neilmehta31 / SheeshLang

A custom language - SheeshLang - compiler made using python. Functionalities such as lexer, parser and 3 addresss code generation is implemented.

SheeshLang

How to run:

Make a virtual environment:

$ virtualenv venv

Then we activate the virtual environment (if on linux or mac):

$ source venv/bin/activate

OR (if on windows)

$ ./venv/Scripts/activate

We install the required dependencies for this project:

$ pip install -r /requirements.txt

Run the following command to run the lexer:

$ python3 parser_main.py <filepath>

Note: ./default.sheesh will run if no arguments are given. The following command will run the lexer_main.py file. It will print a table of the tokens.

IMPLEMENTATION DETAILS

• The lexer and parser are written in Python 3.6.
• Each file and funtion in the code is properly documented and commented about its input arguments, return values and the functionality
• All the keywords, operators and delimeters are defined in the Grammer folder in the hash_maps.py file.
• Sheesh's CFG and corresponding parse table have also been defined in the Grammer folder as Sheesh_CFG.txt and parse_LALR1_table.tsv
• The utils folder provides with all the utility functions required for the lexer and parser.

ADDITIONAL INFORMATION

• Logs for stack, error and input string are stored in the Logs folder. Logs are captured in each iteration of the parsing process.
• Testcases are provided in the Testcases folder. To run each of the testcases run the command:

  $ python3 parser_main.py Testcases/<filename>

Functionalities [updated : 28/04/2022]:

Lexer phase

• Lexes
• 71 keywords.
• Implemented custom keywords like shout, loop, yeet.
• Can distinguish between various datatypes in the lexer phase.
• distinguishes between string literal, integer literal, float literal and new line character.
• Gives error on multiline strings and prints each error after reading through the file.
• Output is properly printed in a table with four categories, namely:
  • Token line
  • Token
  • Token value
  • Token Category

Here's an example.

File:

def main() : void {
    int a = 4c
    return 0
    5
}

Command to test the file:

$ python3 lexer_main.py Testcases/fact.sheesh

Output:

Lexeme seperation successful.

Token line	Token	Token value	Token Category
1	def	38	keyword
1	main	70	keyword
1	(	23	delimeter
1	)	24	delimeter
1	:	10	operator
1	void	71	keyword
1	{	19	delimeter
2	NL	68	new line
2	int	63	keyword
2	a	73	identifier
2	=	12	operator
3	4c	err	invalid
3	NL	68	new line
3	return	57	keyword
4	0	74	integer_lit
4	NL	68	new line
5	5	75	integer_lit
5	NL	68	new line
5	}	20	delimeter

Parser phase

• Parses
• Uses a custom CFG called Sheesh_CFG.
• Parse table created using LALR(1) parsing technique.
• Efficiently uses parse table to process the incoming input string.
• Gives error on multiline strings and prints each error after reading through the file. Parser also prints the line at which the error occurs.
• Errors are automatically dealt with by the parser. The parser will not stop on the first error and will continue finding more errors.
• Output is properly printed in a table with four categories, namely:
  • Token line
  • Token
  • Token value
  • Token Category

Here's the same example.

File:

def main() : void {
    int a = 4c
    return 0
    5
}

Output:

Parsed this bad boy.