Re^3: Ways to implement a closure

I see. Well there you go, yet another way!

Nit: that code works in C (given some imagined Hash class), and the data does stick around:

Hash<...>* demonstration<...>(... a)
   Hash<...>* hash_ptr = new Hash<...>("demo_data", a);
   return new Hash<...>(
      "demonstration_data", hash_ptr
   );
}
[download]

I think you meant this:

sub demonstration {
   my %demonstration_data = ( demo_data => shift );
   return {
      demonstration_data => \%demonstration_data,
   };
}
[download]

Which wouldn't work in C:

Hash<...>* demonstration<...>(... a)
   Hash<...> hash("demo_data", a);
   return new Hash<...>(
      "demonstration_data", \hash    XXX BUG
   );
}
[download]

Comment on Re^3: Ways to implement a closure Select or Download Code

Replies are listed 'Best First'.
Re^4: Ways to implement a closure by SpanishInquisition (Pilgrim) on Oct 18, 2004 at 19:50 UTC
C does not have templates OR classes! (Not that this is a bad thing...)	[reply]
Re^5: Ways to implement a closure by ikegami (Patriarch) on Oct 18, 2004 at 20:05 UTC
Oops, I often call C++ by the wrong name. But while the code was C++, the argument does work for C as well. Here's the C version of the C++ snippets: `HASH* demonstration(void* p) HASH* hash_ptr = hash_new() hash_add(hash_ptr, "demo_data", p); HASH* outer_hash_ptr = hash_new(); hash_add(outer_hash_ptr, "demonstration_data", hash_ptr); return outer_hash_ptr; }` [download] and `HASH* demonstration(void* p) HASH hash; hash_init(&hash); hash_add(&hash, "demo_data", p); HASH* outer_hash_ptr = hash_new(); hash_add(outer_hash_ptr, "demonstration_data", &hash); XXX BUG return outer_hash_ptr; }` [download]	[reply] [d/l] [select]