Quote:
Originally Posted by kiwidude
Of course the downside for people who haven't done this before they add books to their library is how to actually resolve the problem if they then run an external tool, as if they do a direct delete out of the calibre folders they end up with orphaned Calibre database records.
|
Yes, an internal tool integrated into your dupe/detect module is preferred. That's why I mentioned it.
Quote:
I agree with you in that there should be no reason under "normal" circumstances it would be necessary. That "unknown author, unknown title" situation however sounds rather horrific!
|
Now that I think about it, IIRC, the default is actually author=Unknown, title=filename minus extension. In his case, he had the books with some weird filename, but they were originally inside folders with author/title metadata info in the folder names. He lost the author/title folder info on import. I think the recommendation given was to search for Unknown authors, sort by added date, delete them all, fix the filenames to include author/title info with an external program, then re-import, but he still had lots of binary dupes, so he had to address those before import.
It seems to be a pretty rare problem, but it might be useful.
Quote:
I wonder what the best way to integrate that into the dialog would be? Perhaps there should be a new set of radio buttons across the top of "Title/Author", "ISBN", "Binary compare"?
|
Sounds good to me.
Quote:
Would an appropriate implementation be to use the hashlib sha256 stuff or something else?
|
Hash it