pantsbuild / pants / 5407 / 3
0%
master: 0%

DEFAULT BRANCH: master
Ran
Files 332
Run time 12s
Badge
Embed ▾
README BADGES
x

If you need to use a raster PNG badge, change the '.svg' to '.png' in the link

Markdown

Textile

RDoc

HTML

Rst
coverage: 66.168%. Remained the same
CI_FLAGS="-fkmsrcn -u 1/2 'Unit tests for pants and pants-plugins - shard 2'"

push

travis-ci

Tejal Desai 
[junit tool] fix suppress output emits jibberish on console.

At twitter, since org.pantsbuild.junit-runner-0.0.5 upgrade, we saw issues reported with test.junit --no-supress-output flag.

Easily reproducable as mentioned in https://github.com/pantsbuild/pants/issues/1657

This rb is fixing 2 issues:
1. test.junit --no-supress-output emits jibberish on console.
   In https://rbcommons.com/s/twitter/r/2183 @fkorotkov added a MultiOutputStream. However with tests running in paralled the writes were not synchornized.
Deleting this Class and using TeeOutputStream http://grepcode.com/file/repo1.maven.org/maven2/commons-io/commons-io/2.0.1/org/apache/commons/io/output/TeeOutputStream.java

2. test output and errors files only log output from first testmethod.
This rb https://rbcommons.com/s/twitter/r/2257/diff/1#index_header caused a regression. While debugging a TestCase where there are 2 test methods, i found out that only output from one test would get written in the <test>.out.txt.
This is because, registerTests() would return the count of Tests as 1.
testStarted gets invoked with testmethod (hence the number of tests is 1) instead of testclass.
This will set the useCount=1 and when testFinished is fired for the first test, it would close the file stream.

https://github.com/pantsbuild/pants/blob/master/src/java/org/pantsbuild/tools/junit/ConsoleRunner.java#L134 and
https://github.com/pantsbuild/pants/blob/master/src/java/org/pantsbuild/tools/junit/ConsoleRunner.java#L217

The power PowerMockRunNotifier.java in 1.16.2 does use
https://github.com/jayway/powermock/blame/9bd63da35/modules/module-impl/junit4/src/main/java/org/powermock/modules/junit4/internal/impl/PowerMockRunNotifier.java#L96
I am more inclined on reverting this and upgrading to powermock-1.6 if we see any Null pointer exceptions.

Testing Done:
With locally published jar,

03:45:53 00:01     [junit]
03:45:53 00:01       [run]
                     ........
                     Time: 0.002

                     OK (7 tests)

                     .test41
                     .start test42
                     end test42
                     ....
                     Time: 15.075

                     OK (6 tests)

                     .....
                     Time: 0

                     OK (4 tests)

                     ...
                     Time: 0

                     OK (2 tests)

                     ...
                     Time: 0

                     OK (2 tests)

                     ......E
                     Time: 0.007

                     Tests run: 5,  Failures: 1

                     ...
                     Time: 0.006

                     OK (2 tests)

                     ....
                     Time: 0.006

                     OK (3 tests)

                     ......test41
                     .start test42
                     end test42
                     .......
                     Time: 45.392

                     OK (22 tests)

               Waiting for background workers to finish.
03:46:39 00:47   [complete]
               SUCCESS

Bugs closed: 1657

Reviewed at https://rbcommons.com/s/twitter/r/2395/

13147 of 19869 relevant lines covered (66.17%)

0.66 hits per line

Source Files on job 5407.3 (CI_FLAGS="-fkmsrcn -u 1/2 'Unit tests for pants and pants-plugins - shard 2'")
Coverage ∆ File Lines Relevant Covered Missed Hits/Line