Starting from 7.0 Lucene check if offsets in token stream are broken and throws exception if they are. It was added https://issues.apache.org/jira/browse/LUCENE-7626
It causes a lot of issues with combination of word delimiter filters and multi word synonyms. Even with latest version of Elasticsearch I sometimes get "startOffset must be non-negative, and endOffset must be >= startOffset, and offsets must not go backwards" error.
For backward compatibility with already existing filters FixBrokenOffsetsFilter was also added in the same https://issues.apache.org/jira/browse/LUCENE-7626 , but I can't use it since there is no corresponding token filters in Elasticsearch. Are there any plans to add it?