Question

Let’s say we have this whitelist: one two three. And this blacklist: four five. Then:

• three one two is a matching text (contains all whitelist words);
• one three two six is a matching text (contains all whitelist words);
• two one is not a matching text (lacks a whitelist word three);
• one four two three is not a matching text (contains a blacklist word four).

Is there a regex solution to this problem?

Answer 1

This is not something you’d want to use a regex for. Better do it like this (example in Python):

>>> whitelist = ["one", "two", "three"]
>>> blacklist = ["four", "five"]
>>> texts = ["three two one", "one three two six", "one two", "one two three four"]
>>> for text in texts:
...     mytext = text.split()
...     if all(word in mytext for word in whitelist) and \
...        not any(word in mytext for word in blacklist):
...         print(text)
...
three two one
one three two six
>>>

You can do it, though:

^(?=.*\bone\b)(?=.*\btwo\b)(?=.*\bthree\b)(?!.*\bfour\b)(?!.*\bfive\b)

• ^ anchors the search at the start of the string.
• (?=...) ensures that its contents can be matched from the current position
• (?!...) ensures that its contents can’t be matched from the current position
• \bone\b matches one but not lonely.

So you get:

>>> import re
>>> r = re.compile(r"^(?=.*\bone\b)(?=.*\btwo\b)(?=.*\bthree\b)(?!.*\bfour\b)(?!.*\bfive\b)")
>>> for text in texts:
...     if r.match(text):
...         print(text)
...
three two one
one three two six