Metadata-Version: 2.4
Name: symantex
Version: 0.3.0
Summary: Thin python interface for using Large Language Models to convert LaTeX into Sympy for automatic mathematical processing
Author-email: "Nicholas Roy / FormuLearn B.V." <nicholas.roy@formulearn.org>
License-Expression: GPL-3.0-only
Project-URL: Homepage, https://github.com/FormuLearn/Symantex
Project-URL: Issues, https://github.com/FormuLearn/Symantex/issues
Classifier: Programming Language :: Python :: 3
Classifier: Operating System :: OS Independent
Requires-Python: >=3.10
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: sympy<2,>=1.14
Requires-Dist: mirascope[openai]>=1.25
Dynamic: license-file

# Symantex

**Symantex** (pronounced “semantics”) is a modular LaTeX→SymPy converter built on top of Mirascope’s structured-JSON API.  It transforms arbitrary LaTeX snippets into executable SymPy `Expr` and `Eq` objects, with hooks for domain-specific operators, custom prompt engineering, and dynamic parser configuration.

**Documentation:** Available at [FormuLearn Docs](https://docs.formulearn.org/docs/projects)

---

## 🚀 Features

- **Reliable JSON mode** via Mirascope LLM calls  
- **Automatic parsing** into SymPy ASTs (`Eq`, `Sum`, `Matrix`, derivatives, transforms…)  
- **Dynamic locals map** lets you register new functions/symbols (e.g. `argmin`, `softmax`, `attention`)  
- **Pluggable prompt** that enforces valid Sympy code (`Eq(...)`, `Sum(...)`, no ellipses)  
- **Graceful degradation**: partial parse success, intelligent failure logging for diagnostics  
- **Customizable parser** via regular-expression token discovery  

---

## 📦 Installation

```bash
pip install symantex
```
