-
Improvement
-
Resolution: Unresolved
-
Major
When collecting multiple Junit xml files, the test results are imported from files, by file name order.
But usually the filename is not necessarily the actual results order, for example:
- Run tests and create file: The_first_test_suite.xml
- Run more tests and create file: Second_test_suite.xml
When later importing those junit files, "Second_test_suite.xml" will be imported first, and "The_first_test_suite.xml" afterwards.
Adding option to import the junit files by last modification timestamp (reversed) of the files, can fix this issue.
Note 1:
The files should be sorted like `ls -ltrc` so that the first to be imported is the file with the oldest modification timestamp,
and the last to be imported is the file with the newest modification timestamp.
(Sorting by creation time is not good, since it could have been an empty, or an overridden file, but not the first to include test results).
Note 2:
This is related, but not the same feature as JENKINS-12575
[JENKINS-65532] Option to import multiple Junit xml files by last modification date
| Description |
Original:
when seeing JUNIT test results on jenkins (output from a normal mvn job or a result of xUnit plugin), you see passed tests in an alphabetically order. (see attached image). on some scenarios it is important to see the results in the order they were run (sorted by date), similar to the 'Failed Tests List' shown if there were failed tests. it it possible to add an option (checkbox) to choose how the test results will be sorted in the report? (alphabetically or chronologically) thanks. |
New:
When collecting multiple JUNIT xml files, the test results are imported from files, by file name order.
But usually the filename is not necessarily the actual results order, for example: # Run tests and create file: *T*he_first_test_suite.xml # Run more tests and create file: *S*econd_test_suite.xml When later importing those junit files, "*S*econd_test_suite.xml" will be imported first, and "*T*he_first_test_suite.xml" afterwards. Adding option to import the junit files by last modification timestamp of the files, can fix this issue. This is related, but not the same feature as JENKINS-12575 |
| Description |
Original:
When collecting multiple JUNIT xml files, the test results are imported from files, by file name order.
But usually the filename is not necessarily the actual results order, for example: # Run tests and create file: *T*he_first_test_suite.xml # Run more tests and create file: *S*econd_test_suite.xml When later importing those junit files, "*S*econd_test_suite.xml" will be imported first, and "*T*he_first_test_suite.xml" afterwards. Adding option to import the junit files by last modification timestamp of the files, can fix this issue. This is related, but not the same feature as JENKINS-12575 |
New:
When collecting multiple JUNIT xml files, the test results are imported from files, by file name order.
But usually the filename is not necessarily the actual results order, for example: # Run tests and create file: The_first_test_suite.xml # Run more tests and create file: Second_test_suite.xml When later importing those junit files, "Second_test_suite.xml" will be imported first, and "The_first_test_suite.xml" afterwards. Adding option to import the junit files by last modification timestamp of the files, can fix this issue. This is related, but not the same feature as JENKINS-12575 |
| Description |
Original:
When collecting multiple JUNIT xml files, the test results are imported from files, by file name order.
But usually the filename is not necessarily the actual results order, for example: # Run tests and create file: The_first_test_suite.xml # Run more tests and create file: Second_test_suite.xml When later importing those junit files, "Second_test_suite.xml" will be imported first, and "The_first_test_suite.xml" afterwards. Adding option to import the junit files by last modification timestamp of the files, can fix this issue. This is related, but not the same feature as JENKINS-12575 |
New:
When collecting multiple JUNIT xml files, the test results are imported from files, by file name order.
But usually the filename is not necessarily the actual results order, for example: # Run tests and create file: The_first_test_suite.xml # Run more tests and create file: Second_test_suite.xml When later importing those junit files, "Second_test_suite.xml" will be imported first, and "The_first_test_suite.xml" afterwards. Adding option to import the junit files by last modification timestamp (reversed) of the files, can fix this issue. Note 1: The files should be sorted like `ls -ltrc` so that the first to be imported is the file with the oldest modification timestamp, and the last to be imported is the file with the newest modification timestamp. (Sorting by creation time is not good, since it could have been an empty, or an overridden file, but not the first to include test results). Note 2: This is related, but not the same feature as JENKINS-12575 |
| Assignee | Original: Gregory Boissinot [ gbois ] | New: Eric [ eric_dangel_dev ] |
| Summary | Original: Option to import multiple JUNIT xml files by last modification date | New: Option to import multiple Junit xml files by last modification date |
| Description |
Original:
When collecting multiple JUNIT xml files, the test results are imported from files, by file name order.
But usually the filename is not necessarily the actual results order, for example: # Run tests and create file: The_first_test_suite.xml # Run more tests and create file: Second_test_suite.xml When later importing those junit files, "Second_test_suite.xml" will be imported first, and "The_first_test_suite.xml" afterwards. Adding option to import the junit files by last modification timestamp (reversed) of the files, can fix this issue. Note 1: The files should be sorted like `ls -ltrc` so that the first to be imported is the file with the oldest modification timestamp, and the last to be imported is the file with the newest modification timestamp. (Sorting by creation time is not good, since it could have been an empty, or an overridden file, but not the first to include test results). Note 2: This is related, but not the same feature as JENKINS-12575 |
New:
When collecting multiple Junit xml files, the test results are imported from files, by file name order.
But usually the filename is not necessarily the actual results order, for example: # Run tests and create file: The_first_test_suite.xml # Run more tests and create file: Second_test_suite.xml When later importing those junit files, "Second_test_suite.xml" will be imported first, and "The_first_test_suite.xml" afterwards. Adding option to import the junit files by last modification timestamp (reversed) of the files, can fix this issue. Note 1: The files should be sorted like `ls -ltrc` so that the first to be imported is the file with the oldest modification timestamp, and the last to be imported is the file with the newest modification timestamp. (Sorting by creation time is not good, since it could have been an empty, or an overridden file, but not the first to include test results). Note 2: This is related, but not the same feature as JENKINS-12575 |
hi worked on this so hopefully its working now