tab_acc_category and Authority Linking
- Article Type: General
- Product: Aleph
- Product Version: 18.01
Description:
It happens since the recent upgrade to v18.
According to KB 8192-10071, there is possibility for SQL to update the z01_category field works to cause correct linking if we aren't able to do the bib indexing.
1.Can you please check if we have correct settings in cun01 and cun10 tab_acc_category and what SQL can we use to update the records?
a) When you do an author browse search for Dick Philip K why do you get an authority file with 58 records, but below another record with same name with 18 records instead of collapsing into one record.
58 Dick, Philip K - CUN10 Prod AUT, 10010, aab, UPD=N
18 Dick, Philip K
b) the file for the author Steinberg Blema S. which had 8 records split into 2 files.
6 Steinberg, Blema S CUN10 Prod AUT, 10010, aab, UPD=N
2 Steinberg, Blema S
Resolution:
As shown below, there are two z01 Browse Headings records for AUT Dick, Philip. The first has z01_category ZZZ, and the second has z01_category 00:
1st:
01 z01_acc \
02 z01_rec_key \
03 acc_code ..............AUT
03 alpha .................L
03 filing_text ...........DICK PHILIP K
03 filing_sequence .......00001
02 z01_acc_sequence ........001122863
02 z01_hash ................590969258372
02 z01_aut_tag .............10010
02 z01_rec_key_4 \
03 aut_library ...........CUN10
03 aut_doc_number ........000307568
...
02 z01_category ............ZZZ
02 z01_normalized_text .....$$-dick, philip k
02 z01_display_text ........$$aDick, Philip K
2nd:
01 z01_acc \
02 z01_rec_key \
03 acc_code ..............AUT
03 alpha .................L
03 filing_text ...........DICK PHILIP K
03 filing_sequence .......00002
02 z01_acc_sequence ........023039790
02 z01_hash ................590969258372
02 z01_aut_tag .............
02 z01_rec_key_4 \
03 aut_library ...........-CHK-
03 aut_doc_number ........000000000
02 z01_acc_sequence_see ....000000000
02 z01_number_of_doc .......00000
02 z01_cataloger \
03 cataloger_name ........BATCH-UPD
03 cataloger_level .......00
02 z01_open_date ...........20080810
02 z01_update_date .........20080810
02 z01_cataloger_library ...
02 z01_non_filing_char .....00
02 z01_update_doc ..........Y
02 z01_update_z0102 ........
02 z01_ref_type ............
02 z01_category ............00
02 z01_normalized_text .....$$-dick, philip k
02 z01_display_text ........$$aDick, Philip K
Case "b" (Steinberg, Blema) is exactly the same.
We see in tab11_acc that the following tags are included in the AUT index:
apps-r-18(1) CUN01-ALEPH16>>grep AUT tab11_acc
100## AUT -e46
110## AUT -e46
111## AUT -46
700## AUT -efhklmnoprst456
710## AUT -efghklmoprst456
711## AUT -fhklpst456
800## AUT -efhklmnoprstv46
810## AUT -efhklmoprstv46
811## AUT -fhklpstv46
If the "Dick, Philip" is in the 100 or 700 field (fields which end in "00"), then it is connected to the 2nd z01, but if it is in any of the other fields, it is connected to the 1st z01.
If you want records with the heading in the 100 and 700 field to share the same z01 as the records with the heading in other fields, then you should make the 100 and 700 fields in tab_acc_category have "0" in column 2, indicating that they do not use the categories mechanism.
If it is your intention to not use the category mechanism, the above change to tab_acc_category will accomplish this for records which are being sent to the server from this point on.
As far as existing records go, I don't believe that SQL can be used to make it so that just a single heading displays -- such as in this case of Dick, Philip. There are two headings. The only way to get the titles which are linked (via z02) to the 2nd heading (with z01_category "00") to be linked to the 1st heading (with z01_category "ZZZ") is to rerun the Browse Headings indexing jobs.
If you do not want to use the category mechanism at all, then you should do the following:
1. Immediately before running the Headings jobs (step 2), change the $data_tab/tab_acc_category in both cun01 and cun10 so that all column 2 has "0" for all of the lines. (Some of the lines are currently "1".)
2. Run the Headings jobs (p_manage_02, p_manage_17) in cun10.
3. Run the Headings jobs (p_manage_102, p_manage_02, etc.) in cun01.
If the only change you are making is turning off the category mechanism, then rather than doing running the Headings jobs (steps 2 and 3), you could do the following SQL (in abc01 and abc10):
SQL> update z01 set z01_category = 'ZZZ' where z01_category not = 'ZZZ';
After doing this, you will find that the problem described in this SI ("When you do an author browse search for Dick Philip K why do you get an authority file with 58 records, but below another record with same name with 18 records instead of collapsing into one record") will no longer occur.
- Article last edited: 10/8/2013