Lexical closures are an implementation technique in languages with first-class functions.
I'm interested in a simple operational description of function closures. Does anyone know of such a description?
$\begingroup$
$\endgroup$
5
-
5$\begingroup$ Uhm, ask your advisor, the inventor of SOS? $\endgroup$– Andrej BauerSep 7, 2011 at 22:02
-
1$\begingroup$ You are relentless. $\endgroup$– Ohad KammarSep 7, 2011 at 22:36
-
1$\begingroup$ Is there any specific difficulty that prevents you from writing down a small-step SOS yourself? $\endgroup$– Martin BergerSep 8, 2011 at 10:40
-
$\begingroup$ I know what you're up to! $\endgroup$– Andrej BauerMar 8, 2012 at 7:31
-
$\begingroup$ I hope so, although this question is unrelated to that. $\endgroup$– Ohad KammarMar 19, 2012 at 12:14
Add a comment
|
2 Answers
$\begingroup$
$\endgroup$
3
Is chapter 29 of Bob Harper's book what you are looking for?
answered Sep 9, 2011 at 18:20
-
$\begingroup$ No. This is the standard CK-machine style semantics I'm familiar with. From PL-implementation lore, I know that having lexical closures is standard practice. I was wondering whether there is a semantic account which highlights them formally. $\endgroup$ Sep 10, 2011 at 21:13
-
$\begingroup$ Maybe you could edit your title & question to clarify what you are looking for? $\endgroup$ Sep 12, 2011 at 15:57
-
2$\begingroup$ No, he should write his thesis and talk to his advisor. Or make his advisor use StackExchange. $\endgroup$ Sep 13, 2011 at 6:54
$\begingroup$
$\endgroup$
Closures form an important concept that I try to teach in my undergraduate Programming Languages class. You can find my lecture notes online.
The Handout 8 shows big step semantics using environments and closures. The Handout 9 shows my flavour of the SECD machine. (The latter didn't work all that well for teaching purposes because the students got lost in detail and didn't develop good intuitions. So I switched to big-step semantics.)
Neither of these is what you are asking for, but perhaps they could form a starting point?
