Question

Would a regular expression be faster than the following Python code?

myStr.replace(","," ").replace("'"," ").replace("-","").replace("_","").replace("["," ").replace("]"," ")

Does this mean that the str is being iterated over 6 times(the number of replace() calls)? Would a regular experession be faster or is it the same algorithm/methodlogy?

What would the regular expression of this look like?

Answer 1

I would say that a regex would almost certainly be faster, but you should do a simple benchmark to find out for sure. The code you have performs the following separate steps

myStr.replace(","," ")
myStr.replace("'"," ")
myStr.replace("-","")
myStr.replace("_","")
myStr.replace("["," ")
myStr.replace("]"," ")

and could be done in two steps with regexes:

re.sub(r"[,'\[\]]", " ", myStr)
re.sub(r"[-_]", "", myStr)