ChrisHegarty opened a new issue, #13706:
URL: https://github.com/apache/lucene/issues/13706
There are a number of optimization in Elasticsearch that depend upon the
automaton from a `RegExp` being total - accepts all strings - [1] [2]. Changes
in the upcoming Lucene 10, to not minimize automaton returned by `RegExp`, has
broken the assumption that these optimisations were building upon. At least how
they stand today, and I'm not sure how best to replicate the functionality in
Lucene 10.
For example this is fine:
```
RegExp r = new RegExp("@");
Automaton a = r.toAutomaton();
assertTrue(a.isDeterministic());
assertTrue(Operations.isTotal(a));
```
, while this is not:
```
RegExp r = new RegExp(".*");
Automaton a = r.toAutomaton();
assertTrue(a.isDeterministic());
assertTrue(Operations.isTotal(a)); // <<< isTotal returns false
```
Without an API to minimise (since `MinimizationOperations` is now
test-only), I'm not sure how to re-code such optimizations. Or if we should be
attempting to provide our own minimize implementation. Or if RegExp should be
returning a total automaton for `.*`?
[1]
https://github.com/elastic/elasticsearch/blob/0426e1fbd5dbf1eb9dae07f9af3592569165f5de/x-pack/plugin/wildcard/src/main/java/org/elasticsearch/xpack/wildcard/mapper/WildcardFieldMapper.java#L383
[2]
https://github.com/elastic/elasticsearch/blob/0426e1fbd5dbf1eb9dae07f9af3592569165f5de/x-pack/plugin/esql-core/src/main/java/org/elasticsearch/xpack/esql/core/expression/predicate/regex/AbstractStringPattern.java#L30
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]