Browse files

some thoughts on Joy's fold being different from standard functional …

  • Loading branch information...
1 parent 8bbb0ce commit a8115d43f348adf2120bf6dcb611834ba3f2ffce @stevej committed Oct 10, 2012
Showing with 31 additions and 0 deletions.
  1. +31 −0 docs/
@@ -0,0 +1,31 @@
+Joy's Fold is in the opposite direction from every other functional
+language. I'm trying to determine why that is.
+Joy's fold is defined as a left fold, not a right fold. This means
+that it is not general purpose enough to build the other standard
+functional combinators.
+`fold` in Joy is defined as `swapd step`. Using some rewriting, let's
+see how this expands in a small program that adds some chars together
+into a string.
+`["b" "c" "d"] "a" [concat] fold` (expand `fold`)
+`["b" "c" "d"] "a" [concat] swapd step` (expand `swapd`)
+`[concat] ["b" "c" "d"] "a" step`
+The `step` combinator machinery here peels off elements from the head
+of the array and executes the contents of the quotation. I'm hiding
+the `["b" "c" "d"]` array, it's not not available directly to your
+quotation, rather it's elements are peeled off the head (the left
+side) and put on the stack. the results of concat plus the next
+element are what's on top of the stack for the quotation to work with.
+Back to our rewriting:
+`"b" "a" concat`
+`"ab" "c" concat`
+`"abc" "d" concat`
+the final result left on the stack is `"abcd"`

0 comments on commit a8115d4

Please sign in to comment.