Original work is at: https://ankiweb.net/shared/info/1271086873
Spoonfed Chinese Anki Deck Audio Fix
This project addresses issues in the Spoonfed Chinese Anki deck, which contains errors and many missing audio files for Mandarin Chinese sentences.
The deck has 7335 notes with fields like English, Pinyin, Hanzi, and Audio. We focused on generating missing audio using text-to-speech (TTS) technology, ensuring every note's Hanzi field has corresponding spoken audio in the [sound:filename.mp3] format. Using Python scripts in a Python 3.10 environment on WSL, we unzipped the .apkg, analyzed the SQLite database for notes and models, identified missing MP3s, and generated them via Piper TTS (with models like zh_CN-huayan-medium.onnx for natural-sounding Mandarin). AI (grok) assisted in script development, research, and debugging.
Note that original deck errors (e.g., incorrect translations or Pinyin) remain untouched—we only filled audio gaps