pixiv Downloader is a script that batch downloads artworks of a specified artist from pixiv.
Python 2.7 is required.
pixivdl.py artist_id [directory]
artist_id
refers to the unique numerical ID for the artist, which can be
found in the URL. For example, http://www.pixiv.net/member.php?id=92448
would
have an ID of 92448
.
directory
specifies the directory in which works would be downloaded to.
It is optional, and works will be downloaded to a created directory inside the
working directory in the format './img/artist_nickname(artist_ID)'
A metadata.pickle
file is also placed in the download directory. This file
is a Python pickle of the Work
objects from the pixiv_api
module
used by the script. It contains metadata like the titles, descriptions, etc.
of each work, as fetched from the API.
A GUI has been implemented in Tkinter can be found in pixivdl-gui.py
.
The inputs for the GUI are similar to the command line arguments.
Do note that when "Cancel" is clicked, the download process isn't immediately cancelled. Instead, it would only be cancelled when the current work has completed downloading.
This script works without logging in to pixiv. However, if you want to download R18 images or MyPixiv-only images that are made available to you, you must authenticate yourself.
Unfortunately, I could not figure out how logging in works, and thus have used
an alternative method. You must first log in to pixiv in your web browser,
then copy the contents of the PHPSESSID cookie into a file at ~/.config/pixivdl/session
As there isn't an official publically documented API for pixiv, I had to poke at
their iOS app and figure out how the undocumented API for that worked. There is
a basic interface to that API in the pixiv_api
module included in here,
though just enough for this script to work.