Add support for time zone information in tags and commits.

Rick Copeland · Rick Copeland · commit b48e4d3aa853 · 2010-04-27T16:38:35.000-04:00
This commit includes
 - an update to git.objects.utils:parse_actor_and_date to parse the timezone offset
 - updates to the git.objects.Commit and git.objects.Tag objects to support *_tz_offset attributes
 - updates to tests in test.git.test_commit and test.git.test_refs to check for appropriate *_tz_offset attributes
diff --git a/lib/git/objects/commit.py b/lib/git/objects/commit.py
@@ -23,12 +23,14 @@ class Commit(base.Object, Iterable, diff.Diffable, utils.Traversable):
     
     # object configuration 
     type = "commit"
-    __slots__ = ("tree", "author", "authored_date", "committer", "committed_date",
-                    "message", "parents")
+    __slots__ = ("tree",
+                 "author", "authored_date", "author_tz_offset",
+                 "committer", "committed_date", "committer_tz_offset",
+                 "message", "parents")
     _id_attribute_ = "sha"
     
-    def __init__(self, repo, sha, tree=None, author=None, authored_date=None,
-                 committer=None, committed_date=None, message=None, parents=None):
+    def __init__(self, repo, sha, tree=None, author=None, authored_date=None, author_tz_offset=None,
+                 committer=None, committed_date=None, committer_tz_offset=None, message=None, parents=None):
         """
         Instantiate a new Commit. All keyword arguments taking None as default will 
         be implicitly set if id names a valid sha. 
@@ -51,13 +53,19 @@ def __init__(self, repo, sha, tree=None, author=None, authored_date=None,
             is the authored DateTime - use time.gmtime() to convert it into a 
             different format
 
+        ``author_tz_offset``: int_seconds_west_of_utc
+           is the timezone that the authored_date is in
+
         ``committer`` : Actor
             is the committer string
 
         ``committed_date`` : int_seconds_since_epoch
             is the committed DateTime - use time.gmtime() to convert it into a 
             different format
 
+        ``committer_tz_offset``: int_seconds_west_of_utc
+           is the timezone that the authored_date is in
+
         ``message`` : string
             is the commit message
 
@@ -94,8 +102,10 @@ def _set_cache_(self, attr):
             self.tree = temp.tree
             self.author = temp.author
             self.authored_date = temp.authored_date
+            self.author_tz_offset = temp.author_tz_offset
             self.committer = temp.committer
             self.committed_date = temp.committed_date
+            self.committer_tz_offset = temp.committer_tz_offset
             self.message = temp.message
         else:
             super(Commit, self)._set_cache_(attr)
@@ -253,8 +263,8 @@ def _iter_from_process_or_stream(cls, repo, proc_or_stream, from_rev_list):
                 parents.append(parent_line.split()[-1])
             # END for each parent line
             
-            author, authored_date = utils.parse_actor_and_date(next_line)
-            committer, committed_date = utils.parse_actor_and_date(stream.next())
+            author, authored_date, author_tz_offset = utils.parse_actor_and_date(next_line)
+            committer, committed_date, committer_tz_offset = utils.parse_actor_and_date(stream.next())
             
             # empty line
             stream.next()
@@ -276,8 +286,10 @@ def _iter_from_process_or_stream(cls, repo, proc_or_stream, from_rev_list):
             # END message parsing
             message = '\n'.join(message_lines)
             
-            yield Commit(repo, id, parents=tuple(parents), tree=tree, author=author, authored_date=authored_date,
-                          committer=committer, committed_date=committed_date, message=message)
+            yield Commit(repo, id, parents=tuple(parents), tree=tree,
+                         author=author, authored_date=authored_date, author_tz_offset=author_tz_offset,
+                         committer=committer, committed_date=committed_date, committer_tz_offset=committer_tz_offset,
+                         message=message)
         # END for each line in stream
         
         
diff --git a/lib/git/objects/tag.py b/lib/git/objects/tag.py
@@ -15,10 +15,10 @@ class TagObject(base.Object):
     to.
     """
     type = "tag"
-    __slots__ = ( "object", "tag", "tagger", "tagged_date", "message" )
+    __slots__ = ( "object", "tag", "tagger", "tagged_date", "tagger_tz_offset", "message" )
         
     def __init__(self, repo, sha, object=None, tag=None, 
-                tagger=None, tagged_date=None, message=None):
+                tagger=None, tagged_date=None, tagger_tz_offset=None, message=None):
         """
         Initialize a tag object with additional data
         
@@ -40,6 +40,10 @@ def __init__(self, repo, sha, object=None, tag=None,
           ``tagged_date`` : int_seconds_since_epoch
             is the DateTime of the tag creation - use time.gmtime to convert 
             it into a different format
+
+        ``tagged_tz_offset``: int_seconds_west_of_utc
+          is the timezone that the authored_date is in
+
         """
         super(TagObject, self).__init__(repo, sha )
         self._set_self_from_args_(locals())
@@ -58,7 +62,7 @@ def _set_cache_(self, attr):
             self.tag = lines[2][4:]  # tag <tag name>
             
             tagger_info = lines[3][7:]# tagger <actor> <date>
-            self.tagger, self.tagged_date = utils.parse_actor_and_date(tagger_info)
+            self.tagger, self.tagged_date, self.tagger_tz_offset = utils.parse_actor_and_date(tagger_info)
             
             # line 4 empty - it could mark the beginning of the next header
             # in csse there really is no message, it would not exist. Otherwise 
diff --git a/lib/git/objects/utils.py b/lib/git/objects/utils.py
@@ -39,7 +39,7 @@ def get_object_type_by_name(object_type_name):
         
     
 # precompiled regex
-_re_actor_epoch = re.compile(r'^.+? (.*) (\d+) .*$')
+_re_actor_epoch = re.compile(r'^.+? (.*) (\d+) ([+-]\d+).*$')
 
 def parse_actor_and_date(line):
     """
@@ -48,11 +48,11 @@ def parse_actor_and_date(line):
      author Tom Preston-Werner <tom@mojombo.com> 1191999972 -0700
     
     Returns
-        [Actor, int_seconds_since_epoch]
+        [Actor, int_seconds_since_epoch, int_timezone_offset]
     """
     m = _re_actor_epoch.search(line)
-    actor, epoch = m.groups()
-    return (Actor._from_string(actor), int(epoch))
+    actor, epoch, offset = m.groups()
+    return (Actor._from_string(actor), int(epoch), -int(float(offset)/100*3600))
     
     
     
diff --git a/test/git/test_commit.py b/test/git/test_commit.py
@@ -18,6 +18,7 @@ def test_bake(self):
         assert_equal("byronimo@gmail.com", commit.author.email)
         assert commit.author == commit.committer
         assert isinstance(commit.authored_date, int) and isinstance(commit.committed_date, int)
+        assert isinstance(commit.author_tz_offset, int) and isinstance(commit.committer_tz_offset, int)
         assert commit.message == "Added missing information to docstrings of commit and stats module"
 
 
@@ -46,6 +47,8 @@ def check_entries(d):
         assert commit.committer == michael
         assert commit.authored_date == 1210193388
         assert commit.committed_date == 1210193388
+        assert commit.author_tz_offset == 14400, commit.author_tz_offset
+        assert commit.committer_tz_offset == 14400, commit.committer_tz_offset
         assert commit.message == "initial project"
         
     def test_traversal(self):
