convert/git: check status when reading output stream

hgext/convert/git.py

@@ -1,166 +1,170 b''
 # git.py - git support for the convert extension
 #
 #  Copyright 2005-2009 Matt Mackall <[e-mail]> and others
 #
 # This software may be used and distributed according to the terms of the
 # GNU General Public License version 2 or any later version.
 
 import os
 from mercurial import util
 from mercurial.i18n import _
 
 from common import NoRepo, commit, converter_source, checktool
 
 class convert_git(converter_source):
     # Windows does not support GIT_DIR= construct while other systems
     # cannot remove environment variable. Just assume none have
     # both issues.
     if hasattr(os, 'unsetenv'):
         def gitopen(self, s):
             prevgitdir = os.environ.get('GIT_DIR')
             os.environ['GIT_DIR'] = self.path
             try:
                 return util.popen(s, 'rb')
             finally:
                 if prevgitdir is None:
                     del os.environ['GIT_DIR']
                 else:
                     os.environ['GIT_DIR'] = prevgitdir
     else:
         def gitopen(self, s):
             return util.popen('GIT_DIR=%s %s' % (self.path, s), 'rb')
 
     def gitread(self, s):
         fh = self.gitopen(s)
         data = fh.read()
         return data, fh.close()
 
     def __init__(self, ui, path, rev=None):
         super(convert_git, self).__init__(ui, path, rev=rev)
 
         if os.path.isdir(path + "/.git"):
             path += "/.git"
         if not os.path.exists(path + "/objects"):
             raise NoRepo(_("%s does not look like a Git repository") % path)
 
         checktool('git', 'git')
 
         self.path = path
 
     def getheads(self):
         if not self.rev:
             heads, ret = self.gitread('git rev-parse --branches --remotes')
             heads = heads.splitlines()
         else:
             heads, ret = self.gitread("git rev-parse --verify %s" % self.rev)
             heads = [heads[:-1]]
         if ret:
             raise util.Abort(_('cannot retrieve git heads'))
         return heads
 
     def catfile(self, rev, type):
         if rev == "0" * 40:
             raise IOError()
         data, ret = self.gitread("git cat-file %s %s" % (type, rev))
         if ret:
             raise util.Abort(_('cannot read %r object at %s') % (type, rev))
         return data
 
     def getfile(self, name, rev):
         return self.catfile(rev, "blob")
 
     def getmode(self, name, rev):
         return self.modecache[(name, rev)]
 
     def getchanges(self, version):
         self.modecache = {}
         fh = self.gitopen("git diff-tree -z --root -m -r %s" % version)
         changes = []
         seen = set()
         entry = None
         for l in fh.read().split('\x00'):
             if not entry:
                 if not l.startswith(':'):
                     continue
                 entry = l
                 continue
             f = l
             if f not in seen:
                 seen.add(f)
                 entry = entry.split()
                 h = entry[3]
                 p = (entry[1] == "100755")
                 s = (entry[1] == "120000")
                 self.modecache[(f, h)] = (p and "x") or (s and "l") or ""
                 changes.append((f, h))
             entry = None
+        if fh.close():
+            raise util.Abort(_('cannot read changes in %s') % version)
         return (changes, {})
 
     def getcommit(self, version):
         c = self.catfile(version, "commit") # read the commit hash
         end = c.find("\n\n")
         message = c[end + 2:]
         message = self.recode(message)
         l = c[:end].splitlines()
         parents = []
         author = committer = None
         for e in l[1:]:
             n, v = e.split(" ", 1)
             if n == "author":
                 p = v.split()
                 tm, tz = p[-2:]
                 author = " ".join(p[:-2])
                 if author[0] == "<": author = author[1:-1]
                 author = self.recode(author)
             if n == "committer":
                 p = v.split()
                 tm, tz = p[-2:]
                 committer = " ".join(p[:-2])
                 if committer[0] == "<": committer = committer[1:-1]
                 committer = self.recode(committer)
             if n == "parent":
                 parents.append(v)
 
         if committer and committer != author:
             message += "\ncommitter: %s\n" % committer
         tzs, tzh, tzm = tz[-5:-4] + "1", tz[-4:-2], tz[-2:]
         tz = -int(tzs) * (int(tzh) * 3600 + int(tzm))
         date = tm + " " + str(tz)
 
         c = commit(parents=parents, date=date, author=author, desc=message,
                    rev=version)
         return c
 
     def gettags(self):
         tags = {}
         fh = self.gitopen('git ls-remote --tags "%s"' % self.path)
         prefix = 'refs/tags/'
         for line in fh:
             line = line.strip()
             if not line.endswith("^{}"):
                 continue
             node, tag = line.split(None, 1)
             if not tag.startswith(prefix):
                 continue
             tag = tag[len(prefix):-3]
             tags[tag] = node
+        if fh.close():
+            raise util.Abort(_('cannot read tags from %s') % self.path)
 
         return tags
 
     def getchangedfiles(self, version, i):
         changes = []
         if i is None:
             fh = self.gitopen("git diff-tree --root -m -r %s" % version)
             for l in fh:
                 if "\t" not in l:
                     continue
                 m, f = l[:-1].split("\t")
                 changes.append(f)
-            fh.close()
         else:
             fh = self.gitopen('git diff-tree --name-only --root -r %s "%s^%s" --'
                              % (version, version, i + 1))
             changes = [f.rstrip('\n') for f in fh]
-            fh.close()
+        if fh.close():
+            raise util.Abort(_('cannot read changes in %s') % version)
 
         return changes

tests/test-convert-git

@@ -1,173 +1,186 b''
 #!/bin/sh
 
 "$TESTDIR/hghave" git || exit 80
 
 echo "[extensions]" >> $HGRCPATH
 echo "convert=" >> $HGRCPATH
 echo 'hgext.graphlog =' >> $HGRCPATH
 
