Python正则表达式不匹配行

Question

My example log file is big and contains below lines.

<6>[16495.700255]

Memory - START UC1

<4>16495.723327 C0  Memory - START UC1

<4>[16495.723327] C0 [             sh] Memory - START UC1

I am looking for Memory - START UC1 The below regular expression gets the first two lines but not the third.

re.compile("(Memory - +(.*)$)")

Answer 1

Use re.MULTILINE as a flag to re.compile or add (?m) to the start of the Regex. The $ only matches the end of the string unless MULTILINE mode is on, when it matches the end of any line.

Answer 2

I copied the original regex from your question - re.compile("(Memory - +(.*)$)") into the code from your follow-up answer, and ran that against the sample text from your question, and got all three matches.

@Smac89's suggestion of re.compile("(.*?Memory - START UC1)") is only necessary if you are calling the regex with event_regex.match(line) , which is implicitly anchored to the beginning of the string ( ^ ); if you use search(line) or findall(line) then the .*? doesn't do anything except make the regex harder to read: it non-greedily matches zero or more of anything, so if you're not anchored to the start of the string then it will end up matching zero characters anyway.
And I'm afraid that the suggestion of [^.* ]? makes even less sense, unless I'm terribly mistaken (which happens far too often). That says: match zero or one characters from the character group that consists of all characters except a literal . , a literal * , or a space. Which, again, if you're not anchored to the beginning of the string, that part of the regex will end up most likely matching zero characters anyway.

Honestly, if you know that you want to match the exact string Memory - START UC1 , then you're probably better off with a simple line.contains('Memory - START UC1') rather than a regex.
But your initial regex contained + (that's 'space plus') - one or more spaces - and if the number of spaces can vary, then yes you do want a regex. You might also consider \\s+ in that case, which matches both spaces and tabs (and a few other rarer whitespacey characters). If there's a possibility of trailing spaces, then you should put \\s* just before your $ end-of-string anchor. (I actually suspect that trailing space was the reason your initial regex was not matching that third occurrence of your target string.)

A couple of other tips:
In your initial regex, "(Memory - +(.*)$)" you have two capture groups (ie. sets of parentheses) but I suspect that you only actually want one, depending on whether you're interested only in the "UC1" or all of "Memory - UC1".
Also, your if not line: clause never fires, because blank lines still have a linebreak. You could do line.strip() - since you already do a line.strip() later, I would just put a line = line.strip() at the top of the loop and then just use line thereafter, rather than repeating the function call. It's a good thought to early-out, but in this case I'm not sure that it really saves you anything, since it doesn't take the regex engine long to figure out that there's no match on a blank line.
Final thought: It looks like you are only expecting at most one match on a given line. If that's the case, then use search(...) rather than findall(...) . No need to keep looking after you've found what you wanted.

Regexes involve a bit of a learning curve, but they are amazingly powerful once you grok them. Keep at it!

Answer 3

Change your compile to:

re.compile("(.*?Memory - START UC1)")

see if that helps

It seems to work on ideone

If you just want to get the word, replace the regex with:

regex = compile(r'([^.* ]?Memory - START UC1)')