Development: updated 'date' tag matching and removed redundancy

[Luizerko]

• Updated 'date' tag matching to also capture '260__'
• Removed 'date' tag redundancy by not reading '260__' if blob has '269__'
• Type casting correction to handle corner case on 'timestamp' property
• CI updated: python 2.7 -> 3.6 and pip 20 -> 21
• Test files import order corrected to fix 'isort' error
• Handling 'None' and duplicates on tags 260|269
• Handling 'None' on tag '595'
• Updated contributors roles and handling non-default role on tags '100|700|508'
• Changed cds_dojson.marc21.utils.create_record to generate multiple MementoDict for single record with multiple videos
• Tag '8564_' added to the model
• Changed structure of multiple records. Now the 'not_indexed' record is added as one video
• Update on tag '246' to include video parts and video parts names
• New and updated tags: '246', '269__b', '511', '518', '542', '700', '775', '856', '901', '906'
• New property '_digitization' added to the schema
• New tags: '300', '336', '337', '508', '514', '5421_a', '583', '594', '595', '597', '650', '690', '773', '774', '787', '852', '856', '961', '962', '963', '981'
• Updated tests to include new property '_digitization'