Skip to content

Crawl your personal favorite images, photo albums, comics from website. Support pixiv, yande.re, nhentai for now. 爬取你个人收藏的图片、写真集和漫画。

License

Notifications You must be signed in to change notification settings

RyouMon/FavoritesCrawler

Repository files navigation

FavoritesCrawler

Build Version

Crawl your personal favorite images, photo albums, comics from website.

Warning!

  • Not ready for production.
  • Appropriately reduce the crawling speed in the future and provide options to set performance, but your account is still at risk of being disabled by the website.

Plan to support

  • instagram.com

Already support

  • pixiv.net (crawl your liked illust, must login), Thanks for project PixivPy.
  • yande.re (crawl your voted posts, require your username)
  • nhentai.net (crawl your favorite comics, must login)
  • twitter.com (crawl your liked posts, must login)

Requirements

  • Python3.10+

Install

pip install -U favorites_crawler

Config Proxy (Optional)

# on Windows
set https_proxy=http://localhost:8080  # replace with your proxy server
# on Liunx/macOS
export https_proxy=http://localhost:8080

Login

favors login --help

Login Pixiv

Login by command

  1. Run command
    favors login pixiv
    
  2. Pixiv login page will open in browser, input username and password, then press login button.

Login by edit config

  1. Get your access_token and refresh_token by gppt

    gppt l -u your_email -p your_password
    
  2. Get your user_id (Access your pixiv personal page, copy from address bar)

  3. Open config file located in {your_user_home}/.favorites_crawler/config.yml, put access_token, refresh_token and user_id in this file

    pixiv:
      ACCESS_TOKEN: xxxxxxxxxxxxxxxxxxxxxxxxxxxx
      REFRESH_TOKEN: xxxxxxxxxxxxxxxxxxxxxxxxxxxx
      USER_ID: xxxx

Login Yandere

run command:

favors login yandere -u {username}

Login NHentai

  1. Get User-Agent and Cookie File
    1. Open nhentai and login.
    2. Open dev console (F12) and switch to network tab.
    3. Open any comic.
    4. Copy user-agent from any request.
    5. Use "Get cookies.txt" browser extension download cookie file.
  2. Execute command:
    favors login nhentai -u "{User-Agent}" -c "{Cookie File}"
    

Login Twitter

  1. Get Authorization, X-Csrf-Token RequestURL and Cookie File
    1. Open x.com and login, get to your "Likes" page
    2. Open dev console (F12) and switch to network tab.
    3. Enable persistent logging ("Preserve log").
    4. Type into the filter field: Likes?
    5. Refresh Page.
    6. Copy Authorization, X-Csrf-Token and RequestURL from request(Likes?variables...)
    7. Use "Get cookies.txt" browser extension download cookie file.
  2. Execute command:
    favors login x -at "{Authorization}" -ct "{X-Csrf-Token}" -u "{RequestURL}" -c "{Cookie File}"
    

Example:

favors login x -at "Bearer AAAAAAAAAAAAA..." -ct ... -u "https://x.com/i/api/graphql/.../Likes?..." -c "C:\Users\xxx\Downloads\x.com_cookies.txt"

Note: Request URL will make the entire command very long. If you cannot enter such a long command in the macOS terminal, you can write the command in a sh file and execute it.

Crawl

favors crawl --help

Crawl Pixiv

Before run this command, make sure you are already login.

favors crawl pixiv

Crawl Yandere

Before run this command, make sure you are already login.

favors crawl yandere

Crawl NHentai

Before run this command, make sure you are already login.

favors crawl nhantai

Crawl Twitter

Before run this command, make sure you are already login.

favors crawl twitter

Config

Config file config.yml locate on FAVORS_HOME, by default FAVORS_HOME is {your_home}/.favorites_crawler. You can change FAVORS_HOME by set environment variable.

You can set any scrapy built-in settings in this file.

By default, file content likes this:

global:
  ENABLE_ORGANIZE_BY_ARTIST: true
pixiv:
  ACCESS_TOKEN: xxxxxxxxxxxxxxxxxxxxxxxxxxxx
  REFRESH_TOKEN: xxxxxxxxxxxxxxxxxxxxxxxxxxxx
  USER_ID: xxxx
yandere:
  USERNAME: xxxx

Download location

By default, pictures will download to ${FAVORS_HOME}/{site_name}
If you want to change download location, you can update FILES_STORE.
For example, if you want save pixiv files to pictures/a, and want save yandere files to pictures/b, you can modify config file like this:

pixiv:
  ACCESS_TOKEN: xxxxxxxxxxxxxxxxxxxxxxxxxxxx
  REFRESH_TOKEN: xxxxxxxxxxxxxxxxxxxxxxxxxxxx
  USER_ID: xxxx
  FILES_STORE: pictures/a
yandere:
  USERNAME: xxxx
  FILES_STORE: pictures/b

Organize file by artist

if you want to organize pixiv and yandere files by artist, add this line to your config:

global:
  ENABLE_ORGANIZE_BY_ARTIST: true

Store tags to IPTC/Keywords

only support pixiv, yandere and twitter.

global:
   ENABLE_WRITE_IPTC_KEYWORDS: true
   EXIF_TOOL_EXECUTABLE: '<Path to your exiftool executable>'  # default None, If the executable is not in the path, set it manually

Restore your favorites

Vote yandere posts

$ favors restore yandere -h
usage: favors restore yandere [-h] -s {0,1,2,3} -t CSRF_TOKEN -c COOKIE path

positional arguments:
  path                  The location of the post to vote. (Sub-folders are ignored)

optional arguments:
  -h, --help            show this help message and exit
  -s {0,1,2,3}, --score {0,1,2,3}
                        Set 1, 2 or 3 to vote, 0 to cancel vote.
  -t CSRF_TOKEN, --csrf-token CSRF_TOKEN
                        CSRF token. To get it: 
                        1. Open your browser DevTools. 
                        2. Switch to network tab. 
                        3. Vote any post on yandere. 
                        4. Copy x-csrf-token value from request headers.
  -c COOKIE, --cookie COOKIE
                        Cookie. To get it: 
                        1. Open your browser DevTools. 
                        2. Switch to network tab. 
                        3. Vote any post on yandere. 
                        4. Copy cookie value from request headers.

Example:

favors restore yandere -s 3 -t "xxxx" -c "xx=x; xx=x; xx=x" .