MobileRead Forums - View Single Post

lrpirlet · 10-17-2022, 10:16 AM

[Metadata Source Plugin] babelio_db

This plugin gets the metadata for books from the babelio site.

The Babelio website
The address to access Babelio is <https://www.babelio.com/>.

The Babelio homepage says in french: "Babelio est un réseau social dédié aux livres et aux lecteurs. Il permet de créer et d’organiser sa bibliothèque en ligne, d’obtenir des informations sur des oeuvres, de partager et d’échanger ses goûts et impressions littéraires avec d’autres lecteurs."

This translate to: "Babelio is a social network dedicated to books and readers. It allows you to create and organise your online library, get information about works, to share and exchange literary tastes and impressions with other readers."

There is no need to log in to obtain information, (and that avoids the need of the management of an account in the plugin)

Main Features

Can retrieve title, author, comments, publisher, publication date, series and tags.
Starting with test version 0.8.5, oct 2023, can select tags based on pertinence and colour of the tags
Via configuration, adds, or not, a title and the link in the comments (so that this link may be followed from within a catalog with this book)
Via configuration, fetches, or not, the cover related to the volume (same as VdF plugin)
Via configuration, the content of the log may be modulated
Minimum requirement calibre_version = (6, 3, 0)

I developed this plugin based on the existing VdF plugin to solve a series of limitations that I encounter:

There is no reference (id) that allows to find the URL, nor a way to add an id from the url due to the lack of the following two routines:
- get_book_url: used by calibre to convert the identifier to a URL...
- id_from_url : takes an URL and extracts the identifier details...
As soon as the Babelio content is retrieved, ALL the information is translated into utf8, ALL outputs (logs...) and inputs (what is read on the site...) are in utf8... the multiple conversions make the code confusing to me.
Provides only one match even if two titles answer the search query
The compatibility with python 2.x seems obsolete and confusing to me.
Identify() should be invoked with NAME_ID then ISBN and if it doesn't work with title + author... title only.
A config.py is OK but I prefer to use the built-in config, quite sufficient.
Babelio gives in the title the name of the series and its volume number (order in the series), I'll try to introduce it (and check that all my series meet it...)
I much prefer comments in HTML and beautifulsoup (used in calibre by the way)
I hate this search and substitution language (regular expression) probably because I don't understand almost anything... well, I only do a little bit!
Last but not least, the plugin was tuned to avoid DOS detection by the babelio site (I did download 250 metadata titles in 5 queued chunks of 50 without being banned for one week... )

I couldn't find any trace of VdF... His work, under GPL V3 license can be modified. I would have liked to contact him before publishing my work, based on his, on mobileread...

Note: babelio being a site with mostly french books for french talking poeple, I took the liberty to write the interface in french... (well I may have left some comments in french too... luckily, there is no french looking instructions in python

)

Installation Notes:

Download the attached zip file and install the plugin as described in the Introduction to plugins thread.
Alternatively, copy the zip file, unzip it in a dedicated folder, get the needed privileges and execute the runit.bat (linux people can read the file and sudo the content)
Note that this is not a GUI plugin so it is not intended/cannot be added to context menus/toolbars etc.
configure the plugin: it comes with verbosity=3, no cover and no pretty title...
I would advise to read the doc file in the doc folder in the zip file... (if not to get some presentation, at least to make this part of my work useful... )

sources
babelio_db is visible on github https://github.com/lrpirlet/cal-babelio_db

Recommendations:
In order to give the babelio_db plugin time enough for the worse while not getting banned, you need to push the timeout to 45 seconds in the plugins setup (time in second after first match)

last comment:
This work is open source... I had fun writing it, if you think this work should be paid, choose a charity and give them, a little, with a mention like "Thanks to Louis Richard" or "Merci à Louis Richard" or something similar in your language. This will enhance my (unpublished) reputation...

What charity? My worst nightmares involve fire, so I'm giving for deeply burned children... My wife is afraid of cancer so she gives to cancer research, we both feel bad about people starving so we give to the "resto du Coeur"...

Unfortunately, there is always someone who needs help and will not be able to pay back (except perhaps with a thought for the stranger who helped them). So there is a lot of choices.

En Français

Spoiler:

Version History

Spoiler:

The zip
Sorry, because of a bug report, I was made aware that the version 0.8.6 had never been uploaded here... done now.