Implemented main rev-parsing, including long hexshas, tags and refs. Short Shas still to be done

Author: Byron

lib/git/repo.py

@@ -71,7 +71,8 @@ class Repo(object):
 	# precompiled regex
 	re_whitespace = re.compile(r'\s+')
 	re_hexsha_only = re.compile('^[0-9A-Fa-f]{40}$')
-	re_hexsha_shortened = re.compile('^[0-9A-Fa-f]{7:40}$')
+	re_hexsha_shortened = re.compile('^[0-9A-Fa-f]{7,40}$')
+	re_hexsha_domain = re.compile('^[0-9A-Fa-f]{1,40}$')
 	re_author_committer_start = re.compile(r'^(author|committer)')
 	re_tab_full_line = re.compile(r'^\t(.*)$')
 
@@ -724,7 +725,7 @@ def rev_parse(self, rev):
 		def name_to_object(name):
 			hexsha = None
 
-			# is it a hexsha ?
+			# is it a hexsha ? Try the most common ones, which is 7 to 40
 			if self.re_hexsha_shortened.match(name):
 				if len(name) != 40:
 					# find long sha for short sha
@@ -744,6 +745,11 @@ def name_to_object(name):
 
 			# tried everything ? fail
 			if hexsha is None:
+				# it could also be a very short ( less than 7 ) hexsha, which 
+				# wasnt tested in the first run
+				if len(name) < 7 and self.re_hexsha_domain.match(name):
+					raise NotImplementedError()	
+				# END try short name 
 				raise BadObject(name)
 			# END assert hexsha was found
 
@@ -806,7 +812,7 @@ def to_commit(obj):
 					# END exception handling
 				elif output_type in ('', 'blob'):
 					if obj.type == 'tag':
-						obj = deref_tag(tag)
+						obj = deref_tag(obj)
 					else:
 						# cannot do anything for non-tags
 						pass
@@ -815,8 +821,9 @@ def to_commit(obj):
 					raise ValueError("Invalid output type: %s ( in %s )"  % (output_type, rev))
 				# END handle output type
 
-				if obj.type != output_type:
-					raise ValueError("Could not accomodate requested object type %s, got %s" % (output_type, obj.type))
+				# empty output types don't require any specific type, its just about dereferencing tags
+				if output_type and obj.type != output_type:
+					raise ValueError("Could not accomodate requested object type %r, got %s" % (output_type, obj.type))
 				# END verify ouput type
 
 				start = end+1					# skip brace
@@ -849,12 +856,13 @@ def to_commit(obj):
 			parsed_to = start
 			# handle hiererarchy walk
 			try:
-				obj = to_commit(obj)
 				if token == "~":
+					obj = to_commit(obj)
 					for item in xrange(num):
 						obj = obj.parents[0]
 					# END for each history item to walk
 				elif token == "^":
+					obj = to_commit(obj)
 					# must be n'th parent
 					if num:
 						obj = obj.parents[num-1]

test/git/test_repo.py

@@ -381,6 +381,9 @@ def mktiny():
 	def _assert_rev_parse_types(self, name, rev_obj):
 		rev_parse = self.rorepo.rev_parse
 
+		if rev_obj.type == 'tag':
+			rev_obj = rev_obj.object
+		
 		# tree and blob type
 		obj = rev_parse(name + '^{tree}')
 		assert obj == rev_obj.tree
@@ -439,9 +442,6 @@ def _assert_rev_parse(self, name):
 	def test_rev_parse(self):
 		rev_parse = self.rorepo.rev_parse
 
-		# it works with tags !
-		self._assert_rev_parse('0.1.4')
-		
 		# start from reference
 		num_resolved = 0
 		for ref in Reference.iter_items(self.rorepo):
@@ -461,29 +461,53 @@ def test_rev_parse(self):
 		# END for each reference
 		assert num_resolved
 
+		# it works with tags !
+		tag = self._assert_rev_parse('0.1.4')
+		assert tag.type == 'tag'
+		
 		# try full sha directly ( including type conversion )
+		assert tag.object == rev_parse(tag.object.hexsha)
+		self._assert_rev_parse_types(tag.object.hexsha, tag.object)
 
 
 		# multiple tree types result in the same tree: HEAD^{tree}^{tree}:CHANGES
+		rev = '0.1.4^{tree}^{tree}'
+		assert rev_parse(rev) == tag.object.tree
+		assert rev_parse(rev+':CHANGES') == tag.object.tree['CHANGES']
+		
 
 		# try to get parents from first revision - it should fail as no such revision
 		# exists
+		first_rev = "33ebe7acec14b25c5f84f35a664803fcab2f7781"
+		commit = rev_parse(first_rev)
+		assert len(commit.parents) == 0
+		assert commit.hexsha == first_rev
+		self.failUnlessRaises(BadObject, rev_parse, first_rev+"~")
+		self.failUnlessRaises(BadObject, rev_parse, first_rev+"^")
+		
+		# short SHA1
+		commit2 = rev_parse(first_rev[:20])
+		assert commit2 == commit
+		commit2 = rev_parse(first_rev[:5])
+		assert commit2 == commit
+		
 
 		# todo: dereference tag into a blob 0.1.7^{blob} - quite a special one
+		# needs a tag which points to a blob
 
-		# dereference tag using ^{} notation
 
-		# ref^0 returns commit being pointed to, same with ref~0
+		# ref^0 returns commit being pointed to, same with ref~0, and ^{}
 		tag = rev_parse('0.1.4')
-		for token in ('~^'):
-			assert tag.object == rev_parse('0.1.4%s0' % token)
+		for token in (('~0', '^0', '^{}')):
+			assert tag.object == rev_parse('0.1.4%s' % token)
 		# END handle multiple tokens
 
 		# missing closing brace commit^{tree
+		self.failUnlessRaises(ValueError, rev_parse, '0.1.4^{tree')
 
 		# missing starting brace
+		self.failUnlessRaises(ValueError, rev_parse, '0.1.4^tree}')
 
-		# not enough parents ^10
 
 		# cannot handle rev-log for now 
 		self.failUnlessRaises(ValueError, rev_parse, "hi@there")