Writing a code beautifier [closed]

Question

I'd like to write a code beautifier and i thought of using Ruby to do it. Could someone show me a place to get started? I've seen a lot of code beautifiers online but I've never come across any tutorials on how to write one. Is this a very challenging task for someone who's never undertaken any projects such as writing a compiler, parser, etc. before?

(Is there another langauge which would be more well suited for this kind of task, excluding C/C++?)

Gombosi

On this site : a code beautifier for ruby. rubyBeautifier – Kosciusko

Answer 1

Python has an interesting feature - it exposes its own parser to scripts. There are examples that use the AST - abstract syntax tree - and do the pretty printing.

I'm not aware that Ruby exposes its own parser to its scripts in such a way, but there are parsers for Ruby written in Ruby here.

Tobytobye

Answer 2

Well... I think the initial steps are what you'd do for any project.

Write a list of requirements. Describe a user interface to your program, that you like and won't prevent you meeting those requirements. Now you can write down more of a "code" design, and pick the language that would be easiest for you to meet that design.

Here's some requirements off the top of my head:

• Supports code beautifying of these languages: Ruby, Python, Perl
• Output code behaves identically to input
• Output has consistent use of tabs/spaces
• Output has consistent function naming convention
• Output has consistent variable naming convention
• Output has matching braces and indentation

Make as many as you want, it's your program. ;p I was kidding about the Perl, but I think every language you support is going to add a more work.

Overawe

+1. It's all about finding your requirements and then seeing which tool best meets those, just as for any problem. E.g. a requirement of "I'd like to learn Ruby" could be just as important as others and skew things in that direction. :) – Adamsite

Could you tell me what tools I have to use? Do I need to use tools like lex and yacc or are those intended for something else? I did read somewhere that I need to parse the code into tokens or something after which I can handle it but that's all I know about. Thanks. – Gombosi

Tokenizing lines will definitely be part of your program. Look at the rubyBeautifier germanlinux linked to, it's not a long program. It used regular expressions, which is perfect for this. They are hard to learn, but very useful. I don't know Ruby, but it looks like you can use .split(delimiter) function on a string to chop it into tokens. Regular expressions are probably the way to go, though. – Overawe