--- In [email protected], "Sheri" <sheri...@...> wrote: > <ftp://ftp.unicode.org/Public/UNIDATA/CaseFolding.txt> > <ftp://ftp.unicode.org/Public/UNIDATA/UnicodeData.txt> > <http://www.sitepoint.com/blogs/2006/08/10/hot-php-utf-8-tips/>
etc. Chanks. Bottom line, case conversion may change length of a string. I'll have to work with that. BTW one thing currently done by regex code is quite a lot of malloc/free. Some of that could be avioded if it could be guarenteed that one regex operation would go to completion before the next one started. Data structures associated with a pattern handle need to last over time, but in principle buffers to do with subjects and replacement patterns don't...so I could go for one-buffer-for-all operations for e.g. replacement pattern. It may have to grow if an operation threatens to overrun current buffer, but that would be rate, instewad of once-per-op.
