https://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=42044
--- Comment #18 from Martin Renvoize (ashimema) <[email protected]> --- Thanks for this, Janusz. I have attached my signed-off version of the patches (the only change is the Signed-off-by line), but I am deliberately leaving the status at Needs Signoff until we have discussed the points below. What I tested (current main, 80f504edd9b): - Patches apply cleanly; the atomic update runs and is safe to re-run. - t/Koha/Util/MARC.t, t/db_dependent/Koha/Policy/Record.t, t/db_dependent/Biblio.t and t/db_dependent/AuthoritiesMarc.t pass. - The 17 db_dependent tests that call ModBiblio all pass. - The QA script passes (only a spelling false positive, 'acn', in t/db_dependent/Biblio.t). - A deliberate cataloguer change to LDR/5 is respected: n->p, n->d and c->n are all stored as entered. Not tested by me: the preference page in the browser, the authority steps of the test plan by hand, and UNIMARC. Points to discuss (none of them blocking by my reading): 1. Records already marked 'd'. A record whose LDR/5 is already 'd' becomes 'c' on any ordinary edit, because "unchanged from the old value" is treated like any other value. That silently un-deletes a record that has been flagged for misc/cronjobs/delete_records_via_leader.pl (bug 11084). Should 'd' be left alone, so that it is only changed by the cataloguer on purpose? 2. Retyping the same value. If the cataloguer deliberately sets LDR/5 to the value it already had (for example 'n' to 'n'), we cannot tell that from not touching it, so it becomes 'c'. This is probably acceptable, but worth a line in the preference description or the manual. 3. Same-day edits with logging off. With CataloguingLog (or AuthoritiesLog) off, a record created today and edited today still gets 'c', because we cannot establish who created it. That follows the code's stated rule ("if not sure, set LDR/5"), but the preference text does not mention it. 4. Small code nits in Koha/Policy/Record.pm: 'return unless $oldrecord' can never be reached because oldrecord is a mandatory parameter, and the record_type regex (/biblio|authority/) is not anchored. 5. POD typo in Koha/Util/MARC.pm: '=cut;' should be '=cut'. 6. update_record_status only accepts [acdnp]. UNIMARC also defines 'o'. Not used by this bug, but worth keeping in mind if the helper is reused (see also bug 43792). Happy to sign off once you have had a chance to look at 1 and 3 in particular. -- You are receiving this mail because: You are watching all bug changes. _______________________________________________ Koha-bugs mailing list -- [email protected] To unsubscribe send an email to [email protected] website : http://www.koha-community.org/ git : http://git.koha-community.org/ bugs : http://bugs.koha-community.org/
