A Syntax-directed Voice-activated Programming Environment for Java

Coding by voice is an attractive approach for software engineers who suffer from repetitive strain injury, and quite a few work have been done. Most of existing approaches merely provide a speech recognition surface that simply translates program fragment from voice to text. We try to solve this problem from a new perspective by implementing a voice-activated pro-gramming environment based on VSCode. What is more, a Java program is constructed in a syntax-directed manner through a sequence of carefully designed commands recognized from voice, which is different from traditional text-based approach. User does not need to dictate spaces, parentheses, and commas as text-based approach any more which reduces vocal load and syntax errors.