h.flash(msg, category='error')

            raise HTTPNotFound()

        c.changes = OrderedDict()

        c.lines_added = 0  # count of lines added

        c.lines_deleted = 0  # count of lines removes

        c.changeset_statuses = ChangesetStatus.STATUSES

        comments = dict()

        c.statuses = []

        c.inline_comments = []

        c.inline_cnt = 0

        # Iterate over ranges (default changeset view is always one changeset)

        for changeset in c.cs_ranges:

            if method == 'show':

                c.statuses.extend([ChangesetStatusModel().get_status(

                            c.db_repo.repo_id, changeset.raw_id)])

                # Changeset comments

                comments.update((com.comment_id, com)

                                for com in ChangesetCommentsModel()

                                .get_comments(c.db_repo.repo_id,

                                              revision=changeset.raw_id))

                # Status change comments - mostly from pull requests

                comments.update((st.comment_id, st.comment)

                                for st in ChangesetStatusModel()

                                .get_statuses(c.db_repo.repo_id,

                                              changeset.raw_id, with_revisions=True)

                                if st.comment_id is not None)

                inlines = ChangesetCommentsModel() \

                            .get_inline_comments(c.db_repo.repo_id,

                                                 revision=changeset.raw_id)

                c.inline_comments.extend(inlines)

            cs2 = changeset.raw_id

            cs1 = changeset.parents[0].raw_id if changeset.parents else EmptyChangeset().raw_id

            context_lcl = get_line_ctx('', request.GET)

            ign_whitespace_lcl = get_ignore_ws('', request.GET)

            _diff = c.db_repo_scm_instance.get_diff(cs1, cs2,

                ignore_whitespace=ign_whitespace_lcl, context=context_lcl)

            diff_limit = None if c.fulldiff else self.cut_off_limit

            diff_processor = diffs.DiffProcessor(_diff,

                                                 vcs=c.db_repo_scm_instance.alias,

                                                 diff_limit=diff_limit)

            file_diff_data = []

            if method == 'show':

                _parsed = diff_processor.prepare()

                c.limited_diff = False

                if isinstance(_parsed, LimitedDiffContainer):

                    c.limited_diff = True

                for f in _parsed:

                    st = f['stats']

                    c.lines_added += st['added']

                    c.lines_deleted += st['deleted']

                    filename = f['filename']

                    fid = h.FID(changeset.raw_id, filename)

                    url_fid = h.FID('', filename)

                    diff = diff_processor.as_html(enable_comments=enable_comments,

                                                  parsed_lines=[f])

                    file_diff_data.append((fid, url_fid, f['operation'], f['old_filename'], filename, diff, st))

            else:

                # downloads/raw we only need RAW diff nothing else

                diff = diff_processor.as_raw()

                file_diff_data.append(('', None, None, None, diff, None))

            c.changes[changeset.raw_id] = (cs1, cs2, file_diff_data)

        # sort comments in creation order

        c.comments = [com for com_id, com in sorted(comments.items())]

        # count inline comments

        for __, lines in c.inline_comments:

            for comments in lines.values():

                c.inline_cnt += len(comments)

        if len(c.cs_ranges) == 1:

            c.changeset = c.cs_ranges[0]

            c.parent_tmpl = ''.join(['# Parent  %s\n' % x.raw_id

                                     for x in c.changeset.parents])

        if method == 'download':

            response.content_type = 'text/plain'

            response.content_disposition = 'attachment; filename=%s.diff' \

                                            % revision[:12]

            return diff

        elif method == 'patch':

            response.content_type = 'text/plain'

            c.diff = safe_unicode(diff)

            return render('changeset/patch_changeset.html')

        elif method == 'raw':

            response.content_type = 'text/plain'

            return diff

        elif method == 'show':

        c.cs_statuses = c.cs_repo.statuses(raw_ids)

        revs = [ctx.revision for ctx in reversed(c.cs_ranges)]

        c.jsdata = graph_data(c.cs_repo.scm_instance, revs)

        if partial:

            return render('compare/compare_cs.html')

        org_repo = c.a_repo

        other_repo = c.cs_repo

        if merge:

            rev1 = msg = None

            if not c.cs_ranges:

                msg = _('Cannot show empty diff')

