Regex, capture one or more 6-digit numbers AFTER specific pattern

user_13928093 0
Viewed 117

my first question, so bear with me should anything no quite right. (I have read all topics suggested after entering title, but still no luck)

In words I need to capture all (i.e. one or more) 6-digits numbers following after specific keyword (here: "Customer\s"), which begins with 14 or 15

Input string:

Part Numbe

Customer PO

Description

Cust Part No: 325386

EB011036 // 175663/147667/8 ECK

5500003766 / 148266 -

WALK

between the start keyword to look after, and the following suitable 6 digit pattern any signs can be placed.

What I have come up so far: Since the string can contain return, new lines etc. I set the (?s) flag before:

Then a positive lookbehind:

(?s)(?<=Customer )

followed by

.+?

and then

(\D(1(4|5)\d{4})\D)+

So the whole expression:

(?<=Customer ).*?(\D(1(4|5)\d{4})\D)+

https://regex101.com/r/FEKwcY/1

But this only finds the first instance (147667) after Customer, not the second or more following. If I change the wildcard between the groups to gready, it will find the last one, but not the first possible one.

Having played around further I added an alternation in the pos. lookbehind group:

(?s)(?<=Customer |\D\d{6}\D).*?(\D(1(4|5)\d{4})\D)+

This in fact find the two, but I am afraid, that this will also find before the Customer keyword.

powershell regex
2 Answers

You can use

(?s)(?<=\bCustomer\b.*?)\b1[45]\d{4}\b

See the .NET regex demo. Details:

  • (?s) - RegexOptions.Singleline inline modifier flag
  • (?<=\bCustomer\b.*?) - a positive lookbehind that matches a location that is immediately preceded with Customer as a whole word and then any zero or more chars as few as possible
  • \b - a word boundary
  • 1 - a 1 char
  • [45] - 4 or 5
  • \d{4} - any four digits
  • \b - a word boundary

PS demo:

PS C:\Users\admin> $s = "Part Number`r`n`r`nCustomer PO`r`n`r`nDescription`r`n`r`nCust Part No: 325386`r`n`r`nEB011036 // 175663/147667/8 ECK`r`n5500003766 / 148266 -`r`n`r`nWALK"
PS C:\Users\admin> Select-String '(?s)(?<=\bCustomer\b.*?)\b1[45]\d{4}\b' -input $s -AllMatches | Foreach {$_.matches.value}
147667
148266
user_38329700

I feel like this is more of a communal given answer as both @WiktorStribiżew and @TheFourthBird had relevant contributions that lead to the following pattern that should work in PowerShell's .Net regex flavor:

(?:^.*?\bCustomer\b|\G(?!^)).*?\b(1[45]\d{4})\b

See an online demo.

  • (?: - Open non-capture group:
    • ^.*?\bCustomer\b - Start string anchor followed by 0+ characters (lazy) upto "Customer" inside word-boundaries.
    • | - Or:
    • \G(?!^) - Assert position at position of previous match and explicitly avoid to match the start of the string.
    • ) - Close non-capture group.
  • .*?\b - 0+ Characters (lazy) upto a word-boundary.
  • (1[45]\d{4}) - A capture group that start a substring with a 1 followed by either a 4 or 5 and 4 other random digits.
  • \b - Another word-boundary.
user_97581940
Related