Hi all,

I'm working now on the database charset migration as described in
https://twiki.cern.ch/twiki/bin/view/CDS/InvenioUpgrade

My MySQL setting have been always Debian defaults, so I have this double
utf8 conversion.  After doing several tests with iconv and recode (with
and without -c and -f), I've found that this conversion works better
than any other that I have tried:

 iconv -c -f utf8 -t windows-1252 $dbdump >$dbdump.utf8

This works for most of the diacritics, quotes, hyphens, etc. etc,
*except* for all uppercase diacritics (ex., À, Í, etc) and some greek,
math and chemical chars.

I have been hexdumping the original and converted problematic chars
hoping that with a minimal substitution would help me on this
problematic subset of chars.  I've also googled around
(http://www.google.es/search?q=latin1+utf8+double+conversion+mysql) with
many entries, but I haven't found this specific case I'm having.

By chance, has any of you been there and done that?

Thanks,

Ferran

Reply via email to