Subject Problems with IB_IncSearch and German Umlauts
Author Dieter Tremel
Hello,

I want to report that IB_IncSearch does not function well on VARCHAR
fields with WIN1252 Characterset und Collation PXW_INTL. This is
important for german users which have this collation because of
historical reasons coming from BDE. I show example DDL at end of the
post.

In this Collation for instance German umlaut 'ä' is sorted like the
pair 'ae', but the Function AnsiCompareStr that IB_IncSearch uses
orders ä like a separate letter between a and b.

As a result IB_IncSearch does not find record despite they are there.
My workaround was to go to ISO8859_1 with Collation DE_DE which
corresponds to AnsiCompareStr and all is OK, but it is not easy to
transform all customers databases.
Shure, another Workaround would be to use a own
TIB_Connection.OnCustomCompareStr function with is like PXW_INTL, but
should TIB_IncSearch not take care of the collation order? At least
there should be a hint in the help.

Thank You Dieter Tremel

/*--- Metadata report ---*/
SET TERM ^;

/* Started at 27.09.2001 10:24:58 */
/* Connect using owner = SYSDBA */
/* Using server WI-V6.0.0.627 */


SET SQL DIALECT 3^
CREATE DATABASE 'SACDATA.GDB' DEFAULT CHARACTER SET WIN1252^

CREATE DOMAIN NAME1_D AS VARCHAR(39) CHARACTER SET WIN1252 NOT NULL
COLLATE PXW_INTL^

/* Table List. 1 */
CREATE TABLE ADRESSEN (
ADRESSNR PRIMKEY_D /* INTEGER */ NOT NULL,
NAME1 NAME1_D /* VARCHAR(39) CHARACTER SET WIN1252 */ NOT NULL,
...