A web-based LALR(1) parser that takes any context-free grammar and an input string, builds the full ACTION/GOTO table, and runs the parse with a step-by-step trace showing whether the input is accepted or rejected.
- Backend:
Java 21,Spring Boot 3.5 - Build:
Maven - Frontend:
HTML,CSS,JavaScript - Libraries:
AOS scroll animations - Containerization:
Docker,Docker Compose - Deployment:
Fly.io
Grammar Input: Define any context-free grammar via the web interface using a simple production rule syntax.LALR(1) Table Construction: Builds canonical LR(0) item sets, merges them into LALR states, computes FOLLOW sets, and generates the ACTION/GOTO tables.Parse Trace: Step-by-step table showing stack state, remaining input, and action (shift, reduce, accept, error) for every parsing step.Accept / Reject Verdict: Clear badge output on whether the input string belongs to the grammar's language.Conflict Detection: Detects and reports shift/reduce and reduce/reduce conflicts with a merge log.Grammar Presets: One-click presets for common grammars (simple, expression, arithmetic, parentheses) to get started instantly.LALR Badge: Confirms whether the grammar is valid LALR(1) or not.
- Implemented the LALR(1) algorithm from scratch in Java: LR(0) item set construction → canonical state merging → FOLLOW set computation → ACTION/GOTO table generation.
- Built a Spring Boot REST controller to accept grammar + input string as JSON and return the full parse result.
- Wrote a vanilla JS frontend with AOS animations that sends requests and renders the step-by-step parse trace and tables.
- Containerized with Docker and deployed to Fly.io.
-
./mvnw spring-boot:run
-
docker-compose up --build
Opens at:
http://localhost:8080forDockerandMaven
Live at: lalr-parser.fly.dev