[TYPO3-core] RFC #9229: indexing of records containing HTML leads to concatenated words

[Martin Kutschker]

Dmitry Dulepov schrieb:
> Hi!
> 
> Martin Kutschker wrote:
>> But "hi<b>light</b>ing" is one word. IMHO only blocl level elements may be separated with extra
>> white-space prior to a call to strip_tags.
> 
> You are right but I think we can do this absolutely correctly only if we
> refactor indexed search. Right now this trick with str_replace is used
> in several places but not in this one. In my opinion this analyzer
> should be a separate function. But doing it without big refactoring does
> not make sense. Therefore I try to make it at least consistent between
> various parts of the code :)

Yep, makes sense. I thought you might get even more ambitious regarding indexed search :)

Masi