Compare two strings and show results with diff - which allows to quickly spot the difference

When comparing two strings with #kotlintest with the test like this:

    "compare two strings" {
      "1: one\n2: two\n3: three".shouldBe("1: one1\n22: two\n3: three")
    }

, the output looks like this:

expected: "1: one1
22: two
3: three" but was: "1: one
2: two
3: three"
org.opentest4j.AssertionFailedError: expected: "1: one1
22: two
3: three" but was: "1: one
2: two
3: three"

Even with such a short string - it’s quite hard to spot, what exactly differs between those two strings. And it’s much harder if comparing longer strings, or object attributes, as the result becomes just one long horizontally scrollable line.

However - we are all very used to read the “diff” - with its +/- lines, and colors, which instantly allow us to spot the difference.

For example, here’s the result of the same assert, when ran by rspec (test lib for ruby):

expected: "1: one\n22: two\n3: three"
     got: "1: one\n2: two\n3: three"

(compared using ==)

Diff:
@@ -1,4 +1,4 @@
 1: one
-22: two
+2: two
 3: three

As you can see - it helps me in two ways. First - it puts both strings right under each other so that I can spot character level differences. Secondly, it shows nice diff with our familiar coloring so that I instantly see what’s not different, and what’s different.

Is there a way to achieve some test result comparison which would be similar, and give similar benefits to quickly spot the difference?

I tried KotlinTest and JUnit5 - both comparison results are very “horizontally scrollable in one line”.

Am I overlooking something obvious here?

[sksamuel]

kotest will do the diff if a) you’re running on the JVM and b) your string is over 50 lines long. However you can control that 50 line limit. If you set kotest.assertions.multi-line-diff-size=0 as a system property then it should use diff all the time.

[sksamuel]

assert with diff is a good idea

“qweqwe” shouldBeWtihDiff “rtertert” or something.

On Wed, 12 Jan 2022 at 12:52, Holger Brandl @.***> wrote:

My problem was more the other way round: There was no diff and no indication how to enable it. That’s why ended up writing my own extension function (see referenced commit).

So adding some hint why a diff was not reported and how to enable it, would be more useful. A system property is a wrong approach IMHO, because it’s a behaviour the user may want to set on a per test basis (or maybe even within a single test function). So a parameter or a different assert* seems better to me.

— Reply to this email directly, view it on GitHub https://github.com/kotest/kotest/issues/1084#issuecomment-1011351879, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAFVSGSHMCH33HQGMECBSFLUVXEVTANCNFSM4JPPIC4A . Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.

You are receiving this because you commented.Message ID: @.***>