Add Conda meta.yml manifest parser

[pombredanne]

Description

See for example https://github.com/bioconda/bioconda-recipes/blob/master/recipes/abeona/meta.yaml#L42

This is the format used by Conda and bioconda

[shubhscoder]

@pombredanne I am interested in taking this ! Can you guide me on this ? Thank you so much!

[pombredanne]

Thank you. You have to check how this is done for other package manifest parsers in the packagedcode module And first find what is the actual set of data that is there. @epruesse ping

[shubhscoder]

@pombredanne is (scancode-toolkit/src/packagedcode/jar_manifest.py) the right file to look?

[PyAgni]

@pombredanne I'm willing to do this if anyone already hasn't been assigned. I have a few questions:

• We have to use PyYaml for this. Right? Or this has to be done by any other method?
• Once parsed, the collected metadata from the manifest has to be stored in a variable(say 'package'). And make an object that returns that package.
• Do we also need to make unittests for the parser?

And since I'm already working on #295 I'll try to make this python 2/3 compatible.

[shubhscoder]

Hi Agni, I am trying to complete this as my first issue. Would you mind helping me in this ?

…

On Sun, Mar 17, 2019, 12:44 AM Agni Bhattacharyya ***@***.***> wrote: @pombredanne <https://github.com/pombredanne> I'm willing to do this if anyone already hasn't been assigned. I have a few questions: - We have to use PyYaml for this. Right? Or this has to be done by any other method? - Once parsed, the collected metadata from the manifest has to be stored in a variable(say 'package'). And make an object that returns that package. - Do we also need to make unittests for the parser? And since I'm already working on #295 <#295> I'll try to make this python 2/3 compatible. — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#1447 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ARofxQlYzgqnGPCWWQUMa5-M9nH8aJ7Nks5vXUKmgaJpZM4b05J4> .

[PyAgni]

Hello @shubhscoder . Although I'm new here, I would love to help you. I don't have prior experience with yml files but the format seems pretty straight forward if you have worked JSON files or something similar. I think we have to write a script to parse all the contents of the meta.yml file like package details,requirements,about etc. Then we would also probably need to create unittests for that script. (See tests folder)
Apart from this @pombredanne will be able to clarify more about it.

[shubhscoder]

Hello @PyAgni we can work together on this. Thank you!

[pombredanne]

@PyAgni @shubhscoder you can see the many parsers available in src/packagedcode. The main purpose is to normalize the data in our Package model objects. A recent and simple one was contributed by @ritiek in #1426 for Rust Cargo for instance.

About YAML: we do not use PyYAML directly. We use instead the saneyaml parser. Search for "saneyaml" for details.
And we always write automated tests of course.