Proposal for a Function to Generate JSON Schema for Workflow Parameter Files#282
Proposal for a Function to Generate JSON Schema for Workflow Parameter Files#282suecharo wants to merge 2 commits intocommon-workflow-language:mainfrom
Conversation
|
I have not yet been able to address the comment discussed in #273 (comment). |
|
Impeccable timing @suecharo! I've also drafted up an implementation of this too that addresses this comment. It also addresses complex input types, such as the inputs required for this pipeline - https://github.com/umccr/cwl-ica/blob/main/workflows/tso500-ctdna-with-post-processing-pipeline/1.2.0--1.0.0/tso500-ctdna-with-post-processing-pipeline__1.2.0--1.0.0.cwl Or Mind if I put in a PR into this script? |
|
On a different note I've also generated a schema for the workflows themselves! common-workflow-lab/cwl-ts-auto#27 |
Of course, thank you. |
Many correct ways, but yes this one works. I've rebased off your branch onto this PR #288 But can PR this branch onto your branch as well? Rather than PR from main on your remote, I'd recommend making a branch on your repo and putting the PR in from that. My last step is generating a GH actions test to make sure that I can generate a schema for all of the cwl v1.2 tests and then validate the input jsons in the said test directory against the respective workflow input json schemas. |
|
@alexiswl Thank you for your advice.
Indeed, you are correct. I should have created a separate branch. I have now created a new branch https://github.com/suecharo/cwl-utils/commits/enhancement/cwl-inputs-schema-gen . Due to GitHub's constraints, we can't change the source branch in a PR after forking. We'll need to make a new PR instead...
Absolutely, it is possible. In scenarios where both of us are working on similar matters across our forks, it necessitates sending PRs between our forks, doesn’t it? However, given the circumstances with the main branch as mentioned, should there be a need for further modifications on my end, I will send a PR to your Furthermore, feel free to close this PR and shift our discussion to #288 if that is more convenient. |
|
Closing as per #282 (comment) |
From issue #273