Mercurial > evolve
view hgext3rd/topic/revset.py @ 6889:a66cf9008781
obslog: also display patch for rebased changesets
This applies the same logic that is used for "merge-diff" to rebased
changesets. The successors' content is compared to the content of the
predecessors rebased in-memory on the new parents.
This highlights the changes that were actually introduced while rebasing (like
conflict resolution or API adjustment).
As a side effect, obslog now also outputs slightly more diffs for splits,
showing what parts of the original big changeset were moved to the smaller
split components (but for now it only works for the first few changesets).
author | Pierre-Yves David <pierre-yves.david@octobus.net> |
---|---|
date | Sun, 22 Sep 2024 02:58:54 +0200 |
parents | 6518a3b951dc |
children |
line wrap: on
line source
from __future__ import absolute_import from mercurial import ( error, registrar, revset, ) from mercurial.utils import stringutil from . import ( destination, stack, ) revsetpredicate = registrar.revsetpredicate() def getstringstrict(x, err): if x and x[0] == b'string': return x[1] raise error.ParseError(err) @revsetpredicate(b'topicnamespace([string or set])') def topicnamespaceset(repo, subset, x): """All changesets with the specified topic namespace or the topic namespaces of the given changesets. Without the argument, all changesets with any non-empty topic namespace. Pattern matching is supported for `string`. See :hg:`help revisions.patterns`. """ args = revset.getargs(x, 0, 1, b'topicnamespace takes one or no arguments') mutable = revset._notpublic(repo, revset.fullreposet(repo), ()) if not args: return (subset & mutable).filter(lambda r: repo[r].topic_namespace() != b'none') try: tns = getstringstrict(args[0], b'') except error.ParseError: # not a string, but another revset pass else: kind, pattern, matcher = stringutil.stringmatcher(tns) if tns.startswith(b'literal:') and pattern not in repo.topic_namespaces: raise error.RepoLookupError(b"topic namespace '%s' does not exist" % pattern) def matches(r): tns = repo[r].topic_namespace() if tns == b'none': return False return matcher(tns) return (subset & mutable).filter(matches) s = revset.getset(repo, revset.fullreposet(repo), x) namespaces = {repo[r].topic_namespace() for r in s} namespaces.discard(b'none') def matches(r): tns = repo[r].topic_namespace() if tns == b'none': return False return tns in namespaces return (subset & mutable).filter(matches) @revsetpredicate(b'topic([string or set])') def topicset(repo, subset, x): """All changesets with the specified topic or the topics of the given changesets. Without the argument, all changesets with any topic specified. Pattern matching is supported for `string`. See :hg:`help revisions.patterns`. """ args = revset.getargs(x, 0, 1, b'topic takes one or no arguments') mutable = revset._notpublic(repo, revset.fullreposet(repo), ()) if not args: return (subset & mutable).filter(lambda r: bool(repo[r].topic())) try: topic = getstringstrict(args[0], b'') except error.ParseError: # not a string, but another revset pass else: kind, pattern, matcher = stringutil.stringmatcher(topic) if topic.startswith(b'literal:') and pattern not in repo.topics: raise error.RepoLookupError(b"topic '%s' does not exist" % pattern) def matches(r): topic = repo[r].topic() if not topic: return False return matcher(topic) return (subset & mutable).filter(matches) s = revset.getset(repo, revset.fullreposet(repo), x) topics = {repo[r].topic() for r in s} topics.discard(b'') def matches(r): topic = repo[r].topic() if not topic: return False return topic in topics return (subset & mutable).filter(matches) @revsetpredicate(b'ngtip([branch])') def ngtipset(repo, subset, x): """The tip of a branch, ignoring changesets with a topic. Name is horrible so that people change it. """ args = revset.getargs(x, 1, 1, b'ngtip takes one argument') # match a specific topic branch = revset.getstring(args[0], b'ngtip requires a string') if branch == b'.': branch = repo[b'.'].branch() # list of length 1 revs = [repo[node].rev() for node in destination.ngtip(repo, branch)] return subset & revset.baseset(revs) @revsetpredicate(b'stack()') def stackset(repo, subset, x): """All relevant changes in the current topic. This is roughly equivalent to 'topic(.) - obsolete()' with a sorting moving unstable changeset after their future parent (as if evolve were already run). """ err = b'stack takes no arguments, it works on current topic' revset.getargs(x, 0, 0, err) topic = None branch = None if repo.currenttopic: topic = repo.currenttopic else: branch = repo[None].branch() return revset.baseset(stack.stack(repo, branch=branch, topic=topic)[1:]) & subset def stacksubrel(repo, subset, x, rel, z, order): """This is a revset-flavored implementation of stack aliases. The syntax is: rev#stack[n] or rev#s[n]. Plenty of logic is borrowed from topic._namemap, but unlike that function, which prefers to abort (e.g. when stack index is too high), this returns empty set to be more revset-friendly. """ a, b = revset.getintrange( z, b'relation subscript must be an integer or a range', b'relation subscript bounds must be integers', None, None) s = revset.getset(repo, revset.fullreposet(repo), x) if not s: return revset.baseset() def getrange(st, a, b): start = 1 if a is None else a end = len(st.revs) if b is None else b + 1 return range(start, end) revs = [] for r in s: topic = repo[r].topic() if topic: st = stack.stack(repo, topic=topic) else: st = stack.stack(repo, branch=repo[r].branch()) for n in getrange(st, a, b): if abs(n) >= len(st.revs): # also means stack base is not accessible with n < 0, which # is by design continue if n == 0 and b != 0 and a != 0: # quirk: we don't want stack base unless specifically asked # for it (at least one of the indices is 0) continue rev = st.revs[n] if rev == -1 and n == 0: continue if rev not in revs: revs.append(rev) return subset & revset.baseset(revs) revset.subscriptrelations[b'stack'] = stacksubrel revset.subscriptrelations[b's'] = stacksubrel def topicsubrel(repo, subset, x, *args): subset &= topicset(repo, subset, x) return revset.generationssubrel(repo, subset, x, *args) revset.subscriptrelations[b'topic'] = topicsubrel revset.subscriptrelations[b't'] = topicsubrel def stackrel(repo, subset, x, rel, order): z = (b'rangeall', None) return stacksubrel(repo, subset, x, rel, z, order) revset.relations[b'stack'] = stackrel revset.relations[b's'] = stackrel