##// END OF EJS Templates
named branches: server branchmap wire protocol support (issue736)...
named branches: server branchmap wire protocol support (issue736) The repository command, 'branchmap', returns a dictionary, branchname -> [branchheads], and will be implemented for localrepo, httprepo and sshrepo. The following wire format is used for returning data: branchname1 branch1head2 branch1head2 ... branchname2 ... ... Branch names are URL encoded to escape white space, and branch heads are sent as hex encoded node ids. All branches and all their heads are sent. The background and motivation for this command is the desire for a richer named branch semantics when pushing changesets. The details are explained in the original proposal which is included below. 1. BACKGROUND The algorithm currently implemented in Mercurial only considers the graph theoretical heads when determining whether new heads are created, rather than using the branch heads as a count (the algorithm considers a branch head effectively closed when it is merged into another branch or a new named branch is started from that point onward). Our particular problem with the algorithm is that we'd like to see the following case working without forcing a push: Upsteam has: (0:dev) ---- (1:dev) \ `--- (2:stable) Someone merges stable into dev: (0:dev) ---- (1:dev) ------(3:dev) \ / `--- (2:stable) --------´ This can be pushed without --force (as it should). Now someone else does some coding on stable (a bug fix, say): (0:dev) ---- (1:dev) ------(3:dev) \ / `--- (2:stable) ---------´---------(4:stable) This time we need --force to push. We allow this to be pushed without using --force by getting all the remote branch heads (by extending the wire protocol with a new function). We would, furthermore, also prefer if it is impossible to push a new branch without --force (or a later --newbranch option so --force isn't shoe-horned into too many disparate functions, if need be), except of course in the case where the remote repository is empty. This is what our patches accomplish. 2. ALTERNATIVES We have, of course, considered some alternatives to reconstructing enough information to decide whether we are creating new remote branch heads, before we added the new wire protocol command. 2.1. LOOKUP ON REMOTE The main alternative is to use the information from remote.heads() and remote.lookup() to try to reconstruct enough graph information to decide whether we are creating new heads. This is not adequate as illustrated below. Remember that each lookup is typically a request-response pair over SSH or HTTP(S). If we have a simple repository at the remote end like this: (0:dev) ---- (1:dev) ---- (3:stable) \ `--- (2:dev) then remote.heads() will yield [2, 3]. Assume we have nodes [0, 1, 2] locally and want to create a new node, 4:dev, as a descendant from (1:dev), which should be OK as 1:dev is a branch head. If we do remote.lookup('dev') we will get [2]. Thus, we can get information about whether a branch exists on the remote server or not, but this does not solve our problem of figuring out whether we are creating new heads or not. Pushing 4:dev ought to be OK, since after the push, we still only have two heads on branch a. Using remote.lookup() and remote.heads() is thus not adequate to consistently decide whether we are creating new remote heads (e.g. in this situation the latter would never return 1:dev). 2.2. USING INCOMING TO RECONSTRUCT THE GRAPH An alternative would be to use information equivalent to hg incoming to get the full remote graph in addition to the local graph. To do this, we would have to get a changegroup(subset) bundle representing the remote end (which may be a substantial amount of data), getting the branch heads from an instantiated bundlerepository, deleting the bundle, and finally, we can compute the prepush logic. While this is backwards compatible, it will cause a possibly substantial slowdown of the push command as it first needs to pull in all changes. 3. FURTHER ARGUMENTS IN FAVOUR OF THE BRANCHMAP WIRE-PROTOCOL EXTENSION Currently, the commands incoming and pull, work based on the tip of a given branch if used with "-r branchname", making it hard to get all revisions of a certain branch only (if it has multiple heads). This can be solved by requesting the remote's branchheads and letting the revisions to be used with the command be these heads. This can be done by extending the commands with a new option, e.g.: hg pull -b branchname which will be turned into the equivalent of: hg pull -r branchhead1 -r branchhead2 -r branchhead3 We have a simple follow-up patch that can do this ready as well (although not submitted yet as it is pending the acceptance of the branch patch). 4. WRAP-UP We generally find that the branchmap wire protocol extension can provide better named branch support to Mercurial. Currently, some things, like the initial push scenario in this mail, are fairly counter-intuitive, and the more often you have to force push, the more it is likely you will get a lot of spurious and unnecessary merge nodes. Also, restricting incoming and pull to all changes on a branch rather than changes on the tip-most head would be a sensible extension to making named branches a first class citizen in Mercurial. Currently, named branches sometimes feel like a late-coming unwanted step-child. We have run it in a production environment for a while, with fewer multiple heads occurring in our repositories and fewer confused users as a result. Also, it fixes the long-standing issue 736. Co-contributor: Sune Foldager <cryo@cyanite.org>

File last commit:

r8527:f9a80054 default
r8562:e3495c39 default
Show More
copies.py
233 lines | 7.3 KiB | text/x-python | PythonLexer
# copies.py - copy detection for Mercurial
#
# Copyright 2008 Matt Mackall <mpm@selenic.com>
#
# This software may be used and distributed according to the terms of the
# GNU General Public License version 2, incorporated herein by reference.
from i18n import _
import util
import heapq
def _nonoverlap(d1, d2, d3):
"Return list of elements in d1 not in d2 or d3"
return sorted([d for d in d1 if d not in d3 and d not in d2])
def _dirname(f):
s = f.rfind("/")
if s == -1:
return ""
return f[:s]
def _dirs(files):
d = set()
for f in files:
f = _dirname(f)
while f not in d:
d.add(f)
f = _dirname(f)
return d
def _findoldnames(fctx, limit):
"find files that path was copied from, back to linkrev limit"
old = {}
seen = set()
orig = fctx.path()
visit = [(fctx, 0)]
while visit:
fc, depth = visit.pop()
s = str(fc)
if s in seen:
continue
seen.add(s)
if fc.path() != orig and fc.path() not in old:
old[fc.path()] = (depth, fc.path()) # remember depth
if fc.rev() < limit and fc.rev() is not None:
continue
visit += [(p, depth - 1) for p in fc.parents()]
# return old names sorted by depth
return [o[1] for o in sorted(old.values())]
def _findlimit(repo, a, b):
"find the earliest revision that's an ancestor of a or b but not both"
# basic idea:
# - mark a and b with different sides
# - if a parent's children are all on the same side, the parent is
# on that side, otherwise it is on no side
# - walk the graph in topological order with the help of a heap;
# - add unseen parents to side map
# - clear side of any parent that has children on different sides
# - track number of interesting revs that might still be on a side
# - track the lowest interesting rev seen
# - quit when interesting revs is zero
cl = repo.changelog
working = len(cl) # pseudo rev for the working directory
if a is None:
a = working
if b is None:
b = working
side = {a: -1, b: 1}
visit = [-a, -b]
heapq.heapify(visit)
interesting = len(visit)
limit = working
while interesting:
r = -heapq.heappop(visit)
if r == working:
parents = [cl.rev(p) for p in repo.dirstate.parents()]
else:
parents = cl.parentrevs(r)
for p in parents:
if p not in side:
# first time we see p; add it to visit
side[p] = side[r]
if side[p]:
interesting += 1
heapq.heappush(visit, -p)
elif side[p] and side[p] != side[r]:
# p was interesting but now we know better
side[p] = 0
interesting -= 1
if side[r]:
limit = r # lowest rev visited
interesting -= 1
return limit
def copies(repo, c1, c2, ca, checkdirs=False):
"""
Find moves and copies between context c1 and c2
"""
# avoid silly behavior for update from empty dir
if not c1 or not c2 or c1 == c2:
return {}, {}
# avoid silly behavior for parent -> working dir
if c2.node() is None and c1.node() == repo.dirstate.parents()[0]:
return repo.dirstate.copies(), {}
limit = _findlimit(repo, c1.rev(), c2.rev())
m1 = c1.manifest()
m2 = c2.manifest()
ma = ca.manifest()
def makectx(f, n):
if len(n) != 20: # in a working context?
if c1.rev() is None:
return c1.filectx(f)
return c2.filectx(f)
return repo.filectx(f, fileid=n)
ctx = util.cachefunc(makectx)
copy = {}
fullcopy = {}
diverge = {}
def checkcopies(f, m1, m2):
'''check possible copies of f from m1 to m2'''
c1 = ctx(f, m1[f])
for of in _findoldnames(c1, limit):
fullcopy[f] = of # remember for dir rename detection
if of in m2: # original file not in other manifest?
# if the original file is unchanged on the other branch,
# no merge needed
if m2[of] != ma.get(of):
c2 = ctx(of, m2[of])
ca = c1.ancestor(c2)
# related and named changed on only one side?
if ca and (ca.path() == f or ca.path() == c2.path()):
if c1 != ca or c2 != ca: # merge needed?
copy[f] = of
elif of in ma:
diverge.setdefault(of, []).append(f)
repo.ui.debug(_(" searching for copies back to rev %d\n") % limit)
u1 = _nonoverlap(m1, m2, ma)
u2 = _nonoverlap(m2, m1, ma)
if u1:
repo.ui.debug(_(" unmatched files in local:\n %s\n")
% "\n ".join(u1))
if u2:
repo.ui.debug(_(" unmatched files in other:\n %s\n")
% "\n ".join(u2))
for f in u1:
checkcopies(f, m1, m2)
for f in u2:
checkcopies(f, m2, m1)
diverge2 = set()
for of, fl in diverge.items():
if len(fl) == 1:
del diverge[of] # not actually divergent
else:
diverge2.update(fl) # reverse map for below
if fullcopy:
repo.ui.debug(_(" all copies found (* = to merge, ! = divergent):\n"))
for f in fullcopy:
note = ""
if f in copy: note += "*"
if f in diverge2: note += "!"
repo.ui.debug(" %s -> %s %s\n" % (f, fullcopy[f], note))
del diverge2
if not fullcopy or not checkdirs:
return copy, diverge
repo.ui.debug(_(" checking for directory renames\n"))
# generate a directory move map
d1, d2 = _dirs(m1), _dirs(m2)
invalid = set()
dirmove = {}
# examine each file copy for a potential directory move, which is
# when all the files in a directory are moved to a new directory
for dst, src in fullcopy.iteritems():
dsrc, ddst = _dirname(src), _dirname(dst)
if dsrc in invalid:
# already seen to be uninteresting
continue
elif dsrc in d1 and ddst in d1:
# directory wasn't entirely moved locally
invalid.add(dsrc)
elif dsrc in d2 and ddst in d2:
# directory wasn't entirely moved remotely
invalid.add(dsrc)
elif dsrc in dirmove and dirmove[dsrc] != ddst:
# files from the same directory moved to two different places
invalid.add(dsrc)
else:
# looks good so far
dirmove[dsrc + "/"] = ddst + "/"
for i in invalid:
if i in dirmove:
del dirmove[i]
del d1, d2, invalid
if not dirmove:
return copy, diverge
for d in dirmove:
repo.ui.debug(_(" dir %s -> %s\n") % (d, dirmove[d]))
# check unaccounted nonoverlapping files against directory moves
for f in u1 + u2:
if f not in fullcopy:
for d in dirmove:
if f.startswith(d):
# new file added in a directory that was moved, move it
df = dirmove[d] + f[len(d):]
if df not in copy:
copy[f] = df
repo.ui.debug(_(" file %s -> %s\n") % (f, copy[f]))
break
return copy, diverge