Mercurial > kallithea
changeset 6912:e025863fae9c
diffs: avoid extra copy of diff when trying to create a lazy diff line iterator
Make the generator actually be lazy and avoid creating an extra full copy of
all the lines in a file diff at once.
And consistently call it diff_lines.
| author | Mads Kiilerich <mads@kiilerich.com> |
|---|---|
| date | Tue, 03 Oct 2017 00:14:40 +0200 |
| parents | 7dbe020e93fe |
| children | 24a9bec8138c |
| files | kallithea/lib/diffs.py |
| diffstat | 1 files changed, 9 insertions(+), 11 deletions(-) [+] |
line wrap: on
line diff
--- a/kallithea/lib/diffs.py Tue Oct 03 00:14:40 2017 +0200 +++ b/kallithea/lib/diffs.py Tue Oct 03 00:14:40 2017 +0200 @@ -29,8 +29,6 @@ import difflib import logging -from itertools import imap - from tg.i18n import ugettext as _ from kallithea.lib.vcs.exceptions import VCSError @@ -306,8 +304,8 @@ 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|.+$', rest)) # don't split on \r as str.splitlines do - return meta_info, difflines + diff_lines = (self._escaper(m.group(0)) for m in re.finditer(r'.*\n|.+$', rest)) # don't split on \r as str.splitlines do + return meta_info, diff_lines def _parse_gitdiff(self, inline_diff=True): """Parse self._diff and return a list of dicts with meta info and chunks for each file. @@ -321,7 +319,7 @@ starts.append(len(self._diff)) for start, end in zip(starts, starts[1:]): - head, diff = self._get_header(buffer(self._diff, start, end - start)) + head, diff_lines = self._get_header(buffer(self._diff, start, end - start)) op = None stats = { @@ -381,7 +379,7 @@ # a real non-binary diff if head['a_file'] or head['b_file']: try: - chunks, added, deleted = self._parse_lines(diff) + chunks, added, deleted = self._parse_lines(diff_lines) stats['binary'] = False stats['added'] = added stats['deleted'] = deleted @@ -459,7 +457,7 @@ return diff_container(_files) - def _parse_lines(self, diff): + def _parse_lines(self, diff_lines): """ Given an iterator of diff body lines, parse them and return a dict per line and added/removed totals. @@ -469,7 +467,7 @@ try: chunks = [] - line = diff.next() + line = diff_lines.next() while True: lines = [] @@ -500,7 +498,7 @@ 'line': line, }) - line = diff.next() + line = diff_lines.next() while old_line < old_end or new_line < new_end: if not line: @@ -533,7 +531,7 @@ 'line': line[1:], }) - line = diff.next() + line = diff_lines.next() if self._newline_marker.match(line): # we need to append to lines, since this is not @@ -544,7 +542,7 @@ 'action': 'context', 'line': line, }) - line = diff.next() + line = diff_lines.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: