Regex does not contain string without using?

Question

I am trying to put together a piece of regex to match lines that doesn't contain a certain string. So far I have the following:

/^((?!abc).)*$/igm

The issue I have is that the system I am trying to use doesn't support regex fully and so I am not able to use the ?! notation.

How would I accomplish this?

Not sure on the language. It's a web mail client which allows regex to be used in the rules. — Andrew Davis, Dec 22 '15 at 11:42
What are you using that piece of regex for??...Any common task to accomplish??..Because regex for daily tasks are available on the internet. — Mathews Mathai, Dec 22 '15 at 11:44
Its on a mail filter. The regex is fine, it works on testing apps, but just doesn't work on the 3rd party client I'm trying to use it on — Andrew Davis, Dec 22 '15 at 11:46
If the regex engine supports it: use `(*SKIP)(*FAIL)`; see http://stackoverflow.com/questions/24534782/how-do-skip-or-f-work-on-regex — Ruud Helderman, Dec 22 '15 at 16:36

aleroot · Accepted Answer · 2015-12-22T11:48:50.703

0

The notation you are trying to use is a negative lookahead assertion, and there are certain programming languages that don't support lookaround in regex, so if you are using one of those languages try to find an alternative library that support them .

If you can't enable support for lookaround assertions you simply can't negate the part of regex.

edited Dec 22 '15 at 11:48

answered Dec 22 '15 at 11:47

aleroot

71,077
30
176
213

The system is 3rd party and I have no control over it. – Andrew Davis Dec 22 '15 at 11:48
So, I'm afraid you can't do what you want to. – aleroot Dec 22 '15 at 11:49

Ruud Helderman · Answer 2 · 2015-12-22T17:21:02.283

Technically it is possible, but in a very clumsy way. Even something as simple as abc already results in a fairly obscure pattern:

^([^a]|a+[^ab]|a+b[ab]*[^abc])*[ab]*$

Fiddle: https://regex101.com/r/xH8dE9/3

Explanation:

^                    start of string
(
[^a]                 not 'a'
|
a+[^ab]              'a' but not 'ab'
|
a+b[ab]*[^abc]       'ab' but not 'abc'
)*
[ab]*$               left-overs at the end of the string

For longer words the pattern grows quadratically. Becomes particularly complicated for words with the first letter appearing more than once, e.g. adam. It would look something like this (untested):

^([^a]|a+[^ad]|a*(ad)+([^ad]|a[^adm]))*[ad]*$

You may want to use a tool to generate such patterns, rather than write them yourself.

Regex does not contain string without using?

2 Answers2