booru-scripts

Incrementally sync your favorites on moebooru websites to your own danbooru website.

In this repo, moebooru typically means konachan.com and yande.re.

增量同步你在 moebooru 网站上的收藏到你自己的 danbooru 网站。

在这个仓库中，moebooru 一般就指 konachan.com 和 yande.re。

Use-case 案例

As computer nerds and anime/manga art lovers, everyone wants their favorite images on moebooru to be stored locally, so that we can appreciate them later. However, if you just download them to you hard drive, there are no longer tags (a big feature in image boards like moebooru) holding the images together. With no tags, you will have a hard time finding related images by tags. Also, viewing the large original files is a suboptimal experience, especially for mobile devices.

Even when you try to use some photo-management software to organize the images (e.g. use Apple Photos and AppleScript to automate image-tagging), tagging all those images is still a pain in the ass, let alone the experience is still behind *booru for this purpose.

Why not self-hosting a danbooru instance and use this script to sync your favorite images on moebooru to your own website! You will have the same experience as moebooru websites with all the advantages of self-hosted websites. (since moebooru is a fork from danbooru, they are similar.)

作为计算机折腾人士和 anime/manga 作品爱好者，谁不希望能把在 moebooru 上的_大量收藏_的涩图下载到本地，方便以后欣赏呢？但是如果我们直接下载这些图片到电脑，那么维持这些图片之间的关系的图片 tag 就没了（这是 moebooru 这些 image boards 的最大的特点）。没有了 tag，那通过 tag 寻找相关图片就很难受了_{能快速找到同类型/画师/主角的涩图可是一大享受啊}，而且查看这些体积巨大的原图体验也不好（载入慢还烧流量），尤其对于移动设备来说。

就算你想用照片管理工具来管理这些图片（例如用 Apple Photos 并且用 AppleScript 来自动化打标签），给这么多图片打标签写脚本也是无比操蛋的，而且对于这种用途来说使用体验比不上 *booru。

为什么不自建一个 danbooru 实例，然后用这个脚本来同步你在 moebooru 上的收藏到你自己的网站呢！你仍然会拥有与在 moebooru 网站相同的体验，同时又有自建网站的所有优势。（moebooru 是 danbooru 的 fork，它们差不多）

This script will first download the original images and metadata of your favorites posts from moebooru. Then it will create posts on your danbooru instance with the downloaded image and metadata.

这个脚本会先下载你在 moebooru 上收藏的原图，然后会用下载的图片和元数据在你自己的 danbooru 网站上创建 posts。

Since there is no direct access to moebooru websites like konachan.com or yande.re in my country, the script will download the images first (proxy configurable) and then upload them to self-host danbooru instance. Otherwise the self-hosted danbooru instance is able to directly fetch them (although the danbooru instance is able to use a proxy).

因为众所周知的网络原因，我们无法直接访问 konachan.com 和 yande.re，所以脚本需要先下载图片，而不是让自建的 danbooru 直接去获取图片（虽然 danbooru 也能用代理）。

Screenshot of a self-hosting danbooru instance, with metadata of posts synced (see todo for supported metadata):

自建 danbooru 实例的截图，图片元数据已同步（如左边的标签，支持的元数据见 todo）：

How to use 怎么使用

Make sure GNU make, curl, wget, node, and yarn are installed. Steps below are based on a unix-like OS. If you are using Windows, try WSL.

确保你安装了 GNU make, curl, wget, node 和 yarn。以下步骤基于类 Unix 系统，如果你使用 Windows 的话请尝试 WSL。

If it is your first time running this script, run yarn install to install dependencies. 如果你是第一次使用这个脚本，请运行 yarn install 来安装依赖。

1. Prepare environment variables 准备环境变量

   • MOEBOORU_USERNAME: your username on moebooru

     • set it by export MOEBOORU_USERNAME=username

   • MOEBOORU_HOST: moebooru host, e.g. https://konachan.com (don't forget http/https)

     • set it by export MOEBOORU_HOST=https://konachan.com

   • DANBOORU_HOST: your self-hosted danbooru, e.g. https://danbooru.example.com (don't forget http/https)

     • set it by export DANBOORU_HOST=https://konachan.com
     • (recommanded) or putting DANBOORU_HOST=https://konachan.com in a .env file at repo root

   • DANBOORU_USERNAME: username of your self-hosted danbooru account

     • set it by export DANBOORU_USERNAME=username
     • (recommanded) or putting DANBOORU_USERNAME=username in a .env file at repo root

   • DANBOORU_APIKEY: api key of your self-hosted danbooru account (you can create one from account profile page)

     • set it by export DANBOORU_APIKEY=key
     • (recommanded) or putting DANBOORU_APIKEY=key in a .env file at repo root

   • PROXY: proxy to use when accessing moebooru websites (because there is no direct access in my country), e.g. http://127.0.0.1:9999

     • skip this if you don't need a proxy
     • set it by export PROXY=http://yourproxy:1234

   • IMAGE_DIR: path to directory, where to store downloaded images

     • default value: ~/Pictures/${MOEBOORU_HOSTNAME}. For example, if you set MOEBOORU_HOST as https://konachan.com, then IMAGE_DIR will be ~/Pictures/konachan.com
     • set it by export IMAGE_DIR=./your-dir

2. Start syncing 开始同步

   • Run make get_favorites_and_sync_to_danbooru to download and sync your favorites to self-hosted danbooru incrementally (tags will also be synced).
   • Run make get-and-update-tags-all to sync tag metadata from moebooru to self-hosted danbooru. Since this is about tag metadata, not the tags of an image, you don't need to run this frequently (typically they won't change).
   • More actions will be added later after some testing... You can read the targets in Makefile to see what they do.

TODO

• Sync images, tags, rating, and source
• Sync metadata for tags
• Sync parent/child relationships
• Sync pools