I’m the author of MPL. I’ve been working on LLM evaluation and QA architectures for years, but this is my passion project: bridging "Esoteric Intent" with actual executable logic.
*The Problem:*
Writing TradingView strategies (Pine Script) often involves repetitive boilerplate code. I wanted a way to define strategy logic abstractly in a Python environment and have the code generated procedurally.
*The Solution:*
MPL is an interpreted language written in Python (3.8+). It features:
1. *Custom Compiler Architecture:* A hand-rolled Lexer and Recursive Descent Parser (I didn't use external AST tools to keep it educational and dependency-free).
2. *Deterministic Transpilation:* The "Vassago" module takes strategy parameters (RSI limits, Trend length) and compiles them into valid, error-free Pine Script v6.
3. *Experimental Flow Control:* As a constraint experiment, I implemented "Cycle Protocols" (fixed iteration blocks based on 3-6-9 harmonics) instead of standard loops.
hknyrgnc•2mo ago
I’m the author of MPL. I’ve been working on LLM evaluation and QA architectures for years, but this is my passion project: bridging "Esoteric Intent" with actual executable logic.
*The Problem:* Writing TradingView strategies (Pine Script) often involves repetitive boilerplate code. I wanted a way to define strategy logic abstractly in a Python environment and have the code generated procedurally.
*The Solution:* MPL is an interpreted language written in Python (3.8+). It features: 1. *Custom Compiler Architecture:* A hand-rolled Lexer and Recursive Descent Parser (I didn't use external AST tools to keep it educational and dependency-free). 2. *Deterministic Transpilation:* The "Vassago" module takes strategy parameters (RSI limits, Trend length) and compiles them into valid, error-free Pine Script v6. 3. *Experimental Flow Control:* As a constraint experiment, I implemented "Cycle Protocols" (fixed iteration blocks based on 3-6-9 harmonics) instead of standard loops.
It’s open source and fully installable via pip.
Docs: https://hakanovski.github.io/MPL/ Proof of Concept (Live Output): https://www.tradingview.com/script/QOoCRPSF-Vassago-Tesla-Ex...
Happy to answer any questions about the parser implementation!