mercurial_source: add --filemap support

tests/test-convert-filemap

@@ -0,0 +1,124
+#!/bin/sh
+
+HGMERGE=true; export HGMERGE
+
+echo '[extensions]' >> $HGRCPATH
+echo 'hgext.graphlog =' >> $HGRCPATH
+echo 'hgext.convert =' >> $HGRCPATH
+
+glog()
+{
+    hg glog --template '#rev# "#desc#" files: #files#\n' "$@"
+}
+
+hg init source
+cd source
+
+echo foo > foo
+echo baz > baz
+mkdir dir
+echo dir/file >> dir/file
+echo dir/file2 >> dir/file2
+hg ci -d '0 0' -qAm '0: add foo baz dir/'
+
+echo bar > bar
+echo quux > quux
+hg copy foo copied
+hg ci -d '1 0' -qAm '1: add bar quux; copy foo to copied'
+
+echo >> foo
+hg ci -d '2 0' -m '2: change foo'
+
+hg up -qC 1
+echo >> bar
+echo >> quux
+hg ci -d '3 0' -m '3: change bar quux'
+
+hg up -qC 2
+hg merge -qr 3
+echo >> bar
+echo >> baz
+hg ci -d '4 0' -m '4: first merge; change bar baz'
+
+echo >> bar
+echo 1 >> baz
+echo >> quux
+hg ci -d '5 0' -m '5: change bar baz quux'
+
+hg up -qC 4
+echo >> foo
+echo 2 >> baz
+hg ci -d '6 0' -m '6: change foo baz'
+
+hg up -qC 5
+hg merge -qr 6
+echo >> bar
+hg ci -d '7 0' -m '7: second merge; change bar'
+
+echo >> foo
+hg ci -m '8: change foo'
+
+glog
+
+echo '% final file versions in this repo:'
+hg manifest --debug
+hg debugrename copied
+echo
+
+cd ..
+
+splitrepo()
+{
+    msg="$1"
+    files="$2"
+    opts=$3
+    echo "% $files: $msg"
+    prefix=`echo "$files" | sed -e 's/ /-/g'`
+    fmap="$prefix.fmap"
+    repo="$prefix.repo"
+    for i in $files; do
+	echo "include $i" >> "$fmap"
+    done
+    hg -q convert $opts --filemap "$fmap" --datesort source "$repo"
+    glog -R "$repo"
+    hg -R "$repo" manifest --debug
+}
+
+splitrepo 'skip unwanted merges; use 1st parent in 1st merge, 2nd in 2nd' foo
+
+splitrepo 'merges are not merges anymore' bar
+
+splitrepo '1st merge is not a merge anymore; 2nd still is' baz
+
+splitrepo 'we add additional merges when they are interesting' 'foo quux'
+
+splitrepo 'partial conversion' 'bar quux' '-r 3'
+splitrepo 'complete the partial conversion' 'bar quux'
+
+rm -r foo.repo
+splitrepo 'partial conversion' 'foo' '-r 3'
+splitrepo 'complete the partial conversion' 'foo'
+
+splitrepo 'copied file; source not included in new repo' copied
+hg --cwd copied.repo debugrename copied
+
+splitrepo 'copied file; source included in new repo' 'foo copied'
+hg --cwd foo-copied.repo debugrename copied
+
+cat > renames.fmap <<EOF
+include dir
+exclude dir/file2
+rename dir dir2
+include foo
+include copied
+rename foo foo2
+rename copied copied2
+EOF
+hg -q convert --filemap renames.fmap --datesort source renames.repo
+glog -R renames.repo
+hg -R renames.repo manifest --debug
+hg --cwd renames.repo debugrename copied2
+echo 'copied:'
+hg --cwd source cat copied
+echo 'copied2:'
+hg --cwd renames.repo cat copied2