            elif not c.ancestors:

                msg = _('No ancestor found for merge diff')

            elif len(c.ancestors) == 1:

                rev1 = c.ancestors[0]

            else:

                msg = _('Multiple merge ancestors found for merge compare')

            if rev1 is None:

                h.flash(msg, category='error')

                log.error(msg)

                raise HTTPNotFound

            # case we want a simple diff without incoming changesets,

            # previewing what will be merged.

            # Make the diff on the other repo (which is known to have other_rev)

            log.debug('Using ancestor %s as rev1 instead of %s',

                      rev1, c.a_rev)

            org_repo = other_repo

        else: # comparing tips, not necessarily linearly related

            if org_repo != other_repo:

                # TODO: we could do this by using hg unionrepo

                log.error('cannot compare across repos %s and %s', org_repo, other_repo)

                h.flash(_('Cannot compare repositories without using common ancestor'), category='error')

                raise HTTPBadRequest

            rev1 = c.a_rev

        diff_limit = None if fulldiff else self.cut_off_limit

        log.debug('running diff between %s and %s in %s',

                  rev1, c.cs_rev, org_repo.scm_instance.path)

        txtdiff = org_repo.scm_instance.get_diff(rev1=rev1, rev2=c.cs_rev,

                                      ignore_whitespace=ignore_whitespace,

                                      context=line_context)

        _parsed = diff_processor.prepare()

        c.limited_diff = False

        if isinstance(_parsed, LimitedDiffContainer):

            c.limited_diff = True

        c.file_diff_data = []

        c.lines_added = 0

        c.lines_deleted = 0

        for f in _parsed:

            st = f['stats']

            c.lines_added += st['added']

            c.lines_deleted += st['deleted']

            filename = f['filename']

            fid = h.FID('', filename)

            diff = diff_processor.as_html(enable_comments=False,

                                          parsed_lines=[f])

            c.file_diff_data.append((fid, None, f['operation'], f['old_filename'], filename, diff, st))

        return render('compare/compare_diff.html')

                _url = url('files_annotate_home', repo_name=c.repo_name,

                           revision=diff1, f_path=c.f_path)

            else:

                _url = url('files_home', repo_name=c.repo_name,

                           revision=diff1, f_path=c.f_path)

            raise HTTPFound(location=_url)

        try:

            if diff1 not in ['', None, 'None', '0' * 12, '0' * 40]:

                c.changeset_1 = c.db_repo_scm_instance.get_changeset(diff1)

                try:

                    node1 = c.changeset_1.get_node(f_path)

                    if node1.is_dir():

                        raise NodeError('%s path is a %s not a file'

                                        % (node1, type(node1)))

                except NodeDoesNotExistError:

                    c.changeset_1 = EmptyChangeset(cs=diff1,

                                                   revision=c.changeset_1.revision,

                                                   repo=c.db_repo_scm_instance)

                    node1 = FileNode(f_path, '', changeset=c.changeset_1)

            else:

                c.changeset_1 = EmptyChangeset(repo=c.db_repo_scm_instance)

                node1 = FileNode(f_path, '', changeset=c.changeset_1)

            if diff2 not in ['', None, 'None', '0' * 12, '0' * 40]:

                c.changeset_2 = c.db_repo_scm_instance.get_changeset(diff2)

                try:

                    node2 = c.changeset_2.get_node(f_path)

                    if node2.is_dir():

                        raise NodeError('%s path is a %s not a file'

                                        % (node2, type(node2)))

                except NodeDoesNotExistError:

                    c.changeset_2 = EmptyChangeset(cs=diff2,

                                                   revision=c.changeset_2.revision,

                                                   repo=c.db_repo_scm_instance)

                    node2 = FileNode(f_path, '', changeset=c.changeset_2)

            else:

                c.changeset_2 = EmptyChangeset(repo=c.db_repo_scm_instance)

                node2 = FileNode(f_path, '', changeset=c.changeset_2)

        except (RepositoryError, NodeError):

            log.error(traceback.format_exc())

            raise HTTPFound(location=url('files_home', repo_name=c.repo_name,

                                f_path=f_path))

        if c.action == 'download':

