[nycphp-talk] Squashing accented characters
Paul A Houle
paul at devonianfarm.com
Fri Oct 22 14:50:59 EDT 2010
For my site at
http://ookaboo.com/
I'm running into the problem that people are searching for "Dusseldorf"
but the name of the place is "Düsseldorf", so they don't find it.
It seems to me a good answer to this is to have some function that
squashes accented characters down to unaccented forms. I'd index the
unaccented forms and also squash down queries so they'd always match
up. I definitely need to do both ISO-Latin-1 and the
Latin-Extended-A, because fate has given me a lot of place names that
have the Polish dark L in them (?
<http://fileformat.info/info/unicode/char/0142/>). It also seems like
there are a lot of characters in Latin Extended-B that would also map
plausably to unaccented characters.
I can see how to write something like this, I'd need to parse out the
Unicode code points from UTF-8 and run them through a lookup table, but
it's a lot of details and I wonder if anybody has written a PHP function
to do this already.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.nyphp.org/pipermail/talk/attachments/20101022/1d126166/attachment.html>
More information about the talk
mailing list