Request of suggestions for a strategy to create a CLEAN Calibre library

[RotAnal]

Dear Users,
I am quite new in Calibre though I got some experience with other databases (e.g. EndNote).
Here is my starting scenario: a bunch of e-book files (mainly pdf and djvu ones) named in a random fashion (some with ISBNs, others with author/titles in a number of different formats, others with nonsense names). In order to quickly and easily find and manage them, I would like to create a CLEAN Calibre library with metadata by also (auto)renaming the files. The ideal thing would be to let calibre get a pointer field to an external HDD (where the files actually are), perhaps each of them being renamed with its ISBN only.
Which relative fast strategy would you suggest me to accomplish this target?
I tried the "Extract ISBN" plugin (the procedure is rather slow and CPU intensive), but still have no clue of which is the best way to achieve the above result in a reasonable time.
Thanks for attention and a possible answer

[Adoby]

Sorry, I don't think there is any good shortcut.

One way could be to import all books more than once, allowing duplicates, and try to read metadata both from filenames and from inside the books. Then start grouping the books in groups that have correct author, title or ISBN. And use Calibre to fetch missing metadata for each group.

Finally remove duplicates (using binary compare, ISBN and/or Author and Title) and books with bad or missing metadata. Or try to manually fix books with bad metadata if they have not already been added.

You may not have been able to add all books, but at least the ones that had metadata that was good enough to add easily.

What I do:

New books are added to a Calibre Library "Main". As I do this I try to ensure that at least Author and Title is OK. I allow duplicates. I also store all original books with original names in a "original" folder with the date I added the books. I remove DRM first, if present.

I have a custum column "Good". Books that are in a good state. Good metadata, in the formats I need and a nice cover and so on, I set "Good" to Yes.

After I have added books I check for duplicates and try to fix up the books so the new books either becomes "Good" books or if I already have the title, I may delete or replace. Or convert as needed.

Currently I transfer ALL books that are "Good" to my main reading device. About 2000 books, 6 GB. Mostly ePub, some PDF. But I have many more books in the Calibre Library, including duplicates. Originally I never deleted a book from "Main", but they became too many, so I have now started to remove duplicate books that I already have as a better copy or books that obviously are in a state that I don't bother to fix. But I still keep all books in the "original" folder.

Especially I have all my bought books as "Good" books, so I easily can avoid buying the same book twice. But I also have free books from different sources. A lot of those books are also "Good".

If I encounter a book on my device that when reading looks badly formatted or that has bad metadata, I locate it in "Main" and check if there are any alternative copies, and if some of them are better. I might convert using different settings or even edit the book. Sometimes I also do a search, using OS functions, in the original folder to see if I have a better variant there.

I also have a Calibre Library "Problem" where I store some books that need more work, before they can become "Good" books.

I am considering adding another Custom Column, to allow me to filter out books that are in a good state, but that I am not interested in having on my device. But so far that is not many books. I usually only make a book a "Good" book if I like it, want to have it available or if I intend to read it.

Now and then I spend an hour or so trying to make more books in "Main" good. When I am home sick or it is bad weather out. Usually I first try to fix books of authors that already have "Good" books there. It seems that "Main" grows faster than I can make books "Good". But I still tend to make books "Good" faster than I read them. So I am happy that way.

So I don't try to have a pristine Calibre Library. Instead I go for a process that gives me a growing pristine set of books on my reading device, and use Calibre as a tool to achieve that.

[theducks]

I Read EPUB format (and use those to make Mobi for my wife when needed)

So I created 3 custom columns to help track progress :

1) Edited EPUB [Date-Time] that gets set using the 'now' button near it any time I edit the file EXTERNAL to Calibre.

2)OK_FC (OK'd by Flightcrew in Sigil) [T-F-undefined]

3)Master Format [preset tags of formats + Multiple + BAD. Not all formats listed are convertible by Calibre] to help track what I might have started with.

And as a safety net, the real (pre-Calibre) originals are archived.

[travger]

When I first started, I added one folder (files were named rather miscellaneously) to Calibre and then had most damnable time trying to determine which pdf is which. So, if you have duplicates, it's probably faster to add them one by one. AFAIK, Calibre can't grab original filename into comments or custom column, it must be entered manually.
When you have small bunch of books in Calibre, check for duplicates, delete what you want and export renamed files to the new folder (which will eventually replace the folder where they reside now). If you noted down the original filename, you can easily find and delete them from the original location.

[Adoby]

You can use regexes to parse the filenames and get title and author, if present. But if you have a lot of files with differently structured filenames, then that can be labor intensive. It might be better to rename the files first, so their filenames are similar. Title - Author.extension. Or group the files in a few groups with similar filenames. But it could also be a good idea to first attempt to add the books and add title and author from metadata inside the files. Then you don't have to bother with the file names.

[RotAnal]

Thank you very much for your suggestions!
I have been realising that importing books into Calibre may be a tedious task, but I'll be eventually succeeding in doing it. Thanks again.