            _diff = diffs.get_gitdiff(node1, node2,

                                      ignore_whitespace=ignore_whitespace,

                                      context=line_context)

            diff_name = '%s_vs_%s.diff' % (diff1, diff2)

            response.content_type = 'text/plain'

            response.content_disposition = (

                'attachment; filename=%s' % diff_name

            )

            return diff.as_raw()

        elif c.action == 'raw':

            _diff = diffs.get_gitdiff(node1, node2,

                                      ignore_whitespace=ignore_whitespace,

                                      context=line_context)

            response.content_type = 'text/plain'

            return diff.as_raw()

        else:

            fid = h.FID(diff2, node2.path)

            line_context_lcl = get_line_ctx(fid, request.GET)

            ign_whitespace_lcl = get_ignore_ws(fid, request.GET)

            diff_limit = None if fulldiff else self.cut_off_limit

            c.a_rev, c.cs_rev, a_path, diff, st, op = diffs.wrapped_diff(filenode_old=node1,

                                         filenode_new=node2,

                                         diff_limit=diff_limit,

                                         ignore_whitespace=ign_whitespace_lcl,

                                         line_context=line_context_lcl,

                                         enable_comments=False)

            c.file_diff_data = [(fid, fid, op, a_path, node2.path, diff, st)]

            return render('files/file_diff.html')

    @LoginRequired()

    @HasRepoPermissionLevelDecorator('read')

    def diff_2way(self, repo_name, f_path):

        diff1 = request.GET.get('diff1', '')

        diff2 = request.GET.get('diff2', '')

        try:

            if diff1 not in ['', None, 'None', '0' * 12, '0' * 40]:

                c.changeset_1 = c.db_repo_scm_instance.get_changeset(diff1)

                try:

                    node1 = c.changeset_1.get_node(f_path)

                    if node1.is_dir():

                        raise NodeError('%s path is a %s not a file'

                                        % (node1, type(node1)))

                except NodeDoesNotExistError:

                    c.changeset_1 = EmptyChangeset(cs=diff1,

                                                   revision=c.changeset_1.revision,

                                                   repo=c.db_repo_scm_instance)

                    node1 = FileNode(f_path, '', changeset=c.changeset_1)

            else:

                c.changeset_1 = EmptyChangeset(repo=c.db_repo_scm_instance)

                node1 = FileNode(f_path, '', changeset=c.changeset_1)

            if diff2 not in ['', None, 'None', '0' * 12, '0' * 40]:

                c.changeset_2 = c.db_repo_scm_instance.get_changeset(diff2)

                try:

                    node2 = c.changeset_2.get_node(f_path)

                    if node2.is_dir():

                        raise NodeError('%s path is a %s not a file'

                                        % (node2, type(node2)))

                                               avail_revs, revs[0], targethead))

                        c.update_msg = _('The following additional changes are available on %s:') % c.cs_branch_name

                    else:

                        show = set()

                        avail_revs = set() # drop revs[0]

                        c.update_msg = _('No additional changesets found for iterating on this pull request.')

                    # TODO: handle branch heads that not are tip-most

                    brevs = org_scm_instance._repo.revs('%s - %ld - %s', c.cs_branch_name, avail_revs, revs[0])

                    if brevs:

                        # also show changesets that are on branch but neither ancestors nor descendants

                        show.update(org_scm_instance._repo.revs('::%ld - ::%ld - ::%s', brevs, avail_revs, c.a_branch_name))

                        show.add(revs[0]) # make sure graph shows this so we can see how they relate

                        c.update_msg_other = _('Note: Branch %s has another head: %s.') % (c.cs_branch_name,

                            h.short_id(org_scm_instance.get_changeset((max(brevs))).raw_id))

                    avail_show = sorted(show, reverse=True)

            elif org_scm_instance.alias == 'git':

                c.cs_repo.scm_instance.get_changeset(c.cs_rev) # check it exists - raise ChangesetDoesNotExistError if not

                c.update_msg = _("Git pull requests don't support iterating yet.")

