Refine pub_date_isim date extractions for date facet

[eporter23]

Story: As a cataloging specialist, I want our index to exclude certain types of 008 dates, so that users do not experience date facet anomalies due to incorrectly coded data which may come from external sources

This ticket follows work done in #371. While testing the date facet, some records (especially those with multiple dates or complex dates) do not function as expected due to how they are originally coded and how we extract the data from the MARC records. In reviewing some potential cleanup reports with cataloging, we feel it may be more feasible to adjust our indexing process than try to correct large numbers of records with invalid date entries.

• Review the 008 cases in the Date Facet - 008 Extraction Scenarios document:
• Position 06 = c
• Position 06 = d
• Position 06 = n
• Position 06 = r
• Position 06 = s
• Adjust our 008 date extraction logic as needed to implement the changes noted

Examples:

• In some cases, records erroneously have values entered in the date1 and date 2 positions, but in fact are coded in the 06 position to contain a single date only (06 = s)
• In some other cases, a record contains “0000” in the date 1 (position 7-10) or date 2 (11-14) positions of the 008 field. “0000” is not valid for the MARC specification for 008 and we should not store this value in our index.

References and mockups:

• Date Facet - 008 Extraction Scenarios
• MARC 008 reference [note: position counts start at 00, not 1!]
• Metadata and Indexing worksheet
• Data cleanup reports and statistics

SOLR and Indexing

• pub_date_isim - adjust indexing logic based on notes in the extraction scenarios document