Python and RegEx: Repeating Pattern Not Working

Question

I have a text file with a proprietary programming language and I want to extract the relevant information about various function calls.

The structure of the function is:

function name(input1, input2) returns (output1, output2) function body

I'm using Python and RegEx to capture this information, but I've hit a snag. I can capture the name, the inputs and the outputs, but I am unable to grab all of the function body.

I use the following line to capture this info:

re.findall("(function)(.*?)\((.*?)\) returns \((.*?)\)(.*)", file_contents)

However, after the first instance of the word, 'function', this fails. Due to nested statements in the function body, I am unable to use a particular keyword (I've tried different approaches, and I cannot fully grab the entire body) to grab the last group (this would be the function body).

How can I group everything after a particular point and then repeat the pattern?

What I want: 'function', 'name', 'input1, input2', 'output1, output2', 'function body' to repeat indefinitely. I want the last group to grab everything after the outputs and then the pattern to restart when it gets to the next occurrence of the word 'function'. I've tried different variations of the (.?) and (.) quantifiers, but I can't seem to get it.

I am not a programmer by trade, so I am not that adept with RegEx or Python. I know just enough to do the very basics.

Can you give an example of the function? It would be useful to know if it is it always written on a single line; whether there may be variations in the whitespace (new lines versus spaces versus indents); and how the beginning and end of the function body is defined (e.g. until the end of the line, by curly brackets { }, by indents, or by a keyword). — Jason K Lai
– Jason K Lai, Commented Dec 11, 2019 at 20:12
You can see a sample of the format and the most recent RegEx expression here: regex101.com/r/2zqD02/1 — bitmeddler
– bitmeddler, Commented Dec 12, 2019 at 13:47

Jason K Lai · Accepted Answer · 2019-12-12 15:36:17Z

Based on further information from the comments, I tested the following regex code using the re.findall function in Python3.6, which works with the example:

import re

file_contents = "function func1(in1 : bool; in2 : bool; in3 : bool) returns ( out : bool) var L1 : bool; L2 : bool; L5 : bool; L4 : bool; L3 : bool; begin L1 = L3 and L4; L2 = L1 or L5; out = L2; L5 = in3; L4 = in2; L3 = in1; end \n random code \nfunction func2(in1 : bool; in2 : bool; in3 : bool) returns ( out : bool) var L1 : bool; L2 : bool; L5 : bool; L4 : bool; L3 : bool; begin L1 = L3 and L4; L2 = L1 or L5; out = L2; L5 = in3; L4 = in2; L3 = in1;"

pattern = r"(function) (.*?)\((.*?)\) returns \((.*?)\) (.*)"
regex_results = re.findall( pattern, file_contents )

print( regex_results )

Output:

[('function', 'func1', 'in1 : bool; in2 : bool; in3 : bool', ' out : bool', 'var L1 : bool; L2 : bool; L5 : bool; L4 : bool; L3 : bool; begin L1 = L3 and L4; L2 = L1 or L5; out = L2; L5 = in3; L4 = in2; L3 = in1; end '), ('function', 'func2', 'in1 : bool; in2 : bool; in3 : bool', ' out : bool', 'var L1 : bool; L2 : bool; L5 : bool; L4 : bool; L3 : bool; begin L1 = L3 and L4; L2 = L1 or L5; out = L2; L5 = in3; L4 = in2; L3 = in1;')]

bitmeddler · Accepted Answer · 2019-12-12 16:44:49Z

1

I figured out a different way to accomplish what I was trying to do.

I used the following line:

re.split('(function )(.*?)\((.*?)\) returns \((.*?)\)', contents)

This will split up what I wanted into a list. I then chunk the list and assign it to the variables I have.

Thanks for everyone who took the time to answer.

answered Dec 12, 2019 at 16:44

bitmeddler

293 bronze badges

Comments

score 1 · Accepted Answer · 2019-12-12 17:07:12Z

1

This will grab the function up until the next function.
There are 5 capture groups.

If using findall, post-process into a group of 5's to get results.

(?s)(\bfunction\b)(.*?)\((.*?)\)\s+returns\s+\((.*?)\)((?:(?!\bfunction\b).)*)

https://regex101.com/r/PkfofA/1

Expanded

 (?s)
 ( \b function \b )            # (1)
 ( .*? )                       # (2)
 \( 
 ( .*? )                       # (3)
 \) \s+ returns \s+ \( 
 ( .*? )                       # (4)
 \) 
 (                             # (5 start)
      (?:
           (?! \b function \b )
           . 
      )*
 )                             # (5 end)

I guess finditer() is a way to get better control of each set of 5 groups :

iter = re.finditer(r"(?s)(\bfunction\b)(.*?)\((.*?)\)\s+returns\s+\((.*?)\)((?:(?!\bfunction\b).)*)", target)
for result in iter:
    g1 = result.group(1)
    g2 = result.group(2)
    g3 = result.group(3)
    g4 = result.group(4)
    g5 = result.group(5)

edited Dec 12, 2019 at 17:07

answered Dec 11, 2019 at 20:37

user12097764

2 Comments

bitmeddler Over a year ago

Thanks for responding. I tried this bit of code, and unfortunately it still does not want to capture the function body correctly. See here: regex101.com/r/2zqD02/1. Any advice? Again, thanks for taking the time out to respond.

user12097764 Over a year ago

@bitmeddler - My mistake, fixed it. regex101.com/r/PkfofA/1

Collectives™ on Stack Overflow

Python and RegEx: Repeating Pattern Not Working

3 Answers 3

Comments

Comments

2 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

3 Answers 3

Comments

Comments

2 Comments

Your Answer

Sign up or log in

Post as a guest

Related