Bug 6557: Record bib popularity in totalissues

Bug 6557: Record bib popularity in totalissues

Because updating the total issues count associated with a bibliographic
record on issue could cause a significant load on the server, this
commit adds the syspref UpdateTotalIssuesOnCirc (which defaults to OFF
to match existing behavior). The syspref has the following description:

  Do/Do not update a bibliographic record's total issues count whenever
  an item is issued (WARNING! This increases server load significantly;
  if performance is a concern, use the update_totalissues.pl cron job
  to update the total issues count).

Bug 6557: automatically increment totalissues

Adds the ability to automatically increment biblioitems.totalissues
whenever an item is issued.

To test:
1) Choose a record with at least one item that can circulate
2) Check the value of 942$0 (you may need to look at the plain MARC view
   on the OPAC). Most likely there won't be any 942$0 at all
3) Enable UpdateTotalIssuesOnCirc
4) Check out the item you selected
5) Check the value of 942$0 (you may need to look at the plain MARC view
   on the OPAC). That value should now be one greater than before
6) Discharge the item
7) Disable UpdateTotalIssuesOnCirc
8) Check out the item you selected again
9) Check the value of 942$0 (you may need to look at the plain MARC view
   on the OPAC). That value should not have changed

Bug 6557: add script to update totalissues from stats

NAME
       update_totalissues.pl

SYNOPSIS
         update_totalissues.pl --use-stats
         update_totalissues.pl --use-items
         update_totalissues.pl --commit=1000
         update_totalissues.pl --since='2012-01-01'
         update_totalissues.pl --interval=30d

DESCRIPTION
       This batch job populates bibliographic records' total issues count
       based on historical issue statistics.

       --help  Prints this help

       -v|--verbose
               Provide verbose log information (list every bib modified).

       --use-stats
               Use the data in the statistics table for populating total
               issues.

       --use-items
               Use items.issues data for populating total issues. Note that
               issues data from the items table does not respect the --since
               or --interval options, by definition. Also note that if both
               --use-stats and --use-items are specified, the count of biblios
               processed will be misleading.

       -s|--since=DATE
               Only process issues recorded in the statistics table since
               DATE.

       -i|--interval=S
               Only process issues recorded in the statistics table in the
               last N units of time. The interval should consist of a number
               with a one-letter unit suffix. The valid suffixes are h
               (hours), d (days), w (weeks), m (months), and y (years). The
               default unit is days.

       --incremental
               Add the number of issues found in the statistics table to the
               existing total issues count. Intended so that this script can
               be used as a cron job to update popularity information during
               low-usage periods. If neither --since or --interval are
               specified, incremental mode will default to processing the
               last twenty-four hours.

       --commit=N
               Commit the results to the database after every N records are
               processed.

       --test  Only test the popularity population script.

WARNING

If the time on your database server does not match the time on your Koha
server you will need to take that into account, and probably use the
--since argument instead of the --interval argument for incremental
updating.

=== TESTING PLAN ===

NOTE: in order to test this script, you will need to have some sort of
circulation data already existing in your Koha installation.

1) Disable UpdateTotalIssuesOnCirc
2) Run: misc/cronjobs/update_totalissues.pl --use-items -t -v
3) If you have total checkout data in your item records (i.e. anything
   in 952$l), you should see messages like "Processing bib 43 (1 issues)"
4) Choose one of the lines that shows more than 0 issues, and view the
   record with that biblionumber in the staff client, choosing the "Items"
   tab (moredetail.pl). Add up the "Total checkouts" listed for each item,
   and confirm it matches what the script reported
5) Run: misc/cronjobs/update_totalissues.pl --use-stats -t -v
6) If you have any circulation statistics in your database (i.e. any
   'issue' entries in your statistics table), you should see messages
   like "Processing bib 43 (1 issues)";
7) Choose one of the lines and view the record with that biblionumber in
   the staff client, choosing the "Items" tab (moredetail.pl). If you
   count the number of checkouts listed in each item's checkout history,
   the total should match what the script reported.
8) Check out an item
9) Run: misc/cronjobs/update_totalissues.pl --use-stats
   --incremental --interval=1h -t -v
10) You should see one line reporting a single circ for the bib record
    associated with the item you just checked out (there may be more if
    you checked out any books in the hour prior to running these tests
11) If the results in steps 4, 7, and 10 match the predictions, the
    script worked

This patch to Koha was sponsored by the Arcadia Public Library and the
Arcadia Public Library Foundation in honor of Jackie Faust-Moreno, late
director of the Arcadia Public Library.

Signed-off-by: Liz Rea <wizzyrea@gmail.com>
Tested this with my test data - numbers are correct and updated appropriately.

More importantly - if I do a popularity search, the most popular items *come up first*. Amazing.