rocky/OSGKeyboard
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

chore(typing): snapshot local pinyin WIP before syncing cloud branch

Browse Source 
Preserve the in-progress local typing/pinyin implementation so feat/pinyin
can safely reset to origin/feat/pinyin (cloud English + Chinese typing).
This commit is contained in:
Rocky
2026-08-03 13:14:01 +08:00
parent c037be3654
commit d1e5fed964
58 changed files with 370539 additions and 121 deletions
Download Patch File Download Diff File Expand all files Collapse all files
README.en.md
+18 -1
View File
@@ -47,6 +47,7 @@ Voice input for iPhone, iPad, and Mac. Speak in any app — polished text lands
| | iOS / iPadOS | macOS |
|---|:---:|:---:|
| Keyboard / global hotkey | ✅ | ✅ hold Option |
| Full / Microsoft / Sogou double pinyin | ✅ optional fuzzy pairs | — |
| Local speech recognition | ✅ SpeechAnalyzer | ✅ SenseVoice / Qwen3 |
| AI polish | ✅ | ✅ |
| Post-polish translation | ✅ | ✅ |
@@ -58,7 +59,7 @@ Voice input for iPhone, iPad, and Mac. Speak in any app — polished text lands
## Privacy
Speech is transcribed on-device by default. Polish sends **text only** — the transcript and a small amount of nearby cursor text for continuity, never raw audio. Cursor context is not logged or saved to voice history, and secure fields are never captured. See the [Privacy Policy](https://hkgood.github.io/OSGKeyboard/privacy/).
Speech is transcribed on-device by default. Polish sends **text only** — the transcript and a small amount of nearby cursor text for continuity, never raw audio. Chinese candidate learning stays in the on-device App Group and is not uploaded. Cursor context is not logged or saved to voice history, and secure fields are never captured. See the [Privacy Policy](https://hkgood.github.io/OSGKeyboard/privacy/).
---
@@ -122,6 +123,22 @@ Append a preset in `OSGKeyboardShared/Models/LLMProvider.swift` — any OpenAI-c
---
## Acknowledgements
OSGKeyboard's voice and keyboard features build on these projects and platforms:
- [Typeless](https://typeless.com) — product inspiration for voice-first input
- [Apple SpeechAnalyzer](https://developer.apple.com/documentation/speech) — on-device iOS speech recognition
- [librime](https://github.com/rime/librime) and [librime-xcframework](https://github.com/ghostflyby/librime-xcframework) — Chinese input engine and static iOS packaging
- [NanoMouse](https://github.com/xjwhnxjwhn/nanomouse) and [Hamster](https://github.com/imfuxiao/Hamster) — references for iOS Rime lifecycle and architecture
- [rime-pinyin-simp](https://github.com/rime/rime-pinyin-simp), [Jieba](https://github.com/fxsjy/jieba), [phrase-pinyin-data](https://github.com/mozillazg/phrase-pinyin-data), and [pinyin-data](https://github.com/mozillazg/pinyin-data) — pinyin, phrase, and frequency data
- [mlx-audio-swift](https://github.com/Blaizzy/mlx-audio-swift) — local Qwen3 MLX streaming ASR on macOS
- [Google Material Icons](https://github.com/google/material-design-icons) — icon font used by the iOS app
See [Third-Party Notices](./NOTICE-TYPING.md) for exact versions and licenses, or open Settings → About → Third-Party Licenses in the app.
---
## License
[Source Available License](./LICENSE) — personal, non-commercial use only. Commercial licensing: [rocky.hk@gmail.com](mailto:rocky.hk@gmail.com).