Non-latin characters in find/replace with Gumbo

Full-Text RSS
1

Hello,
I have met such feature.
Processing with gumbo works like a charm, but…
using non-latin character in function, e.g.
strip: //p[contains{text(),‘some non-latin text’)]
doesn’t find the string, while html5 processor processer the rule correctly without any problem.
What I do wrong with gumbo? Should I use any additional functions or settings?
P.S. I have tried to convert character into utf-8 string with the same result.
Added: find_string: / replace_string works.

Thank a lot for a fast reply.

2

I think the problems here is with the way those characters are encoded. I don’t think it’s UTF-8 encoding, but HTML character encoding. We had some issues around inconsistency between Gumbo and HTML5-PHP which was taking a lot of time to debug and resulting in quite a few problem reports like this one, so in the latest release we disabled Gumbo. I suggest you do the same and rely on HTML5-PHP too.