1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
|
Testing regressions
-------------------
This explains how to test regressions.
This assumes you have a properly setup gst-qa environment (i.e. that
PYTHONPATH is pointing to the proper location).
1. Choosing what tests you want
If you have been modifyin a gst element used in standard playback
(like a decoder, demuxer, source), you will want to use these tests
(ranging from basic test to more advanced tests).
* typefind-test : checks that we can handle the various formats
* playbin-test : Tests playbin(2) on the given file
* full-gnlfilesource-scenario : checks the file with gnlfilesource
(you'll need GNonLin installed). The advantage of this test is that
due to how GNonLin works, it will be able to test seekability in
various location of the file on all streams.
2. First run
You'll want to run gst-media-test with your reference system
(i.e. what you're comparing with).
Ex : gst-media-test -t typefind-tes playbin-test -f /data/medias/
-t typefind-test playbin-test : run those two tests
-f /data/medias : on any file present in /data/medias
3. Switch to code to test against
By changing env variables or whatever technique you use
4. Second run
Do the same as in 2. First Run
5. Comparing the runs
# insanity-compare testrun.db <first-testrun-id> <second-testrun-id> >
regressions
You will end up with a list of the regressions.
The test results will be paired by old and new test.
You can then read the information, results and debug logs to figure
out what went wrong.
|
