Re: COPY FROM wish list

Scott Bailey <artacus(at)comcast(dot)net> writes:
> PgFoundry has http://pgfoundry.org/projects/pgloader/
> It is a step in the right direction but definitely not as powerful as
> sql*loader.

Yeah, it's only offering what I needed and what I've been requested to
add. So far there's support for INFORMIX UNLOAD files, mysqldump -t
files, fixed with files, ragged files, and some more.

And it also support python threading for some parallel workload, either
loading several files at once or several chunks of the same file, and
then 2 modes are possible.

If you need more, try asking, you never know. I'm still on the hook to
maintaining it, though I've not received any bug report in a long
while. I guess it's not much used anymore.

I've been proposed to replace the pgloader.conf INI file with a custom
COPY command parser exposing all the options, and will consider that
sometime in the future.

> I've been writing a Postgres equivalent that does Oracle SQL*Loader/DB2 Load
> style import, and external tables using CSV, fixed with and XML data
> sources. But its not production ready. I'm hoping SQL/MED makes it in to the
> next release of Postgres so I can throw it all away :)

XML support in pgloader will certainly take the form of applying user
given XSLT filter that outputs CSV from the XML. That's the option
requiring the less code in pgloader I guess :)

I'd be happy to see pgloader deprecated by PostgreSQL offering its
features.

Meantime, do you want to join the fun on the pgloader front?
--
dim