English | 中文

Kards Card Crawler

Automatically download all card images from Kards, the WWII CCG.

How it works

1. GraphQL API — queries https://herokuapi.kards.com/graphql for card metadata (name, faction, kredit cost, image path)
2. Image download — fetches AVIF images from the CDN via curl_cffi (browser fingerprint impersonation)
3. Format conversion — converts AVIF to PNG using Pillow

The crawler iterates over all factions (11 nations including Neutral) and kredit costs (0–7), with pagination and deduplication.

Folder structure

imgs/
├── 苏联/
│   ├── 0k/
│   │   └── 步兵第13步兵团_13th_rifles.png
│   └── ...
├── 美国/
│   └── ...
├── 中立/
│   ├── 0k/
│   ├── 1k/
│   └── ...
└── ...

Each file is named {卡牌中文名}_{cardId}.png. Falls back to unknown_{cardId}.png if no Chinese name is available.

Usage

git clone https://github.com/Tuning-Luna/kards-decks-collection-scraper.git
cd kards-decks-collection-scraper
pip install -r requirements.txt
python main.py

Configuration

Edit src/config.py:

Key	Description	Default
NATION_IDS	Faction IDs (0 = Neutral)	[1..10, 0]
KOSTS	Kredit costs to crawl	[0, 1, 2, 3, 4, 5, 6, 7]
IMAGE_BASE_URL	Change language (e.g. zh-Hans -> en-EN)	.../zh-Hans/
PROXIES	HTTP/HTTPS proxy	http://127.0.0.1:7897

Project structure

main.py                 # Entry point
src/
    config.py           # API config, headers, query, nation/kredit params
    scraper.py          # Crawl orchestration (pagination, dedup)
    image.py            # AVIF → PNG download (curl_cffi + Pillow)

Disclaimer

For personal study only. Please control crawl frequency and respect the game developer's copyright.