miloignis/kraken
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Make drop_redundent_veval recursive, but realized we now need to re-partial eval in the cases where it does change. Added fail points with error messages for things I know I still need to do - veval is in a tricky space now because it *can't* have the main calling code re-evaluate it's parameters, but with the new prim comb works like regular change, it does. Maybe a negative evaluation level?

Browse Source
This commit is contained in:
Nathan Braswell
2022-02-09 01:11:13 -05:00
parent 5a67c704e0
commit 6e18a66e3b
1 changed files with 15 additions and 8 deletions
Download Patch File Download Diff File Expand all files Collapse all files

23
partial_eval.csc
View File

@@ -508,16 +508,21 @@
(drop_redundent_veval (rec-lambda drop_redundent_veval (env_id x) (dlet ((r (if
(and (marked_array? x)
(not (.marked_array_is_val x))
(prim_comb? (idx (.marked_array_values x) 0))
(= 'veval (.prim_comb_sym (idx (.marked_array_values x) 0)))
(= 3 (len (.marked_array_values x)))
(not (marked_env_real? (idx (.marked_array_values x) 2)))
(= env_id (.marked_env_idx (idx (.marked_array_values x) 2)))) (drop_redundent_veval env_id (idx (.marked_array_values x) 1))
(not (.marked_array_is_val x)))
(if (and (prim_comb? (idx (.marked_array_values x) 0))
(= 'veval (.prim_comb_sym (idx (.marked_array_values x) 0)))
(= 3 (len (.marked_array_values x)))
(not (marked_env_real? (idx (.marked_array_values x) 2)))
(= env_id (.marked_env_idx (idx (.marked_array_values x) 2)))) (drop_redundent_veval env_id (idx (.marked_array_values x) 1))
(marked_array false
(.marked_array_is_attempted x)
(map (lambda (it) (drop_redundent_veval env_id it))
(.marked_array_values x))))
x)))
(begin (print_strip "result of drop_redundent_veval (with " env_id ") (problem was " (cond
(begin (error "if we do a drop_redundent_veval and it does, we need to re-evaluate because the veval might have been blocking")
(print_strip "result of drop_redundent_veval (with " env_id ") (problem was " (cond
((not (marked_array? x)) "(marked_array? x)")
((not (not (.marked_array_is_val x))) "(not (.marked_array_is_val x))")
((not (prim_comb? (idx (.marked_array_values x) 0))) "(prim_comb? (idx (.marked_array_values x) 0))")
@@ -761,16 +766,17 @@
; If our env was implicit, then our unval'd code can be inlined directly in our caller
(implicit_env (array pectx nil (drop_redundent_veval (.marked_env_idx de) ebody)))
((combiner_return_ok ebody (.marked_env_idx eval_env)) (array pectx nil (drop_redundent_veval (.marked_env_idx de) ebody)))
(true (error "FIXME - veval needs to re-val it's body, including? somehow? any env references or the main call will re-partial-eval them in the wrong env"))
(true (array pectx nil (drop_redundent_veval (.marked_env_idx de) (marked_array false true (array (marked_prim_comb recurse 'veval 0 true) ebody eval_env)))))
))))
) (veval_inner only_head de env_stack pectx (if implicit_env (array unval_body) (array unval_body eval_env)) indent))))
) 'eval 1 true))
; This will have to evaluate the other sides?
(array 'cond (marked_prim_comb ((rec-lambda recurse (first_evaled_already) (lambda (only_head de env_stack pectx params indent)
(mif (!= 0 (% (len params) 2)) (array pectx (str "partial eval cond with odd params " params) nil)
(dlet (
(_ (error "This will have to evaluate the other sides? Also, if we figure out veval re-val, maybe this can collapse back into cond"))
(eval_helper (lambda (to_eval pectx)
(dlet (((ok unvald) (try_unval to_eval (lambda (_) nil))))
(mif (not ok)
@@ -3151,6 +3157,7 @@
; ctx is (datasi funcs memo env pectx)
; return is (value? code? error? (datasi funcs memo env pectx))
(compile-inner (rec-lambda compile-inner (ctx c need_value) (cond
(true (_ (error "The entire compiler needs to support our new value-default thing, esp for cond and unval'ing function call params")))
((val? c) (let ((v (.val c)))
(cond ((int? v) (array (<< v 1) nil nil ctx))
((= true v) (array true_val nil nil ctx))