Thunderstone Support Forums

I am now running a Spanish version of my site. The search functionality appears to work fine. However I've noticed that Webinator treats "crédito" different than "credito" when returning its search results. Ideally I would like for Webinator to be smart enough to handle this correctly. In this case the page contains "crédito". I would expect it to be archived that way when indexing the site. However when searching the db, both "crédito" and "credito" would return the same page results. Is this possible?

You would need to obtain or create a custom locale that maps the accented characters to their non-accented versions when doing upper/lower case conversions. since they are different characters with different meanings, not just different cases of the same character. The default C locale, and most other locales, keep the accents when upper/lower-casing.

Then you would need to change Storage Charset in Webinator's All Walk Settings to ISO-8859-1 so that single-byte characters are stored instead of UTF-8.

You would also need to edit the dowalk and search scripts and manually set the locale to your custom locale. Uncomment the lines like this:

<sql "set locale='es_ES'"></sql>

and change `es_ES' to your custom locale.

Unfortunately I am not familiar with creating custom locales. However after doing a quick search, it appears a locale is an OS attribute. Is it possible to assign a locale to just Webinator at runtime instead of adjusting it for the entire OS?

Can you point me to any help documentation for creating a custom locale?

Sorry, no, that's an OS specific thing. Check the docs for your OS.