Re: How to implent CONVERT ( data_type [ ( length ) ] , expression ) function in postgreSQL - Mailing list pgsql-hackers

From Tom Lane
Subject Re: How to implent CONVERT ( data_type [ ( length ) ] , expression ) function in postgreSQL
Date
Msg-id 32443.1414945719@sss.pgh.pa.us
Whole thread Raw
In response to Re: How to implent CONVERT ( data_type [ ( length ) ] , expression ) function in postgreSQL  (rohtodeveloper <rohtodeveloper@outlook.com>)
List pgsql-hackers
rohtodeveloper <rohtodeveloper@outlook.com> writes:
> I  need the convert function because that Our application will be switched from SQL Server to PostgreSQL.
>> For the options you suggested:

>> 1) Pass in datatype as string and deparse and process in the function.
>> 2) Are you referring to pg_convert here?

> 1) is yes. but I want to use the CONVERT ( data_type [ ( length ) ] , expression ) just as same as in the SQLServer,
SO,that doesn't  work.2)  I mean modifying the 'src\backend\parser\gram.y'  file.  There will be a grammer conflict
withthe PostgreSQL self's convert(data,src_encoding_name,dest_encoding_name) function.
 

So what?  Presumably your SQL-Server-based app doesn't use that function.

You could probably make it work anyway by introducing two new productions,
one that implements the CAST-equivalent syntax and one that defines
extract() with a regular func_arg_list argument list.  But I'm not sure
I see the point if you're building a private fork.

On the whole I agree with the other commenters suggesting that fixing your
app to use SQL-standard syntax would be a better answer in the long run.
It's quite unlikely that you're going to be able to hack Postgres to be
bug-compatible with SQL Server in every last respect, so trying to run
your app totally unmodified from its present state seems like a fool's
errand.  Anyplace where you can dodge the problem by switching to
spec-mandated syntax that both DBMSes understand, you're way ahead of
the game if you fix it that way.
        regards, tom lane



pgsql-hackers by date:

Previous
From: Andrew Dunstan
Date:
Subject: Re: Let's drop two obsolete features which are bear-traps for novices
Next
From: Tom Lane
Date:
Subject: Re: how to handle missing "prove"