Project 1: Pretty Printing
Due: Sunday, September 23, 2012 at 11:59pm
In class, we have discussed the use of regular expressions to express the lexical tokens that make up a programming language. In this project, we will use the JFlex tool to express the tokens that comprise the MiniJava language presented in Appendix A of the textbook and then simply print the program back out in a nicely formatted way.
A pretty printer is a program that formats a source code file according to some standard rules. Our program should take an arbitrary MiniJava program as input and produce a nicely-formatted source code file as output.
1. Curly braces should follow the Java convention where the opening curly brace is the last character on the line the block structure begins (class, method, if, else, while, …) and the closing curly brace is on a separate line by itself, indented to the same level as the line that opened the block was. The opening curly brace should be preceded by one space.
2. Indentation should be done with tabs, not spaces, with one tab per level of nesting.
3. Binary operators (+, -, *, etc.) should be surrounded by one space.
4. Commas should be attached to the previous identifier and followed by a single space.
5. Unary operators should be directly attached to the expression they precede.
6. Parenthesis () and brackets [] should have no spaces before and one space following the opening one and one space prior to the closing one, except when empty. There should be no space between empty parenthesis or brackets.
7. Data types should be separated by one space from the identifier they precede.
8. Block comments should be indented to the current indentation level and their text wrapped at the end of the word preceding column 80, unless this is impossible (one long “word” that is greater than the remaining space can spill over, but the next “word” must be on the next line).
9. Line comments should be discarded. (This is crazy for a pretty printer, but will be useful for using this lexer in our compiler later).
10. Method signatures should all be on one line ending with an opening curly brace as per rule 1.
11. Class declarations should all be on one line ending with an opening curly brace as per rule 1.
· You may assume that the input programs are all valid, compiling MiniJava programs.
· Your output program must also be a valid MiniJava program and be functionally equivalent to the input program.
· You may not use or write a parser. All of this must be done using the built-in capabilities of JFlex plus some additional code that you will need to track indentation level, etc. You will want to look at the lexer states and lookahead operators to help.
· You will find that because of the previous requirement, this project is impossible to do perfectly for any complex language.
class Factorial {
public static void main(String[] a) {
System.out.println(new Fac().ComputeFac(10));
}}
class Fac {
public int ComputeFac(int num) {
int num_aux;if (num < 1)num_aux = 1;
else num_aux = num* (this.ComputeFac(num-1));
return num_aux;
}}
class Factorial {
public static void main( String[] a ) {
System.out.println( new Fac().ComputeFac( 10 ) );
}
}
class Fac {
public int ComputeFac( int num ) {
int num_aux;
if( num < 1 )
num_aux = 1;
else
num_aux = num * ( this.ComputeFac( num – 1 ) );
return num_aux;
}
}
The regular expressions that JFlex recognizes are not powerful enough to do everything that we might need to do without using or writing a parser. If you find something that you cannot express in JFlex, please write up what structure you cannot deal with and why in a README text file to be submitted with your project.
By the deadline, you need to submit:
1. Your JFlex file containing your lexer
2. Your java files containing main() and any auxiliary java files you have used
3. Your README text file describing what was impossible to do
4. Any examples that you have tested your program on
Create a zip file of the above files which you will upload to a website whose URL will be posted on the class website shortly before the due date.