The software in this package is a Python module for generating objects that compute the Cyclic Redundancy Check (CRC). It includes a (optional) C extension for fast calculation, as well as a pure Python implementation.
There is no attempt in this package to explain how the CRC works. There are a number of resources on the web that give a good explanation of the algorithms. Just do a Google search for “crc calculation” and browse till you find what you need. Another resource can be found in chapter 20 of the book “Numerical Recipes in C” by Press et. al.
This package allows the use of any 8, 16, 24, 32, or 64 bit CRC. You can generate a Python function for the selected polynomial or an instance of the crcmod.Crc class which provides the same interface as the hashlib, md5 and sha modules from the Python standard library. A crcmod.Crc class instance can also generate C/C++ source code that can be used in another application.
Documentation is available here as well as from the doc strings.
It is up to you to decide what polynomials to use in your application. Some common CRC algorithms are predefined in crcmod.predefined. If someone has not specified the polynomials to use, you will need to do some research to find one suitable for your application. Examples are available in the unit test script test.py.
If you need to generate code for another language, I suggest you subclass the crcmod.Crc class and replace the method crcmod.Crc.generateCode(). Use crcmod.Crc.generateCode() as a model for the new version.
The package has separate code to support the 2.x and 3.x Python series.
For the 2.x versions of Python, these versions have been tested:
It may still work on earlier versions of Python 2.x, but these have not been recently tested.
For the 3.x versions of Python, these versions have been tested:
To build the C extension, the appropriate compiler tools for your platform must be installed. Refer to the Python documentation for building C extensions for details.
python setup.py install
If the extension module builds, it will be installed. Otherwise, the installation will include the pure Python version. This will run significantly slower than the extension module but will allow the package to be used.
For Windows users who want to use the mingw32 compiler, run this command:
python setup.py build --compiler=mingw32 install
For Python 3.x, the install process is the same but you need to use the 3.x interpreter.
When you first install crcmod, you should run the unit tests to make sure everything is installed properly. The test script performs a number of tests including a comparison to the direct method which uses a class implementing polynomials over the integers mod 2.
To run the unit tests on Python >=2.5:
python -m crcmod.test
Alternatively, in the test directory run:
The crcmod package is capable of generating C functions that can be compiled with a C or C++ compiler. In the test directory, there is an examples.py script that demonstrates how to use the code generator. The result of this is written out to the file examples.c. The generated code was checked to make sure it compiles with the GCC compiler.