actor: added __eq__, __ne__ and __hash__ methods including simple test

commit: Fixed long-standing issue during message parsing that would fail to parse properly in case we were created from data. Also it would strip white space from the messages although it shouldn't

Author: Byron

lib/git/actor.py

@@ -18,6 +18,15 @@ def __init__(self, name, email):
 		self.name = name
 		self.email = email
 
+	def __eq__(self, other):
+		return self.name == other.name and self.email == other.email
+		
+	def __ne__(self, other):
+		return not (self == other)
+		
+	def __hash__(self):
+		return hash((self.name, self.email))
+
 	def __str__(self):
 		return self.name
 

lib/git/objects/commit.py

@@ -83,7 +83,7 @@ def _set_cache_(self, attr):
 			# prepare our data lines to match rev-list
 			data_lines = self.data.splitlines()
 			data_lines.insert(0, "commit %s" % self.id)
-			temp = self._iter_from_process_or_stream(self.repo, iter(data_lines)).next()
+			temp = self._iter_from_process_or_stream(self.repo, iter(data_lines), False).next()
 			self.parents = temp.parents
 			self.tree = temp.tree
 			self.author = temp.author
@@ -111,7 +111,8 @@ def count(self, paths='', **kwargs):
 			to commits actually containing the paths
 
 		``kwargs``
-			Additional options to be passed to git-rev-list
+			Additional options to be passed to git-rev-list. They must not alter
+			the ouput style of the command, or parsing will yield incorrect results
 		Returns
 			int
 		"""
@@ -144,9 +145,8 @@ def iter_items(cls, repo, rev, paths='', **kwargs):
 		options = {'pretty': 'raw', 'as_process' : True }
 		options.update(kwargs)
 
-		# the test system might confront us with string values - 
 		proc = repo.git.rev_list(rev, '--', paths, **options)
-		return cls._iter_from_process_or_stream(repo, proc)
+		return cls._iter_from_process_or_stream(repo, proc, True)
 
 	def iter_parents(self, paths='', **kwargs):
 		"""
@@ -191,7 +191,7 @@ def stats(self):
 		return stats.Stats._list_from_string(self.repo, text)
 
 	@classmethod
-	def _iter_from_process_or_stream(cls, repo, proc_or_stream):
+	def _iter_from_process_or_stream(cls, repo, proc_or_stream, from_rev_list):
 		"""
 		Parse out commit information into a list of Commit objects
 
@@ -201,17 +201,20 @@ def _iter_from_process_or_stream(cls, repo, proc_or_stream):
 		``proc``
 			git-rev-list process instance (raw format)
 
+		``from_rev_list``
+			If True, the stream was created by rev-list in which case we parse 
+			the message differently
 		Returns
 			iterator returning Commit objects
 		"""
 		stream = proc_or_stream
 		if not hasattr(stream,'next'):
 			stream = proc_or_stream.stdout
 
-		
 		for line in stream:
-			id = line.split()[1]
-			assert line.split()[0] == "commit"
+			commit_tokens = line.split() 
+			id = commit_tokens[1]
+			assert commit_tokens[0] == "commit"
 			tree = stream.next().split()[1]
 
 			parents = []
@@ -231,13 +234,20 @@ def _iter_from_process_or_stream(cls, repo, proc_or_stream):
 			stream.next()
 
 			message_lines = []
-			next_line = None
-			for msg_line in stream:
-				if not msg_line.startswith('    '):
-					break
-				# END abort message reading 
-				message_lines.append(msg_line.strip())
-			# END while there are message lines
+			if from_rev_list:
+				for msg_line in stream:
+					if not msg_line.startswith('    '):
+						# and forget about this empty marker
+						break
+					# END abort message reading 
+					# strip leading 4 spaces
+					message_lines.append(msg_line[4:])
+				# END while there are message lines
+			else:
+				# a stream from our data simply gives us the plain message
+				for msg_line in stream:
+					message_lines.append(msg_line)
+			# END message parsing
 			message = '\n'.join(message_lines)
 
 			yield Commit(repo, id=id, parents=tuple(parents), tree=tree, author=author, authored_date=authored_date,

test/git/test_actor.py

@@ -13,6 +13,14 @@ def test_from_string_should_separate_name_and_email(self):
 		a = Actor._from_string("Michael Trier <mtrier@example.com>")
 		assert_equal("Michael Trier", a.name)
 		assert_equal("mtrier@example.com", a.email)
+		
+		# base type capabilities
+		assert a == a
+		assert not ( a != a )
+		m = set()
+		m.add(a)
+		m.add(a)
+		assert len(m) == 1
 
 	def test_from_string_should_handle_just_name(self):
 		a = Actor._from_string("Michael Trier")

test/git/test_commit.py

@@ -16,6 +16,9 @@ def test_bake(self):
 
 		assert_equal("Sebastian Thiel", commit.author.name)
 		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 commit.message == "Added missing information to docstrings of commit and stats module"
 
 
 	def test_stats(self):
@@ -37,6 +40,14 @@ def check_entries(d):
 			check_entries(d)
 		# END for each stated file
 
+		# assure data is parsed properly
+		michael = Actor._from_string("Michael Trier <mtrier@gmail.com>")
+		assert commit.author == michael
+		assert commit.committer == michael
+		assert commit.authored_date == 1210193388
+		assert commit.committed_date == 1210193388
+		assert commit.message == "initial project"
+		
 	@patch_object(Git, '_call_process')
 	def test_rev_list_bisect_all(self, git):
 		"""
@@ -52,7 +63,7 @@ def test_rev_list_bisect_all(self, git):
 									  bisect_all=True)
 		assert_true(git.called)
 
-		commits = Commit._iter_from_process_or_stream(self.rorepo, ListProcessAdapter(revs))
+		commits = Commit._iter_from_process_or_stream(self.rorepo, ListProcessAdapter(revs), True)
 		expected_ids = (
 			'cf37099ea8d1d8c7fbf9b6d12d7ec0249d3acb8b',
 			'33ebe7acec14b25c5f84f35a664803fcab2f7781',