I recently criticized Swift 3 for allowing many expressions with no side 
effects as statements, in this blog post: 
https://blogs.synopsys.com/software-integrity/2017/03/24/swift-programming-language-design-part-2/
 (search for "that expression").  I've seen some related discussion such as 
"Make non-void functions @warn_unused_result by default", but not quite this.

For emphasis, let me include a real example (from open-source software, 
simplified) of defective JavaScript code that happens to be legal Swift as well:

var html = "<table class='modification'>" +
      "<tbody>" +
      trs
      "</tbody>" +
      "</table>";

DEFECT SPOILER:                                               (There is a 
missing '+')

Part of my argument is that people commonly ignore compiler warnings.  We see 
lots of defective code that would be (or is) caught by compiler warnings but 
people don't pay attention.

I have not formulated this into a detailed proposal, but I suspect that because 
of Swift's pervasive overloading, implicit constructor calls, etc., it would 
involve introducing a new @error_unused_result annotation and using that in 
many places in the standard library.  I also suggest that user overloads of 
traditionally non-side-effecting operators, as well as non-mutating struct 
methods, be @error_unused_result by default, or perhaps by mandate.  Our 
experience also suggests this @error_unused_result feature could also be useful 
for ordinary methods of classes, as we find a number of defects where a method 
call is expecting a side effect but there is none, because the method is only 
useful for its return value.

If you would like to see more defect examples from open-source software (other 
languages for the moment), I should be able to dig up some more.

Thanks

--
Peter Dillinger, Ph.D.
Software Engineering Manager, Coverity Analysis, Software Integrity Group | 
Synopsys
www.synopsys.com/software

_______________________________________________
swift-evolution mailing list
[email protected]
https://lists.swift.org/mailman/listinfo/swift-evolution

Reply via email to