != text

    filelog.filelog = kwfilelog
    repo.__class__ = kwrepo


def pretxnkw(ui, repo, hooktype, **args):
    '''pretxncommit hook that collects candidates for keyword expansion
    on commit and expands keywords in working dir.'''
    from mercurial import cmdutil, commands
    import sys

    if hooktype != 'pretxncommit':
        return True

    cmd, sysargs, globalopts, cmdopts = commands.parse(ui, sys.argv[1:])[1:]
    if repr(cmd).split()[1] in ('tag', 'import_'):
        return False

    files, match, anypats = cmdutil.matchpats(repo, sysargs, cmdopts)
    modified, added = repo.status(files=files, match=match)[:2]
    candidates = [f for f in modified + added if not f.startswith('.hg')]
    if not candidates:
        return False

    fmatchers = kwfmatchers(ui, repo)
    for f in candidates:
        for mf in fmatchers:
            if mf(f):
                data = repo.wfile(f).read()
                if not util.binary(data):
                    data, kwct = re_kw.subn(lambda m:
                            kwexpand(m, repo, f, changeid=args['node']),
                            data)
                    if kwct:
                        ui.debug(_('overwriting %s expanding keywords\n' % f))
                        repo.wfile(f, 'w').write(data)
                break