您如何在python中的字符串中查找文本，然后在其后查找数字？

Question

I have a log file and at the end of each line in the file there is this string: Line:# where # is the line number.

I am trying to get the # and compare it to the previous line's number. what would be the best way to do that in python?

Answer 1

I would probably use str.split because it seems easy:

with open('logfile.log') as fin:
    numbers = [ int(line.split(':')[-1]) for line in fin ]

Now you can use zip to compare one number with the next one:

for num1,num2 in zip(numbers,numbers[1:]):
    compare(num1,num2)  #do comparison here.

---

Of course, this isn't lazy (you store every line number in the file at once when you really only need 2 at a time), so it might take up a lot of memory if your files are HUGE . It wouldn't be hard to make it lazy though:

def elem_with_next(iterable):
    ii = iter(iterable)
    prev = next(ii)
    for here in ii:
        yield prev,here
        prev = here

with open('logfile.log') as fin:
    numbers = ( int(line.split(':')[-1]) for line in fin )
    for num1,num2 in elem_with_next(numbers):
        compare(num1,num2)

Answer 2

I'm assuming that you don't have something convenient to split a string on, meaning a regular expression might make more sense. That is, if the lines in your log file are structured like:

date: 1-15-2013, error: mildly_annoying, line: 121
date: 1-16-2013, error: err_something_bad, line: 123

Then you won't be able to use line.split('#') as mgilson as suggested, although if there is always a colon, line.split(':') might work. In any case, a regular expression solution would look like:

import re
numbers = []
for line in log:
    digit_match = re.search("(\d+)$", line)
    if digit_match is not None:
        numbers.append(int(digit_match.group(1)))

Here the expression "(\\d+)$" is matching some number of digits and then the end of the line. We extract the digits with the group(1) method on the returned match object and then add them to our list of line numbers.

If you're not confident that the "Line: #" will always come at the end of the log, you could replace the regular expression used above with something akin to "Line:\\s*(\\d+)" which checks for the string "Line:" then some (or no) whitespace, and then any number of digits.