Closed colinbrislawn closed 5 months ago
Thanks for the report! Looks like there's disagreement in URI encoding between JS and Python:
>>> urllib.parse.quote("+")
'%2B'
> encodeURI("+")
'+'
Which explains why this is happening. The visualizer seems to have escaped it with the expectation that q2view would do a similar escaping, which... normally it would, if not for the disagreement about +
. Python's definitely wrong here, as + is one of the older "safe" characters to use in a URL.
Maybe we should transfer this issue to q2-diversity, that way we can address the quoting behavior there?
Actually, my memory is stale, q2view no longer URI encodes: https://github.com/qiime2/q2view/commit/ea899cc0945e730ca2796328403aa3d547fca280
Yeah perhaps, if q2view shouldn't do anything here.
It might make sense to attempt a search for the "python escaped" version of the file if we can't find it in the zip the first time. But I'm not sure if that's a good idea or a bad one.
I'm good with q2view remaining unchanged here - we know there have been issues with the quoting of various visualizations in q2-diversity (at my hand), so I'm good with pointing the blame there, lol.
I'm closing this in favor of #345
Bug Description A metadata subcategory called
HF+
caused the generated .pdf and png files within the .qzv file to have different names, breaking the embedded figure and link on view.qiime2.org.Steps to reproduce the behavior See the attached file for a broken output: VIEWABLE_braycurtis-norm_feature-table_HF_Status_missingBoxPlot.zip
Expected behavior Links are preserved, through consistent file names or sanitized metadata categories.
Computation Environment