使用正则表达式从使用 Python 的日志文件中提取文件名

Question

looking to create a list of files accessed from a log file. Two examples of strings from the file are shown below.

.... [08/Mar/2020:19:11:15 -0700] "GET /socview/15Ragged.htm HTTP/1.1" 200 6564 .........

.... [08/Mar/2020:19:11:31 -0700] "GET /socview/?C=D;O=A HTTP/1.1" 200 13443 ..............

/socview/15Ragged.htm is what i'm looking to extract ie. ending in a .htm .log .txt etc.

/socview/?C=D;O=A is what i'm trying to avoid extracting.

It seems that the "." is whats causing issues, as when I run the code without searching for it, ie. the RE below runs perfectly as part of the loop shown at the bottom of this post.

unique = re.search(r'GET (\S+)', x)

However it is extracting strings I do not want. Below is the loop and RE that I'm trying to use, it makes sense to me and I cant figure out whats wrong, when ran the message below is displayed. Any help would be greatly appreciated

"if unique.group(1) not in unilist:

AttributeError: 'NoneType' object has no attribute 'group'"

for x in input:
     unique = re.search(r'GET (\S+\.\S+)', x)

     if unique.group(1) not in unilist:
           unilist.append(unique.group(1))

Answer 1

The GET (\\S+\\.\\S+) is fine. The problem is that the re.search() returns None if the match has failed, so for the second string you provided the unique is None which does not have a group property.

Try the following:

for x in input:
    unique = re.search(r'GET (\S+\.\S+)', x)

    if unique is None:
        continue

    if unique.group(1) not in unilist:
           unilist.append(unique.group(1))

I do recommend you to use better variable names. For example input is a build-in in Python, avoid shadowing it. If the loop body grows, it will be hard to follow names like x .

Also, I recommend to pre-compile the regex like this, otherwise it compiles it in every cycle which is very time-consuming:

matcher = re.compile("GET (\S+\.\S+)")

for line in lines:
    # your loop body here