[wip] [feat] distinct analyzers for indexing and querying with Lucene

[line-o]

Description:

Original from @sten1ee :
In our usage of eXist, we're utilising a 'State of the art' LuceneAnalyzer for Sanskrit language developed by our colleagues from BDRC It has several modes of operation, but the one that we use - 'lenient mode' - requires one Analyzer to be used during the indexing phase, but then a different (or a differently configured) one to be used during querying.
Currently, eXist does not provide support for this use case, but this PR is aiming to achieve this feat.

• rebased on current develop
• tests for full-text indices
• tests for fields
• tests for new Lucene fields
• documentation

Reference:

based on #2908, #2991 opened by @sten1ee

Type of tests:

testsuite in XML

[sten1ee]

@line-o : how's it going with the additional tests and the documentation of the feature?

[line-o]

@sten1ee the tests are part of this PR - documentation is still TODO (see the checkboxes in the description)

[line-o]

@sten1ee would you be able to describe your use-case in the referenced documentation ticket. A link to the indexer and a code example would be great, too.

[sten1ee]

There's some documentation, (who volunteers to review it?):
eXist-db/documentation#445

[sten1ee]

The corresponding docs PR#445 was approved and merged.
So perhaps it is about time to approve and merge this too, @adamretter ? :)

[adamretter]

@sten1ee @line-o I think this one has to be reviewed by @wolfgangmm instead of me, as how the Lucene stuff works in eXist-db is his domain.

[wolfgangmm]

Thanks to the additional tests @line-o provided, I'm entirely happy with this PR now. Thanks!