tests/test-convert-filemap.out

@@ -0,0 +1,154
+@  8 "8: change foo" files: foo
+|
+o    7 "7: second merge; change bar" files: bar baz
+|\
+| o  6 "6: change foo baz" files: baz foo
+| |
+o |  5 "5: change bar baz quux" files: bar baz quux
+|/
+o    4 "4: first merge; change bar baz" files: bar baz
+|\
+| o  3 "3: change bar quux" files: bar quux
+| |
+o |  2 "2: change foo" files: foo
+|/
+o  1 "1: add bar quux; copy foo to copied" files: bar copied quux
+|
+o  0 "0: add foo baz dir/" files: baz dir/file dir/file2 foo
+
+% final file versions in this repo:
+9463f52fe115e377cf2878d4fc548117211063f2 644 bar
+94c1be4dfde2ee8d78db8bbfcf81210813307c3d 644 baz
+6ca237634e1f6bee1b6db94292fb44f092a25842 644 copied
+3e20847584beff41d7cd16136b7331ab3d754be0 644 dir/file
+75e6d3f8328f5f6ace6bf10b98df793416a09dca 644 dir/file2
+9a7b52012991e4873687192c3e17e61ba3e837a3 644 foo
+bc3eca3f47023a3e70ca0d8cc95a22a6827db19d 644 quux
+copied renamed from foo:2ed2a3912a0b24502043eae84ee4b279c18b90dd
+
+% foo: skip unwanted merges; use 1st parent in 1st merge, 2nd in 2nd
+o  3 "8: change foo" files: foo
+|
+o  2 "6: change foo baz" files: foo
+|
+o  1 "2: change foo" files: foo
+|
+o  0 "0: add foo baz dir/" files: foo
+
+9a7b52012991e4873687192c3e17e61ba3e837a3 644 foo
+% bar: merges are not merges anymore
+o  4 "7: second merge; change bar" files: bar
+|
+o  3 "5: change bar baz quux" files: bar
+|
+o  2 "4: first merge; change bar baz" files: bar
+|
+o  1 "3: change bar quux" files: bar
+|
+o  0 "1: add bar quux; copy foo to copied" files: bar
+
+9463f52fe115e377cf2878d4fc548117211063f2 644 bar
+% baz: 1st merge is not a merge anymore; 2nd still is
+o    4 "7: second merge; change bar" files: baz
+|\
+| o  3 "6: change foo baz" files: baz
+| |
+o |  2 "5: change bar baz quux" files: baz
+|/
+o  1 "4: first merge; change bar baz" files: baz
+|
+o  0 "0: add foo baz dir/" files: baz
+
+94c1be4dfde2ee8d78db8bbfcf81210813307c3d 644 baz
+% foo quux: we add additional merges when they are interesting
+o  8 "8: change foo" files: foo
+|
+o    7 "7: second merge; change bar" files:
+|\
+| o  6 "6: change foo baz" files: foo
+| |
+o |  5 "5: change bar baz quux" files: quux
+|/
+o    4 "4: first merge; change bar baz" files:
+|\
+| o  3 "3: change bar quux" files: quux
+| |
+o |  2 "2: change foo" files: foo
+|/
+o  1 "1: add bar quux; copy foo to copied" files: quux
+|
+o  0 "0: add foo baz dir/" files: foo
+
+9a7b52012991e4873687192c3e17e61ba3e837a3 644 foo
+bc3eca3f47023a3e70ca0d8cc95a22a6827db19d 644 quux
+% bar quux: partial conversion
+o  1 "3: change bar quux" files: bar quux
+|
+o  0 "1: add bar quux; copy foo to copied" files: bar quux
+
+b79105bedc55102f394e90a789c9c380117c1b4a 644 bar
+db0421cc6b685a458c8d86c7d5c004f94429ea23 644 quux
+% bar quux: complete the partial conversion
+o  4 "7: second merge; change bar" files: bar
+|
+o  3 "5: change bar baz quux" files: bar quux
+|
+o  2 "4: first merge; change bar baz" files: bar
+|
+o  1 "3: change bar quux" files: bar quux
+|
+o  0 "1: add bar quux; copy foo to copied" files: bar quux
+
+9463f52fe115e377cf2878d4fc548117211063f2 644 bar
+bc3eca3f47023a3e70ca0d8cc95a22a6827db19d 644 quux
+% foo: partial conversion
+o  0 "0: add foo baz dir/" files: foo
+
+2ed2a3912a0b24502043eae84ee4b279c18b90dd 644 foo
+% foo: complete the partial conversion
+o  3 "8: change foo" files: foo
+|
+o  2 "6: change foo baz" files: foo
+|
+o  1 "2: change foo" files: foo
+|
+o  0 "0: add foo baz dir/" files: foo
+
+9a7b52012991e4873687192c3e17e61ba3e837a3 644 foo
+% copied: copied file; source not included in new repo
+o  0 "1: add bar quux; copy foo to copied" files: copied
+
+2ed2a3912a0b24502043eae84ee4b279c18b90dd 644 copied
+copied not renamed
+% foo copied: copied file; source included in new repo
+o  4 "8: change foo" files: foo
+|
+o  3 "6: change foo baz" files: foo
+|
+o  2 "2: change foo" files: foo
+|
+o  1 "1: add bar quux; copy foo to copied" files: copied
+|
+o  0 "0: add foo baz dir/" files: foo
+
+6ca237634e1f6bee1b6db94292fb44f092a25842 644 copied
+9a7b52012991e4873687192c3e17e61ba3e837a3 644 foo
+copied renamed from foo:2ed2a3912a0b24502043eae84ee4b279c18b90dd
+o  4 "8: change foo" files: foo2
+|
+o  3 "6: change foo baz" files: foo2
+|
+o  2 "2: change foo" files: foo2
+|
+o  1 "1: add bar quux; copy foo to copied" files: copied2
+|
+o  0 "0: add foo baz dir/" files: dir2/file foo2
+
+e5e3d520be9be45937d0b06b004fadcd6c221fa2 644 copied2
+3e20847584beff41d7cd16136b7331ab3d754be0 644 dir2/file
+9a7b52012991e4873687192c3e17e61ba3e837a3 644 foo2
+copied2 renamed from foo2:2ed2a3912a0b24502043eae84ee4b279c18b90dd
+copied:
+foo
+copied2:
+foo

