Changeset - 7b8cbcb927b2
[Not reviewed]
default
0 4 0
Mads Kiilerich - 11 years ago 2014-12-15 13:47:36
madski@unity3d.com
diff: refactor diff parsing - fail on any parse error instead of showing an incorrect diff
4 files changed with 25 insertions and 18 deletions:
0 comments (0 inline, 0 general)
kallithea/lib/diffs.py
Show inline comments
 
@@ -485,132 +485,138 @@ class DiffProcessor(object):
 
            for chunk in diff_data['chunks']:
 
                lineiter = iter(chunk)
 
                try:
 
                    while 1:
 
                        line = lineiter.next()
 
                        if line['action'] not in ['unmod', 'context']:
 
                            nextline = lineiter.next()
 
                            if nextline['action'] in ['unmod', 'context'] or \
 
                               nextline['action'] == line['action']:
 
                                continue
 
                            self.differ(line, nextline)
 
                except StopIteration:
 
                    pass
 

	
 
        return diff_container(_files)
 

	
 
    def _parse_udiff(self, inline_diff=True):
 
        raise NotImplementedError()
 

	
 
    def _parse_lines(self, diff):
 
        """
 
        Parse the diff and return data for the template.
 
        """
 

	
 
        lineiter = iter(diff)
 
        stats = [0, 0]
 
        (old_line, old_end, new_line, new_end) = (None, None, None, None)
 

	
 
        try:
 
            chunks = []
 
            line = lineiter.next()
 
            line = diff.next()
 

	
 
            while line:
 
            while True:
 
                lines = []
 
                chunks.append(lines)
 

	
 
                match = self._chunk_re.match(line)
 

	
 
                if not match:
 
                    break
 
                    raise Exception('error parsing diff @@ line %r' % line)
 

	
 
                gr = match.groups()
 
                (old_line, old_end,
 
                 new_line, new_end) = [int(x or 1) for x in gr[:-1]]
 
                old_line -= 1
 
                new_line -= 1
 

	
 
                context = len(gr) == 5
 
                old_end += old_line
 
                new_end += new_line
 

	
 
                if context:
 
                    # skip context only if it's first line
 
                    if int(gr[0]) > 1:
 
                        lines.append({
 
                            'old_lineno': '...',
 
                            'new_lineno': '...',
 
                            'action':     'context',
 
                            'line':       line,
 
                        })
 

	
 
                line = lineiter.next()
 
                line = diff.next()
 

	
 
                while old_line < old_end or new_line < new_end:
 
                    command = ' '
 
                    if line:
 
                        command = line[0]
 
                    if not line:
 
                        raise Exception('error parsing diff - empty line at -%s+%s' % (old_line, new_line))
 

	
 
                    affects_old = affects_new = False
 

	
 
                    # ignore those if we don't expect them
 
                    if command in '#@':
 
                        continue
 
                    elif command == '+':
 
                    command = line[0]
 
                    if command == '+':
 
                        affects_new = True
 
                        action = 'add'
 
                        stats[0] += 1
 
                    elif command == '-':
 
                        affects_old = True
 
                        action = 'del'
 
                        stats[1] += 1
 
                    else:
 
                    elif command == ' ':
 
                        affects_old = affects_new = True
 
                        action = 'unmod'
 
                    else:
 
                        raise Exception('error parsing diff - unknown command in line %r at -%s+%s' % (line, old_line, new_line))
 

	
 
                    if not self._newline_marker.match(line):
 
                        old_line += affects_old
 
                        new_line += affects_new
 
                        lines.append({
 
                            'old_lineno':   affects_old and old_line or '',
 
                            'new_lineno':   affects_new and new_line or '',
 
                            'action':       action,
 
                            'line':         self._clean_line(line, command)
 
                        })
 

	
 
                    line = lineiter.next()
 
                    line = diff.next()
 

	
 
                    if self._newline_marker.match(line):
 
                        # we need to append to lines, since this is not
 
                        # counted in the line specs of diff
 
                        lines.append({
 
                            'old_lineno':   '...',
 
                            'new_lineno':   '...',
 
                            'action':       'context',
 
                            'line':         self._clean_line(line, command)
 
                        })
 

	
 
                        line = diff.next()
 
                if old_line > old_end:
 
                        raise Exception('error parsing diff - more than %s "-" lines at -%s+%s' % (old_end, old_line, new_line))
 
                if new_line > new_end:
 
                        raise Exception('error parsing diff - more than %s "+" lines at -%s+%s' % (new_end, old_line, new_line))
 
        except StopIteration:
 
            pass
 
        if old_line != old_end or new_line != new_end:
 
            raise Exception('diff processing broken when old %s<>%s or new %s<>%s line %r' % (old_line, old_end, new_line, new_end, line))
 

	
 
        return chunks, stats
 

	
 
    def _safe_id(self, idstring):
 
        """Make a string safe for including in an id attribute.
 