        except ChangesetDoesNotExistError:

            c.update_msg = _('Error: some changesets not found when displaying pull request from %s.') % c.cs_rev

        c.avail_revs = avail_revs

        c.avail_cs = [org_scm_instance.get_changeset(r) for r in avail_show]

        c.avail_jsdata = graph_data(org_scm_instance, avail_show)

        raw_ids = [x.raw_id for x in c.cs_ranges]

        c.cs_comments = c.cs_repo.get_comments(raw_ids)

        c.cs_statuses = c.cs_repo.statuses(raw_ids)

        ignore_whitespace = request.GET.get('ignorews') == '1'

        line_context = safe_int(request.GET.get('context'), 3)

        c.ignorews_url = _ignorews_url

        c.context_url = _context_url

        fulldiff = request.GET.get('fulldiff')

        diff_limit = None if fulldiff else self.cut_off_limit

        # we swap org/other ref since we run a simple diff on one repo

        log.debug('running diff between %s and %s in %s',

                  c.a_rev, c.cs_rev, org_scm_instance.path)

        try:

            txtdiff = org_scm_instance.get_diff(rev1=safe_str(c.a_rev), rev2=safe_str(c.cs_rev),

                                                ignore_whitespace=ignore_whitespace,

                                                context=line_context)

        except ChangesetDoesNotExistError:

            txtdiff = _("The diff can't be shown - the PR revisions could not be found.")

        _parsed = diff_processor.prepare()

        c.limited_diff = False

        if isinstance(_parsed, LimitedDiffContainer):

            c.limited_diff = True

        c.file_diff_data = []

        c.lines_added = 0

        c.lines_deleted = 0

        for f in _parsed:

            st = f['stats']

            c.lines_added += st['added']

            c.lines_deleted += st['deleted']

            filename = f['filename']

            fid = h.FID('', filename)

            diff = diff_processor.as_html(enable_comments=True,

                                          parsed_lines=[f])

            c.file_diff_data.append((fid, None, f['operation'], f['old_filename'], filename, diff, st))

        # inline comments

        c.inline_cnt = 0

        c.inline_comments = cc_model.get_inline_comments(

                                c.db_repo.repo_id,

                                pull_request=pull_request_id)

        # count inline comments

        for __, lines in c.inline_comments:

            for comments in lines.values():

                c.inline_cnt += len(comments)

        # comments

        c.comments = cc_model.get_comments(c.db_repo.repo_id, pull_request=pull_request_id)

        # (badly named) pull-request status calculation based on reviewer votes

        (c.pull_request_reviewers,

         c.pull_request_pending_reviewers,

         c.current_voting_result,

         ) = cs_model.calculate_pull_request_result(c.pull_request)

        c.changeset_statuses = ChangesetStatus.STATUSES

        c.is_ajax_preview = False

        c.ancestors = None # [c.a_rev] ... but that is shown in an other way

        return render('/pullrequests/pullrequest_show.html')

    @LoginRequired()

    @NotAnonymous()

    @HasRepoPermissionLevelDecorator('read')

    @jsonify

    def comment(self, repo_name, pull_request_id):

from itertools import imap

from tg.i18n import ugettext as _

from kallithea.lib.vcs.exceptions import VCSError

from kallithea.lib.vcs.nodes import FileNode, SubModuleNode

from kallithea.lib.vcs.backends.base import EmptyChangeset

from kallithea.lib.helpers import escape

from kallithea.lib.utils2 import safe_unicode

log = logging.getLogger(__name__)

def wrap_to_table(html):

    """Given a string with html, return it wrapped in a table, similar to what

    DiffProcessor returns."""

    return '''\

              <table class="code-difftable">

                <tr class="line no-comment">

                <td class="lineno new"></td>

                <td class="code no-comment"><pre>%s</pre></td>

                </tr>

              </table>''' % html

def wrapped_diff(filenode_old, filenode_new, diff_limit=None,

                ignore_whitespace=True, line_context=3,

                enable_comments=False):

    """

    Returns a file diff wrapped into a table.

    Checks for diff_limit and presents a message if the diff is too big.

