mjhjmtu

I am using a regular expression in a python script to search through a file to find variable declarations. This is what I have so far:

ret1 = re.compile(r'^s*(volatile|register|typedef)?s?([w<>:*, ]+?)s+([w:]+)[w:[]]*s*;s*$')

ret2 = ret1.match(nextLineInFile)

print("Group 2: ", ret2.group(2))#variable type
print("Group 3: ", ret2.group(3))#variable name

Later in the code, I'm using groupings to capture the variable type and the variable name. I have the following input:

long myArray1[2];
long myArray2 [2];
long long myArray3[2];
long long myArray4 [2];

My RegEx is only finding myArray1 and myArray3. I need it find all four declarations. I've tried the following:

ret = re.compile(r'^s*(volatile|register|typedef)?s?([w<>:*, ]+?)s+([w:]+)s*[w:[]]*s*;s*$')

This catches myArray1, myArray2, and myArray4 perfectly. But, now myArray3 is coming back with a variable type of "long" and variable name of "long". What am I doing wrong?

After tinkering a bit with your input and current regex, I've come to the following expression : ^s*(volatile|register|typedef)?s?([w<>:*, ]+?)s+([w:]+)s?[d+]s*;s*$.

Instead of Looking for arrayName[size] and arrayName [size] as two separate elements, i'm just looking to find the name of the array ([w<>:*, ]+?) followed or not by a space s?, and then by the brackets with the size [d+].

You can test the expression at https://regex101.com/r/6KDHjj/1

Do not hesitate if you have any question about my expression, or if it does not work as expected.

Another possibility I just thought of would be to have the part matching the type of the array be greedy : ^s*(volatile|register|typedef)?s?([w<>:*, ]+)s+([w:]+)s*[w:[]]*s*;s*$.

I just removed the interrogation mark at the end of ([w<>:*, ]+?) to get this behaviour.

Here is a test of this second possibility : https://regex101.com/r/WMWC93/1