Mercurial > hg
view mercurial/graphmod.py @ 16696:d1afbf03e69a
rebase: allow collapsing branches in place (issue3111)
We allow rebase plus collapse, but not collapse only? I imagine people would
rebase first then collapse once they are sure the rebase is correct and it is
the right time to finish it.
I was reluctant to submit this patch for reasons detailed below, but it
improves rebase --collapse usefulness so much it is worth the ugliness.
The fix is ugly because we should be fixing the collapse code path rather than
the merge. Collapsing by merging changesets repeatedly is inefficient compared
to what commit --amend does: commitctx(), update, strip. The problem with the
latter is, to generate the synthetic changeset, copy records are gathered with
copies.pathcopies(). copies.pathcopies() is still implemented with merging in
mind and discards information like file replaced by the copy of another,
criss-cross copies and so forth. I believe this information should not be lost,
even if we decide not to interpret it fully later, at merge time.
The second issue with improving rebase --collapse is the option should not be
there to begin with. Rebasing and collapsing are orthogonal and a dedicated
command would probably enable a better, simpler ui. We should avoid advertizing
rebase --collapse, but with this fix it becomes the best shipped solution to
collapse changesets.
And for the record, available techniques are:
- revert + commit + strip: lose copies
- mq/qfold: repeated patching() (mostly correct, fragile)
- rebase: repeated merges (mostly correct, fragile)
- collapse: revert + tag rewriting wizardry, lose copies
- histedit: repeated patching() (mostly correct, fragile)
- amend: copies.pathcopies() + commitctx() + update + strip
author | Patrick Mezard <patrick@mezard.eu> |
---|---|
date | Thu, 03 May 2012 15:14:58 +0200 |
parents | 6e4de55a41a4 |
children | 0849d725e2f9 |
line wrap: on
line source
# Revision graph generator for Mercurial # # Copyright 2008 Dirkjan Ochtman <dirkjan@ochtman.nl> # Copyright 2007 Joel Rosdahl <joel@rosdahl.net> # # This software may be used and distributed according to the terms of the # GNU General Public License version 2 or any later version. """supports walking the history as DAGs suitable for graphical output The most basic format we use is that of:: (id, type, data, [parentids]) The node and parent ids are arbitrary integers which identify a node in the context of the graph returned. Type is a constant specifying the node type. Data depends on type. """ from mercurial.node import nullrev import util CHANGESET = 'C' def dagwalker(repo, revs): """cset DAG generator yielding (id, CHANGESET, ctx, [parentids]) tuples This generator function walks through revisions (which should be ordered from bigger to lower). It returns a tuple for each node. The node and parent ids are arbitrary integers which identify a node in the context of the graph returned. """ if not revs: return cl = repo.changelog lowestrev = min(revs) gpcache = {} knownrevs = set(revs) for rev in revs: ctx = repo[rev] parents = sorted(set([p.rev() for p in ctx.parents() if p.rev() in knownrevs])) mpars = [p.rev() for p in ctx.parents() if p.rev() != nullrev and p.rev() not in parents] for mpar in mpars: gp = gpcache.get(mpar) if gp is None: gp = gpcache[mpar] = grandparent(cl, lowestrev, revs, mpar) if not gp: parents.append(mpar) else: parents.extend(g for g in gp if g not in parents) yield (ctx.rev(), CHANGESET, ctx, parents) def nodes(repo, nodes): """cset DAG generator yielding (id, CHANGESET, ctx, [parentids]) tuples This generator function walks the given nodes. It only returns parents that are in nodes, too. """ include = set(nodes) for node in nodes: ctx = repo[node] parents = set([p.rev() for p in ctx.parents() if p.node() in include]) yield (ctx.rev(), CHANGESET, ctx, sorted(parents)) def colored(dag, repo): """annotates a DAG with colored edge information For each DAG node this function emits tuples:: (id, type, data, (col, color), [(col, nextcol, color)]) with the following new elements: - Tuple (col, color) with column and color index for the current node - A list of tuples indicating the edges between the current node and its parents. """ seen = [] colors = {} newcolor = 1 config = {} for key, val in repo.ui.configitems('graph'): if '.' in key: branch, setting = key.rsplit('.', 1) # Validation if setting == "width" and val.isdigit(): config.setdefault(branch, {})[setting] = int(val) elif setting == "color" and val.isalnum(): config.setdefault(branch, {})[setting] = val if config: getconf = util.lrucachefunc( lambda rev: config.get(repo[rev].branch(), {})) else: getconf = lambda rev: {} for (cur, type, data, parents) in dag: # Compute seen and next if cur not in seen: seen.append(cur) # new head colors[cur] = newcolor newcolor += 1 col = seen.index(cur) color = colors.pop(cur) next = seen[:] # Add parents to next addparents = [p for p in parents if p not in next] next[col:col + 1] = addparents # Set colors for the parents for i, p in enumerate(addparents): if not i: colors[p] = color else: colors[p] = newcolor newcolor += 1 # Add edges to the graph edges = [] for ecol, eid in enumerate(seen): if eid in next: bconf = getconf(eid) edges.append(( ecol, next.index(eid), colors[eid], bconf.get('width', -1), bconf.get('color', ''))) elif eid == cur: for p in parents: bconf = getconf(p) edges.append(( ecol, next.index(p), color, bconf.get('width', -1), bconf.get('color', ''))) # Yield and move on yield (cur, type, data, (col, color), edges) seen = next def grandparent(cl, lowestrev, roots, head): """Return all ancestors of head in roots which revision is greater or equal to lowestrev. """ pending = set([head]) seen = set() kept = set() llowestrev = max(nullrev, lowestrev) while pending: r = pending.pop() if r >= llowestrev and r not in seen: if r in roots: kept.add(r) else: pending.update([p for p in cl.parentrevs(r)]) seen.add(r) return sorted(kept)