Using Calibre Python API example

[sigizmund]

Hey folks,

I am trying to understand if I can write a Python code running not as a plugin which will be manipulating Calibre without UI. Example: add book to the specified Calibre library. Obviously starting calibredb command is one option, but I'm interested if I can just `import` some libraries and manipulate database in-process. In pseudocode:

import calibre.database as db

database = db.OpenDb('/path/to/library')
database.AddBook('/path/to/book.epub')
database.Close()

Thanks!

[kovidgoyal]

https://manual.calibre-ebook.com/db_api.html

https://manual.calibre-ebook.com/develop.html

[chaley]

Originally Posted by sigizmund

Hey folks,

I am trying to understand if I can write a Python code running not as a plugin which will be manipulating Calibre without UI. Example: add book to the specified Calibre library. Obviously starting calibredb command is one option, but I'm interested if I can just `import` some libraries and manipulate database in-process. In pseudocode:

import calibre.database as db

database = db.OpenDb('/path/to/library')
database.AddBook('/path/to/book.epub')
database.Close()

Thanks!

Yes, you can do what you want. Use the calibre-debug command, and in particular calibre-debug -e. The entire calibre api is available.

This example prints all the tags in a library, sorted.

from calibre.utils.icu import sort_key
def init_cache(library_path):
	from calibre.db.backend import DB
	from calibre.db.cache import Cache
	backend = DB(library_path)
	cache = Cache(backend)
	cache.init()
	return cache

cache = init_cache(library_path = sys.argv[1])
all_tags = sorted(list(cache.all_field_names('tags')), key=sort_key)
for t in all_tags:
	print(t)

You execute it with

calibre-debug -e all_tags.py PATH-TO-LIBRARY

[sigizmund]

Thanks folks! Are there any examples of set up of how I can set up development environement for this (ideally using something like pyvenv so it depends less on external configuration)?

[kovidgoyal]

There's no external configuration and no need for pyenv. Write your script, run it with calibre-debug -e. If your script requires python packages that are not bundles with calibre, simply add the path to them at the top of the script with


sys.path.append('/path/to/wherever your python package is')

[sigizmund]

Got it - thanks for help!

[Xwang]

Is it possible to use this method even if an instance of calibre is already running or is it necessary (or better) to close calibre before running calibre-debug?
I'm going to use it in a shell script and I'm sure that I will not be doing any modification in calibre while the shell script will be running.

[kovidgoyal]

depends on what you plan to do with calibre-debug. Any kind of database/library access should not be done while calibre is running. Anything else should mostly be fine

[Xwang]

Since I would like to modify a metadata, I suppose it is better to do that when calibre is not running, then. Is it correct?

[kovidgoyal]

if youa re modifying metadata in a file use ebook-meta.exe if you are modifying it in the calibre library use calibredb.exe

[Xwang]

OK, thanks again