Re: exposing COPY API - Mailing list pgsql-hackers

From Andrew Dunstan
Subject Re: exposing COPY API
Date
Msg-id 4D4B6FC5.2060000@dunslane.net
Whole thread Raw
In response to Re: exposing COPY API  (Itagaki Takahiro <itagaki.takahiro@gmail.com>)
Responses Re: exposing COPY API  (Itagaki Takahiro <itagaki.takahiro@gmail.com>)
List pgsql-hackers

On 02/03/2011 09:43 PM, Itagaki Takahiro wrote:
> On Fri, Feb 4, 2011 at 11:32, Andrew Dunstan<andrew@dunslane.net>  wrote:
>> Umm, where? I can't find this in the documentation
>> <http://pgbulkload.projects.postgresql.org/pg_bulkload.html>
> Here:
> http://pgbulkload.projects.postgresql.org/pg_bulkload.html#filter
>
>> The object, as I have explained previously, is to have a FDW that returns a
>> text array from a (possibly irregularly shaped) file.
> I remember the text array proposal, but if the extension is written in C,
> it can only handle one kind of input files. If another file is broken
> in a different way, you need to rewrite the C code, no?


AFAICT, this doesn't support ragged tables with too many columns, which 
is my prime use case. If it supported variadic arguments in filter 
functions it might come closer.

But where does the COPY syntax you showed come in?

Also, a FDW allows the COPY to be used as a FROM target, giving it great 
flexibility. AFAICT this does not.

cheers

andrew


pgsql-hackers by date:

Previous
From: Itagaki Takahiro
Date:
Subject: Re: exposing COPY API
Next
From: Greg Smith
Date:
Subject: Re: [PERFORM] pgbench to the MAXINT