Обсуждение: BUG #1860: Insert failed due to unique index

The following bug has been logged online:

Bug reference:      1860
Logged by:          Viswanath Ramineni
Email address:      v_ramineni@hotmail.com
PostgreSQL version: 8.1 Beta
Operating system:   Windows 2000 Professional
Description:        Insert failed due to unique index
Details:

I have a table with a primary key on id (integer) column and a unique index
on varchar(10) column. Insertion fails when the value of the unique column
is substring of an existing value. For example, if I have a row with the
unique column value 'TEST1', it will not allow me insert another row with
'TEST' as the value for this column.

Viswanath Ramineni wrote:
> I have a table with a primary key on id (integer) column and a unique
> index on varchar(10) column. Insertion fails when the value of the
> unique column is substring of an existing value. For example, if I
> have a row with the unique column value 'TEST1', it will not allow me
> insert another row with 'TEST' as the value for this column.

Please show the exact sequence of commands that leads to this situation.

--
Peter Eisentraut
http://developer.postgresql.org/~petere/

Peter Eisentraut <peter_e@gmx.net> writes:
> Viswanath Ramineni wrote:
>> I have a table with a primary key on id (integer) column and a unique
>> index on varchar(10) column. Insertion fails when the value of the
>> unique column is substring of an existing value. For example, if I
>> have a row with the unique column value 'TEST1', it will not allow me
>> insert another row with 'TEST' as the value for this column.

> Please show the exact sequence of commands that leads to this situation.

This is probably the bug that was introduced here:

2005-08-24 13:50  tgl

    * src/backend/utils/adt/: oracle_compat.c, varlena.c: Add small
    hack to support use of Unicode-based locales on WIN32.    This is not
    adequately tested yet, but let's get it into beta1 so it can be
    tested.  Magnus Hagander and Tom Lane.

and fixed here:

2005-08-26 13:40  tgl

    * src/backend/utils/adt/varlena.c: The idea of using _strncoll() on
    Windows doesn't work.  Revert to same code as we use on other
    platforms when encoding is not UTF8.

Non-unicode comparison on Windows was pretty thoroughly broken during
that interval :-(

            regards, tom lane