What is the reason to replace existing rule with new rule when longer?

[eddyw]

Hi 👋

I'm failing to understand this particular piece of code here:

          if (errorMessageByKey[key]) {
            const existingRule = errorMessageByKey[key];

            /**
             * If the new rule is longer than the existing rule,
             * replace the existing rule with the new rule.
             */
            errorMessageByKey[key] =
              newError.length > existingRule.error.length
                ? errorObj
                : existingRule;
          } else {
            errorMessageByKey[key] = errorObj;
          }

What is the intention?
Is there an example diagnostic? I'm failing to reproduce it and it's driving me crazy 😄

[mattpocock]

The reason is that some errors are actually a subset of other errors - i.e:

"{0} is bad"
"{0} is bad because of something"

In that case, we want to choose the longer version of the same rule.

HOWEVER - this code might not be needed any more, we should add tests in this area to check it out.

[eddyw]

Oh that's interesting 👍