hgext/convert/hg.py

@@ -1,228 +1,243
 # hg backend for convert extension
 
 # Note for hg->hg conversion: Old versions of Mercurial didn't trim
 # the whitespace from the ends of commit messages, but new versions
 # do.  Changesets created by those older versions, then converted, may
 # thus have different hashes for changesets that are otherwise
 # identical.
 
 
 import os, time
 from mercurial.i18n import _
 from mercurial.node import *
 from mercurial import hg, lock, revlog, util
 
 from common import NoRepo, commit, converter_source, converter_sink
 
 class mercurial_sink(converter_sink):
     def __init__(self, ui, path):
         self.path = path
         self.ui = ui
         self.branchnames = ui.configbool('convert', 'hg.usebranchnames', True)
         self.clonebranches = ui.configbool('convert', 'hg.clonebranches', False)
         self.tagsbranch = ui.config('convert', 'hg.tagsbranch', 'default')
         self.lastbranch = None
         try:
             self.repo = hg.repository(self.ui, path)
         except:
             raise NoRepo("could not open hg repo %s as sink" % path)
         self.lock = None
         self.wlock = None
         self.filemapmode = False
 
     def before(self):
         self.wlock = self.repo.wlock()
         self.lock = self.repo.lock()
         self.repo.dirstate.clear()
 
     def after(self):
         self.repo.dirstate.invalidate()
         self.lock = None
         self.wlock = None
 
     def revmapfile(self):
         return os.path.join(self.path, ".hg", "shamap")
 
     def authorfile(self):
         return os.path.join(self.path, ".hg", "authormap")
 
     def getheads(self):
         h = self.repo.changelog.heads()
         return [ hex(x) for x in h ]
 
     def putfile(self, f, e, data):
         self.repo.wwrite(f, data, e)
         if f not in self.repo.dirstate:
             self.repo.dirstate.normallookup(f)
 
     def copyfile(self, source, dest):
         self.repo.copy(source, dest)
 
     def delfile(self, f):
         try:
             util.unlink(self.repo.wjoin(f))
             #self.repo.remove([f])
         except OSError:
             pass
 
     def setbranch(self, branch, pbranch, parents):
         if (not self.clonebranches) or (branch == self.lastbranch):
             return
 
         self.lastbranch = branch
         self.after()
         if not branch:
             branch = 'default'
         if not pbranch:
             pbranch = 'default'
 
         branchpath = os.path.join(self.path, branch)
         try:
             self.repo = hg.repository(self.ui, branchpath)
         except:
             if not parents:
                 self.repo = hg.repository(self.ui, branchpath, create=True)
             else:
                 self.ui.note(_('cloning branch %s to %s\n') % (pbranch, branch))
                 hg.clone(self.ui, os.path.join(self.path, pbranch),
                          branchpath, rev=parents, update=False,
                          stream=True)
                 self.repo = hg.repository(self.ui, branchpath)
 
     def putcommit(self, files, parents, commit):
         seen = {}
         pl = []
         for p in parents:
             if p not in seen:
                 pl.append(p)
                 seen[p] = 1
         parents = pl
         nparents = len(parents)
         if self.filemapmode and nparents == 1:
             m1node = self.repo.changelog.read(bin(parents[0]))[0]
             parent = parents[0]
 
         if len(parents) < 2: parents.append("0" * 40)
         if len(parents) < 2: parents.append("0" * 40)
         p2 = parents.pop(0)
 
