“配对词提取器” - 正则表达式

Question

I am using Regular Expression in Python to extract "And" Words. Meaning, words that are separated by and.

For example

• banking and finance
• profit and loss

Effort so far:

import re
read = open("sample.txt", "r")
regex = re.compile('(?:\S+\s)?\S*and\S*(?:\s\S+)?')
f=open('write.txt','w')
for line in read:
    words = regex.findall(line)
    for word in words:
        f.write(str(word)+'\n')
f.close()

This code seems to work well but finds and inside words such as commands.

So I used this Regular Expression

regex = re.compile('a-zA-Z]+\s?\S*and\S*\s+[a-zA-Z]+')

which works well in website but returns only word and without the preceding word and succeeding word as output inside python.

My intention is to find words separated by and inside a document.

Input

This is a sample text to find profit and loss. It should also find banking and finance. But it should not find commands.

Current output

• profit and loss.
• banking and finance.
• find commands.

Expected out put

• profit and loss
• banking and finance

Answer 1

You're making this more complicated than it needs to be. Just use the following regex:

\S+\sand\s\S+

See it in action

The issue was the \\S* you added around the and . That matches any number of non-whitespace characters around the "and", which would match words like "brandy".

Answer 2

You could try this:

\w+(?=\sand\s)|(?<=\sand\s)\w+

Which is:

• Some word ( \\w+ ) matched only where it precedes \\sand\\s with a positive lookahead assertion, OR
• Some workd ( \\w+ ) matched only where it follows \\sAnd\\s with a positive look-behind assertion

The positive lookbehind needs a string of fixed length so you can't do (?<=\\s+and\\s+) so this solution assumes all the spacing is single spaces.

Tested at regex101.com

Edit

Further to the update in the question to get the something and something else as a three-word phrase you can try:

\w+(?:\s+and\s+)\w+

Tested with this output: