ISA-tools / mzml2isa-galaxy

mzml2isa for galaxy
6 stars 5 forks source link

Discovers ISA files as individual files and as Zip file #4

Closed pcm32 closed 7 years ago

pcm32 commented 8 years ago

This PR addresses issue #3 by providing both the individual files (collections for a and s files) and all the ISA files in a Zip. The second alternative is provided to keep the file names of the ISA files (Galaxy remanes them when "transforming" them recognized datasets).

pcm32 commented 8 years ago

Can @RJMW or someone who has done the conda testing part advise what else is needed for the test to be passed? We don't use conda at our side, but I can confirm that the tool works on Galaxy.

Tomnl commented 7 years ago

Have you tried to test with planemo?

I think it is failing because we now have different output files to the original test files

<test>
            <param name="name_of_study" value="test" />
            <param name="inputzip" value="metabolomics_study.zip" />
            <output name="output" value="metabolomics_study/test/index.html" />
</test>

I will look into it now

Tomnl commented 7 years ago

I have made a few changes and it seems to pass the unit tests now.

Also, I have hidden the old html output file as it had become a bit redundant

pcm32 commented 7 years ago

Thanks @Tomnl! Besides the testing, is there a real need for the inputzip datatype? I'm under the impression that this was introduced so that Galaxy wouldn't expand a zip file. However when using current versions, when you upload a zip file, if you label it as a zip file, it won't unzip it for you (which I think makes the use of inputzip redundant). This is of course if my understanding of the usage of inputzip is correct.

Tomnl commented 7 years ago

Hi @pcm32,

inputzip is just being used as the name of the param the format is actually "zip".

<param format="zip" name="inputzip" type="data" label="mzML zip file" help="A zipped folder of mzML files"/>

But you are right that we were using a redundant dependency on theno_unzip_datatype. I have now removed this.