According to Marco Scheurer: > > In that case, please try changing Configuration::AddParsed() to > > follow the setlocale(LC_ALL, ps->get(dict)) with a > > setlocale(LC_CTYPE, ps->get(dict)), and see if that fixes the problem. > > String *str = new String(setlocale(LC_ALL, ps->get(dict))); > setlocale(LC_CTYPE, ps->get(dict)); > > (did I get this right?) does not get read of the warning. > > > If not, try without the additional setlocale() call, but change the > > original LC_ALL to LC_CTYPE and see if that works. > > String *str = new String(setlocale(LC_CTYPE, ps->get(dict))); > > does remove the warning, and it looks like accented characters are > indexed too! What are the chances of breaking something else by using > LC_CTYPE instead of LC_ALL?
Yay! Glad to help. I think the chances of actually breaking something are pretty slim. The only thing you need to be aware of is that the locale setting won't be as far-reaching as it would otherwise, which may be seen as good or bad depending on your perspective. E.g., normally with LC_ALL set to your locale, it affects number formats, so that things like your *_factor settings and the weights for fuzzy algorithms in search_algorithm may need a decimal comma rather than a decimal point, if that's what your locale specifies (and if your C library's atof() is locale-aware). Changing LC_ALL to LC_CTYPE means the locale won't have this effect on number formats, which I guess is a good thing as long as you're not expecting it to. In either case, htsearch forces LC_TIME _back_ to your current locale setting, so it should still set the default date format as appropriate for your locale. The default format can also be overridden via iso_8601 or date_format. I can't think of any other side-effects of this change to setlocale(), so I think you're reasonably safe. Now, from a developer perspective, I guess we need to decide whether we make this change permanently to the source, and force it for all systems (which I imagine may break things on other systems), or do we come up with a configure test for this? -- Gilles R. Detillieux E-mail: <[EMAIL PROTECTED]> Spinal Cord Research Centre WWW: http://www.scrc.umanitoba.ca/ Dept. Physiology, U. of Manitoba Winnipeg, MB R3E 3J7 (Canada) ------------------------------------------------------- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf _______________________________________________ htdig-general mailing list <[EMAIL PROTECTED]> To unsubscribe, send a message to <[EMAIL PROTECTED]> with a subject of unsubscribe FAQ: http://htdig.sourceforge.net/FAQ.html
