Omitting certain style string labels in QGIS 3 rule-based labeling

Question

I’m labeling airport taxiways. The taxiway labels are all strings but they tend to be one of three varieties and I would only like to show one of those. I’ve been trying to mess around with Rule-Based Labeling for this layer and haven’t had any success yet.

The strings can be one of the following format:

1. A, B, C...AA,BB,CC… A1, B1 etc
2. A.B, A.B.C, A1.B2.A, etc <— the use of the “.” In these makes them different from the first variety
3. $UNK<— this is when the feature has no known value but the field must still be filled in

Below I've attached an example of what the labels look like in QGIS.

I would like to only show the labels of the first variety. I think the best way would be to create a rule that doesn’t print anything that equals $UNK or has “.” in it. I’m just unsure of how to write that as a rule or even if rule-based labeling is the right way to achieve my goal.

Answer 1

In the label expression settings you can set up a case statement that basically states

case 
when label in ('list', 'of', 'values', 'to', 'omit') then '' 
else 'label_column' 
end

You wouldn't need to set a rule for every pattern you want to omit, and you can also use LIKE statements as well, which would extend the above to something like:

case 
when label_column in ('list', 'of', 'values', 'to', 'omit') then '' 
when label_column like 'BADVALUE%' THEN ''
else 'label_column' 
end

The case statement is converting all the labels in your IN or LIKE statements to '' (nothing), otherwise, using the existing value in the label_column

Answer 2

To add onto @DPSSpatial's reply you could use an expression as a data-defined override for the 'Show label' option under the Label Rendering tab (see screenshot - click on the circled button and select Edit... then paste one of the below expressions adapted to your data)

A simple expression for your case would be: "label_column" NOT LIKE '%.%' AND "label_column" != '$UNK'

This filters out all labels with a full stop in them as well as any matching '$UNK'

Another approach is to use a regex expression to only match labels that have only alphanumeric characters (which excludes the $ and . characters in the labels you don't want). An example for your case would be: regexp_match("label_column",'^[\\w]+$').

This matches any combination of one or more alphanumeric and whitespace characters ([\\w]+) that are the only contents between the beginning and end of the string (^ and $ in regex)