Doesn't that mean typing try a million times instead of putting potentially throwable calls inside one try block and a bunch of different catch handlers?
I wonder why they insist on that. Perhaps it is easier to parse?
I suspect it's useful to visually signal that the call can fail, and make it obvious what's happening; after all, the "throws" annotation is at the target site, not on the caller end.
You need to prefix anything that throws with a "try":
Other than that, it looks exactly like C++ exceptions: Propagation happens along the chain of functions marked as "throws", until a "catch" handler intercepts it.
"try" is an expression, by the way, so you can do things like:
let line = try file.readline()
There's also a "try!" statement that throws a runtime exception if the inner statement throws.
I wonder why they insist on that. Perhaps it is easier to parse?