• Article Type: General
• Product: Aleph
• Product Version: 18.01

Description:
Site had a problem where they didn't have enough Oracle space to create the z00p table in the $usr_library for the entire database. However, they wanted to make a subset of exported records that spanned the entire database. They have more than 2 million records but only wanted 30,000 out, and that 30,000 was from all record number keys.

First, I thought they could run the selected records through ue_21 (ue_01) but this seems only to update existing records - it doesn't put new ones in.

Publish-04, the job that puts records into z00p, doesn't take a file of bib keys -- it only takes a publishing set (from $data_tab/tab_publish) or a straight range of records.

Publish-06, which exports from z00p and makes a tar file, does take a file of bib keys but all of the bib keys need to have been PUT in z00p first with publish-04. Since site had space problems, this was a problem.

Resolution:
I created this workaround for making a logical base based on a file of record numbers. This process will create a subset of 100 records for the Primo export on the Aleph test server:

1. in /exlibris/aleph/u18_1/abc01/tab/import ($data_tab/import), make a fix file which will add the tag PRM $aPRIMO to records I choose:
>> cat make_subset.fix
!-!!!!!-!!-!-!!!-!!!-!!!!!-!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!-!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
1 LDR ADD-FIELD PRM ,L,$$aPRIMO

2. Add this fix to the tab_fix file (abc data_tab)
>> grep PRIMO tab_fix
PRIMO fix_doc_do_file_08 make_subset.fix

3. index the tag PRM in the WPR keyword index:
>> grep WPR tab11_word tab00.eng
tab11_word:PRM 03 WPR
tab00.eng:H WPR W-030 00 00 W-PRIMO subset

4. Add a logical base to $alephe_tab/tab_base that will reference this keyword index:
>> grep PRIMO $alephe_tab/tab_base.eng
ABCPRIMO ABC PRIMO Subset ABC01 ABC01 Y wpr=primo

5. Restart lib_batch, ue_01, pc_server

6. Run p_manage_37 to add the PRM tag to your set of 100 records. Input file: abc01/scratch/primonums, Fix section PRIMO. The records will have the PRM $a primo tag added to them, and then go through ue_01 and be indexed. Becaues they are indexed, they will be available to that logical base ABCPRIMO.

7. Change abc01/tab/publish to have a PRIMO-SUBSET with the ABCPRIMO logical base:
>> grep PRIMO tab_publish
PRIMO-FULL ABC01PUB N PFULL MARC_XML
PRIMO-SUBSET ABCPRIMO N PFULL MARC_XML
PRIMO-AVAIL N AVA MARC_XML

8. Drop the z00p table in abc50 with UTIL-A-17-1-z00p to clean out the table.

9. Run p_publish_04 with PRIMO-SUBSET as the set name.

10. Run p_publish_06 with either PRIMO-SUBSET or all as the set

• Article last edited: 10/8/2013