ScanCode: Extract more information

[abraemer]

ScanCode has a lot of options of what it should output. On one hand, we need to think about what options we absolutely need and document that. OTOH, if there is additional information (such as the license information) it would be nice to also use this. This has the potential that there could be a lot of combinations of information. It would be nice if our tool was "smart" and always could use the information present in the most sensible way.

In Opossum we have the additional fields: frequentLicenses and baseUrlsForSources

Example: If ScanCode was run with -license-text --license-references then it would be easy to populate frequentLicenses

Steps:

• Learn about the different options of ScanCode
• Create a test catalog of files using as many different argument combinations as possible
• Ensure that the pydantic models can handle all of these properly, i.e. add them to the Options model and introduce aditional model if necessary
• adapt the faker models accordingly - depending on the chosen options, we need to generate different fields
• Go option by option and consider what value the information has and whether it is sensible to include into the opossum output

Note collection:

• ScanCode provides copyright and license information separately, i.e. a file could have detected copyrights but no license_detections. Can we preserve the copyright information in this case?