pylitex 0.2.0

Python package for Litex core

Pylitex

A Python API library for Litex core, designed to help Python users interact with Litex core seamlessly. This library provides both local and online execution capabilities, persistent environments, and multi-process support.

Installation

💡 Install Litex core before using pylitex, visit our website and read the Installation of Litex core.

After installing Litex core on your machine, install pylitex in the same way as installing other Python packages:

# remember to install Litex core before install pylitex
# change your Python env to which your are using
# then run following commands
pip install pylitex

pylitex is under active development. Update to the latest version:

pip install -U pylitex

Usage

Import pylitex to get started:

import pylitex

Basic Functions

Local Execution

Execute Litex code using your local Litex installation:

# Run a single code snippet
result = pylitex.run("1 + 1 = 2")
print(result)
# Output: {"success": True, "payload": "1 + 1 = 2", "message": "..."}

Online Execution

Execute Litex code using the online Litex service (no local installation required):

# Run code online
result = pylitex.run_online("1 + 1 = 2")
print(result)
# Output: {"success": True, "payload": "1 + 1 = 2", "message": "..."}

LaTeX Conversion

Convert Litex code to LaTeX format using the -elatex flag:

# Convert code to LaTeX
result = pylitex.convert_to_latex("1 + 1 = 2")
print(result)
# Output: {"success": True, "payload": "1 + 1 = 2", "message": "LaTeX output..."}

Persistent Execution with Runner

For multiple code executions that need to maintain state:

# Create a persistent runner
runner = pylitex.Runner()

try:
    # Variables persist between calls
    result1 = runner.run("let a R: a = 1")
    result2 = runner.run("let b R: b = 2")
    result3 = runner.run("let c R: c = a + b")
    print("Final result:", result3)
finally:
    runner.close()  # Always close the runner

Multi-Process Execution with RunnerPool

For parallel execution with separate environments per session:

# Create a runner pool with 2 workers and 30-second timeout
pool = pylitex.RunnerPool(max_workers=2, timeout=30)

try:
    # Each ID maintains its own environment
    pool.inject_code({"id": "session1", "code": "let x R: x = 5"})
    pool.inject_code({"id": "session2", "code": "let y R: y = 10"})
    pool.inject_code({"id": "session1", "code": "let result R: result = x * 2"})
    pool.inject_code({"id": "session2", "code": "let result R: result = y / 2"})
    
    # Get all results grouped by session ID
    results = pool.get_results()
    print("Session 1 results:", results["session1"])
    print("Session 2 results:", results["session2"])
finally:
    pool.close()  # Always close the pool

Utility Functions

Version Information

Get version information for both pylitex and Litex core:

# Get pylitex version
version = pylitex.get_version()
print(f"Pylitex version: {version}")

# Get Litex core version (requires local installation)
litex_version = pylitex.get_litex_version()
print(f"Litex core version: {litex_version}")

API Reference

Core Functions

run(code: str) -> dict

Execute Litex code locally using subprocess with the -e flag.

• Parameters: code - The Litex code string to execute
• Returns: Dictionary with success, payload, and message keys
• Requires: Local Litex installation

run_online(code: str) -> dict

Execute Litex code using the online Litex service at https://litexlang.com/api/litex.

• Parameters: code - The Litex code string to execute
• Returns: Dictionary with success, payload, and message keys
• Requires: Internet connection

convert_to_latex(code: str) -> dict

Convert Litex code to LaTeX format using the -elatex flag.

• Parameters: code - The Litex code string to convert
• Returns: Dictionary with success, payload, and message keys
• Requires: Local Litex installation

Utility Functions

get_version() -> str

Get the current version of pylitex package.

• Returns: Version string (currently "0.2.0")

get_litex_version() -> str

Get the version of the installed Litex core.

• Returns: Litex version string or error message
• Requires: Local Litex installation with --version support

Classes

Runner()

Persistent Litex execution environment using REPL wrapper.

• Methods:
  • run(code: str) -> dict: Execute code in persistent environment with automatic formatting
  • close(): Close the REPL wrapper and cleanup resources
• Features: Maintains state between executions, automatic environment reset on errors

RunnerPool(max_workers: int = 1, timeout: int = 10)

Multi-process runner pool for parallel execution (experimental).

• Parameters:
  • max_workers: Number of worker processes
  • timeout: Timeout in seconds for runner operations
• Methods:
  • inject_code(code_info: dict): Add code to execution queue
    • code_info should contain {"id": "session_id", "code": "litex_code"}
  • get_results() -> dict: Get all execution results grouped by session ID
  • close(): Close all worker processes
• Note: This feature is experimental and may have limitations

Return Format

All execution functions return a dictionary with the following structure:

{
    "success": bool,      # True if execution succeeded
    "payload": str,       # The original code that was executed
    "message": str        # Output message or error details
}

Error Handling

The library handles common errors gracefully:

• Local execution: Returns error message if Litex core is not installed or not in PATH
• Online execution: Returns error message for network issues or API failures
• Runner errors: Automatically resets environment on unexpected failures

Requirements

• Python 3.9+
• Dependencies: pexpect >= 4.8.0, requests >= 2.22.0
• For local execution: Litex core must be installed and accessible via litex command
• For online execution: Internet connection required

Version

Current version: 0.2.0

Check version programmatically:

import pylitex
print(f"Pylitex version: {pylitex.get_version()}")
print(f"Litex core version: {pylitex.get_litex_version()}")

Important Notes

RunnerPool Limitations

The current RunnerPool implementation has some limitations:

• The multiprocess execution is still under development
• Some methods may not work as expected in the current version
• For production use, consider using the Runner class for persistent execution

LaTeX Conversion

The convert_to_latex function uses the -elatex flag internally to generate LaTeX output from Litex code.

License

MIT License - see LICENSE file for details.