GETTING STARTED WITH DEYIN

Deyin is designed for immediate developer velocity. You can run your first pipeline locally within 2 minutes.

1. INSTALLATION

# Clone repository
git clone https://github.com/ayodejiades/deyin.git
cd deyin

# Install parser dependencies (Lark 1.1.8)
pip install -r requirements.txt

# Generate reproducible offline datasets
python bench/generate.py

2. YOUR FIRST PIPELINE

Create a file named signals.deyin:

stream prices from file "bench/prices.csv"
  | window 1m
  | filter price > 20.0
  | filter rsi(14) < 30.0
  | map buy 10%
  | sink to stdout

3. RUNNING VIA INTERPRETER

# Execute the pipeline with limit
python cli.py run signals.deyin --limit 50

# Output will stream evaluated records and summary stats:
# [interpreter] 14 rows in 0.012s (1,166 rows/s) - in:50 filtered:36 windows:1

4. INTERACTIVE REPL

Launch the interactive shell for real-time expression testing:

python cli.py repl
deyin> stream s from file "bench/prices.csv" | filter price > 100 | sink to stdout
[ok] 82 rows (from 100 in, filtered 18)

5. COMPILATION

# Transpile to standalone Python
python cli.py compile signals.deyin --target python -o dist/signals.py
python dist/signals.py 200000

# Emit WebAssembly (WAT)
python cli.py compile signals.deyin --target wasm