Problems with colldef tool

[agata]

Since most latin utf8 locales have the collation file linked to us-ascii, and I really need good sorting at least for polish language in utf8, I tried to make my own.

I made the file for the colldef tool according to manpage and I keep getting this error:

colldef: Char 0x00 not found near line 2

I have no idea what it means, I tried to google it, nada

The file is very simple, just arbitrary order of a selection of letters. I'm attaching it here.

If somebody knows what I did wrong or can point me in the right direction, I would be very grateful.

 

[agata]

Okay, my bad. Those were not actual utf8 codes in the file.

For future use, if anybody finds this thread. This is a correct utf8 file for use with colldef. It has all the basic codes plus all the letters that have utf8 description matching "LATIN CAPITAL LETTER.*" and "LATIN SMALL LETTER.*" sorted in a "reasonable order" (i.e. after a letter "a" there are all variations of "a" with "add-ons" like acute, dot, ogonek, sorted in alphabetical order of descriptions). Not perfect, but better than US_ASCII. Use at your own risk.