On Wed, Jun 19, 2013 at 6:00 PM, Albe Laurenz <laurenz.albe@wien.gv.at> wrote:
> sachin kotwal wrote:
>> While migrating my application from DB2 to PostgreSQL.
>>
>> I want to migrate following functions in PostgreSQL.
>>
>> TO_CHAR() in DB2 which can take three arguments as follows:
>>
>> SELECT TO_CHAR(CURRENT_DATE,'YYYY-MM-DD',112.50) FROM SYSIBM.SYSDUMMY1
>>
>> I am not sure what is the purpose of third argument in TO_CHAR() function of
>> DB2.
>>
>> I think we can create one user define function with name TO_CHAR which can
>> take three argument like DB2.
>>
>> I want to know that if we want to create user define function TO_CHAR().
>> What should be the data type of third argument?
>>
>> which can handle maximum data types and more suitable for real time
>> scenarios.
>
> Look at IBM's documentation at
> http://pic.dhe.ibm.com/infocenter/db2luw/v10r1/index.jsp?topic=%2Fcom.ibm.db2.luw.sql.ref.doc%2Fdoc%2Fr0007110.html
>
> TO_CHAR is a synonym for VARCHAR_FORMAT.
> The third argument is an optional locale name, like 'en_US',
> which will be used to generate the text parts of the
> format you chose.
>
> For a format like 'YYYY-MM-DD' this has no effect at all
> and can be omitted.
>
> PostgreSQL has no such capability. Unless you need that and
> want to code it yourself, the best solution would be to
> write a function that just ignores the third argument.
>
Just wondering:
If this particular function is to be used repeatedly in a single
query, would the cost of having a wrapper function around the original
function be too large? For example, if this function appears in a
WHERE clause against a table containing millions of rows.
--
Amit Langote
