1 Revision history for TAP::Harness::Junit
2 Format inspired by WWW::Mechanize's revision log
4 See GIT history for more detailed listing:
5 http://repo.or.cz/w/TAP-Harness-JUnit.git
8 ======================================
10 If an output of test was of length zero, TAP::Parser would not
11 recognize it as TAP output and would panic. We help it by adding
14 0.32 Mon Jul 13 17:31:25 2009 +0200
15 ======================================
17 Tests were timed not time separately, but aggregated duration
18 of the whole test run was used.
21 Recommend using the perl name mangling
23 0.31 Mon Jul 13 08:59:27 2009 +0200
24 ======================================
26 Make test name mangling configurable
27 More recent Hudson versions cope with this well, we don't want to
28 arbitrarily change the test names now, to retain compatibility.
30 Make up timing information for tests
31 Hudson ignores the testcase duration, computes it as sum of
34 0.30 Thu Apr 16 17:46:03 2009 +0200
35 ======================================
37 Use TAP::Harness::Aggregator->elapsed to set the elapsed time for testcases
39 Added "notimes" paramter that prevents timing of test cases
40 made use of that parameter in t/outputs.t
42 Remove SPEC file from here -- master copy is Fedora devel
44 0.26 Sat Jan 3 00:53:44 2009 +0100
45 ======================================
47 Do not call /bin/cat to get TAP output into parser (rt#41344)
49 Make 'xmlfile' optional (rt#42069)
51 Ensure valid UTF-8 output
52 Do not report failed plain and bad return value as two failures
55 Iteratively choose name for tests with ambigious names to prevent deep recursion warnings for long tests
56 Improve the output for tests with broken plan a bit