It happens that librarians, by mistake, open a biblio editor putting in the
URL, by hand, the biblionumber prefixed with a blank (e.g.
.../addbiblio.pl?biblionumber= 123 -- mind the space before 123).
In such a case the editor opens with the right biblio record (i.e. 123)
but, after saving the record, the content of the biblionumber MARC
field (999 $c for a standard MARC 21 installation) results modified and
contains additional initial blanks.
Moreover, while using ES and making a search for the record (with title,
author etc.) we get two records on the result list (instead of one).
This is because in the addbiblio.pl script $biblionumber is taken (and
continuously used) directly from CGI parameter, without any
validation and/or correction.
Test plan:
==========
0. Have a test installation with ES.
1. Open a biblio record in the editor with an added space before
biblionumber value, e.g.:
http://ktd:8081/cgi-bin/koha/cataloguing/addbiblio.pl?biblionumber= 123
Save the record.
2. From the Normal view choose Save -> MARCXML. Open the saved file in
your favourite editor. You should see, at the end, something like:
<datafield tag="999" ind1=" " ind2=" ">
<subfield code="c"> 123</subfield>
(mind the space before 123).
This is not right.
3. Make a search with the title or author's name from the record (e.g.
Henning Mankell for the record 123 from the default ktd data set).
You should get two records instead of one (while using ES).
4. Apply the patch, restart_all. Repeat p. 1 and 2 with a different
biblionumber. Notice the unchanged (i.e. without spaces) value
of 999 $c subfield in the exported record and only one record
as a result of a search.
WNC amended patch - rebased, added conditional in case no bib, moved comments to their own lines
Sponsored-by: Ignatianum University in Cracow
Signed-off-by: Roman Dolny <roman.dolny@jezuici.pl>
Signed-off-by: Nick Clemens <nick@bywatersolutions.com>
Signed-off-by: Katrin Fischer <katrin.fischer@bsz-bw.de>
c74169ba27
