Assorted bug fixes for May #521
Conversation
Fixes #10884.
Fixes #10843.
It's very common to have similarly-named files in the same directory that have very different dimensions. To prevent weird problems with grouping, and because the workflow of dealing with slides means that grouping isn't really necessary, it is now disabled.
Noticed by Stéphane Dallongeville.
|
--test cellworx |
Conflicts: components/bio-formats/src/loci/formats/in/SlidebookTiffReader.java
|
A couple of the builds are red; please hold off on fully reviewing until they are fixed. |
| } | ||
| catch (NumberFormatException e) { | ||
| LOGGER.debug("", e); | ||
| } |
There was a problem hiding this comment.
If an exception is caught here, then we won't have added a value to fieldPosX (or Y, below) and then on the next loop iteration, the indexing of fieldPosX will be off-by-one, and will further off for each additional Double parse error.
Do we want to add a dummy value to the list in the catch block; possibly an invalid one like Double.NaN if the code can cope with it, or null? We can then check for this dummy value in the code above and fall back to posX[](or Y, accordingly).
There was a problem hiding this comment.
Should be fixed now.
|
10887 appears fixed. |
This allows the log files to be picked up if isGroupFiles() returns false.
|
OK, should all be sorted out pending the builds going green then (which should happen tonight). The size rounding in Metamorph isn't really something we can do much about, and is not uncommon for other formats where we have to calculate the size instead of reading it. The number of instruments in Mike.lif is typical behavior for .lif files; I think it is correct (since each series stored in the file has its own instrument definitions), but as this PR doesn't introduce that behavior I'd leave any changes there for a separate PR. |
|
The builds are all green, and the latest changes all look fine. Please merge. |
See:
http://trac.openmicroscopy.org.uk/ome/ticket/10887
http://trac.openmicroscopy.org.uk/ome/ticket/10884
http://trac.openmicroscopy.org.uk/ome/ticket/10843
http://www.openmicroscopy.org/community/viewtopic.php?f=13&t=4710#p9964
Pretty much everything else was noticed while doing the ground work for a complete repository testing job for Jenkins.
These are all straightforward reader fixes, so just need verifying that the above listed issues are resolved and that the Jenkins jobs are all green.