regex with group names in python

Question

I started to play with python3.8 re lib and I find strange things. First I'm making my regex online to test how it works and when I found out that its already working I try in python.

test_string = [Tues Jan 20 11:35:13.405644 2020] [access_compat:error] [pid 1871:tid 140301098780416] [client 192.168.123.9:59662] AH01797: client denied by server configuration: /var/www/website/401.html

res = re.compile(r"(?P<Time>\w+\s\w+\s\d+\s\d+:\d+:\d{2})|(?P<Type>[a-zA-Z]+_\w+:\w+)", re.VERBOSE)

for line in logfile:
    line = res.search(line)
    print(line.groupdict())

I'm trying to parse log line like that like that. but I get the following result:

type : none

I dont know why, or how to fix it , any ideas?:

{'time': Mon Jan 20 11:34:13, 'type': access_compat:error}

Try `r"(?P\w+\s\w+\s\d+\s\d+:\d+:\d{2}).*?(?P[a-zA-Z]+_\w+:\w+)"`, see https://ideone.com/HimnIT — Wiktor Stribiżew, Jan 21 '20 at 14:22
AttributeError: 'NoneType' object has no attribute 'groupdict' — robotiaga, Jan 21 '20 at 14:24
Do all the lines in the log file contain this pattern? You should put an if statement in before the print to see if line is not a NoneType. — tomgalpin, Jan 21 '20 at 14:28
My link works and shows the same solution as posted by Valdi (just the quantifier is `*?`, not `+?`) — Wiktor Stribiżew, Jan 21 '20 at 14:43

score 1 · Accepted Answer · answered Jan 21 '20 at 14:36

1

You used a pattern with 2 alternatives, whereas you should use a pattern matching both alternatives at once, e.g.:

(?P<Time>\w+\s\w+\s\d+\s\d+:\d+:\d{2}).+?(?P<Type>[a-zA-Z]+_\w+:\w+)

or

'time': (?P<Time>\w+\s\w+\s\d+\s\d+:\d+:\d{2}), 'type': (?P<Type>[a-zA-Z]+_\w+:\w+)

For a working example see https://regex101.com/r/v6g791/1

answered Jan 21 '20 at 14:36

Valdi_Bo

30,023
4
23
41

Thank You!!! :)) – robotiaga Jan 21 '20 at 14:37

RomainD · Answer 2 · 2020-01-21T14:58:21.120

This regex seems to work :

r"\[(?P<Time>\w+\s\w+\s\d+\s\d+:\d+:\d{2}).*\].*\[(?P<Type>[a-zA-Z]+_\w+:\w+)\]"

The code :

import re

test_string = "[Tues Jan 20 11:35:13.405644 2020] [access_compat:error] [pid 1871:tid 140301098780416] [client 192.168.123.9:59662] AH01797: client denied by server configuration: /var/www/website/401.html"
res = re.compile(r"\[(?P<Time>\w+\s\w+\s\d+\s\d+:\d+:\d{2}).*\].*\[(?P<Type>[a-zA-Z]+_\w+:\w+)\]", re.VERBOSE)
print(res.search(test_string).groupdict())

Result :

{'Time': 'Tues Jan 20 11:35:13', 'Type': 'access_compat:error'}

Explanation :

You forgot to match characters [ and ]. You must use : \[<pattern>\]
You have used character | between your 2 pattern. And this character means : match either A or B (e.g.: r"A|B")

Source : https://docs.python.org/3/library/re.html

Thank you too, i am still learning :)) – robotiaga Jan 21 '20 at 14:42 — robotiaga, Jan 21 '20 at 14:42

regex with group names in python

2 Answers2