I could see this being the difference between expr and po. jingham@apple.com (JIRA User), what do you think?

Comment by Jim Ingham (JIRA)

The result of "expr" for a string should show the \n , \t, etc. Since you need something to compare against what you typed into your code, and expr is the debugger view of the object, it seems right that expr should show the characters as typed.

The output of "po" for swift is the result of the debugDescription for String, which comes from the swift standard library. lldb just outputs the resultant text to the console, so real newlines would be newlines, etc. I don't have a strong opinion about how that behaves, but that's something to take that up with the standard library folks.

Note that if you want to render the string - as opposed to getting its debug description, you can do:

(lldb) expr print(someString)

That will show the rendered string and it's actually more efficient than po for printing strings - since `po` has to evaluate the expression, then pass it the expression's result to the mirror printing routine.

lldb just outputs the [ result of the debugDescription for String ] to the console, so real newlines would be newlines, etc.

So, String.debugDescription os escaping itself then? I guess we can't change how po works, can we just make String.debugDescription not escape itself? Seems redundant since expr is going to do the same thing for us (@ standard library folks). Maybe we could make a special case for String?

(lldb) expr print(someString)

That's a neat workaround

Comment by Jim Ingham (JIRA)

Yes, probably the Standard Library folks thought the distinction between "description" - which doesn't escape - and which BTW is what print uses - and debugDescription which shows the string as typed in code was useful. I can actually see the sense behind that, especially given that you can get the rendered version from "print" in the debugger pretty easily...