Reuse query parser

[tobiasdiez]

Instead of parsing a search query using a custom grammar we should reuse some existing query parser. For example, Lucene's query parser with QueryParser.parse.

Triggered by #1633 (comment).

• Description of Lucene's syntax: http://www.solrtutorial.com/solr-query-syntax.html and https://lucene.apache.org/core/2_9_4/queryparsersyntax.html
• Lucene provides a few classes which let us modify the search syntax: https://lucene.apache.org/core/4_1_0/queryparser/index.html?org/apache/lucene/queryparser/flexible/core/parser/SyntaxParser.html
• Or we use the default Lucene syntax: https://lucene.apache.org/core/4_0_0/queryparser/org/apache/lucene/queryparser/classic/QueryParser.html

[simonharrer]

Be aware that a lot of groups rely on the current search query grammar, and they will break if this is changed.

[Le-wi]

So introducing the new parser would require to migrate all the old queries as well?

[tobiasdiez]

The only groups effected are the "search based groups". I doubt that many users use them (and then probably only experienced users). Thus, I would say that displaying a warning dialog of the kind "some groups need to be updated by hand to work properly" is sufficient. @JabRef/developers opinions?

[koppor]

Steps:

1. Dive into our powerful search syntax: https://docs.jabref.org/finding-sorting-and-cleaning-entries/search
2. Compare with lucene
3. Evaluate whether a switch is possible
4. Get rid off ANTLRv4
5. Add an ADR

[koppor]

@DominikVoigt implemented a lucene parsing at #6799. This could be a good basis to solve this issue.

[koppor]

The follow-up ticket koppor#341 should also be linked here.