WELCOME TO LARYNXCODE : A CLI application that enables you to code using your voice
Coding is an integral part of a developer's life. Sometimes, due to continuous and rigorous coding without breaks, a person may develop a condition known as Repetitive Strain Injury(RSI), that restricts all finger movements. How will you,then, finish that project that's due in 2 days ?
LarynxCode, based on the functionalities of the dragonfly Python framework, is a CLI appliaction, that fires up voice enabled coding in you system. It also comes with a linked web application that is basic guide on the functions of a few basic commands.
COMMANDS :
enable-voice-coding --voice --start: Starts Caster and enables you to code via Windows Speech Recognition
Executing the command :
Coding using your voice : Writing print('name') via voice coding :
enable-voice-coding help menu :
2. chelp --casterhelp --google : Launches the voice coding commands guide
Executing the command :
The webapp :
chelp help menu :
SYSTEM REQUIREMENTS :
- Python 2.7.18, 64/32 bit, Python 3.7.4 64 bit
- Windows OS
- Pre-trained Windows Speech Recognition GUI : Make sure you train your WSR 2-3 times properly.
SETTING UP CASTER :
- Install dragonfly :
pip install dragonfly2 - Follow the official docs only till step 2. LarynxCode automates the remaining setup and runs the algorithm. NOTE : Also, instead of installing at C:/Documents, install it in C:/ directly
- If you want to change rules/grammar, head over to C:/Caster/castervoice/rules to change the rules.
SETTING UP LARYNXCODE :
- Download this repository as zip, Extract it at C:/
- cd to the directory :
cd LarynxCode\cli_scripts - Setup a virtual environment :
py -m virtualenv venv, and activate it :venv\Scripts\activate - Install required libraries :
pip install -r requirements.txt - Run LarynxCode :
py voice_coding_cli.py enable-voice-coding --voice --start(refer to the commands explained above)
SHORTCOMINGS AND FUTURE SCOPE :
- Curently runs only on Windows OS.
- Windows Speech Recognition is poor in terms of identifying dictation, so it takes a lot of time and 5-6 attempts to get the correct command executed. In the future, we aim to shift to Dragon Naturally Speaking (DNS) speech recognition engine, which works the best in voice coding.
- Voice coding is heavily dependant on Python 2.7.18 till date.
SHOUTOUTS :
Leshna Balara, for collaborating with me to build the CLI. Shoutout to one of the most dedicated Pod mates ever.
Gabriel Cruz, for being an exceptional Pod leader.
CONTRIBUTE :
Feel free to raise an Issue if you feel like improving our CLI. LarynxCode is open sourced under the MIT License.
To contribute to the web application (frontend/backend) : Raise an Issue at this repository
~ LarynxCode, developed by Balaka Biswas and Leshna Balara