关于 Python 中正则表达式的问题

Question

I am learning about regular expressions, But I'm having trouble making sense of certain things. I am working on an assignment to use regex's to find various characters, and words in a string:

Using the findall function, get all of the instances of non alphanumeric characters in the string assigned to 'lorem_ipsum'

Output to the console, the number of non-alphanumeric characters. Hint: use the len function. Use the ^ and [] regular expression operator along with the findall() regular expression function.

pattern = re.compile(r'sit-:amet')
occurrance_sit_amet = pattern.findall(lorem_ipsum)
for match in occurrance_sit_amet:
  print (match)

Why would I use the len function? and even more puzzling, why would I use ^ and [], when they can only be used to find characters at the start of the string and characters in brackets? Also my code gave me this error:

Cannot read property 'toISOString' of undefined - 9b2bb9d0-119a-11e8-95f3-4351563e5e1b

can someone explain what that means?

Answer 1

I think you misunderstand what the hints are (though I admit they are quite misleading). When it says "use ^ and [] ", it's not telling you to use a character class ( [] ) and a start of string anchor ( ^ ). It's telling you to combine the characters [] and ^ together to form a reverse character class [^] . Everything inside a reverse character class will not be matched.

The regex you need is this:

[^a-zA-Z0-9]

It means "everything but az, AZ or 0-9"

Regarding the the len function, your task is to find how many non alphanumeric characters there are, isn't it? findall returns a list of the matches. That's why you need to find the length of the list to find out how many such characters there are.

Here is some code:

pattern = re.compile(r'[^a-zA-Z0-9]')
allMatches = pattern.findall(lorem_ipsum)
print(len(allMatches))

Answer 2

To answer to your questions:

1. Using findall returns matches , which is a list of all substrings that match your pattern. So len(matches) should give you the "number of non-alphanumeric characters."
2. ^ can be used in conjunction with [] to do a search for patterns NOT containing certain characters .

A few issues with your regex that might be causing that error:

• The hyphen. Hyphens denote a range when placed between two characters, so right now the regex is trying to find characters ranging from t to : , which doesn't mean anything.
• The alphabetical characters in the regex. Aside from the colon, right now r'sit-:amet' matches a string composed of specific lowercase alphabetical letters one after another. Even if the regex worked, this would not match your desired "non-alphanumeric" pattern.

What you need is a negative search for anything alphanumeric:

[^A-Za-z0-9]

PS I highly recommend using regexr whenever you have regex assingments. It's a great way to check how pattern matching syntax works and test your regexs :)