\a.\b.\c.\d.ac(bcd); Y\a.\b.\c.\d.\e.b(cd\f.\g.e)\f.\g.ce\h.\i.f(h=)(agide)e; Y\a.\b.\c.bc\d.\e.:d(aec); \a.\b.\c.cab; \a.\b.\c.ca; \a.\b.@$(@#ab); \a.\b.bK\c.\d.ac(@%cd)K; \a.\b.bK\c.ca; \a.\b.\c.@'(\d.\e.@'(\f.\g.@%(df)g)(be))(ac); \a.\b.@((@%a)b; \a.\b.\c.ac(bc)@$; Y\a.\b.\c.\d.dc\e.\f.be(abcf); \a.\b.\c.@((@)ab)c; Y\a.\b.@*(@,:b(ab))(@%K); \a.@,:a(@-a); \a.@&\b.b(a=); @,(KI); @,K; @0(@/#-)(@0(@/#-)(@0(@-(@&\a.C(a(# =))))(@/# ))); @-(@*(@&\a.@ (a(# =))(a(# =)))@2); \a.@1a@3; B@4@/; \a.\b.\c.\d.Cad(bcd); @4(@.(@&\a.@6(#z(aL))(a(#aL)))); \a.\b.\c.\d.\e.ba; \a.\b.\c.\d.\e.ca; \a.\b.\c.\d.\e.\f.eab; \a.\b.\c.\d.\e.\f.fab; @)(C:K)(@4(@&(KK))); @*(@0(@/#@)@<)(@,:(@/##)@<); \a.@0(@5#\)(@,(C(@+@;))(@.@7)(@0(@/#-)(@0(@5#>)a))); \a.@*(@*(@*(@0(@5#()(@1a(@5#))))(@>a))(@)@8@=))(@)@9@7); Y\a.\b.@*(@,T(@?b)(@)(\c.\d.\e.c(@:ed))(ab)))(@%I); Y\a.@,T(@?a)(@@a); @,@#@7(@,(C(@+@;))(@-@7)(@0(@5#=)@A)); @0@3(@.(@1@B(@5#;))); \a.\b.@+(\c.\d.@!a(cK)(\e.:#@(:eK))(Bd\e.# (#!-)(e+)))(Ka)b# ; Y\a.\b.\c.cI(\d.@Ddb)(\d.\e.:#`(@"(abd)(abe)))?; Y\a.\b.\c.c(\d.KI)(\d.@!bd)(\d.\e.@ (abd)(abe))\d.\e.C(@ (@!bd)(C(abe))); Y\a.\b.\c.@Fbc(c?(K(@8(:#IK)))(\d.\e.@:(@:(@9(:#SK))(abd))(abe))?)(@:(@9(:#KK))c); Y\a.\b.b@8@9(\c.\d.@:(ac)(ad))\c.\d.@Gc(ad); Y\a.\b.\c.cK\d.\e.@"(@Eb(@H(d(KI))))(:#;(abe)); \a.@Ca(:#?K)(B(\b.@Ibb)(TK));

# The Compiler Singularity

I have a confession. Behind the scenes, I’ve been running a computer program
to convert readable code to combinators or lambdas with one-character
variables, and to deal with `@`

addresses.

Ideally, to avoid "trusting trust" issues, all this should have been done manually. Nonetheless, each of the our terms seem short enough that a sufficiently motivated human could verify them.

Now that the truth is out, instead of pretending to derive terms by hand, we show the program-generated source of our next compiler up front:

(Again, for clarity we have placed a newline after each semicolon.)

Hopefully, it’s plausible that the above can be verified by hand to be equivalent to the following compiler. We have upgraded our previous parser to support comments, whitespace, and variables consisting of lowercase letters. We also added code to look up the index of a top-level definition.

Prefixing a character with `@`

gives us direct access to the primitive
combinators. (Not to be confused with the `@`

operator of ION assembly.)

or f g x y = f x (g x y); lsteq = @Y \r xs ys a b -> xs (ys a (\u u -> b)) (\x xt -> ys b (\y yt -> x(y(@=)) (r xt yt a b) b)); append = @Y \r xs ys -> xs ys (\x xt -> @: x (r xt ys)); pair x y f = f x y; just x f g = g x; pure x inp = just (pair x inp); sat f inp = inp @K (\h t -> f h (pure h t) @K); bind f m = m @K (\x -> x f); ap x y = \inp -> bind (\a t -> bind (\b u -> pure (a b) u) (y t)) (x inp); fmap f x = ap (pure f) x; alt x y = \inp -> (x inp) (y inp) just; foldr = @Y \r c n l -> l n (\h t -> c h(r c n t)); liftaa f x y = ap (fmap f x) y; many = @Y \r p -> alt (liftaa @: p (r p)) (pure @K); some p = liftaa @: p (many p); char c = sat (\x -> x(c(@=))); liftki = liftaa (@K @I); liftk = liftaa @K; com = liftki (char #-) (liftki (char #-) (liftki (many (sat (\c -> @C (c(# (@=)))))) (char # ))); sp = many (alt (sat (\c -> or (c(# (@=))) (c(# (@=))))) com); spc f = liftk f sp; spch = @B spc char; and f g x y = @C f y (g x y); var = spc ( some (sat (\x -> and (#z(x(@L))) (x(#a(@L))) ))); lcr s = \a b c d -> a s; lcv v = \a b c d -> b v; lca x y = \a b c d -> c x y; lcl x y = \a b c d -> d x y; anyone = fmap (@C @: @K) (spc (sat (@K @K))); pre = alt (liftki (char #@) anyone) (liftaa @: (char ##) anyone); lam r = liftki (spch #\) (liftaa (@C (foldr lcl)) (some var) (liftki (char #-) (liftki (spch #>) r))); atom r = alt (alt (alt (liftki (spch #() (liftk r (spch #)))) (lam r)) (fmap lcr pre)) (fmap lcv var); apps = @Y \rr r -> alt (liftaa @T (atom r) (fmap (\vs v x -> vs (lca x v)) (rr r))) (pure @I); expr = @Y \r -> liftaa @T (atom r) (apps r); def = liftaa pair var (liftaa (@C (foldr lcl)) (many var) (liftki (spch #=) expr)); program = liftki sp (some (liftk def (spch #;))); rank v ds = foldr (\d t -> lsteq v (d @K) (\n -> @: #@ (@: n @K)) (@B t \n -> # (#!(@-))(n(@+)) )) (@K v) ds # ; show = @Y \r ds t -> t @I (\v -> rank v ds) (\x y -> @:#`(append (r ds x) (r ds y))) @?; occurs = @Y(\r v t -> t (\x -> @K @I) (\x -> lsteq v x) (\x y -> or (r v x) (r v y)) @?); unlam = @Y(\r v t -> occurs v t (t @? (@K (lcr (@:#I@K))) (\x y -> lca (lca (lcv (@:#S@K)) (r v x)) (r v y)) @?) (lca (lcv (@:#K@K)) t)); babs = @Y(\r t -> t lcr lcv (\x y -> lca (r x) (r y)) (\x y -> unlam x (r y))); dump = @Y \r tab ds -> ds @K \h t -> append (show tab (babs (h (@K @I)))) (@: #; (r tab t)); main s = program s (@:#?@K) (@B (\ds -> dump ds ds) (@T @K));

Our language is now friendly enough that we are willing to work in it with our bare hands. No more cheating.

In other words, we have reached a singularity because this compiler is self-hosting: it can read its 100% organic artisanal handmade source and produce the corresponding ION assembly.

*blynn@cs.stanford.edu*💡