regular expression improvement to dected a specific set of numbers

aristosv · 10-23-2022, 10:09 AM

I have a list of strings containing text and numbers

Code:

test 1 99 435 18 1 more text
test 2 97 123 1 81 more text2
test 3 96 4 3 5567 more text3
test 4 99 43 5181 more text4

I am using this regular expression to extract numbers starting with 9, followed by 4 or 5 or 6 or 7 or 9, followed by 6 more numbers

Code:

9*[45679]( *[0-9]){6}

So basically I want to end up with something like this

Code:

I was told that the regular expression I'm using is not good enough though because it will also match 5 8 8 8 8 8 8 and 6888888.

Can you suggest an improved regular expression?
Thanks

boughtonp · 10-23-2022, 11:55 AM

The answer appears to be in this thread: https://www.linuxquestions.org/questions/programming-9/tell-regex-to-run-as-if-there-were-no-spaces-4175712277

aristosv · 10-23-2022, 11:58 AM

No that was just for removing spaces. I know I cant do that with regex. This is for improving the way the matching of the query happens

boughtonp · 10-23-2022, 12:13 PM

Review that thread and compare the patterns to what you have.

If that isn't enough for you to figure it out, spend some time learning the basics of regex - in particular what "*" means.

MadeInGermany · 10-26-2022, 12:34 AM

You are missing a space that was given in your earlier thread

Code:

9 *[45679]( *[0-9]){6}

Without the first space it is 9 any times - even zero times.
With the space it is a 9 followed by a space any times.

grail · 10-26-2022, 04:44 AM

Maybe you could look at something like https://rubular.com/ and terst until your hearts content and you learn what some of thos symbols mean

lucmove · 11-07-2022, 04:02 AM

The error is right in the beginning:

Code:

9*[45679]( *[0-9]){6}

9* says that 9 is optional. The string may or may not start with a 9.

Also, note that the "4" in "test 4" is also matched incorrectly because of that.

This regex works better:

Code:

9[45679]( *[0-9]){6}

But I'm not too sure because the specification is not clear. When you say "followed by 6 more numbers," the distribution of numbers and spaces is not predictable. You provide four different models in your test lines.

Maybe this will work even better:

Code:

9[45679] [0-9 ]{7,8}

But again, I can't be sure because I'm not sure of the specification.