[PYTHON] Regular expression look-ahead, after-yomi

As a prerequisite, the library re is being imported.

import re

All of the following notations extract only the text part, and the conditions at the time of extraction are shown in groups (). Affirmation is extracted when the text conditions in the group also match. Negation is extracted when the text conditions in the group do not match.

Positive look-ahead

Extract the text part when the text you want to extract matches by inserting (? = xxx) after the text you want to extract and the xxx part also matches.

re.findall('AB(?=CDEF)', 'ABCDEF') #['AB']
re.findall('AB(?=DEF)', 'ABCDEF') #[]
re.findall('AB(?=CD)', 'ABCDEF') #['AB']
re.findall('.+(?=CD)', 'ABCDEF') #['AB']
re.findall('AB(?=[A-Z]{2,3})', 'ABCDEF') #['AB']

Negative look-ahead

In negative look-ahead, write (?! Xxx) after the text you want to extract. Extract the text part when the text part matches and the xxx part does not match

re.findall('AB(?!CDEF)', 'ABCDEF') #[]
re.findall('AB(?!DEF)', 'ABCDEF') #['AB']
re.findall('AB(?!CD)', 'ABCDEF') #[]

Since it is denied, the result is the opposite of lookahead.

Affirmative look-behind

By writing the text you want to extract after (? <= xxx), the text part is extracted when the xxx part matches and the text also matches the part.

re.findall('(?<=ABCD)EF', 'ABCDEF') #['EF']
re.findall('(?<=BC)EF', 'ABCDEF') #[]
re.findall('(?<=)EF', 'ABCDEF') #['EF']

Negative look-ahead

By writing the text you want to extract after (? <! Xxx), the text part is extracted when the xxx part does not exist and the text matches the part.

re.findall('(?<!ABCD)EF', 'ABCDEF') #[]
re.findall('(?<!BC)EF', 'ABCDEF') #['EF']
re.findall('(?<!)EF', 'ABCDEF') #[]

Application of affirmative look-ahead

text = 'Python python PYTHON'

re.findall('py(?=thon)', text) #['py']
re.findall('Py|py(?=thon)', 'Python python PYTHON') #['Py', 'py']
re.findall('py(?=thon)', 'Python python PYTHON', re.IGNORECASE) #['Py', 'py', 'PY']

Usage that is likely to be scraping

text = 'Street address:Shinjuku-ku, Tokyo 〇〇〇〇 Tower\r Click here for map'
re.findall('(?<=:).*(?=\r)', text)[0] #Shinjuku-ku, Tokyo 〇〇〇〇 Tower

[Reference site] Master the look-ahead and look-behind of regular expressions!