Regular expression to filter list of strings matching a pattern

I use R a lot more and it is easier for me to do it in R:

> test <- c('bbb', 'ccc', 'axx', 'xzz', 'xaa')
> test[grepl("^x",test)]
[1] "xzz" "xaa"

But how to do it in python if test is a list?

P.S. I am learning python using google's python exercise and I prefer using regular expression.

Solution 1:

In general, you may use

import re                                  # Add the re import declaration to use regex
test = ['bbb', 'ccc', 'axx', 'xzz', 'xaa'] # Define a test list
reg = re.compile(r'^x')                    # Compile the regex
test = list(filter(reg.search, test))      # Create iterator using filter, cast to list 
# => ['xzz', 'xaa']

Or, to inverse the results and get all items that do not match the regex:

list(filter(lambda x: not reg.search(x), test))
# >>> ['bbb', 'ccc', 'axx']

See the Python demo.

USAGE NOTE:

• re.search finds the first regex match anywhere in a string and returns a match object, otherwise None
• re.match looks for a match only at the string start, it does NOT require a full string match. So, re.search(r'^x', text) = re.match(r'x', text)
• re.fullmatch only returns a match if the full string matches the pattern, so, re.fullmatch(r'x') = re.match(r'x\Z') = re.search(r'^x\Z').

If you wonder what the r'' prefix means, see Python - Should I be using string prefix r when looking for a period (full stop or .) using regex? and Python regex - r prefix.

Solution 2:

You can use the following to find if any of the strings in list starts with 'x'

>>> [e for e in test if e.startswith('x')]
['xzz', 'xaa']
>>> any(e.startswith('x') for e in test)
True

Solution 3:

You could use filter. I am assuming you want a new list with certain elements from the old one.

new_test = filter(lambda x: x.startswith('x'), test)

Or if you want to use a regular expression in the filter function you could try the following. It requires the re module to be imported.

new_test = filter(lambda s: re.match("^x", s), test)