redferret / growdy Goto Github PK

Pull from Rowdy the test utils to help with testing the grammar

GRowdy should be able to load in a grammar resource and build/compile source files defined for the grammar. Deserialization of the GRBuilder to grab the resources and tables should be built out. This resource would be packaged inside the JAR file of the user's language/interpreter.

Write Unit tests for all the current objects, GRBuilder, and GRowdy

For a production rule => NonTerm_1* NonTerm_2 | NonTerm3*;
The OR should give ids to the groups so that if one group is built out the other groups aren't expanded. This will force the user to improvise though on how to handle two nonterminals expanding on the same terminal
For example given

NonTerminal(ID) => NonTerm_1* NonTerm_2 | NonTerm3*;
NonTerm_1(ID) =>// This one will always take the token ID starving NonTerm3
NonTerm3(ID) =>

The fix for this is not easy and will require new language changes

• Implement the parse tree so that a grammar can be constructed as a GRBuilder object containing all the data for Language, RowdyLexer, and RowdyBuilder to create a parse tree.

• For users to identify terminals and non-terminals GRowdyGrammar will generate a Java file containing all the constants such as BOOL_TERM to identify nodes.
  Output file would be named <Grammar name>GrammarConstants.java

public class <Grammar Name>GrammarConstants {
 public static final int <List of constants>;
}

A simple example:

public class RowdyGrammarConstants {
 public static final int 
 ID = 0, 
 CONST = 1, 
 BOOL_TERM = 100, 
 BOOL_TERM_TAIL = 101;
}

GRowdy offers build methods that wraps the RowdyLexer and RowdyBuilder. You can build using the path to the source file with the file name or build with a String of code.
Similar to the builder you can get the program which wraps RowdyBuilder's get program method. It's possible that the builder will merge with GRowdy to flatten these calls out to just one method call

GRowdy gr = GRowdy.build(nameOfGrammarResource);
gr.build(programFilePathWithFileName)
gr.buildLine(code)
gr.getProgram()

Below is the grammar for building grammars.
Add the non terminals and PRules

GR => ID GRAMMAR

GRAMMAR(LEFTCARET) => TERMINAL_BODY GRAMMAR_BODY

TERMINAL_BODY(LEFTCARET) => LEFTCARET TERMINAL RIGHTCARET TERMINAL_DEFS SPECIAL_DEF(CONST)
SPECIAL_DEF(LEFTCARET) => LEFTCARET SPECIAL RIGHTCARET CONST SEMICOLON
TERMINAL_DEFS(ID) => TERMINAL_DEF SEMICOLON TERMINAL_DEFS
TERMINAL_DEF(ID) => ID IS ATOMIC
ATOMIC(id constant CONST) => id | constant | CONST

GRAMMAR_BODY(LEFTCARET) => LEFTCARET BEGIN RIGHTCARET NONTERMINAL_DEFS
NONTERMINAL_DEFS(ID) => NONTERMINAL_DEF SEMICOLON NONTERMINAL_DEFS
OR_OPT(OR) => OR ID OR_OPT
ID_LIST(ID) => ID ID_LIST 
NONTERMINAL_DEF(ID) => ID NONTERMINAL_PARAMS IS ID_LIST OR_OPT
NONTERMINAL_PARAMS(LEFTPAREN) => LEFTPAREN TERMINAL_PARAMS RIGHTPAREN
TERMINAL_PARAMS(ID) => ID TERMINAL_PARAMS