`JavaCFG` - Installation and Usage

After unzipping or gunzipping and untar'ing, a directory called "JavaCFG" will be created, and you should cd into it.
Create the Java bytecode with the command
javac CFGParse.java
Write a grammar file such as the racecar grammar or sample from some examples provided in the grammars directory.
attempt to parse an input-string with the command
java CFGParse grammar-file-path input-string
if input-string is generated by the grammar, two trees are created:
- the first tree is represented by a file system style tree that is viewable and clickable in a Java window. Directory icons () represent variables in the grammar, while file icons () represent terminals. You can see which productions were used in the derivation tree by opening up the directories to reveal the right hand side of the corresponding production.
- the second tree is represented by a LaTeX file called "input-string.tex" which contains information for drawing the derivation tree in a more standard fashion than the first representation
if input-string is not generated by the grammar this fact is reported in a window created by the program. No LaTeX file is created in this case.
If you want convert the LaTeX tree into PDF you'll need LaTeX installed. The command
pdflatex filename.tex
creates a file called "filename.pdf" which contains a picture of the derivation tree.

Grammar Files

Grammars are stored as text files which should end with the .cfg extension. They are specified as follows:

Each line consists of white-space separated strings which represent productions from a common variable
The first string of any line must have length 1 and is a variable
Subsequent strings on a given line are the right hand sides of productions from the variable for that line
The first variable on the first line is the start variable
Any character which isn't a variable, is a terminal
THE FOLLOWING IS NOT RECOMMENDED (EARLEY'S PARSING ALGORITHM MAKES NO GUARANTEES ABOUT EPSILONS): A single double-quote symbol " represents the empty string ε (we think of " as two single quotes with nothing in between)

For example the language

{aⁿbⁿ | n > 0} ∪ {bⁿaⁿ | n > 0}

has the context free grammar:

S → X | Y
X → aXb | ab
Y → bYa | ba

which is represented in the grammar file by:

S X Y
X aXb ab
Y bYa ba

After untaring or unzipping you will find the following files

CFG.java: A class the defines the context free grammar object used by JavaCFG. Embedded in this class is some additional functionality, added by Yoav Hirsch for de-epsilonizing, and converting grammars into Chomsky Normal Form. Instructions for how to call on the conversion algorithms is embedded in the file.
CFGParse.java
: A class created for simplifying the parsing command.
EarleyAgent.java: A class used in the parsing algorithm.
EarleyRule.java: A class used in the parsing algorithm
MalformedGrammarException.java: An exception class used to let the user know that their grammar has not been defined properly, has a typo, or is not the right type for the given operation.
Rule.java: A class which defines the productions of a CFG.
TreeModel2LaTeX.java: A class for converting JTree objects into LaTeX trees.
grammars: A directory containing some examples of grammars NOTE: some of these grammars contain epsilons and that the success of the parsing algorithm on such grammars is not guaranteed.
instructions.html: This file