Timing re operations

_untimed_code.py

# code that I would write in an interactive session to perform a re.search or re.sub

# setup code
>>> import re
>>> text = open('lorem_ipsum.txt').read()  # filled with https://loremipsum.io/generator/?n=99&t=p
>>> regex = re.compile('foo')

# test code for re.search
>>> regex.search(text)  # returns None, there is no foo in lorem ipsum it seems

# test code for re.sub
>>> regex.sub('bar', text)  # returns lorem ipsum unchanged
Lorem ipsum dolor sit amet, ...
[...]
... sagittis vitae et leo duis ut diam.

actual_timings.sh

# calling timeit from the command line, to ensure clean interpreter state

# timing call for re.search
$ python -m timeit -s "import re; text=open('lorem_ipsum.txt').read(); reg = re.compile('foo')" "reg.search(text)"
10000 loops, best of 5: 26.3 usec per loop

# timing call for re.sub
$ python -m timeit -s "import re; text=open('lorem_ipsum.txt').read(); reg = re.compile('foo')" "reg.sub('bar', text)"
10000 loops, best of 5: 25.4 usec per loop

# looks to be about the same speed, definiteliy not worth it to re.search in order to know if re.sub does something.