[lug] question on Relational database record loading

[Bear Giles]

Lee Woodworth wrote:

>I think you mean the COPY command. For Postgresql 8.1.2
>\h LOAD displays:
>   Command:     LOAD
>   Description: load or reload a shared library file
>   Syntax:
>   LOAD 'filename'
>
>While \h COPY displays:
>   COPY tablename [ ( column [, ...] ) ]
>       FROM { 'filename' | STDIN }
>       [ [ WITH ]
>             [ BINARY ]
>             [ OIDS ]
>             [ DELIMITER [ AS ] 'delimiter' ]
>             [ NULL [ AS ] 'null string' ]
>             [ CSV [ HEADER ]
>                   [ QUOTE [ AS ] 'quote' ]
>                   [ ESCAPE [ AS ] 'escape' ]
>                   [ FORCE NOT NULL column [, ...] ]
>
>
>Haven't personally used the COPY command, so can't
>really help with the particulars.
>
You're right.  I was thinking of LOAD since I've written user-defined 
types and you need that to load your, uh, user-defined types.  In my 
case there was one 'load' statement followed by a kiloline of type and 
function declarations... but I ended up with a database that had native 
support for X.509 digital certificates and encrypted private keys.  Too 
bad that was for PostgreSQL 6.x.

I've used COPY on a few occasions.  It's faster than individual INSERT 
statements.  It's far faster.  IIRC it was something like 6x or 8x 
faster for my datasets.  Your mileage may vary.