PICA: Extend classification/subject headings schemes from config file

[nichtich]

The list of classification/subject headings schemes seems to be hard-coded at several places:

• https://github.com/pkiraly/metadata-qa-marc/blob/0162996dea4662c02589c83195ce0a61162ba1cc/src/main/java/de/gwdg/metadataqa/marc/analysis/ClassificationAnalyzer.java#L88-L104
• https://github.com/pkiraly/metadata-qa-marc/blob/0162996dea4662c02589c83195ce0a61162ba1cc/src/main/java/de/gwdg/metadataqa/marc/dao/record/BibliographicRecord.java#L58

As discussed in #154 there are more of these fields.

• 041A keywords
• 044. all subject indexing fields starting with 044
• 045. all subject indexing fields starting with 045

Maybe better not hardcode each field but get it from the Avram schema?

Current list of subject indexing fields from PICA+ K10plus Avram schema via

curl https://format.k10plus.de/avram.pl?profile=k10plus-title | jq -r '.fields[]|select(.tag|match("04[45]|041A"))|[.tag,.occurrence,.label]|@tsv' -r

The following fields are also subject indexing but only on level 1, so not needed so far:

• 144Z local library keywords
• 145S local library classification
• 145Z local library classification

[nichtich]

For current development, hard-coding the fields is ok.

[pkiraly]

@nichtich I am working on this issue and I found that we have a vocabularies.json file with the similar purpose (https://github.com/pkiraly/qa-catalogue/blob/0162996dea4662c02589c83195ce0a61162ba1cc/src/main/resources/pica/vocabularies.json). There are two main differences between JSON and TSV files:

• the JSON file provides a number of properties for each fields, while the TSV does only the title
• the JSON file have 15 fields, while the TSV does 40.

The result is depending on which file we use as a starting point. If we choose the JSON, it should be extended with those fields that are available in the TSV.

What do you think?