	
 
        The HTML spec says that id attributes 'must begin with
 
        a letter ([A-Za-z]) and may be followed by any number
 
        of letters, digits ([0-9]), hyphens ("-"), underscores
 
        ("_"), colons (":"), and periods (".")'. These regexps
 
        are slightly over-zealous, in that they remove colons
 
        and periods unnecessarily.
 

	
 
        Whitespace is transformed into underscores, and then
 
        anything which is not a hyphen or a character that
 
        matches \w (alphanumerics and underscore) is removed.
 

	
 
        """
 
        # Transform all whitespace to underscore
 
        idstring = re.sub(r'\s', "_", '%s' % idstring)
 
        # Remove everything that is not a hyphen or a member of \w
 
        idstring = re.sub(r'(?!-)\W', "", idstring).lower()
 
        return idstring
 

	
 
    def prepare(self, inline_diff=True):
kallithea/tests/fixtures/diff_with_diff_data.diff
Show inline comments
 
@@ -393,24 +393,25 @@ index e34033e29fa9b3d3366b723beab129cee7
 
+-FOOBAR
 
+\ No newline at end of file
 
+diff --git a/foobar3 b/foobar3
 
+--- a/foobar3
 
++++ b/foobar3
 
+@@ -1,1 +1,3 @@
 
+-foobar3
 
+\ No newline at end of file
 
++FOOBAR
 
++FOOBAR
 
++FOOBAR
 
+''')
 
+
 
+
 
 # For each backend create test case class
 
 for alias in SCM_TESTS:
 
     attrs = {
 
@@ -38,7 +211,6 @@ for alias in SCM_TESTS:
 
     bases = (RepositoryBaseTest, unittest.TestCase)
 
     globals()[cls_name] = type(cls_name, bases, attrs)
 
 
 
-
 
 if __name__ == '__main__':
 
     unittest.main()
 
 
kallithea/tests/fixtures/hg_diff_mod_single_file_and_rename_and_chmod.diff
Show inline comments
 
diff --git a/README.rst b/README
 
old mode 100755
 
new mode 100644
 
rename from README.rst
 
rename to README
 
--- a/README.rst
 
+++ b/README
 
@@ -1,4 +1,7 @@
 
 readme2
 
 line 1
 
  line2
 

	
 
 
 
+line 1
 
+ line2
 
+
 
\ No newline at end of file
 
+
kallithea/tests/models/test_diff_parsers.py
Show inline comments
 
@@ -234,31 +234,31 @@ DIFF_FIXTURES = {
 
        ('file4', 'M',
 
         {'added': 0,
 
          'deleted': 0,
 
          'binary': True,
 
          'ops': {COPIED_FILENODE: 'file copied from file3 to file4',
 
                  CHMOD_FILENODE: 'modified file chmod 100644 => 100755'}}),
 
    ],
 
    'hg_diff_copy_chmod_and_edit_file.diff': [
 
        ('file5', 'M',
 
         {'added': 2,
 
          'deleted': 1,
 
          'binary': False,
 
          'ops': {COPIED_FILENODE: 'file copied from file4 to file5',
 
                  CHMOD_FILENODE: 'modified file chmod 100755 => 100644',
 
                  MOD_FILENODE: 'modified file'}}),
 
    ]
 
}
 

	
 

	
 
class DiffLibTest(BaseTestCase):
 

	
 
    @parameterized.expand([(x,) for x in DIFF_FIXTURES])
 
    def test_diff(self, diff_fixture):
 

	
 
        diff = fixture.load_resource(diff_fixture)
 
        diff = fixture.load_resource(diff_fixture, strip=False)
 

	
 
        diff_proc = DiffProcessor(diff)
 
        diff_proc_d = diff_proc.prepare()
 
        data = [(x['filename'], x['operation'], x['stats']) for x in diff_proc_d]
 
        expected_data = DIFF_FIXTURES[diff_fixture]
 
        self.assertListEqual(expected_data, data)
0 comments (0 inline, 0 general)