         text = commit.desc
         extra = {}
         if self.branchnames and commit.branch:
             extra['branch'] = commit.branch
         if commit.rev:
             extra['convert_revision'] = commit.rev
 
         while parents:
             p1 = p2
             p2 = parents.pop(0)
             a = self.repo.rawcommit(files, text, commit.author, commit.date,
                                     bin(p1), bin(p2), extra=extra)
             self.repo.dirstate.clear()
             text = "(octopus merge fixup)\n"
             p2 = hg.hex(self.repo.changelog.tip())
 
         if self.filemapmode and nparents == 1:
             man = self.repo.manifest
             mnode = self.repo.changelog.read(bin(p2))[0]
             if not man.cmp(m1node, man.revision(mnode)):
                 self.repo.rollback()
                 self.repo.dirstate.clear()
                 return parent
         return p2
 
     def puttags(self, tags):
         try:
             old = self.repo.wfile(".hgtags").read()
             oldlines = old.splitlines(1)
             oldlines.sort()
         except:
             oldlines = []
 
         k = tags.keys()
         k.sort()
         newlines = []
         for tag in k:
             newlines.append("%s %s\n" % (tags[tag], tag))
 
         newlines.sort()
 
         if newlines != oldlines:
             self.ui.status("updating tags\n")
             f = self.repo.wfile(".hgtags", "w")
             f.write("".join(newlines))
             f.close()
             if not oldlines: self.repo.add([".hgtags"])
             date = "%s 0" % int(time.mktime(time.gmtime()))
             extra = {}
             if self.tagsbranch != 'default':
                 extra['branch'] = self.tagsbranch
             try:
                 tagparent = self.repo.changectx(self.tagsbranch).node()
             except hg.RepoError, inst:
                 tagparent = nullid
             self.repo.rawcommit([".hgtags"], "update tags", "convert-repo",
                                 date, tagparent, nullid)
             return hex(self.repo.changelog.tip())
 
     def setfilemapmode(self, active):
         self.filemapmode = active
 
 class mercurial_source(converter_source):
     def __init__(self, ui, path, rev=None):
         converter_source.__init__(self, ui, path, rev)
         try:
             self.repo = hg.repository(self.ui, path)
         except:
             raise NoRepo("could not open hg repo %s as source" % path)
         self.lastrev = None
         self.lastctx = None
+        self._changescache = None
 
     def changectx(self, rev):
         if self.lastrev != rev:
             self.lastctx = self.repo.changectx(rev)
             self.lastrev = rev
         return self.lastctx
 
     def getheads(self):
         if self.rev:
             return [hex(self.repo.changectx(self.rev).node())]
         else:
             return [hex(node) for node in self.repo.heads()]
 
     def getfile(self, name, rev):
         try:
             return self.changectx(rev).filectx(name).data()
         except revlog.LookupError, err:
             raise IOError(err)
 
     def getmode(self, name, rev):
         m = self.changectx(rev).manifest()
         return (m.execf(name) and 'x' or '') + (m.linkf(name) and 'l' or '')
 
     def getchanges(self, rev):
         ctx = self.changectx(rev)
+        if self._changescache and self._changescache[0] == rev:
+            m, a, r = self._changescache[1]
+        else:
         m, a, r = self.repo.status(ctx.parents()[0].node(), ctx.node())[:3]
         changes = [(name, rev) for name in m + a + r]
         changes.sort()
         return (changes, self.getcopies(ctx, m + a))
 
     def getcopies(self, ctx, files):
         copies = {}
         for name in files:
             try:
                 copies[name] = ctx.filectx(name).renamed()[0]
             except TypeError:
                 pass
         return copies
 
     def getcommit(self, rev):
         ctx = self.changectx(rev)
         parents = [hex(p.node()) for p in ctx.parents() if p.node() != nullid]
         return commit(author=ctx.user(), date=util.datestr(ctx.date()),
                       desc=ctx.description(), parents=parents,
                       branch=ctx.branch())
 
     def gettags(self):
         tags = [t for t in self.repo.tagslist() if t[0] != 'tip']
         return dict([(name, hex(node)) for name, node in tags])
+
+    def getchangedfiles(self, rev, i):
+        ctx = self.changectx(rev)
+        i = i or 0
+        changes = self.repo.status(ctx.parents()[i].node(), ctx.node())[:3]
+
+        if i == 0:
+            self._changescache = (rev, changes)
+
+        return changes[0] + changes[1] + changes[2]
+