Spaces in Python Regular Expressions

Question

I am currently writing a Python script that will search an entire .c or .h file and find all the function definitions so I can then make some slight edits. To find all of the function definitions I am trying to use Regular Expressions.

What I currently have is:

"r'\w+?\s+?\w+?\s*?\(.*?\)[\n\s]*?{"

The problem with this logic is that it currently will accept certain if statements in some cases. For example:

else
   if(//herpderp){}

It does this because that \s includes \n. I feel that I wouldn't have this issue if I had my expression only look for spaces instead of any whitespace, but I can't test my theory out as it seems there is no \(insert letter here) for just a simple space.

So there is the problem. If you have any advice on how to fix my regular expression, or if there is a better way of writing the script in general please let me know.

Martijn Pieters · Accepted Answer · 2014-06-23 15:15:33Z

2

A single space can be matched by using a single space, the same way you'd match any other character that isn't a metacharacter:

"r'\w+? +?\w+? *?\(.*?\)\s*?{"

The ' +?' sequence matches only one or more spaces, non-greedily. I replaced [\n\s] with \s as the \n is included already.

You can expand to a character class with more types of whitespace:

[ \t]

which would match a space or a tab.

edited Jun 23, 2014 at 15:15

answered Jun 23, 2014 at 15:10

Martijn Pieters

1.1m326 gold badges4.2k silver badges3.4k bronze badges

Sign up to request clarification or add additional context in comments.

Comments

Casimir et Hippolyte · Accepted Answer · 2014-06-23 15:44:38Z

1

It does this because that \s includes \n

I'm not sure that this is a good theory since writing something like this in C is allowed:

int


    main()

A possible way can be to use a black or whitelist to ensure that what you obtain is a function. Example:

r'\b(int|float|double|char)\s+(\w+)\s*\([^)]*\)\s*{'   // whitelist

or

r'(?!(?:else)\b)\b(\w+)\s+(\w+)\s*\([^)]*\)\s*{'    // blacklist

Note: no need to use lazy quantifiers.

edited Jun 23, 2014 at 15:44

answered Jun 23, 2014 at 15:20

Casimir et Hippolyte

90k5 gold badges102 silver badges131 bronze badges

4 Comments

Adam Smith Over a year ago

To be fair, I don't think C is a regular language so regex is poorly suited to parse it. It would probably be easier to refactor it in his editor than to write a Python program to do it :)

Casimir et Hippolyte Over a year ago

@AdamSmith: I'm not a C expert, but I'm not sure that nested parenthesis are allowed inside the params. In any event, the fact that a language is regular or not, is no more a problem with modern regex engines (in particular with the new regex module).

Adam Smith Over a year ago

I haven't used the regex module -- is it worth picking up and learning?

Casimir et Hippolyte Over a year ago

@AdamSmith: yes, it's easy to install (no problem), relatively easy to learn and to use (backward compatibility with re module), and has very interesting features you can't find nowhere in other script languages (like the fuzzy search)

Collectives™ on Stack Overflow

Spaces in Python Regular Expressions

2 Answers 2

Comments

4 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

Comments

4 Comments

Your Answer

Sign up or log in

Post as a guest

Related