Re: uniq i18n implementation

Paul Eggert wrote: > Pádraig Brady <[EMAIL PROTECTED]> writes: > > >>There seems to be serious overhead with strcoll on glibc-2.3.5-10 at least. > > > We can fix the performance problem for that particular test case as > follows. I installed this (diff -pubw format): > > 2006-08-14 Paul Egge

Re: uniq i18n implementation

Pádraig Brady <[EMAIL PROTECTED]> writes: > There seems to be serious overhead with strcoll on glibc-2.3.5-10 at least. We can fix the performance problem for that particular test case as follows. I installed this (diff -pubw format): 2006-08-14 Paul Eggert <[EMAIL PROTECTED]> * memc

Re: uniq i18n implementation

Pádraig Brady wrote: > Paul Eggert wrote: > >Using strcoll is inefficient anyway Don't we know it! If we can avoid it, we'd like to. >>> >>>Well, the mbstowcs+wcscoll solution I presented >>>should be equivalent to strcoll on any platform, >>>and it's much faster in my tests. >> >> >

Re: uniq i18n implementation

Pádraig Brady <[EMAIL PROTECTED]> writes: > I was also using the string length comparison > shortcut on the wide string. I'm unsure whether > this is valid (on all platforms). Me too, which is why the current code is cautious about this sort of thing. >> Sorry, I'm not familiar with the ICU code

Re: uniq i18n implementation

Paul Eggert wrote: Using strcoll is inefficient anyway >>> >>>Don't we know it! If we can avoid it, we'd like to. >> >>Well, the mbstowcs+wcscoll solution I presented >>should be equivalent to strcoll on any platform, >>and it's much faster in my tests. > > > That's good to know, though I'm

Re: uniq i18n implementation

>>>Using strcoll is inefficient anyway >> >> Don't we know it! If we can avoid it, we'd like to. > > Well, the mbstowcs+wcscoll solution I presented > should be equivalent to strcoll on any platform, > and it's much faster in my tests. That's good to know, though I'm puzzled as to why it's true.

Re: uniq i18n implementation

Paul Eggert wrote: > Pádraig Brady <[EMAIL PROTECTED]> writes: > > >>memcoll does 2 errno accesses per call, which shows up significantly >>in profiles. Does strcoll even set errno? > > > says it's > allowed to. I assume some platforms do

Re: uniq i18n implementation

Pádraig Brady <[EMAIL PROTECTED]> writes: > memcoll does 2 errno accesses per call, which shows up significantly > in profiles. Does strcoll even set errno? says it's allowed to. I assume some platforms do. I wouldn't be surprised if errno