Intial checkin of layout test analyzer.

media/tools/layout_tests/test_expectations_history.py

+#!/usr/bin/python
+# Copyright (c) 2011 The Chromium Authors. All rights reserved.
+# Use of this source code is governed by a BSD-style license that can be
+# found in the LICENSE file.
+
+"""A Module for the history of the test expectation file."""
+
+import pysvn

[dennis_jeffrey, 2011/08/25 00:36:46]

If this is not a standard python library but is instead a third-party library,
it should come down below, after the standard library imports.

[imasaki1, 2011/08/25 23:57:03]

Done.

+import re
+import time
+
+from datetime import datetime
+from datetime import timedelta
+
+# Default Webkit SVN location for chromium test expectation file.
+# TODO(imasaki): support multiple test expectations files.

[dennis_jeffrey, 2011/08/25 00:36:46]

'expectations' --> 'expectation'

[imasaki1, 2011/08/25 23:57:03]

Done.

+DEFAULT_TEST_EXPECTATION_LOCATION = (
+    'http://svn.webkit.org/repository/webkit/trunk/'
+    'LayoutTests/platform/chromium/test_expectations.txt')
+
+
+class TestExpectationsHistory:
+  """A class to represent history of the test expectation file.
+
+  The history is obtained by calling PySVN.log()/diff() APIs.
+
+  TODO(imasaki): Add more functionalities here like getting some statistics
+      about the test expectation file.
+  """
+
+  @staticmethod
+  def GetDiffBetweenTimes(start, end, testname_list,
+                          te_location=DEFAULT_TEST_EXPECTATION_LOCATION):
+    """Get difference between time perios for the specified test names.

[dennis_jeffrey, 2011/08/25 00:36:46]

'perios' --> 'periods'

[imasaki1, 2011/08/25 23:57:03]

Done.

+
+    Given the time period, this method first gets the revision number. Then,
+    it gets the diff for each revision. Finally, it keeps the diff relating to
+    the test names and returns them along with other information about revision.
+
+    Args:
+      start: A float time specifying start of the time period to be
+          looked at.
+      end: A float time object specifying end of the time period to be

[dennis_jeffrey, 2011/08/25 00:36:46]

is it really a "float time object", or just a floating-point value?

[imasaki1, 2011/08/25 23:57:03]

Actually it is timestamp. change here.

+          looked at.
+      testname_list : A list of the test name string that you are interested

[dennis_jeffrey, 2011/08/25 00:36:46]

'A list of strings representing test names of interest.'

Also remove the space right before the ':'

[imasaki1, 2011/08/25 23:57:03]

Done.

+          in.
+      te_location: A location of the test expectation file.
+
+    Returns:
+      A list of tuples (old_rev, new_rev, author, date, message, lines). The
+          |lines| contains the diff of the tests that I am interested in.

[dennis_jeffrey, 2011/08/25 00:36:46]

'that I am interested in' --> 'of interest'

[imasaki1, 2011/08/25 23:57:03]

Done.

+    """
+    # Get directory name which is necesary to call PySVN.checout().

[dennis_jeffrey, 2011/08/25 00:36:46]

'checout' --> 'checkout'

[imasaki1, 2011/08/25 23:57:03]

Done.

+    te_location_dir = te_location[0:te_location.rindex('/')]
+    client = pysvn.Client()
+    client.checkout(te_location_dir, 'tmp', recurse=False)
+    logs = client.log('tmp/test_expectations.txt',
+                      revision_start=pysvn.Revision(
+                          pysvn.opt_revision_kind.date, start),
+                      revision_end=pysvn.Revision(
+                          pysvn.opt_revision_kind.date, end))
+    result_list = []
+    # Find the last revision outside of time period and
+    # append it to preserve the last change before entering the time period.
+    # I am using 1 day back as |start2| (called start2) from original |start|.

[dennis_jeffrey, 2011/08/25 00:36:46]

'I am using' --> 'This uses'.

Also, you may want to explain why you're using start2 as being 1 day back.

[imasaki1, 2011/08/25 23:57:03]

Done.

+    start2 = time.mktime(
+        (datetime.fromtimestamp(start) - timedelta(days=1)).timetuple())
+    logs2 = client.log('tmp/test_expectations.txt',
+                      revision_start=pysvn.Revision(
+                          pysvn.opt_revision_kind.date, start2),
+                      revision_end=pysvn.Revision(
+                          pysvn.opt_revision_kind.date, start))

[dennis_jeffrey, 2011/08/25 00:36:46]

nit: indent the above 4 lines by 1 more space each

[imasaki1, 2011/08/25 23:57:03]

Done.

+    if len(logs2) > 0:

[dennis_jeffrey, 2011/08/25 00:36:46]

if logs2:

[imasaki1, 2011/08/25 23:57:03]

Done.

+        logs.append(logs2[len(logs2)-2])

[dennis_jeffrey, 2011/08/25 00:36:46]

only indent this line by 2 spaces, not 4

[dennis_jeffrey, 2011/08/25 00:36:46]

put spaces around the '-'

[imasaki1, 2011/08/25 23:57:03]

Done.

[imasaki1, 2011/08/25 23:57:03]

Done.

+    for i in xrange(len(logs)-1):

[dennis_jeffrey, 2011/08/25 00:36:46]

put spaces around the '-'

[imasaki1, 2011/08/25 23:57:03]

Done.

+        # PySVN.log() returns logs in reverse chronological order.

[dennis_jeffrey, 2011/08/25 00:36:46]

Indent this line and the lines below by only 2 spaces, not 4

[imasaki1, 2011/08/25 23:57:03]

Done.

+        new_rev = logs[i].revision.number
+        old_rev = logs[i + 1].revision.number
+        # Getting information about new revision.

[dennis_jeffrey, 2011/08/25 00:36:46]

'# Get information about the new revision.'

[imasaki1, 2011/08/25 23:57:03]

Done.

+        author = logs[i].author
+        date = logs[i].date
+        message = logs[i].message

[dennis_jeffrey, 2011/08/25 00:36:46]

I think the above 3 variables probably don't need to be defined; we could just
inline the code where needed below.

[imasaki1, 2011/08/25 23:57:03]

Done.

+        # Parsing the actual diff.
+        text = client.diff('/tmp', 'tmp/test_expectations.txt',
+                           revision1=pysvn.Revision(
+                               pysvn.opt_revision_kind.number, old_rev),
+                           revision2=pysvn.Revision(
+                               pysvn.opt_revision_kind.number, new_rev))
+        lines = text.split('\n')
+        target_lines = []
+        for line in lines:
+          for testname in testname_list:
+            matches = re.findall(testname, line)
+            if matches:
+              if line[0] == '+' or line[0] == '-':
+                target_lines.append(line)
+        if len(target_lines) > 0:

[dennis_jeffrey, 2011/08/25 00:36:46]

if target_lines:

[imasaki1, 2011/08/25 23:57:03]

Done.

+          result_list.append((
+              old_rev, new_rev, author,
+              # Needs to convert to normal date string for presentation.

[dennis_jeffrey, 2011/08/25 00:36:46]

I think we probably don't need this comment.  If you'd prefer to keep it, I
recommend moving it right before the current statement (function call), rather
than putting it in the middle of the parameter list of the function call.

[imasaki1, 2011/08/25 23:57:03]

Done.

+              datetime.fromtimestamp(date).strftime('%Y-%m-%d %H:%M:%S'),
+              message, target_lines))
+    return result_list