exonerator/gerrit_urls.py

# Copyright 2017 The Chromium OS Authors. All rights reserved.
# Use of this source code is governed by a BSD-style license that can be
# found in the LICENSE file.
"""Utility for parsing Gerrit URLs into CL numbers."""

from __future__ import absolute_import
from __future__ import division
from __future__ import print_function

import re

from chromite.lib import clactions


_URL_PREFIX = r'(?:https?://)?'

_EXTERNAL_PATTERNS = tuple(
    re.compile(_URL_PREFIX + url_pattern) for url_pattern in (
        r'crrev\.com/c.*/(\d+)',
        r'crosreview\.com/[^i].*/(\d+)',
        r'crosreview\.com/(\d+)',
        r'chromium-review\.googlesource\.com.*/(\d+)')
)

_INTERNAL_PATTERNS = tuple(
    re.compile(_URL_PREFIX + url_pattern) for url_pattern in (
        r'crrev.com/i.*/(\d+)',
        r'crosreview.com/i.*/(\d+)',
        r'chrome-internal-review\.googlesource\.com.*/(\d+)')
)


def CLFromBlameURL(blame_url):
  """Parses an annotation's blame url into a GerritChangeTuple.

  External URLs are of the form (possibly prefixed by http(s?)://)
      crrev.com/c/.../<gerrit CL number>
      crosreview.com/.../<gerrit CL number>
      chromium-review.googlesource.com/.../<gerrit CL number>

  Internal URLs are of the form:
      crrev.com/i/.../<gerrit CL number>
      crosreview.com/i/.../<gerrit CL number>
      chrome-internal-review.googlesource.com/.../<gerrit CL number>

  E.g. _CLFromBlameURL('crrev.com/i/anything/12345') == '*12345'
          _CLFromBlameURL('crrev.com/anything/54321') == '54321'

  Args:
      blame_url: The URL of the CL

  Returns:
      The matched CL, or None.
  """
  for pattern in _EXTERNAL_PATTERNS:
    m = pattern.match(blame_url)
    if m:
      return clactions.GerritChangeTuple(int(m.group(1)), internal=False)
  for pattern in _INTERNAL_PATTERNS:
    m = pattern.match(blame_url)
    if m:
      return clactions.GerritChangeTuple(int(m.group(1)), internal=True)
  raise ValueError("Malformed Blame URL; can't parse a gerrit CL.")