Adding books recursively...

[OtherOkyay]

Calibre can add books recursively from a directory hierarchy, and this works
fine. However, during flattening out of the directories, directory information is lost.

Say I have a book in a directory dirA/dirB/dirC/book.pdf. I would like
dirA, dirB, dirC to be entered as tags while adding the book into the library.
Later on searching the tags for dirA would list all books under the directory
dirA. Searching for dirC would list only the books under the directory dirC.

Is there an easy way to do this? I didn't see any feature that would allow
me to do this. Would it be possible to do this with a plug-in? I am thinking
that after adding all the books, perhaps adding entries into the tags table
would accomplish this.

Thanks for any help
-Dave

[theducks]

Originally Posted by OtherOkyay

Calibre can add books recursively from a directory hierarchy, and this works
fine. However, during flattening out of the directories, directory information is lost.

Say I have a book in a directory dirA/dirB/dirC/book.pdf. I would like
dirA, dirB, dirC to be entered as tags while adding the book into the library.
Later on searching the tags for dirA would list all books under the directory
dirA. Searching for dirC would list only the books under the directory dirC.

Is there an easy way to do this? I didn't see any feature that would allow
me to do this. Would it be possible to do this with a plug-in? I am thinking
that after adding all the books, perhaps adding entries into the tags table
would accomplish this.

Thanks for any help
-Dave

No
Directory paths can not supply Metadata, only the title.
Calibre can get those books, just not any path data

[wallcraft]

Originally Posted by OtherOkyay

I am thinking that after adding all the books, perhaps adding entries into the tags table
would accomplish this.

You can manually add the tags using Edit Metadata -> Edit Metadata in Bulk -> Add tags, providing you have a way to select the ebooks once in Calibre. Adding the folders one at a time would be one way to make selection easy.

[OtherOkyay]

Tags can be updated manually, but I think it will be tedious. There are tens of
nested directories with hundreds of books, already categorized nicely. Maybe
it is possible to accomplish this automatically.

I was looking at the simple HelloWorld plugin based on FileTypePlugin. I
changed on_postprocess to on_import so that it gets called when a book
is imported.

However when I added a simple line print "Hello", I got a syntax error.
I wanted to print path_to_ebook to see if that variable contains the path
to the book BEFORE or AFTER the import.

[transmitthis]

I'm sure there is a way to do this,
probably involving a simple "folder to text" program
that will give you a list of the folders and files in a directory.
as a text file.
Then some inport function with regular expression..


Anyway good luck

[OtherOkyay]

theducks, wallcraft, transmitthis: thank you for your comments.

There is some progress. The example plugin is called before the book is
imported and contains the full file path to the book.

Tags work like this. There is a tag table is defined as(id int, name text), each
entry is a tag. Each book also has an entry in the books table with a unique
id int. books_tags_link table links the two table with (id int, book int, tag int).
I can update the tags and books_tags_link table from the plugin.

But, here is the problem: when the plugin is called, the book has not been
entered into the book table yet and therefore does not have an id. Without the
book id, it is not possible to make an entry into the books_tags_link table.

I thought of getting the max(id) from books table and incrementing it, but the
system rejects the request because the book ID is not in the known books range.
It also is not reliable that the next ID will always be exactly one more from the
previous one. What is needed is that the calibre calls the plugin after the book
is entered into the books table (as well as before).

I don't know how to go from here. If the book titles are unique, then it is
possible to build a file with two columns: path and book title. and then
run an external program to generate tags and update books_tags_link table.