[OPEN-ILS-DEV] Tsearch2 error, setup instructions

[Josh Stompro]

Thank for the response Mike.

I dumped the database and set it back up with the C locale and now that 
error is no longer happening, but I am still not successful with 
importing a record.  The errors I see now in the

[2006-5-29 10:2:53] open-ils.storage [ERR :18918:::] / Eval failed:
[2006-5-29 10:2:53] open-ils.storage [ERR :18918:::] Fingerprint script 
died!
[2006-5-29 10:2:54] open-ils.storage [:18918:::] Returning 65 
Fieldmapper nodes
from metabib xml
[2006-5-29 10:2:55] open-ils.storage [ERR :18918:::] Data extraction 
failed for
record 5: DBD::Pg::db pg_release failed: ERROR:  current transaction is 
aborted,
 commands ignored until end of transaction block

I'm not sure how to diagnose this one. I hope everyone has a good weekend.
Josh


Mike Rylander wrote:
> Sorry for the delay...
>
> On 5/30/06, Josh Stompro <stomproj at larl.org> wrote:
>> When I try to import a record from oclc I get this error.
>>
>> [2006-5-28 10:14:20] open-ils.storage [ERR :9098:::] Wormization failed
>> : DBD::Pg::db pg_endcopy failed: ERROR:  could not find tsearch config
>> by locale
>> CONTEXT:  COPY full_rec, line 1: "4     LDR     \N      \N      \N
>> 01296cam a2200325 a 4500"
>>
>> The problem seems to be that tsearch2 needs some setup work done before
>> it will work with my locale 'en_US'.  There is an easy way around this
>> of just running the command.
>> select set_curcfg('default');
>
> That will work fine.
>
>> But I don't know if that is the correct way, if everything would work if
>> I just inserted that command before the copy.  Could someone tell me how
>> it is setup on the test server?  Do I need to setup tsearch2 to
>> recognize 'en_US'?
>>
>
> I actually create the DB in the C locale, since the only real sorting
> we do is case insensitive anyway (sort by title for search, sort
> library names, etc).
>
>> One other question, what is Wormization?  Wikipedia doesn't know that
>> term, and google thinks that someone by that name has a blog.  Is it
>> converting a MARC record to a long string, or something like that?
>>
>
> It's a made up word ... and I'm open to suggestions.  The OpenSRF
> application that handles indexing of records is called the WoRM (for
> "Write-fired Record Mangler"), and what it does is called wormization.
> ;)
>
>> Thanks
>> Josh
>>
>>
>>
>
>

-- 
--
Lake Agassiz Regional Library - Moorhead MN larl.org
Josh Stompro               | Office 218.233.3757 EXT-139
LARL Network Administrator | Mobile 701.371.3857