Matching backslash with regex in QGIS Field Calculator

Question

I'm attempting to remove a part of string in an attribute column of a shapefile. The part in question is Snapshots\\ from Snapshots\390_measure_1.jpeg. So I want the output to be 390_measure_1.jpeg.

Right now I'm using the Field Calculator like so to do this:

regexp_replace("Snapshot_1", '([a-zA-Z])+\\ ([0-9])+_([a-zA-Z])+\.([a-zA-Z])+', '\\2\\3')

This however does nothing to the output. I don't understand what I'm doing wrong. Also there seems to be no way to escape \\ (The backslash character) in order to match it separately.

Answer 1

The correct way to do this seems to be

regexp_replace("Snapshot_1",'Snapshots\\\\','')

This removes "Snapshots\" from the regex as it somehow escapes the backslash.

Answer 2

You could just split on the \ character and return the last element of the array:

split( "Snapshots_1", '\\', -1)

The custom split function can be defined in the Function Editor:

from qgis.core import *

@qgsfunction(args='auto', group='Custom')
def split(value, char, index, feature, parent):
    """
    Splits the parameter on char and returns split[index].
    <h2>Example usage:</h2>
    <ul>
      <li>split("a_b", "_", 0) -> a</li>
      <li>split("a b c", " ", 1) -> b</li>
    </ul>
    """
    return value.split(char)[index]

Answer 3

Two other options how to do it: one without, the other with Regular Expressions.

1. Without using regular expressions:

   replace (
       "Snapshot_1",
       left ("Snapshot_1",10),
       ''
   )
   

   This looks for the 10 characters at the left (first 10 characters, line 3) of the input string (line 2) and replaces them with an empty string (line 4).

   If \ (backslash character) is part of the input string, replace 10 with 11 in the expression.

---

2. With regular expressions

   substr( 
       "Snapshot_1",
       regexp_match( 
           "Snapshot_1",
           '\\\\'
       )+1
   )
   

   This looks for the position of the first backslash \ and returns the substring based on the input starting with the next character: from the first character after the \ till the end.

Answer 4

Another option is to use substr() and strpos(). strpos(string, char) returns the position of the first match for char in the string. substr(string, start[, optional length]) returns the part of the string starting at position start and an optional length, which isn't needed here.

Position of the first backslash: strpos(Snapshot_1, '\\')

Part of the string starting after the backslash. (Note: we need to add 1 position to set the start after the backslash.) substr(Snapshot_1, strpos(Snapshot_1, '\\')+1)