Preprocess and clean data

[gracwng]

• [x] Fix the links in the source column by removing the "api." substring so anyone can be directed to the source file without any error
• [x] Do some data analysis. Check how many cwl files had issues being loaded into the CSV
• [ ] Compare with data fields / attributes with the Workflow Hub data source
• [ ] Clean / standardize the data
  • [ ] If a document's cwl_doc field is empty (null), check to see if the cwl_label field is empty. If not, then set cwl_doc to be cwl_label

[gracwng]

I'm considering using OpenAI as a way to generate metadata like the description of each of our CWL file using the OpenAI metadata tagger