Hello,
in the Diamond-Exercise the predefined Function looks like this:
rows(Letter) -> undefined.
So, I expected a single Letter, a Char, something like $A. But to my surprise the tests calling the function with a String “A”.
?_assertEqual(Expected, diamond:rows("A"))
This raises for me the question: How single letters are handled in Erlang? Are they handled as Numbers like $A or as Lists like “A”?
Regards,
Eisenaxt
I don’t know Erlang, but for this particular exercise, the tests indicate that single char strings are used to represent letters. That’s how they’ve chosen to represent them when writing the tests, and that’s perfectly fine.
Note that using strings for letters comes directly from the problem specs, but each track is free to use whatever they believe it is more appropriate. For example, in Clojure we use characters (\A) instead of strings (“A”).
A single letter is really a “codepoint” or integer which represents the letter. the string notation, “…”, is a shortcut for a list of codepoints, so things like:
WordAsString = "Hello"
WordAsChars = [$H, $e, $l, $l, $o]
WordAsCodePoints = [72, 101, 108, 108, 111]
are all equivalent. The $ prefix is really just to make it easier than working with raw numbers, but it is an integer and not a separate datatype.
Thank you for the link. I looked into it.
The problem specs looking very general. But yes, they have to match all languages. They differentiate also not between Lists and Arrays.
Thank you for clarification neenjaw,
exactly what I have thought about Erlang. This means the diamond-Exercise is not idiomatic Erlang. Because: If a function expects a List, but only uses the Head of the List, why the function than not simply expect a single value?
So, the Erlang-Track is simply automated from the problem specs without finetune. Good to know.