Skip to main content

Python package for analyzing and transforming netlists

Project description

https://img.shields.io/pypi/v/spydrnet.svg https://img.shields.io/pypi/pyversions/spydrnet.svg https://travis-ci.com/byuccl/spydrnet.svg?branch=master

A flexible framework for analyzing and transforming netlists. Built to fill an important gap in FPGA research and reliability. Currently available as a pure Python package.

Simple Examples

SpyDrNet can be used to create netlists from scratch. Because it is a low-level framework, manual netlist creation can be tedious (much like writting a high level application in assembly). To assist in rapid productivity, parsers and composers are provided for common netlist formats. Currently only EDIF is supported, but the roadmap includes structural Verilog, structural VHDL, Verilog Quartus Mapping Files (Intel’s VQM), and JSON.

Loading Example Netlists

Several example netlists are included with the package to introduce the framework, its features, and functionality. To list and load these netlists, modify the following example:

>>> import spydrnet as sdn
>>> sdn.example_netlist_names
['4bitadder', '8051', ... , 'zpu4']
>>> netlist = sdn.load_example_netlist_by_name('4bitadder')

Parsing a Netlist

>>> netlist = sdn.parse('<netlist_filename>.edf')

View Data Associated with any Netlist Element

>>> netlist.data
{'.NAME': 'adder', 'EDIF.identifier': 'Z4bitadder', ... }

List Libraries in a Netlist

>>> list(library.name for library in netlist.libraries)
['VIRTEX', 'UNILIB', 'work']

List Definitions in a Library

>>> library = netlist.libraries[2]
>>> list(definition.name for definition in library.definitions)
['adder']

List Ports, Cables, and Instances in a Definition

>>> definition = library.definitions[0]
>>> list(port.name for port in definition.ports)
['data1(3:0)', 'data2(3:0)', 'answer(3:0)', 'clk', 'reset', 'enable']
>>> list(cable.name for cable in definition.cables)
['answer_1(0)', 'answer_1(1)', 'answer_1_(2)', 'answer_1(3)', ... ]
>>> list(instance.name for instance in definition.children)
['un3_answer1_axbxc3', 'un2_answer2_axbxc3', 'reset_c_i', ... ]

Compose a Netlist

This example exports a netlist into an EDIF formatted netlist file by the given name.

>>> sdn.compose(netlist, '<filename>.edf')

The following equivalent code may also be used.

>>> netlist.compose('<filename>.edf')

Additional Examples

Additional examples are available in the documentation for netlist creation, analysis, and transformation.

Install

The stable release of SpyDrNet can be installed using pip:

> pip install spydrnet

To install from PyPI with all optional dependicies use:

> pip install spydrnet[all]

For more installation instruction, see INSTALL.rst.

Bugs

Bugs can be reported on the issues page or they can be fixed through a fork / pull request. All changes are welcome. Discussion of ideas for new features is available on the mailing list.

A Brief History

The BYU Configurable Computing Lab actively maintains the BYU EDIF Tools - a Java API for creating, modifying, or analyzing EDIF netlists. These tools are tied to the EDIF netlist format and provide JEDIF tools capable of flattening a circuit (by removing hierarchical organization) and applying fault-tolerance techniques such as triple modular redundancy (TMR). Development of SpyDrNet began back in 2016 with the idea of creating an accessible, format independent, tool for netlist analysis and transformation. The underlying intermediate data structure is designed preserve proper netlist relationship as a generic netlist while allowing for the preservation of format specific constructs. A language agnostic prototype was developed and this prototype soon became useful in the lab for netlist analysis and reliability transformation studies. A more mature (though still having room for growth) tool is presented here.

Design Notes

We have tried to build this tool around the principles of expandability and modularity. Care has been taken to separate different parts of the program in an organized fashion.

How to contribute

If this tool has been useful to you, or have new feature ideas that you would like to implement, feel free to make a pull request, or take a look at the issues to see how to contribute. New ideas, bug fixes and suggestions are also welcome (See CONTRIBUTING.rst).

Special Thanks

Special thanks is given to NetworkX - “a python package for the creation, manipulation, and study of the structure, dynamics and functions of complex networks.” This mature project has been used as a template for much of SpyDrNet’s documentation and code structure. It also has saved enormous effort in heavy graph analysis as a robust and complete library used to analyze the relationships between circuit nodes.

License

Released under the BSD 3-Clause License (see LICENSE):

Copyright (C) 2019, Brigham Young University
All rights reserved.

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

spydrnet-1.6.0.tar.gz (323.6 kB view hashes)

Uploaded Source

Built Distribution

spydrnet-1.6.0-py3-none-any.whl (397.3 kB view hashes)

Uploaded Python 3

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page