    """

    if filenode_old is None:

        filenode_old = FileNode(filenode_new.path, '', EmptyChangeset())

    op = None

    a_path = filenode_old.path # default, might be overriden by actual rename in diff

    if filenode_old.is_binary or filenode_new.is_binary:

        diff = wrap_to_table(_('Binary file'))

        stats = (0, 0)

    elif diff_limit != -1 and (

            diff_limit is None or

            (filenode_old.size < diff_limit and filenode_new.size < diff_limit)):

        f_gitdiff = get_gitdiff(filenode_old, filenode_new,

                                ignore_whitespace=ignore_whitespace,

                                context=line_context)

        _parsed = diff_processor.prepare()

        if _parsed: # there should be exactly one element, for the specified file

            f = _parsed[0]

            op = f['operation']

            a_path = f['old_filename']

        diff = diff_processor.as_html(enable_comments=enable_comments)

        stats = diff_processor.stat()

    else:

        diff = wrap_to_table(_('Changeset was too big and was cut off, use '

                               'diff menu to display this diff'))

        stats = (0, 0)

    if not diff:

        submodules = filter(lambda o: isinstance(o, SubModuleNode),

                            [filenode_new, filenode_old])

        if submodules:

            diff = wrap_to_table(escape('Submodule %r' % submodules[0]))

        else:

            diff = wrap_to_table(_('No changes detected'))

    cs1 = filenode_old.changeset.raw_id

    cs2 = filenode_new.changeset.raw_id

    return cs1, cs2, a_path, diff, stats, op

def get_gitdiff(filenode_old, filenode_new, ignore_whitespace=True, context=3):

    """

    Returns git style diff between given ``filenode_old`` and ``filenode_new``.

    """

    # make sure we pass in default context

    context = context or 3

    submodules = filter(lambda o: isinstance(o, SubModuleNode),

                        [filenode_new, filenode_old])

    if submodules:

        return ''

    for filenode in (filenode_old, filenode_new):

        if not isinstance(filenode, FileNode):

            raise VCSError("Given object should be FileNode object, not %s"

                % filenode.__class__)

    repo = filenode_new.changeset.repository

    old_raw_id = getattr(filenode_old.changeset, 'raw_id', repo.EMPTY_CHANGESET)

    new_raw_id = getattr(filenode_new.changeset, 'raw_id', repo.EMPTY_CHANGESET)

    Give it a unified or git diff and it returns a list of the files that were

    mentioned in the diff together with a dict of meta information that

    can be used to render it in a HTML template.

    """

    _chunk_re = re.compile(r'^@@ -(\d+)(?:,(\d+))? \+(\d+)(?:,(\d+))? @@(.*)')

    _newline_marker = re.compile(r'^\\ No newline at end of file')

    _git_header_re = re.compile(r"""

        # has already been split on this:

        # ^diff[ ]--git

            [ ]a/(?P<a_path>.+?)[ ]b/(?P<b_path>.+?)\n

        (?:^old[ ]mode[ ](?P<old_mode>\d+)\n

           ^new[ ]mode[ ](?P<new_mode>\d+)(?:\n|$))?

        (?:^similarity[ ]index[ ](?P<similarity_index>\d+)%\n

           ^rename[ ]from[ ](?P<rename_from>.+)\n

           ^rename[ ]to[ ](?P<rename_to>.+)(?:\n|$))?

        (?:^new[ ]file[ ]mode[ ](?P<new_file_mode>.+)(?:\n|$))?

        (?:^deleted[ ]file[ ]mode[ ](?P<deleted_file_mode>.+)(?:\n|$))?

        (?:^index[ ](?P<a_blob_id>[0-9A-Fa-f]+)

            \.\.(?P<b_blob_id>[0-9A-Fa-f]+)[ ]?(?P<b_mode>.+)?(?:\n|$))?

        (?:^(?P<bin_patch>GIT[ ]binary[ ]patch)(?:\n|$))?

        (?:^---[ ](a/(?P<a_file>.+?)|/dev/null)\t?(?:\n|$))?

        (?:^\+\+\+[ ](b/(?P<b_file>.+?)|/dev/null)\t?(?:\n|$))?

    """, re.VERBOSE | re.MULTILINE)

    _hg_header_re = re.compile(r"""

        # has already been split on this:

        # ^diff[ ]--git

            [ ]a/(?P<a_path>.+?)[ ]b/(?P<b_path>.+?)\n

        (?:^old[ ]mode[ ](?P<old_mode>\d+)\n

           ^new[ ]mode[ ](?P<new_mode>\d+)(?:\n|$))?

