3 ms·

is it really the case that parsers can't tell the greater than sign in "a, b = w < x, y > (z)" without type information? or is it simply difficult to parse for

• nemothekid 5y ago
  The doc says it's actually impossible, and I'm inclined to believe so. Rust has a similar problem in their case they adopted the "turbo fish" operator, where the generic function call would take the following form: let (a,b) = w::<x,y>(z);
• mseepgood 5y ago
  Can you as a human? If you can't do it a parser can't do it either.
• arp242 5y ago
  It's a problem all languages with <..> need to solve one way or the other; see e.g. https://blog.dyvil.org/syntax/2016/04/19/angle-bracket-generics.html https://blog.dyvil.org/syntax/2016/04/19/angle-bracket-gener... So for C# it's resolved "by examining the token after the closing >: If it is one of (, ), ], :, ;, ,, ., ?, == or !=, the expression is parsed as a generic method call." I assume this works reasonably well; however, it's not hard to see how this might be a problem in some cases where the parser gets it wrong.
  • blondin 5y ago
    thanks for explaining :)