cvs: pear /MDB2 package.php package_ibase.php package_mssql.php package_my

Lorenzo Alberton wrote:
> quipo		Fri Feb 23 19:57:36 2007 UTC
> 
>   Modified files:              
>     /pear/MDB2/MDB2/Driver/Reverse	Common.php ibase.php
mssql.php 
>                                   	mysql.php mysqli.php
oci8.php 
>                                   	pgsql.php sqlite.php

>     /pear/MDB2	package.php package_ibase.php
package_mssql.php 
>               	package_mysql.php package_mysqli.php
package_oci8.php 
>               	package_pgsql.php package_sqlite.php 
>   Log:
>   added a new boolean parameter to
getTableIndexDefinition() and getTableConstraintDefinition()
in the Reverse module to disable the 'idxname_format' option
and use the index name as provided

do you think this is a good idea?
should the users simply set the option to %s if they want to
disable it?
i think you may be opening the flood gates for a lot of
parameters that 
just complicate the code ...

regards,
Lukas

-- 
PEAR Development Mailing List (http://pear.php.net/)
To unsubscribe, visit: http://www.php.net/unsub
.php

Hi Lukas,

>> quipo        Fri Feb 23 19:57:36 2007 UTC
>>
>>   added a new boolean parameter to
getTableIndexDefinition() and 
>> getTableConstraintDefinition() in the Reverse
module to disable the 
>> 'idxname_format' option and use the index name as
provided
> 
> do you think this is a good idea?
> should the users simply set the option to %s if they
want to disable it?
> i think you may be opening the flood gates for a lot of
parameters that 
> just complicate the code ...

maybe I was a bit too hasty in examining a bug report,
and was misled by a statement that I took for granted,
so I implemented the solution that seemed the lesser evil.

I still believe that the root cause of all this mess
is the 'idxname_format' option itself, that maybe
should have been implemented in a transparent way
only in the driver that had problems (oci8 or pgsql,
I don't remember now), just like it's done in mysql
for PKs.

Anyway, I've already told you that I think that a
fallback mechanism should be implemented because not
everyone created the tables and indices within MDB2.
I just realized that I can check if the index exists
with an extra query, and eventually falling back
to the original name without suffix, thus avoiding
the extra parameter... I hadn't thought about that.

If you don't have a better solution, I'll revert
my latest changes and implement the "extra query"
workaround.

BTW: what do you think about this one?
http://pear.php.net/bu
gs/10024

Best regards,
-- 
Lorenzo Alberton
http://pear.php.net/us
er/quipo
____________________________________________________________
_______
Quipo Free Internet - 2 email, 150 Mb di spazio web e molto
di pił.
ADSL, Hardware & Software Online Store

-- 
PEAR Development Mailing List (http://pear.php.net/)
To unsubscribe, visit: http://www.php.net/unsub
.php