        (?:^similarity[ ]index[ ](?P<similarity_index>\d+)%(?:\n|$))?

        (?:^rename[ ]from[ ](?P<rename_from>.+)\n

           ^rename[ ]to[ ](?P<rename_to>.+)(?:\n|$))?

        (?:^copy[ ]from[ ](?P<copy_from>.+)\n

           ^copy[ ]to[ ](?P<copy_to>.+)(?:\n|$))?

        (?:^new[ ]file[ ]mode[ ](?P<new_file_mode>.+)(?:\n|$))?

        (?:^deleted[ ]file[ ]mode[ ](?P<deleted_file_mode>.+)(?:\n|$))?

        (?:^index[ ](?P<a_blob_id>[0-9A-Fa-f]+)

            \.\.(?P<b_blob_id>[0-9A-Fa-f]+)[ ]?(?P<b_mode>.+)?(?:\n|$))?

        (?:^(?P<bin_patch>GIT[ ]binary[ ]patch)(?:\n|$))?

        (?:^---[ ](a/(?P<a_file>.+?)|/dev/null)\t?(?:\n|$))?

        (?:^\+\+\+[ ](b/(?P<b_file>.+?)|/dev/null)\t?(?:\n|$))?

    """, re.VERBOSE | re.MULTILINE)

    # Used for inline highlighter word split, must match the substitutions in _escaper

    _token_re = re.compile(r'()(&amp;|&lt;|&gt;|<u>\t</u>|<u class="cr"></u>| <i></i>|\W+?)')

    _escape_re = re.compile(r'(&)|(<)|(>)|(\t)|(\r)|(?<=.)( \n| $)')

        """

        :param diff:   a text in diff format

        :param vcs: type of version control hg or git

        :param diff_limit: define the size of diff that is considered "big"

            based on that parameter cut off will be triggered, set to None

            to show full diff

        """

        if not isinstance(diff, basestring):

            raise Exception('Diff must be a basestring got %s instead' % type(diff))

        self._diff = diff

        self.adds = 0

        self.removes = 0

        # calculate diff size

        self.diff_size = len(diff)

        self.diff_limit = diff_limit

        self.cur_diff_size = 0

        self.parsed = False

        self.parsed_diff = []

        self.vcs = vcs

    def _escaper(self, string):

        """

        Do HTML escaping/markup and check the diff limit

        """

        self.cur_diff_size += len(string)

        # escaper gets iterated on each .next() call and it checks if each

        # parsed line doesn't exceed the diff limit

        if self.diff_limit is not None and self.cur_diff_size > self.diff_limit:

            raise DiffLimitExceeded('Diff Limit Exceeded')

        def substitute(m):

            groups = m.groups()

            if groups[0]:

                return '&'

            if groups[1]:

                return '<'

            if groups[2]:

                return '>'

            if groups[3]:

                return '<u>\t</u>'

            if groups[4]:

                return '<u class="cr"></u>'

            if groups[5]:

                return ' <i></i>'

            assert False

        return self._escape_re.sub(substitute, safe_unicode(string))

        """

        Highlight simple add/remove in two lines given as info dicts. They are

        modified in place and given markup with <del>/<ins>.

        """

        assert old['action'] == 'del'

        assert new['action'] == 'add'

        oldwords = self._token_re.split(old['line'])

        newwords = self._token_re.split(new['line'])

        sequence = difflib.SequenceMatcher(None, oldwords, newwords)

        oldfragments, newfragments = [], []

        for tag, i1, i2, j1, j2 in sequence.get_opcodes():

            oldfrag = ''.join(oldwords[i1:i2])

            newfrag = ''.join(newwords[j1:j2])

            if tag != 'equal':

                if oldfrag:

                    oldfrag = '<del>%s</del>' % oldfrag

                if newfrag:

                    newfrag = '<ins>%s</ins>' % newfrag

            oldfragments.append(oldfrag)

            newfragments.append(newfrag)

        old['line'] = "".join(oldfragments)

        new['line'] = "".join(newfragments)

    def _get_header(self, diff_chunk):

        """

        Parses a Git diff for a single file (header and chunks) and returns a tuple with:

