C++ Parser

A C++ based parser designed around the different stages involved.

• The link for this project will be changed to "C_Parser" starting April 15, 2017 (nick1au/C_Parser)

About

• Course-based Project - ENSC 251 (Software Design and Analysis for Engineers)
• Active during September to December 2016
• Worked with Win Aung
• Professor: Craig Scratchley

Project Info

• C++ based Parser for C++
• Created on:
  i. Eclipse IDE Mars 2
  ii. MinGW-W64 4.9.3
  iii. BOOST 1.55
• Automated test cases were used with BOOST library and debugging using Eclipse.
• Developed with Object-Oriented Programming

Project Breakdown

Part 1

• Tokenize an input into a vector
• Keep float numbers together (eg. 1232456E-3 is a token)
• Must separate string into tokens even if there are no whitespaces between them.

Part 2

• Classifies the tokens as a class type
• Types include:
  i. ID
  ii. Assignment operator
  iii. numeric_const (where T is a template)
  iv. unary_operator v. additive_operator
  vi. punctuator
  vii. div_operator
  viii. type_spec
  ix. conditional_operator
  x. shift_operator
  xi. string
  xii. postfix_operator
  xiii. incorrect

Part 3

• Forms a tree structure that will be used in Part 4
• Implements function pointers (passing itself to other functions)
• Uses the traversal methods given:
  i. In-order if = 2 children
  ii. Pre-order if > 2 children
  iii. Post-order if < 2 children

Part 4

• Creates an Abstract Syntax Tree (AST) using the tokens based on the Backus-Naur Form (BNF) i. If int a; a = a + 5; , then the AST is
  ; ~ int ~ a
  _ ~ = ~ a
  ______~ + ~ a
  ___________~ 5

Part 5

• Evaluate the tree given using the AST
• Uses the evaluation_map (map class from std library).