GENERATOR-README

0.  The grammar file 'grammar.txt' should consist of two sections- token declarations and production declarations. Each section must be terminated by a % symbol. Token declaration should done using token = token1 [token2 ...]. Everything to the right of '=' will be treated as a token. Individual tokens must be separated by atleast one white space. Production declaration should be  of the type variable : production1 | production2 | production3 .... ;. Two productions of the same variable must be separated by the pipe '|' symbol. There must be a semicolon ';' after the last production of each variable. The symbols in each production must be separated by atleast one whitespace.
 
1.  The parser generator uses the symbol set  : ; | and % for punctuation. Please do not use these symbols in your grammar. Instead use something like 'colon' or 'modulus' for these grammar symbols.

2.  'epsilon' is reserved as the token for empty strings. It must be used for this purpose only.

3. $ is hard coded as end of input marker. Do not use this symbol in your grammar.

4.  All varibles must appear on the LHS of atleast one production. Otherwise the behavior of the parser generator is unpredictable. 

5.  Any symbol that is not declared as a terminal will be treated as as a variable. 

6.  Sometimes a runtime error [LL-Parser: ../common.c:411: isNullable: Assertion 'j>=0' failed.] may occur because the parser generator tried searching for a variable but did not find it in the grammar table. No productions starting with this varibale were found. Either the grammar uses a terminal symbol that is not declared as a token or it uses a variable symbol that has no associated productions.

7.  There must be atleast one white space between any two grammar symbols. Something like  '(expr)' will be treated as a single unit (symbol) by the parser generator. Store it as   ( expr ) , if you want the symbols to be identified individually.

8.	Some of the files in this directory are used to build the generator while others are used to build the parser. A shell script named AUTOMAKE exists in this directory that can build the two binaries. For a manual build, dependencies can be looked up in AUTOMAKE.
Please use the switches --std=gnu99 and -Wall when building manually.

9. Binary 'generator' can be invoked to read the grammar from 'grammar.txt', build a parsing table and save it to 'parsing-table.txt ' if no fatal errors are encountered.

10. Binary 'parser' can be invoked to read input from a file and parse it.