I'm finalizing work on improvements to "put". The main
improvement is that put will now be able to transcode on the fly,
allowing it to do things such as putting a dchar into char sink.
What this means is that "(const(char)[]){}" is now considered an
output range for dchar. This wasn't the case before, and a source
of bugs in functions like formattedWrite: They didn't get much
visibility, since Appender (the sink of choice for tests) can do
it natively. But passing a delegate char sink to formattedWrite
often ended in error.
In any case, my main question is:
*currently*, "char[]" isn't considered an output range. Shouldn't
it though? The rationale is that it contains dchar elements, and
we don't know how to "put" a dechar in a char[]'s front. But we
do now.
Was this just an implementation restriction? Or is there a real
good reason to not allow it?
In my code, it's a one line tweak to "unlock" char[] as a full
fledged output range for char/wchar/dchar/string/wstring/dstring.
Should I do it?
