QGIS Field Calculator - Remove text inside multipe parentheses from a string

Question

On the QGIS Field calculator, I am trying to remove from a string the text that is inside parentheses using regexp_replace.

While it is not a problem to remove the text from a string that contains only one parentheses group (using

regexp_replace('Steep (61° to 84°)','[(](.*)?[)]','')

it gives "Steep" as a result.

I am having some issues with extracting the text from a string where are present multiple parentheses, e.g.,

'Steep (61° to 84°), Vertical / Near Vertical (85° to 90°)'

From the above string, I would like to obtain

'Steep, Vertical / Near Vertical'

but if I try to use the same regex I obtain again only "steep", because for the system the closing parenthesis is the one at the end of the string.

What happens when you run your regexp_replace() on the whole string? — Erik, Commented Oct 12, 2022 at 11:06
Hi, the result is the same as the previous: steep , because for the expression the last ) is at the end of the string. — Val P, Commented Oct 12, 2022 at 11:17
Have you tried running the regexp_replace() on substrings, which you define by whether they are before or after the comma? — Erik, Commented Oct 12, 2022 at 11:20

Matt · Accepted Answer · 2022-10-12 11:20:24Z

8

You can expand your group to include the [(] and [)] by moving the inner parentheses outwards, like so:

'([(].*?[)])'

The expression becomes:

regexp_replace('Steep (61° to 84°), Vertical / Near Vertical (85° to 90°)','([(].*?[)])','')

answered Oct 12, 2022 at 11:20

Matt

17.9k4 gold badges23 silver badges59 bronze badges

Thanks a lot for your answer. It solves my issue.
– Val P
Commented Oct 12, 2022 at 11:24
Glad to help. There is a trailing space after Steep but it should be fairly simple to remove if you need to.
– Matt
Commented Oct 12, 2022 at 11:26
Prefixing the pattern with a space seems to do the trick: ' ([(].*?[)])'
– Matt
Commented Oct 12, 2022 at 11:29
I was using a simple replace(xxx, ' ,',',') but this is definitely a better solution. Thanks again.
– Val P
Commented Oct 12, 2022 at 11:38
As long as the data is formatted consistently for every row, with both sets of parentheses having a space before them. Otherwise, your method might be safer :)
– Matt
Commented Oct 12, 2022 at 11:45

Add a comment |

pigreco · Accepted Answer · 2022-10-12 18:32:16Z

6

With this expression also take into account the space after Steep:

regexp_replace(
  'Steep (61° to 84°), Vertical / Near Vertical (85° to 90°)',
  ' \\(\\d+° to \\d+°\\)','')

answered Oct 12, 2022 at 18:32

pigreco

4,5201 gold badge17 silver badges33 bronze badges

Add a comment |

2 Answers 2