hgext/fetch.py

bdiff: deal better with duplicate lines

The longest_match code compares all the possible positions in two
files to find the best match. Given a pair of sequences, it
effectively searches a grid like this:

  a b b b c . d e . f
  0 1 2 3 4 5 6 7 8 9
a 1 - - - - - - - - -
b - 2 1 1 - - - - - -
b - 1 3 2 - - - - - -
b - 1 2 4 - - - - - -
. - - - - - 1 - - 1 -


Here, the 4 in the middle says "the first four lines of the
file match", which it can compute be comparing the fourth lines and
then adding one to the result found when comparing the third lines in
the entry to the upper left.

We generally avoid the quadratic worst case by only looking at lines
that match, which is precomputed. We also avoid quadratic storage by
only keeping a single column vector and then keeping track of the best
match.

Unfortunately, this can get us into trouble with the sequences above.
Because we want to reuse the '3' value when calculating the '4', we
need to be careful not to overwrite it with the '2' we calculate
immediately before. If we scan left to right, top to bottom, we're
going to have a problem: we'll overwrite our 3 before we use it and
calculate a suboptimal best match.

To address this, we can either keep two column vectors and swap
between them (which significantly complicates bookkeeping), or change
our scanning order. If we instead scan from left to right, bottom to
top, we'll avoid ever overwriting values we'll need in the future.

This unfortunately needs several changes to be made simultaneously:

- change the order we build the initial hash chains for the b sequence
- change the sentinel values from INT_MAX to -1
- change the visit order in the longest_match inner loop
- add a tie-breaker preference for earlier matches

This last is needed because we previously had an implicit tie-breaker
from our visitation order that our test suite relies on. Later matches
can also trigger a bug in the normalization code in diff().

# fetch.py - pull and merge remote changes
#
# Copyright 2006 Vadim Gelfer <[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.

'''pull, update and merge in one command (DEPRECATED)'''

from mercurial.i18n import _
from mercurial.node import short
from mercurial import commands, cmdutil, hg, util, error
from mercurial.lock import release
from mercurial import exchange

cmdtable = {}
command = cmdutil.command(cmdtable)
# Note for extension authors: ONLY specify testedwith = 'internal' for
# extensions which SHIP WITH MERCURIAL. Non-mainline extensions should
# be specifying the version(s) of Mercurial they are tested with, or
# leave the attribute unspecified.
testedwith = 'internal'

@command('fetch',
    [('r', 'rev', [],
     _('a specific revision you would like to pull'), _('REV')),
    ('e', 'edit', None, _('invoke editor on commit messages')),
    ('', 'force-editor', None, _('edit commit message (DEPRECATED)')),
    ('', 'switch-parent', None, _('switch parents when merging')),
    ] + commands.commitopts + commands.commitopts2 + commands.remoteopts,
    _('hg fetch [SOURCE]'))
def fetch(ui, repo, source='default', **opts):
    '''pull changes from a remote repository, merge new changes if needed.

    This finds all changes from the repository at the specified path
    or URL and adds them to the local repository.

    If the pulled changes add a new branch head, the head is
    automatically merged, and the result of the merge is committed.
    Otherwise, the working directory is updated to include the new
    changes.

    When a merge is needed, the working directory is first updated to
    the newly pulled changes. Local changes are then merged into the
    pulled changes. To switch the merge order, use --switch-parent.

    See :hg:`help dates` for a list of formats valid for -d/--date.

    Returns 0 on success.
    '''

    date = opts.get('date')
    if date:
        opts['date'] = util.parsedate(date)

    parent, _p2 = repo.dirstate.parents()
    branch = repo.dirstate.branch()
    try:
        branchnode = repo.branchtip(branch)
    except error.RepoLookupError:
        branchnode = None
    if parent != branchnode:
        raise error.Abort(_('working directory not at branch tip'),
                         hint=_("use 'hg update' to check out branch tip"))

    wlock = lock = None
    try:
        wlock = repo.wlock()
        lock = repo.lock()

        cmdutil.bailifchanged(repo)

        bheads = repo.branchheads(branch)
        bheads = [head for head in bheads if len(repo[head].children()) == 0]
        if len(bheads) > 1:
            raise error.Abort(_('multiple heads in this branch '
                               '(use "hg heads ." and "hg merge" to merge)'))

        other = hg.peer(repo, opts, ui.expandpath(source))
        ui.status(_('pulling from %s\n') %
                  util.hidepassword(ui.expandpath(source)))
        revs = None
        if opts['rev']:
            try:
                revs = [other.lookup(rev) for rev in opts['rev']]
            except error.CapabilityError:
                err = _("other repository doesn't support revision lookup, "
                        "so a rev cannot be specified.")
                raise error.Abort(err)

        # Are there any changes at all?
        modheads = exchange.pull(repo, other, heads=revs).cgresult
        if modheads == 0:
            return 0

        # Is this a simple fast-forward along the current branch?
        newheads = repo.branchheads(branch)
        newchildren = repo.changelog.nodesbetween([parent], newheads)[2]
        if len(newheads) == 1 and len(newchildren):
            if newchildren[0] != parent:
                return hg.update(repo, newchildren[0])
            else:
                return 0

        # Are there more than one additional branch heads?
        newchildren = [n for n in newchildren if n != parent]
        newparent = parent
        if newchildren:
            newparent = newchildren[0]
            hg.clean(repo, newparent)
        newheads = [n for n in newheads if n != newparent]
        if len(newheads) > 1:
            ui.status(_('not merging with %d other new branch heads '
                        '(use "hg heads ." and "hg merge" to merge them)\n') %
                      (len(newheads) - 1))
            return 1

        if not newheads:
            return 0

        # Otherwise, let's merge.
        err = False
        if newheads:
            # By default, we consider the repository we're pulling
            # *from* as authoritative, so we merge our changes into
            # theirs.
            if opts['switch_parent']:
                firstparent, secondparent = newparent, newheads[0]
            else:
                firstparent, secondparent = newheads[0], newparent
                ui.status(_('updating to %d:%s\n') %
                          (repo.changelog.rev(firstparent),
                           short(firstparent)))
            hg.clean(repo, firstparent)
            ui.status(_('merging with %d:%s\n') %
                      (repo.changelog.rev(secondparent), short(secondparent)))
            err = hg.merge(repo, secondparent, remind=False)

        if not err:
            # we don't translate commit messages
            message = (cmdutil.logmessage(ui, opts) or
                       ('Automated merge with %s' %
                        util.removeauth(other.url())))
            editopt = opts.get('edit') or opts.get('force_editor')
            editor = cmdutil.getcommiteditor(edit=editopt, editform='fetch')
            n = repo.commit(message, opts['user'], opts['date'], editor=editor)
            ui.status(_('new changeset %d:%s merges remote changes '
                        'with local\n') % (repo.changelog.rev(n),
                                           short(n)))

        return err

    finally:
        release(lock, wlock)