1Nevov
Posting this as its own bug report, originally mentioned at: https://www.librarything.com/topic/335019#9053333 but hasn't picked up attention yet.
Following the recent change to preserve series information through work combination (in the above topic), there is one thing not working quite right.
If the lesser work being combined has the series information on it, its label, eg. "1", gets erased in the combination. The series, and the grouping, eg. "Core", do transfer onto the main work, but it ends up with a blank label.
Following the recent change to preserve series information through work combination (in the above topic), there is one thing not working quite right.
If the lesser work being combined has the series information on it, its label, eg. "1", gets erased in the combination. The series, and the grouping, eg. "Core", do transfer onto the main work, but it ends up with a blank label.
2SimoneA
I hadn't noticed this yet, but see it now on https://www.librarything.com/nseries/65735/Zagor. It would be great if the label is preserved!
3knerd.knitter
>1 Nevov: >2 SimoneA: I'm working on this, but I'm going to need an example of two works that need to be combined and would have this problem so that I can verify if my fix works. Can anyone provide one or two examples that can remain unfixed that I can use to test?
4AnnieMod
>3 knerd.knitter: If you cannot find one in the wild, it is easy to build one either with a test series (or even test works) or with a live one that will get fixed quickly once you are done testing. All you need is:
- Bigger work not in the series
- Smaller work in the series with a number
- Bigger work not in the series
- Smaller work in the series with a number
5knerd.knitter
>4 AnnieMod: Thanks!
6knerd.knitter
I think this is fixed now. If someone can let me know when they find a real one in the wild, that would be great!
7Nevov
Thanks! It seems to be fine for the usual cases. Also tried some multiple combinations; preserving the smaller of three, the middle of three; and work 1 in series A and series B, with work 2 also in series A; etc. and it did what was expected.
I managed to break it a little bit:
https://www.librarything.com/nseries/groups/62287/
The group 'Test' that's listed on the above – I combined a work which was in group Test, with a work that was in Core. Chose to preserve the Core one, meaning no works remained in Test. Usually a group would disappear when it has no works left in it (if editing via the Add to/Organize... menu option), but this one has stuck around. Tried various ways to get it to vanish (adding a further work to Test then removing) but no joy.
Same with: https://www.librarything.com/nseries/groups/405463/
Maybe it's not a problem since that Sort Groups page is 'under the hood' for series, and a link to it only even shows up on the main series page when there are multiple custom groups in use. But FYI in case it's something fixable. I don't think we want a delete group option outright since the capacity to wrongly delete correct groupings could be dangerous without a restore/last good edit option. (Straying into RSI but maybe with the complexity we now have for series, being able to do a wiki-style restore would be great.)
I managed to break it a little bit:
https://www.librarything.com/nseries/groups/62287/
The group 'Test' that's listed on the above – I combined a work which was in group Test, with a work that was in Core. Chose to preserve the Core one, meaning no works remained in Test. Usually a group would disappear when it has no works left in it (if editing via the Add to/Organize... menu option), but this one has stuck around. Tried various ways to get it to vanish (adding a further work to Test then removing) but no joy.
Same with: https://www.librarything.com/nseries/groups/405463/
Maybe it's not a problem since that Sort Groups page is 'under the hood' for series, and a link to it only even shows up on the main series page when there are multiple custom groups in use. But FYI in case it's something fixable. I don't think we want a delete group option outright since the capacity to wrongly delete correct groupings could be dangerous without a restore/last good edit option. (Straying into RSI but maybe with the complexity we now have for series, being able to do a wiki-style restore would be great.)
