README
1Layout test analyzer script.
2
3* Prerequisite:
4
5This script requires PySVN (http://pysvn.tigris.org/) to be installed on the
6machine.
7
8* How to execute
9
10python layouttest_analyzer.py (Please run with '-h' for available command-line
11 options)
12
13Example: python layouttest_analyzer.py -r imasaki@chromium.org
14-t /var/www/analyzer/graph.html
15
16* File/Directory structure
17
18layout_tests/
19 README: this file.
20 *.py: python scripts.
21 tmp/: temp files.
22 test_data/: data for unit tests.
23 graph/: the default location for graph files.
24 anno/: the default location for bug annotation files.
25 result/: the default location for the analyzer results.
26 testname/: the default location for the test group definition files in CSV.
27
28* Execution overview.
29
301) using PySVN, the script collects layout test of interest
31(specified in command-line parameter and testnames/*.csv file) from
32Webkit SVN repository as well as test description. Then, it gets the
33latest test expectation file and does the simple parsing. The
34parsed test expectation is joined with the layout tests using test name
35as its join key. The tests are classified into 'whole', 'skip',
36'nonskip' test groups. 'Whole' contains all tests, 'skip' contains
37tests that are skipped as specified in the test expectation file, 'nonskip'
38tests are in the test expectation file and not skipped (these are tests
39that need attention).
40
412) the script reads the previous latest analyzer results, then,
42compares it with current results for each test group ('whole', 'skip',
43and 'nonskip'). It also looks into the SVN test expectation history
44diff for the time period for the test names of interest.
45
463) all obtained information is sent out to a email list specified
47in the parameters in pretty HTML format.
48
494) the trend graph is updated using the current results.
50
51* Components
52
53** LayoutTests:
54
55A class to store test names in layout tests. The test names (including
56regular expression patterns) are read from a CSV file and used for
57getting layout test names from Webkit SVN as well as test description.
58
59** TestExpectations:
60
61A class to model the content of test expectation file for analysis.
62The location of the test expectations file can be found in
63|TEST_EXPECTATIONS_LOCATIONS|. It is necessary to parse this
64file and store meaningful information for the analysis (joining with
65existing layout tests using a test name). Instance variable
66|all_test_expectation_info| is used. A test name such as
67'media/video-source-type.html' is used for the key to store
68information. However, a test name can appear multiple times in the
69test expectation file. So, the map should keep all the occurrence
70information. For example, the current test expectation file has the
71following two entries:
72
73BUGWK58587 LINUX DEBUG GPU : media/video-zoom.html = IMAGE
74BUGCR86714 MAC GPU : media/video-zoom.html = CRASH IMAGE
75
76In this case, all_test_expectation_info['media/video-zoom.html'] will
77have a list with two elements, each of which is the map of the test
78expectation information.
79
80** TestExpectationsHistory:
81
82A class to represent history of the test expectation file. The history
83is obtained by calling PySVN.log()/diff() APIs using the specified time
84period.
85
86** TrendGraph:
87
88A class to manage trend graph which is using Google Visualization
89 APIs. Google Visualization API
90 (http://code.google.com/apis/chart/interactive/docs/gallery/annotatedtimeline.html)
91 is used to present the historical analyzer result. Currently, data
92 is directly written to JavaScript file using file in-place
93 replacement for simplicity.
94