Using Regex, how do I extract the numbers from this serial code?

Question

I have over 1,000 serial codes I need to enter into a database but they have to be completely numerical for conversion identification purposes. They all look similar to this format but contain different characters/numbers:

d47a3c06-r188-4203-n838-fefd32082fd9

I've been trying to figure out how to use regex to remove all letters and dashes but I'm now at a loss.

I need to know how to turn this: d47a3c06-a188-4203-b838-fefd32082fc9

Into this: 473061884203838320829

Using regex. Then possibly trim it down to a 5 digit number using the first 5 numbers.

Thank you so much!

http://stackoverflow.com/questions/1533659/how-do-i-remove-the-non-numeric-character-from-a-string-in-java - just `\D` is enough. — Wiktor Stribiżew, Apr 18 '16 at 19:39
Any feedback? Tried `\D`? Please post what you tried so far. — Wiktor Stribiżew, Apr 18 '16 at 20:25
Are you aware that that is a UUID, and as such, is in fact a 32-digit hexadecimal number? (I notice where the first one has 'r' and 'n', the other has 'a' and 'b'...) — PJSCopeland, Apr 19 '16 at 04:25

score 0 · Answer 1 · answered Apr 18 '16 at 20:25

0

Depending on your programming language, you can easily filter digits and join them afterwards.
Here's an example in Python with the help of the re module and list comprehensions:

import re

serials = ['d47a3c06-r188-4203-n838-fefd32082fd9', 'e48a3c08-r199-4203-n838-fefd32082fd0']
corrected_serials = []
for serial in serials:
    numbers = re.findall(r'\d+', serial)
    corrected_serials.append(''.join(numbers))

corrected_abbreviated = [item[0:5] for item in corrected_serials]

print corrected_serials
print corrected_abbreviated

# output
# ['473061884203838320829', '483081994203838320820']
# ['47306', '48308']

See a demo on ideone.com

answered Apr 18 '16 at 20:25

Jan

42,290
8
54
79

So far.. Thank you all for your help, I am still struggling with this. I'm using the Regex tool here: http://regexr.com/v1/ and using the expression : [a-z/-] I am so far able to get all the non-numerical characters removed, I am still stumped on how to trim to 5 characters. Eventually this regex expression will be going into a Drupal site that uses feeds to import data from a CSV file, inside the CSV file there are the serials which will be parsed using Drupal feeds CSV. I am using the listed regex tool listed above to make sure I get the right expression. – Jeremy Womack Apr 18 '16 at 21:29

score 0 · Answer 2 · answered Apr 18 '16 at 20:47

Using a first regex with s (search and replace) command, all non digit can be removed s/[^0-9]//g

The result is used with a second regex with s command, only the digits before the fith one are printed "/^$.\{5\}$.*$/\1/.

Use these with bash shell and the sed command.

If the serial numbers are in serials.txt file:

cat serials.txt
d47a3c06-r188-4203-n838-fefd32082fd9

sed -e "s/[^0-9]//g" -e "s/^\(.\{5\}\).*$/\1/" serials.txt
47306

Using printf:

printf d47a3c06-r188-4203-n838-fefd32082fd9 | sed -e "s/[^0-9]//g" -e "s/^\(.\{5\}\).*$/\1/"
47306

The serials are in a CSV file which are then imported intro Drupal using the Feeds module. I've come close to what I need with a bit of your help by using [a-z/-] which removed all letters and dashes but I'm still clueless on how to trim the results. — Jeremy Womack, Apr 18 '16 at 21:50

score 0 · Accepted Answer · edited May 11 '16 at 15:55

0

Since you are using Drupal, if what you need is an answer in PHP, then a PHP translation of the answer made by @jay-jargot is like this:

$input = "d47a3c06-r188-4203-n838-fefd32082fd9";
$str = preg_replace("/[^0-9]/", "", $input);
$str = substr($str, 0, 5);
echo $str, "\n";        ## output: 47306

edited May 11 '16 at 15:55

Mogsdad

44,709
21
151
275

answered Apr 19 '16 at 04:02

Ray

36
4

Using Regex, how do I extract the numbers from this serial code?

3 Answers3