I noticed that HighlightPhase.java hardcodes the use of
DefaultEncoder, which does not escape the text before adding highlight
I know that Lucene supports HTML encoding, so is there a chance to
support it in ElasticSearch?
The problem is critical for highlighting plain-text fields. If you
escape a field after highlighting (on client), highlight tags will be
escaped as well. Note that all user-provided content HAS to be escaped
if you display it inside HTML page.
Am I the only one experiencing this problem?