 GIT_AUTHOR_NAME='test'; export GIT_AUTHOR_NAME
 GIT_AUTHOR_EMAIL='[e-mail]'; export GIT_AUTHOR_EMAIL
 GIT_AUTHOR_DATE="2007-01-01 00:00:00 +0000"; export GIT_AUTHOR_DATE
 GIT_COMMITTER_NAME="$GIT_AUTHOR_NAME"; export GIT_COMMITTER_NAME
 GIT_COMMITTER_EMAIL="$GIT_AUTHOR_EMAIL"; export GIT_COMMITTER_EMAIL
 GIT_COMMITTER_DATE="$GIT_AUTHOR_DATE"; export GIT_COMMITTER_DATE
 
 count=10
 commit()
 {
     GIT_AUTHOR_DATE="2007-01-01 00:00:$count +0000"
     GIT_COMMITTER_DATE="$GIT_AUTHOR_DATE"
     git commit "$@" >/dev/null 2>/dev/null || echo "git commit error"
     count=`expr $count + 1`
 }
 
 mkdir git-repo
 cd git-repo
 git init-db >/dev/null 2>/dev/null
 echo a > a
 mkdir d
 echo b > d/b
 git add a d
 commit -a -m t1
 
 # Remove the directory, then try to replace it with a file
 # (issue 754)
 git rm -f d/b
 commit -m t2
 echo d > d
 git add d
 commit -m t3
 
 echo b >> a
 commit -a -m t4.1
 
 git checkout -b other HEAD~ >/dev/null 2>/dev/null
 echo c > a
 echo a >> a
 commit -a -m t4.2
 
 git checkout master >/dev/null 2>/dev/null
 git pull --no-commit . other > /dev/null 2>/dev/null
 commit -m 'Merge branch other'
 cd ..
 
 hg convert --datesort git-repo
 hg up -q -R git-repo-hg
 hg -R git-repo-hg tip -v
 
 count=10
 mkdir git-repo2
 cd git-repo2
 git init-db >/dev/null 2>/dev/null
 
 echo foo > foo
 git add foo
 commit -a -m 'add foo'
 
 echo >> foo
 commit -a -m 'change foo'
 
 git checkout -b Bar HEAD~ >/dev/null 2>/dev/null
 echo quux >> quux
 git add quux
 commit -a -m 'add quux'
 
 echo bar > bar
 git add bar
 commit -a -m 'add bar'
 
 git checkout -b Baz HEAD~ >/dev/null 2>/dev/null
 echo baz > baz
 git add baz
 commit -a -m 'add baz'
 
 git checkout master >/dev/null 2>/dev/null
 git pull --no-commit . Bar Baz > /dev/null 2>/dev/null
 commit -m 'Octopus merge'
 
 echo bar >> bar
 commit -a -m 'change bar'
 
 git checkout -b Foo HEAD~ >/dev/null 2>/dev/null
 echo >> foo
 commit -a -m 'change foo'
 
 git checkout master >/dev/null 2>/dev/null
 git pull --no-commit -s ours . Foo > /dev/null 2>/dev/null
 commit -m 'Discard change to foo'
 
 cd ..
 
 glog()
 {
     hg glog --template '{rev} "{desc|firstline}" files: {files}\n' "$@"
 }
 
 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 git-repo2 "$repo"
     hg up -q -R "$repo"
     glog -R "$repo"
     hg -R "$repo" manifest --debug
 }
 
 echo '% full conversion'
 hg -q convert --datesort git-repo2 fullrepo
 hg up -q -R fullrepo
 glog -R fullrepo
 hg -R fullrepo manifest --debug
 
 splitrepo 'octopus merge' 'foo bar baz'
 
 splitrepo 'only some parents of an octopus merge; "discard" a head' 'foo baz quux'
 
 echo
 echo '% test binary conversion (issue 1359)'
 mkdir git-repo3
 cd git-repo3
 git init-db >/dev/null 2>/dev/null
 python -c 'file("b", "wb").write("".join([chr(i) for i in range(256)])*16)'
 git add b
 commit -a -m addbinary
 cd ..
 
 echo '% convert binary file'
 hg convert git-repo3 git-repo3-hg
 
 cd git-repo3-hg
 hg up -C
 python -c 'print len(file("b", "rb").read())'
 cd ..
 