        1. A dict with meta info:

            a_path, b_path, similarity_index, rename_from, rename_to,

            old_mode, new_mode, new_file_mode, deleted_file_mode,

            a_blob_id, b_blob_id, b_mode, a_file, b_file

        2. An iterator yielding lines with simple HTML markup.

        """

        match = None

        if self.vcs == 'git':

            match = self._git_header_re.match(diff_chunk)

        elif self.vcs == 'hg':

            match = self._hg_header_re.match(diff_chunk)

        if match is None:

            raise Exception('diff not recognized as valid %s diff' % self.vcs)

        meta_info = match.groupdict()

        rest = diff_chunk[match.end():]

        if rest and not rest.startswith('@') and not rest.startswith('literal ') and not rest.startswith('delta '):

            raise Exception('cannot parse %s diff header: %r followed by %r' % (self.vcs, diff_chunk[:match.end()], rest[:1000]))

        difflines = imap(self._escaper, re.findall(r'.*\n|.+$', rest)) # don't split on \r as str.splitlines do

        return meta_info, difflines

    def _clean_line(self, line, command):

        """Given a diff line, strip the leading character if it is a plus/minus/context line."""

        if command in ['+', '-', ' ']:

            line = line[1:]

        return line

    def _parse_gitdiff(self, inline_diff=True):

        """Parse self._diff and return a list of dicts with meta info and chunks for each file.

        If diff is truncated, wrap it in LimitedDiffContainer.

        Optionally, do an extra pass and to extra markup of one-liner changes.

        """

        _files = [] # list of dicts with meta info and chunks

        diff_container = lambda arg: arg

        # split the diff in chunks of separate --git a/file b/file chunks

        for raw_diff in ('\n' + self._diff).split('\ndiff --git')[1:]:

            head, diff = self._get_header(raw_diff)

            op = None

            stats = {

                'added': 0,

                'deleted': 0,

                # a way of seeing deleted content could perhaps be nice - but

                # not with the current UI

                chunks = []

            chunks.insert(0, [{

                'old_lineno': '',

                'new_lineno': '',

                'action':     'context',

                'line':       msg,

                } for _op, msg in stats['ops'].iteritems()

                  if _op not in [MOD_FILENODE]])

            _files.append({

                'old_filename':     head['a_path'],

                'filename':         head['b_path'],

                'old_revision':     head['a_blob_id'],

                'new_revision':     head['b_blob_id'],

                'chunks':           chunks,

                'operation':        op,

                'stats':            stats,

            })

        if not inline_diff:

            return diff_container(_files)

        # highlight inline changes when one del is followed by one add

        for diff_data in _files:

            for chunk in diff_data['chunks']:

                lineiter = iter(chunk)

                try:

                    peekline = lineiter.next()

                    while True:

                        # find a first del line

                        while peekline['action'] != 'del':

                            peekline = lineiter.next()

                        delline = peekline

                        peekline = lineiter.next()

                        # if not followed by add, eat all following del lines

                        if peekline['action'] != 'add':

                            while peekline['action'] == 'del':

                                peekline = lineiter.next()

                            continue

                        # found an add - make sure it is the only one

                        addline = peekline

                        try:

                            peekline = lineiter.next()

                        except StopIteration:

                            # add was last line - ok

                            raise

                        if peekline['action'] != 'add':

                            # there was only one add line - ok

                except StopIteration:

                    pass

        return diff_container(_files)

    def _parse_lines(self, diff):

        """

        Given an iterator of diff body lines, parse them and return a dict per

        line and added/removed totals.

        """

        added = deleted = 0

        old_line = old_end = new_line = new_end = None

        try:

            chunks = []

            line = diff.next()

            while True:

                lines = []

                chunks.append(lines)

                match = self._chunk_re.match(line)

                if not match:

                    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 = diff.next()

                while old_line < old_end or new_line < new_end:

                    if not line:

                        raise Exception('error parsing diff - empty line at -%s+%s' % (old_line, new_line))

                    affects_old = affects_new = False

                    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, added, deleted

    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', "_", 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):

        """

        Prepare the passed udiff for HTML rendering. It'll return a list

        of dicts with diff information

        """

        self.parsed = True

        self.parsed_diff = parsed