如何在 python 中使用正则表达式获取文件名

Question

I have a python script that list folders and files in a given path i want to find a pdf file that that start with number and with the current date. For this i used regular expression but the problem is that the system does not match the requested file.

regular expression = [0-9]+_[0-3]?[0-9]-[0-3]?[0-9]-(?:[0-9]{2})?[0-9]{2}$

example of file name: 10204_09-03-2021.pdf

where is the error in my code?

code:

for file in files:
   if file.endwith("pdf"):
      if file == re.findall("^[0-9]+_[0-3]?[0-9]-[0-3]?[0-9]-(?:[0-9]{2})?[0-9]{2}$",file)
         shutil.copy(filename,destination)

Answer 1

When your file name is 10204_09-03-2021.pdf . Your regex pattern won't match it because it's supposed to end with [0-9]{2}$ because of the $. But your file name will be 10204_09-03-2021.pdf, so it will always end with .pdf .

Adding a.pdf between and [0-9]{2} and $ will match it. Eg -

>>> import re
>>> file = "10204_09-03-2021.pdf"
>>> pattern = re.compile(r"[0-9]+_[0-3]?[0-9]-[0-3]?[0-9]-(?:[0-9]{2})?[0-9]{2}\.pdf$")
>>> pattern.findall(file)
['10204_09-03-2021.pdf']

Also note that in file == re.findall(...) file is a string and re.findall(...) is a list. So you'll get a False even if it matches. You can try using re.fullmatch which returns a Match Object (where boolean value is always True) or None if it doesn't match. Or, you can do file in re.findall(...)

Answer 2

You can use

import os, re, shutil
destination = r'new/folder/path'
files = ['folder/10204_09-03-2021.pdf', 'folder/04_09-03-2021.pdf']
rx = re.compile(r'^[0-9]+_[0-3]?[0-9]-[0-3]?[0-9]-(?:[0-9]{2})?[0-9]{2}\.pdf$', re.I)
for file in files:
    if rx.match(os.path.basename(file)):
        #shutil.copy(filename,destination)
        print(file) # Just print here
# => folder/10204_09-03-2021.pdf, folder/04_09-03-2021.pdf

See the Python demo and the regex demo .

NOTE : The regex is applied only to the basename of the file path (the file name itself, with extension). Thus, the \.pdf is added to the regex pattern, and ^ / $ anchors make sure the entire string must match the regex. Using re.match here yields the same results as re.search or re.fullmatch .

Regex details :

• ^ - start of string
• [0-9]+ - one or more digits
• _ - a _ char
• [0-3]?[0-9] - an optional digit from 0 to 3 and then any one digit
• - - a hyphen
• [0-3]?[0-9]- - an optional digit from 0 to 3 and then any one digit and then a hyphen
• (?:[0-9]{2})? - an optional sequence of two digits
• [0-9]{2} - two digits
• \.pdf - .pdf string
• $ - end of string.

The regex is case insensitive due to re.I flag.