Somef

This release of SOMEF includes new features:
- Better recognition for license files.
- Readthedocs URLs
- Arxiv links
- Code of conduct files
- Improved documentation (for new fields and older fields)

In addition, codemeta, ttl and JSON exports have been updated, classification pipelines have been updated to Python 3.9 (version of Scikitlearn 1.0.0) and multiple bugs and redundancies have been reduced. Check https://github.com/KnowledgeCaptureAndDiscovery/somef/milestone/4?closed=1 for more information about the issues addressed.

This release addresses small bug fixes for making the codemeta export more robust. For example, when a repository does not have a code release, now the program will not fail.

This release addresses the following issues:
- Automated evaluation reports for header analysis, so we can compare improvements without having to re-annotate corpus.
- Fixed errors in annotation of fields being GitHub API (they were a custom file exploration)
- Now we export a new category: Acknowledgements
- Fixed errors on empty exported metadata fields.
- Disambiguated category `Issues`
- See https://github.com/KnowledgeCaptureAndDiscovery/somef/milestone/3?closed=1 for more details.

This release improves on the following aspects of SOMEF:
- Notebooks and Dockerfiles are recognized.
- Docs folders are extracted.
- License is discovered through file inspection.
- New Codemeta specific export
- Fixed the `technique` in the JSON output to accommodate the new types of techniques supported
- MyBinder links extraction
- Zenodo DOIs extraction
- Tests
- Attempt at extracting a long title of a repo

This version of SOMEF includes full documentation and provides the following features:
- Cleanup and reorganization of repository contents.
- Synchronized package with GitHub releases
- Prepared package release.
- Bug fixes and code refactoring
- Export in JSON-LD and Turtle

This version of the system is more robust and includes:
* Answer in JSON format
* Configuration allowing to make authentication optional
* Initial integration of the header analysis.