On Wed, 2012-03-14 at 21:52 -0700, Mark Phillips wrote:
> I am not familiar with sed, except for some trivial bits I nicked off
> the web. Enough to know it works, and to be dangerous. Nonetheless,
> using SED may be the way to go as there are two tables that contain a
> bit over 3,000,000 rows eac
On Mar 14, 2012, at 6:32 PM, Andy Colson wrote:
> On 03/14/2012 08:16 PM, Scott Marlowe wrote:
>>
>>
>> Can you run it through sed and replace the "-00-00 00:00:00" to
>> NULL (no quotes) ? That should work.
>>
>
> I think COPY (depending on arguments) uses \N by default.
>
> Another op
On Wed, Mar 14, 2012 at 7:37 PM, Andy Colson wrote:
> On 03/14/2012 08:32 PM, Andy Colson wrote:
>>
>> On 03/14/2012 08:16 PM, Scott Marlowe wrote:
>>>
>>> On Wed, Mar 14, 2012 at 6:47 PM, Mark Phillips
>>> wrote:
I am migrating a data set from Oracle 8i to PG 9.1. The process is to
>>>
On 03/14/2012 08:32 PM, Andy Colson wrote:
On 03/14/2012 08:16 PM, Scott Marlowe wrote:
On Wed, Mar 14, 2012 at 6:47 PM, Mark Phillips
wrote:
I am migrating a data set from Oracle 8i to PG 9.1. The process is to export data into
csv files, then use the pg "copy table from file csv header" sta
On 03/14/2012 08:16 PM, Scott Marlowe wrote:
On Wed, Mar 14, 2012 at 6:47 PM, Mark Phillips
wrote:
I am migrating a data set from Oracle 8i to PG 9.1. The process is to export data into
csv files, then use the pg "copy table from file csv header" statement to load
the tables.
There are a nu
On Wed, Mar 14, 2012 at 6:47 PM, Mark Phillips
wrote:
> I am migrating a data set from Oracle 8i to PG 9.1. The process is to export
> data into csv files, then use the pg "copy table from file csv header"
> statement to load the tables.
>
> There are a number of date columns in the tables that
