No subject

Owen and I did some testing of the item editing problem I reported (bug 599) 
using the default templates, with inconsistent results.  We were able to 
reproduce the error sometimes, but not every time.  So now, I think it would be 
valuable to take a closer look at what actually happened when we first noticed 
the error to see if this is a database problem or a code problem.

First, remember that the problem started when staff had added an item to an 
existing biblio record without a barcode, and then edited that item to add the 
barcode.  The barcode was 31000000109284.  After the edit, this barcode 
appeared correctly when searching the MARC database (bibid 130229), but was 
attached to the wrong biblio (number 37717) in the old Koha DB.  (At this point 
in the database, the bibid and the biblionumber are the same.)

So here's the bibid/biblionumber 37717 from the MARC database:

mysql> select * from marc_subfield_table where bibid=37717;
29 rows in set (0.02 sec)

And here's bibid/biblionumber 37717 from the items table:

mysql> select * from items where biblionumber=37717;
3 rows in set (0.00 sec)

Notice that it look like the edited barcode (31000000109284) replaced barcode 
37000000025523 that appears in the MARC record.  But also notice that the MARC 
record has an accession date (tag 952v) attached to barcode 37000000025523 that 
is actually the date when the edit to barcode 31000000109284 was done!  It 
should actually have no accession date, because no date was in the imported 
MARC record loaded on September 1.

So what happened to barcode 37000000025523 in the old Koha DB?:

mysql> select * from items where barcode='37000000025523';
Empty set (0.00 sec)

It's gone, been replaced by the edited barcode number.
Here's the edited barcode number in the MARC table:

mysql> select * from marc_subfield_table where subfieldvalue='31000000109284';
1 row in set (0.02 sec)

mysql> select * from marc_subfield_table where bibid=130229;
31 rows in set (0.03 sec)

So what happened here?  It looks like the editing process saved the barcode to 
the correct item ('A') in the MARC table, but also saved the accession date to 
an item ('B') attached to a completely different bibid in the MARC table.  And 
it also looks like the barcode and date for item A overwrote the information 
for item B stored in the old Koha DB.  So if that is what happened, why did it 
happen?  (Does the tag and subfield order in bibid 130229 look odd to you?)



------- You are receiving this mail because: -------
You are the QA contact for the bug, or are watching the QA contact.