Re: Kurdish locale

[barbier@linuxfr.org (Denis Barbier)]

On Sun, Apr 24, 2005 at 06:03:49PM +0200, Erdal Ronahi wrote:
> Hi,
> 
> I have fixed some small bugs and changed the collation chars. I hope the 
> rest is correct. I cannot read the <U+XXXX> chars, so I don't know 
> whether the month names and so on are correct.

Hi again,

your changes in collation rules are wrong.  Lines like
   collating-symbol <e/>>
define a named symbol.  It contains only ASCII characters for
portability reasons.  A circumflex accent is represented by a
right angle bracket, which needs to be escaped (by a slash)
because symbols appear within angle brackets.  The line above
defines a symbol which is used in collation rules.
Lines like
   reorder-after <e>
   <e/>>
means that e+circumflex is a letter in your language, and is sorted
after e.  One could have defined instead
  collating-symbol <ecirc>
or any other name, but using non-ASCII characters is a bad idea.

Apart from these symbol renames, you also removed dotless-i (and
uppercase dotted i) from collation rules, are these characters not
used in your language?

Your other changes in LC_NAME contain invalid characters, here is
a patch.

Denis

--- ku_TR2	2005-04-24 19:04:08.091455464 +0200
+++ ku_TR2	2005-04-24 19:03:46.060804632 +0200
@@ -177,14 +177,14 @@
 <U0025><U006D><U0025><U0074><U0025><U0066>"
 % TODO
 % Birêz
-name_gen    "<U0042><U0061><U0072><U0EA><U007A>"
+name_gen    "<U0042><U0061><U0072><U00EA><U007A>"
 % "M."
-name_mr     "<U0042><U0061><U0072><U0EA><U007A>"
+name_mr     "<U0042><U0061><U0072><U00EA><U007A>"
 % "Birêz"
-name_mrs    "<U0042><U0061><U0072><U0EA><U007A>"
-name_ms     "<<U0042><U0061><U0072><U0EA><U007A>"
+name_mrs    "<U0042><U0061><U0072><U00EA><U007A>"
+name_ms     "<U0042><U0061><U0072><U00EA><U007A>"
 % "Birêz"
-name_miss   "<U0042><U0061><U0072><U0EA><U007A>"
+name_miss   "<U0042><U0061><U0072><U00EA><U007A>"
 END LC_NAME
 
 LC_ADDRESS