cullmann added a comment.
In D19367#424714 <https://phabricator.kde.org/D19367#424714>, @loh.tar wrote: > > The batch processing approach is sane. > > hm, thanks > > > My real issue here is: What to do with multi-line expressions that might span more than one line? Won't we miss matches if the ranges just overlap at that locations? > > Well, yes...that should be the case. I had also some idea but ... > > > not compute the batch ranges beforehand > > ...that sound not so nice. What if I would keep it but find some solution? I am not sure that the precomputation saves you a lot of things, you can your compute the next batch in the slot that searches. But if the issue with the multi-line stuff is solved, I won't t reject this for the precomputation, this still can be "optimized" later on if wanted. > OTH is that multiline search for me a slightly rare case Yes, that is true. It should just work reasonable well, if would be bad if e.g. even a trivial case like lala\nlala would break because of badly choosen boundaries. REPOSITORY R39 KTextEditor REVISION DETAIL https://phabricator.kde.org/D19367 To: loh.tar, #ktexteditor, dhaumann, cullmann Cc: cullmann, kwrite-devel, kde-frameworks-devel, #ktexteditor, domson, michaelh, ngraham, bruns, demsking, sars, dhaumann
