Auto-Merge Question re: Duplicate Formats

[barium]

Is it possible to prevent the auto-merge on import option from silently ignoring formats if a duplicate of the format exists in the library? This would be helpful when you have different versions of the same book in the same format.

If that doesn't make sense: suppose you have an RTF copy of The Origin of Species by Charles Darwin in your library, and you import a set of books, one of which is a different RTF copy of The Origin of Species by Charles Darwin. The auto-merge feature will ignore the second copy. Sometimes, the second copy is a different edition, and you don't want it to be ignored.

[itimpi]

Unfortunately not.

There has been quite a lot of discussion recently about enhancing the merge capabilities to support this sort of requirement. However whether anything will result in a future Calibre release is not clear.

[DoctorOhh]

Originally Posted by barium

Is it possible to prevent the auto-merge on import option from silently ignoring formats if a duplicate of the format exists in the library? This would be helpful when you have different versions of the same book in the same format.

You can't get the automerge feature to ignore it but you can uncheck the automerge feature. With the automerge off calibre should pop up a dialog saying it found duplicates do you want to add them. Saying yes will create new entries for these books. They end up has different book entries but both versions are now in your library and you can tag them as appropriate to keep track of them.

[barium]

Originally Posted by dwanthny

You can't get the automerge feature to ignore it but you can uncheck the automerge feature. With the automerge off calibre should pop up a dialog saying it found duplicates do you want to add them. Saying yes will create new entries for these books. They end up has different book entries but both versions are now in your library and you can tag them as appropriate to keep track of them.

This is what I do, and what I'll stick with. Thanks for the responses. Also, looks like I posted to the wrong forum - sorry about that.

[Starson17]

Originally Posted by barium

Sometimes, the second copy is a different edition, and you don't want it to be ignored.

What did you want done with it?
1) Did you want it to ask you what to do for every book?
a) it's no longer automerge - it's sit-around-and-wait-for-it-to-ask-merge
b) don't you want to know what the other "similar author/title" book is and don't you want to pen it and view it to compare for quality?

2) Should it automatically ignore some, but not others? - How do you automatically decide which?

3) Should you change some setting before you add the books so this time it ignores, but next time it overwrites, and the next time it creates a new record?

[barium]

The more I think about it, the more I think the behavior that would work best for me doesn't make sense for the program. What I'd like is for an auto-merge to occur where there's no duplicate format, but for a new/duplicate book entry to be created where there is a duplicate format.

Then, in a separate process (e.g. when I went to read the title in question), I could review duplicate entries to determine whether they're different editions or the same thing. If they're the same, I could delete one of them. If they're different, I could manually update the metadata accordingly. This seems like inconsistent behavior in a way, though. If you use the automerge feature at all, you can end up with different editions in the same book entry if the different editions happened to be in different formats.

[Starson17]

Originally Posted by barium

The more I think about it, the more I think the behavior that would work best for me doesn't make sense for the program.

?? Why?

What I'd like is for an auto-merge to occur where there's no duplicate format,

That happens now.

but for a new/duplicate book entry to be created where there is a duplicate format.

That's one of the three options being discussed (user would get to choose the desired option). (The other two are to ignore duplicate formats, as we do now or to overwrite a duplicate format with the new one.)

Then, in a separate process (e.g. when I went to read the title in question), I could review duplicate entries to determine whether they're different editions or the same thing. If they're the same, I could delete one of them. If they're different, I could manually update the metadata accordingly.

Yes, that's how that option would be used. All books would be added, instead of ignoring matching formats as we do now.

This seems like inconsistent behavior in a way, though. If you use the automerge feature at all, you can end up with different editions in the same book entry if the different editions happened to be in different formats.

Yes, that's a risk of doing it automatically, but I see no way to avoid that.

[barium]

That's one of the three options being discussed (user would get to choose the desired option). (The other two are to ignore duplicate formats, as we do now or to overwrite a duplicate format with the new one.)

That sounds brilliant. Ideally, you'd also have a way to search for duplicate book entries created during auto-merge.

Originally Posted by barium

I think the behavior that would work best for me doesn't make sense for the program.

Originally Posted by Starson17

?? Why?

Only because of the inconsistency I described. With this behavior, you avoid merging because you may have a different edition in one case (where duplicate formats get separate book entries) but you ignore the possibility that a different edition exists in the other (where different formats are merged). The auto-merge feature is consistent in this regard right now - it always ignores the possibility that a different edition exists. (Assuming that the program is designed to have different editions stored in separate book entries, which I think it is, because you can only have one of each format per book entry.)