Partial Matches

If kotest fails to match a String or an instance of a data class, it may try to find something similar. For instance, in the following example two fields out of three match, so kotest considers sweetGreenApple to be 66.6% similar to sweetRedApple:

listOf(sweetGreenApple, sweetGreenPear) shouldContain (sweetRedApple)

(snip)

PossibleMatches:
 expected: Fruit(name=apple, color=red, taste=sweet),
  but was: Fruit(name=apple, color=green, taste=sweet),
  The following fields did not match:
    "color" expected: <"red">, but was: <"green">

By default, kotest will only consider pairs of objects that have more than 50% matching fields. If needed, we can change similarityThresholdInPercent in configuration.

Likewise, if kotest does not detect an exact match, it may try to find a similar String. In the output, the matching part of String is indicated with plus signs:

listOf("sweet green apple", "sweet red plum") shouldContain ("sweet green pear")

(snip)

PossibleMatches:
Match[0]: part of slice with indexes [0..11] matched actual[0..11]
Line[0] ="sweet green apple"
Match[0]= ++++++++++++-----

By default, searching for similar strings is only enabled when both expected and actuals strings' lengthes are between 8 and 1024.

If we need to consider shorter or longer expected values, we can change configuration values named minSubstringSubmatchingSize and maxSubtringSubmatchingSize.

Likewise, should we need to consider shorter or longer actual values, we can change configuration values named minValueSubmatchingSize and maxValueSubmatchingSize.

By default, possible matches that are less than 66% similar are dismissed, and that default can be changed via similarityThresholdInPercentForStrings value in configuration.

To disable searching for similar strings altogether, set enabledSubmatchesInStrings to false in configuration.