An Earley Algorithm Parser toolkit.

This package uses Jay Earley's algorithm for parsing context-free grammars, and comes with some generic Abstract Syntax Tree routines. There is also a prototype scanner that does its job by combining Python regular expressions.

(SPARK stands for Scanning, Parsing, and Rewriting Kit. It is a poor name since it conflicts with a more popular package of the same name. In the future we will rename this.)

The original version of this was written by John Aycock for his Ph.d thesis and was described in his 1998 paper: "Compiling Little Languages in Python" at the 7th International Python Conference. The current incarnation of this code is maintained (or not) by Rocky Bernstein.

Note: Earley algorithm parsers are almost linear when given an LR grammar. These are grammars that are left-recursive.

Installation

This package is available from PyPI:

$ pip install spark-parser

However if you want to install from the github source:

$ pip install       # creates wheel and install

To run from the source tree:

$ pip install -e .  # set up to run from source tree

Features

Many of the added features are directly related to using it in the Python decompiler uncompyle6.

One unique feature of this code is the ability to have it profile which grammar rules have been used over a series of parses. This can inform which grammar rules are not needed.

Another unusual feature is to be able to remove grammar rules after a rule has been added. This is used in uncompyle6 where grammar rules inherited from one version of Python to another.

The non-standard grammar generator system feature is the ability to perform a callback check just before a reduction rule occurs. This allows for checking a token stream or partial parse tree by means other than through the grammar language.

NEW-FEATURES describes these features and others in more detail.

Example

The github example directory has worked-out examples; The PyPI package uncompyle6 uses this and contains a much larger example.

Support of older versions of Python

We support running this from older versions of Python in various git branches:

• python-2.4-to-2.7 has code for Python 2.4 to 2.7
• python-3.0-to-3.2 has code for Python 3.0 to 3.2
• python-3.3-to-3.5 has code for Python 3.3 to 3.5
• python-3.6-to-3.10 has code for Python 3.6 to 3.10
• master has code for Python 3.11 to the current version of Python

See Also

• http://pages.cpsc.ucalgary.ca/~aycock/spark/ (Old and not very well maintained)
• https://pypi.python.org/pypi/uncompyle6/