Open dannon opened 7 years ago
Fixed same-name different-value tag issue preventing tagging in the UI with #3760
Tagging for collections in https://github.com/galaxyproject/galaxy/pull/3868
@dannon It looks like tags on collections do not propagate through tool executions
Yeah, it looks like where I added the tag addition code, handle_output
only touches the datasets. I can add it to collections too pretty easily though, it looks like. Good catch!
Redraw also handled in #3868 now.
Testing TAGs on test.galaxyproject.org getting this:
Datasets 2 - 5 are labelled with tags. Tags propagate correctly to 6 - 9. In step 10 I merge datasets 6 - 9 into one. Tag A gets duplicated. This history can be accessed here: https://test.galaxyproject.org/u/anton/h/hiv-r01-example-tagged
Clicking on dataset 10 does not show duplicated tag within tag list:
Hashtags in #3979
@nekrut Was the recent addition (sometimes not propagating tags) yours? Can you elaborate? I'd lean towards doing this in the job execution context (at runtime) rather than doctoring the tool XMLs, if that'd handle your use-case.
@dannon: I can't come up with an example for this
Sort/unique done in https://github.com/galaxyproject/galaxy/pull/4139
Are workflow-set hashtags supposed to overwrite (reset) the dataset-derived hashtags? If so, is there a way to combine both? I'd like the workflow to propagate which tools and reference genome was used to derived datasets, while the original input dataset hashtags propagate the sample identifier.
Resultant tag should be the combination of both, my guess is a bug in the way PJA (workflow set) tags are set is overwriting. I'll check and get a fix out shortly.
Hi everyone, I see another potential issue related to tagging workflow step in "configure output" / "Tags". If at a step X I add a tag #lovegalaxy and that in the step x+1 I add #loveiuc, then executing the wf on a history will result for the x+1 output step on a uniq tag (here #loveiuc). It seems that the previous one is erased...
(#4188 fixed @yvanlebras problem)
Nothing here seems pressing, removing milestone for now so we do not have to keep pushing it back.
#hashtags
are a tag that is propagated to derived datasets and displayed alongside history items even when collapsed.In history collections should display tags just like individual datasets. When collection is expanded:
we need to display tags somewhere around
a list of paired datasets
element.