Skip to main content

Python package and script collection to manage szurubooru.

Project description

szurubooru-toolkit icon

szurubooru-toolkit

Python package and script collection to manage your szurubooru image board.

Requirements

In order to run the included scripts, a Python release >=3.8 and the configuratrion file config.toml is required.

The config.toml file needs to be present in your current working directory. You can find a sample config file in the git repository of this package.

Installation

This package is available on PyPI and can be installed with pip: pip install szurubooru-toolkit

Alternatively, you can clone the package from GitHub and set everything up with Poetry. In the root directory of this repository, execute poetry install.

Please note this package requires Tensorflow for Deepbooru tagging and other required packages which do take up quite a lot of disk space (>400MB). A future release of this package will offer an option without the need to install Tensorflow if you don't want to use Deepbooru tagging.

User configuration

Make your changes in the config_sample.toml file provided in the git repo and rename it to config.toml afterwards.

Note that path names have to be specified with forward slashes (/) if you're using Windows.

config.toml reference
Section Option Description Examples/Default
szurubooru url The URL of your szurubooru "https://szuru.example.com"
szurubooru username Username which connects to the szuruboori API "my_szuru_user"
szurubooru api_token API token of username. Generate one in szurubooru from Account > Login tokens > Create token "my_api_token"
szurubooru public If your szurubooru is reachable over the internet false
auto_tagger saucenao_api_token In case you want to increase your daily query limit "my_saucenao_api_token"
auto_tagger saucenao_enabled Set this to false and deepbooru_enabled to true if you only want to tag with Deepbooru true
auto_tagger deepbooru_enabled If enabled, tag the post with Deepbooru if no tags with SauceNAO were found false
auto_tagger deepbooru_model Path to the Deepbooru model "./misc/deepbooru/model-resnet_custom_v3.h5"
auto_tagger deepbooru_threshold Define how accurate the matched tag from Deepbooru has to be "0.7"
auto_tagger deepbooru_forced Always tag with SauceNAO and Deepbooru false
auto_tagger hide_progress Set this to true to hide the progress bar false
auto_tagger tmp_path Local path where media files get downloaded temporarily if you szurubooru is not public. /tmp, C:/Users/Foo/Desktop
danbooru user Danbooru user "None"
danbooru api_key Danbooru api key "None"
gelbooru user Gelbooru user "None"
gelbooru api_ley Gelbooru api key "None"
konachan user Konachan user "None"
konachan password Konachan password "None"
yandere user Yandere user "None"
yandere password Yandere password "None"
pixiv user Pixiv user. Currently not being used. "None"
pixiv password Pixiv password. Currently not being used. "None"
pixiv token Pixiv token. Currently not being used. "None"
upload_media src_path Every valid media file under this dir (recursively) will get uploaded "/local/path/to/upload/dir"
upload_media hide_progress Set this to true to hide the progress bar false
upload_media cleanup Set this to true if images in the src_path should be deleted after upload false
upload_media tags These tags will get set for all uploaded posts. Separate them by a comma. ["tagme", "tag1", "tag2", "tagN"]
upload_media auto_tag Set this to true if you want your post to be automatically tagged after upload false
logging log_enabled If logging to a log file should be enabled false
logging log_file Specify the path of the log file "C:/Users/Foo/Desktop/szurubooru_toolkit.log"
logging log_level Specify the log level. DEBUG logs the most information "DEBUG"|"INFO"|"WARNING"|"ERROR"|"CRITICAL"
logging log_colorized If the log file should be colorized. Requires compatible viewer (e.g. less -r <log_file>). true

Creating a SauceNAO account and an API key is recommended. Please consider supporting the SauceNAO team as well by upgrading your plan. With a free plan, you can request up to 200 posts in 24h.

For Deepbooru support, download the current release here (v3-20211112-sgd-e28) and extract the contents of the zip file. Specify the path of the folder with the extracted files in deepbooru_model. Please note that you have to set deepbooru_enabled if you want to use it.

Scripts

auto-tagger

This script accepts a szurubooru query as a user input, fetches all posts returned by it and attempts to tag it using SauceNAO/Deepbooru.

If no matches from SauceNAO were found, the script keeps the previously set tags of the post and additionally appends the tag tagme.

You can set deepbooru_enabled to true in your config.toml file. In that case, the script falls back to tag posts with the supplied Deepbooru model. If you only want to use Deepbooru, set deepbooru_enabled to true and saucenao_enabled to false. If you want to use SauceNAO and Deepbooru, set following options to true: saucenao_enabled, deepbooru_enabled and deepbooru_forced.

Usage

After editing and renaming the sample config file to config.toml, we can just execute the script with our query:

  • auto-tagger "date:today tag-count:0"
  • auto-tagger "date:2021-04-07"
  • auto-tagger "tagme"
  • auto-tagger "id:100,101"

If we want to tag a single post, we can omit the keyword id in our query:

  • auto-tagger 100

Alternatively, we can tag a single post and specify --sankaku_url to fetch the tags from the supplied URL (untested with current release):

auto-tagger --sankaku_url https://chan.sankakucomplex.com/post/show/<id> 100

This is especially useful since Sankaku has changed their API and aggregator sites like SauceNAO don't have the latest results there.

If you cloned the repo from GitHub, prefix the above commands with poetry run, e.g. poetry run auto-tagger "date:today". Note that your current working directory has to be the the root of the GitHub project.

upload-media

This script searches through your specified upload folder in the config file for any image/video files and uploads them to your szurubooru.

Usage

After editing the config file, we can just execute the script.

If you installed it with pip, execute upload-media. Note that config.toml has to be in your current working directory.

If you cloned the repo from GitHub, execute poetry run upload-media. Note that your current working directory has to be the the root of the GitHub project.

create-tags (Currently not working, WIP)

This script reads the file ./misc/tags/tags.txt, parses its contents and creates the tags in your szurubooru.

If the tag already exists, it will get updated with your changes.

You can use tools like Grabber to download a tag list from common boorus.

The file has to be in following format:

<tag_a>,<category_number>
<tag_b>,<category_number>
<tag_..n>,<category_number>
Category Number
default 0
artist 1
series 2
character 3
meta 4

Usage

After editing the config file, we can just execute the script.

If you installed it with pip, execute create-tags. Note that config.toml has to be in your current working directory.

If you cloned the repo from GitHub, execute poetry run create-tags. Note that your current working directory has to be the the root of the GitHub project.

Image credit

GitHub repo icon: Code icons created by Smashicons - Flaticon

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

szurubooru-toolkit-0.4.0.tar.gz (33.3 kB view hashes)

Uploaded Source

Built Distribution

szurubooru_toolkit-0.4.0-py3-none-any.whl (34.3 kB view hashes)

Uploaded Python 3

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page