 echo
 echo '% test author vs committer'
 mkdir git-repo4
 cd git-repo4
 git init-db >/dev/null 2>/dev/null
 echo >> foo
 git add foo
 commit -a -m addfoo
 echo >> foo
 GIT_AUTHOR_NAME="nottest"
 commit -a -m addfoo2
 cd ..
 
 echo '% convert author committer'
 hg convert git-repo4 git-repo4-hg
 hg -R git-repo4-hg log -v
 
 echo '% --sourceorder should fail'
 hg convert --sourcesort git-repo4 git-repo4-sourcesort-hg
 
+echo '% damage git repository and convert again'
+cat > damage.py <<EOF
+import os
+for root, dirs, files in os.walk('git-repo4/.git/objects'):
+    if files:
+        path = os.path.join(root, files[0])
+        os.remove(path)
+        break
+EOF
+python damage.py
+hg convert git-repo4 git-repo4-broken-hg 2>&1 | \
+    sed 's/fatal:.*/fatal: git error/g'
+
 true

tests/test-convert-git.out

@@ -1,129 +1,137 b''
 rm 'd/b'
 assuming destination git-repo-hg
 initializing destination git-repo-hg repository
 scanning source...
 sorting...
 converting...
 5 t1
 4 t2
 3 t3
 2 t4.1
 1 t4.2
 0 Merge branch other
 changeset:   5:c78094926be2
 tag:         tip
 parent:      3:f5f5cb45432b
 parent:      4:4e174f80c67c
 user:        test <[e-mail]>
 date:        Mon Jan 01 00:00:15 2007 +0000
 files:       a
 description:
 Merge branch other
 
 
 % full conversion
 @    9 "Discard change to foo" files: foo
 |\
 | o  8 "change foo" files: foo
 | |
 o |  7 "change bar" files: bar
 |/
 o    6 "(octopus merge fixup)" files:
 |\
 | o    5 "Octopus merge" files: baz
 | |\
 o | |  4 "add baz" files: baz
 | | |
 +---o  3 "add bar" files: bar
 | |
 o |  2 "add quux" files: quux
 | |
 | o  1 "change foo" files: foo
 |/
 o  0 "add foo" files: foo
 
 245a3b8bc653999c2b22cdabd517ccb47aecafdf 644   bar
 354ae8da6e890359ef49ade27b68bbc361f3ca88 644   baz
 9277c9cc8dd4576fc01a17939b4351e5ada93466 644   foo
 88dfeab657e8cf2cef3dec67b914f49791ae76b1 644   quux
 % foo bar baz: octopus merge
 @    8 "Discard change to foo" files: foo
 |\
 | o  7 "change foo" files: foo
 | |
 o |  6 "change bar" files: bar
 |/
 o    5 "(octopus merge fixup)" files:
 |\
 | o    4 "Octopus merge" files: baz
 | |\
 o | |  3 "add baz" files: baz
 | | |
 +---o  2 "add bar" files: bar
 | |
 | o  1 "change foo" files: foo
 |/
 o  0 "add foo" files: foo
 
 245a3b8bc653999c2b22cdabd517ccb47aecafdf 644   bar
 354ae8da6e890359ef49ade27b68bbc361f3ca88 644   baz
 9277c9cc8dd4576fc01a17939b4351e5ada93466 644   foo
 % foo baz quux: only some parents of an octopus merge; "discard" a head
 @  6 "Discard change to foo" files: foo
 |
 o  5 "change foo" files: foo
 |
 o    4 "Octopus merge" files:
 |\
 | o  3 "add baz" files: baz
 | |
 | o  2 "add quux" files: quux
 | |
 o |  1 "change foo" files: foo
 |/
 o  0 "add foo" files: foo
 
 354ae8da6e890359ef49ade27b68bbc361f3ca88 644   baz
 9277c9cc8dd4576fc01a17939b4351e5ada93466 644   foo
 88dfeab657e8cf2cef3dec67b914f49791ae76b1 644   quux
 
 % test binary conversion (issue 1359)
 % convert binary file
 initializing destination git-repo3-hg repository
 scanning source...
 sorting...
 converting...
 0 addbinary
 1 files updated, 0 files merged, 0 files removed, 0 files unresolved
 4096
 
 % test author vs committer
 % convert author committer
 initializing destination git-repo4-hg repository
 scanning source...
 sorting...
 converting...
 1 addfoo
 0 addfoo2
 changeset:   1:d63e967f93da
 tag:         tip
 user:        nottest <[e-mail]>
 date:        Mon Jan 01 00:00:21 2007 +0000
 files:       foo
 description:
 addfoo2
 
 committer: test <[e-mail]>
 
 
 changeset:   0:0735477b0224
 user:        test <[e-mail]>
 date:        Mon Jan 01 00:00:20 2007 +0000
 files:       foo
 description:
 addfoo
 
 
 % --sourceorder should fail
 initializing destination git-repo4-sourcesort-hg repository
 abort: --sourcesort is not supported by this data source
+% damage git repository and convert again
+fatal: git error
+initializing destination git-repo4-broken-hg repository
+scanning source...
+sorting...
+converting...
+1 addfoo
+abort: cannot read changes in 6a101ac3f6d8b2524a64295ffd9be87ed927bfeb