Added a test case showing that the outcome of using an object for the result of a conditional transform is incorrect.
FLUID-5250: Added a test case
This pull request should go in with a fix for FLUID-5250, which is waiting to be fixed.
There is a pull request which fixes this: #455
This network of pull requests is a mess, and we should abandon them with the exception of #454 - two entirely unrelated issues have become conflated together as a result of sloppy pull handling, that is, FLUID-5250 and FLUID-5251. FLUID-5251 is a legitimate issue that kasparmarkus is working on - FLUID-5250, at face value, is a nonissue since the use of "literalValue" is the intended way for the user to achieve the effect they want. However, we need to review the behaviour of the MT system since this behaviour often confuses people.
The test case supplied by jobara is unrelated to the fix which is supplied for FLUID-5251.