In "plain" mode, skip tags that contain characters not valid in Git.

3 files changed, 68 insertions, 0 deletions

diff --git a/NEWS b/NEWS
index 45c601a..f1f6ca5 100644
--- a/NEWS
+++ b/NEWS
@@ -23,6 +23,9 @@ Bug fixes
   These options are not supported by the underlying svn-fast-export.py yet.
   (Jelmer Vernooij, #513747)
 
+* In "plain" mode, skip tags that contain characters not valid in Git.
+  (Jelmer Vernooij, #386986)
+
 0.10 08-Mar-2011
 
 Changes
diff --git a/exporter.py b/exporter.py
index 6d1a9c0..cc9452b 100644
--- a/exporter.py
+++ b/exporter.py
@@ -61,6 +61,39 @@ def _get_output_stream(destination):
     else:
         return open(destination, 'wb')
 
+# from dulwich.repo:
+def check_ref_format(refname):
+    """Check if a refname is correctly formatted.
+
+    Implements all the same rules as git-check-ref-format[1].
+
+    [1] http://www.kernel.org/pub/software/scm/git/docs/git-check-ref-format.html
+
+    :param refname: The refname to check
+    :return: True if refname is valid, False otherwise
+    """
+    # These could be combined into one big expression, but are listed separately
+    # to parallel [1].
+    if '/.' in refname or refname.startswith('.'):
+        return False
+    if '/' not in refname:
+        return False
+    if '..' in refname:
+        return False
+    for c in refname:
+        if ord(c) < 040 or c in '\177 ~^:?*[':
+            return False
+    if refname[-1] in '/.':
+        return False
+    if refname.endswith('.lock'):
+        return False
+    if '@{' in refname:
+        return False
+    if '\\' in refname:
+        return False
+    return True
+
+
 
 class BzrFastExporter(object):
 
@@ -507,6 +540,10 @@ class BzrFastExporter(object):
                     'revision %s' % (tag, revid))
             else:
                 git_ref = 'refs/tags/%s' % tag.encode("utf-8")
+                if self.plain_format and not check_ref_format(git_ref):
+                    self.warning('not creating tag %r as its name would not be '
+                                 'valid in git.', git_ref)
+                    continue
                 self.print_cmd(commands.ResetCommand(git_ref, ":" + str(mark)))
 
     def _next_tmp_branch_name(self):
diff --git a/tests/test_exporter.py b/tests/test_exporter.py
index fe50e3b..24fedb1 100644
--- a/tests/test_exporter.py
+++ b/tests/test_exporter.py
@@ -24,6 +24,7 @@ from bzrlib import tests
 
 from bzrlib.plugins.fastimport.exporter import (
     _get_output_stream,
+    check_ref_format,
     )
 
 from bzrlib.plugins.fastimport.tests import (
@@ -60,3 +61,30 @@ class TestOutputStream(tests.TestCase):
         f = open(filename, 'r')
         self.assertEquals("foo", f.read())
         f.close()
+
+
+# from dulwich.tests.test_repository:
+class CheckRefFormatTests(tests.TestCase):
+    """Tests for the check_ref_format function.
+
+    These are the same tests as in the git test suite.
+    """
+
+    def test_valid(self):
+        self.assertTrue(check_ref_format('heads/foo'))
+        self.assertTrue(check_ref_format('foo/bar/baz'))
+        self.assertTrue(check_ref_format('refs///heads/foo'))
+        self.assertTrue(check_ref_format('foo./bar'))
+        self.assertTrue(check_ref_format('heads/foo@bar'))
+        self.assertTrue(check_ref_format('heads/fix.lock.error'))
+
+    def test_invalid(self):
+        self.assertFalse(check_ref_format('foo'))
+        self.assertFalse(check_ref_format('heads/foo/'))
+        self.assertFalse(check_ref_format('./foo'))
+        self.assertFalse(check_ref_format('.refs/foo'))
+        self.assertFalse(check_ref_format('heads/foo..bar'))
+        self.assertFalse(check_ref_format('heads/foo?bar'))
+        self.assertFalse(check_ref_format('heads/foo.lock'))
+        self.assertFalse(check_ref_format('heads/v@{ation'))
+        self.assertFalse(check_ref_format('heads/foo\bar'))