- Product: Aleph
- Product Version: 20, 21, 22, 23
- Relevant for Installation Type: Dedicated-Direct, Direct, Local, Total Care
We are making decisions about OPAC statistics reports and table clearing, and it would be helpful if you could answer the following questions about xxx01 z35 and z69.
KB 5872 seems to imply that z69 replaces z35 in xxx01. However, z35's event types (10 = Search done, 11 = Search above the maximum hit limit, 12 = Empty search) are not included in z69's events. Our z35 file only contains event types 10 and 12. Therefore I assume that 11 is not functional. Correct? I have totaled various event types in z69 in an effort to get a correlation with the number of event type 10s in z35, but have not been successful.
(For example, a total of event types 20 through 25 (those that include the word "search" in their descriptions), inclusive, in z69 is only about half the total of event type 10 in z35. However, a total of all z69 event types is more than double the total of event type 10 in z35.) Is there a way to correlate the two?
Also, what exactly is an "empty search?" One with no search key entered? And, why was it not included in z69?
Finally, is the mechanism that populates z35 being maintained as ALEPH evolves, so that we can count on the accuracy of the data in it?
The handle_event "10" which writes z35 event_type "10" is written by the ./set_server/set_find program.
The www_handle_event's which write z69 records are written by a variety of www_f programs.
We find that the following www_f programs call set_find (and, therefore, write z35 event type "10"s) but do not do any www_handle_event (and, therefore, do not write any z69 record): find_b_permute, find_f (full-text), find_r (refine), find_ug, locate, save_history_find, service_find_aut_bib, and tree_find_b.
Therefore, there are www_server activities which are not included in the z69. It seems that these are, in general, less common, less important activities.
Event-types 20-25 do not include type 26 (Cross sets) or 29 (Scan/Browse).
Therefore, the z69 includes Browses, while the z35 does not. In many OPACs over 50% of all searches could be Browses.
The z35 event-type "12" (empty search) is a zero-hits condition. This should roughly correspond to the z69 records with Z69_NO_HITS = '0'.
The z35 event-type "11" ("Search above the maximum hit limit") should correspond to the z69 records with Z69_NO_HITS > set_result_set_limit . As you have noted, the z35 event-type 11 does not seem to work, therefore, you should rely on the Z69_NO_HITS > set_result_set_limit instead.
Aleph does not read or use the bib z35 in any way. It can grow quite large. If you have SQL routines which read this table, you should run them and delete the records for past years. If you don't have such routines, then you should drop the table every week or so. Or, you can do “s+ abc01” … “drop table z35” (note: make sure you are in the xxx01; you definitely *do* need the ADM z35, which is used to produce circulation statistics), and then remove the z35 from the $data_root/file_list.
In addition, the 10, 11, 12, and 40 lines in the abc01 tab_events.eng table need to have column 2 ("Activate") changed to "N".
Note: Though the bib z35 is has been supplanted by the z69 for *Web* search events the bib z35 may still be needed for *GUI* search events. See KB 16384-48353.
See KB 1892-2130 in regard to the z34 and z69.
- Article last edited: 12-Mar-2016