How to extract part of string after '*', '~', '-' or '#'

Question

I have strings like this:

234/1-1
3345#5
11/2~10
6754*15

It needed only rightmost part of string after '-', '#', '~'or '*' . I figured how to extract when there is only one sign for example '~' with right and strpos. Can I do that with all at once with field calculator?

• Fezter: Have you looked at regular expressions? regexp_substr in the field calculator?

Answer 1

ArcGIS Method

In field calculator switch the parser to Python and check the "Show Codeblock" box.

In the Pre-Logic Script Code box enter

import re
def splitField(inField):
 return re.split('[-|#|~|\*]', inField)[1]

In the TEXT box enter

splitField(!nameofField!)

QGIS Method

Open Field Calculator and go to the Function Editor tab.

In the code pane enter

import re
from qgis.core import *
from qgis.gui import *
@qgsfunction(args="auto", group='Custom')
def splitField(inField, feature, parent):
 return re.split('[-|#|~|\*]', inField)[1]

Click the load button.

Go to the Expression tab and enter

splitField("NAMEOFFIELD")

Set your output field.

Then click OK to run.

• Aaron: +1 Regular expressions can tackle the most challenging string manipulation.
• draganche85: Sorry, I didn't mention that I use Qgis... and I can't reach how this script work in a qgis field calculator
• Jacob F: I updated my answer with a method for QGIS.
• Nathan W: No need for a custom function this works: regexp_substr('11/2~10','[-|#|~|\*](.*)') regexp_substr is a built in function.
• Jacob F: Good to know, although I had to modify the statement to regexp_substr('11/2~10','[-|#|~|\\*](.*)') to get it to work with the asterisk.

Answer 2

Hmm not sure how to do it all at once but you can do it in a couple of steps. I made an example using the values you specified:

Example

---

1. First we need to replace all forward slashes with an underscore _, the reason for this will be explained in the next step. You can use the expression:

   replace( "name", '/', '_')
   

   Field calculator to replace

---

2. Now we can extract the strings after any non-ASCII symbols such as those you mentioned. We can do this a varierty of operations:

   • substr - returns a part of a string
   • strpos - return first matching position of a substring

   We will also use the special regular expression operation \\W which matches any non-alphanumeric character (i.e. anything other than letters, numbers, ^ and _). This is why we needed to replace / with _. Combining all the above allows us to extract the string after a non-alphanumeric symbol:

   substr( "replace", strpos( "replace", '\\W+')+1, strpos( "replace", '\\W+'))
   

   Field calculator to extract

We also use +1 because we do not want to include the non-alphanumeric symbol itself, only everything after it.

---

3. Result:

   Result

• draganche85: This was method I was thinking about, but I used Jacobs code which is shorter and work great with my data. Thank you on your answer.
• Joseph: @draganche85 - Most welcome but I also agree in that Jacob's answer is much better =)