(module shared mzscheme (require "my-macros.ss" (lib "contract.ss") (lib "list.ss") (lib "etc.ss")) ; CONTRACTS (define varref-set? (listof identifier?)) (define binding-set? (union varref-set? (symbols 'all))) (define (arglist? v) (or (null? v) (identifier? v) (and (pair? v) ((flat-contract-predicate (cons/c identifier? arglist?)) v)) (and (syntax? v) (null? (syntax-e v))) (and (syntax? v) ((flat-contract-predicate (cons/c identifier? arglist?)) (syntax-e v))))) (provide/contract ;[varref-set-remove-bindings (-> varref-set? varref-set? varref-set?)] ;[binding-set-varref-set-intersect (-> binding-set? varref-set? binding-set?)] ;[binding-set-union (-> (listof binding-set?) binding-set?)] ;[varref-set-union (-> (listof varref-set?) varref-set?)] [skipto-annotate (-> (listof procedure?) syntax? (-> syntax? syntax?) syntax?)] [skipto-reconstruct (-> (listof procedure?) syntax? (-> syntax? any/c) any/c)] [in-closure-table (-> any/c boolean?)] [sublist (-> number? number? list? list?)] [attach-info (-> syntax? syntax? syntax?)] [transfer-info (-> syntax? syntax? syntax?)] [arglist->ilist (-> arglist? any)] [arglist-flatten (-> arglist? (listof identifier?))]) (provide binding-set-union binding-set-pair-union varref-set-union varref-set-pair-union varref-set-remove-bindings binding-set-varref-set-intersect step-result? (struct before-after-result (finished-exprs exp post-exp after-exprs kind)) (struct before-error-result (finished-exprs exp err-msg after-exprs)) (struct error-result (finished-exprs err-msg)) (struct finished-result (finished-exprs)) (struct finished-stepping ()) list-take list-partition (struct closure-record (name mark constructor? lifted-index)) *unevaluated* no-sexp skipped-step struct-flag multiple-highlight flatten-take closure-table-put! closure-table-lookup get-lifted-var get-arg-var zip let-counter syntax-pair-map make-queue ; -> queue queue-push ; queue val -> queue-pop ; queue -> val queue-length ; queue -> num rebuild-stx ; datum syntax-object -> syntax-object break-kind? ; predicate varref-set? ; predicate binding-set? ; predicate ; get-binding-name ; bogus-binding? if-temp ; get-lifted-gensym ; expr-read ; set-expr-read! values-map a...b ; a list of numbers from a to b reset-profiling-table ; profiling info get-set-pair-union-stats ; profiling info re-intern-identifier finished-xml-box-table) ; A step-result is either: ; (make-before-after-result finished-exprs exp redex reduct) ; or (make-before-error-result finished-exprs exp redex err-msg) ; or (make-error-result finished-exprs err-msg) ; or (make-finished-result finished-exprs) (define-struct before-after-result (finished-exprs exp post-exp after-exprs kind) (make-inspector)) (define-struct before-error-result (finished-exprs exp err-msg after-exprs) (make-inspector)) (define-struct error-result (finished-exprs err-msg) (make-inspector)) (define-struct finished-result (finished-exprs) (make-inspector)) (define-struct finished-stepping () (make-inspector)) (define step-result? (union before-after-result? before-error-result? error-result? finished-result? finished-stepping?)) ; the closure record is placed in the closure table (define-struct closure-record (name mark constructor? lifted-index)) ; bogus-binding is used so that we can create legal bindings for temporary variables (define (create-bogus-binding name) (let* ([gensymed-name (gensym name)] [binding (datum->syntax-object #'here gensymed-name)]) binding)) ; make-binding-source creates a pool of bindings, indexed by arbitrary keys. These bindings ; not eq? to any other bindings[*], but a client can always get the same binding by ; invoking the resulting procedure with the same key (numbers work well). make-binding-source ; also takes a string which will be part of the printed representation of the binding's ; name; this makes debugging easier. ; [*] actually, this is not true if you don't use a one-to-one function as the binding-maker ; make-gensym-source : (string -> (key -> binding)) (define (make-binding-source id-string binding-maker key-displayer) (let ([assoc-table (make-hash-table 'weak)]) (lambda (key) (let ([maybe-fetch (hash-table-get assoc-table key (lambda () #f))]) (or maybe-fetch (begin (let* ([new-binding (binding-maker (string-append id-string (key-displayer key) "-"))]) (hash-table-put! assoc-table key new-binding) new-binding))))))) ; get-arg-var maintains a list of bindings associated with the non-negative ; integers. These symbols are used in the elaboration of applications; the nth ; in the application is evaluated and stored in a variable whose name is the nth ; gensym supplied by get-arg-var. (define get-arg-var (make-binding-source "arg" create-bogus-binding number->string)) ; test cases: (returns #t on success) ; (printf "test of get-arg-binding: ~a\n" ; (let* ([arg3 (get-arg-var 3)] ; [arg2 (get-arg-var 2)] ; [arg1 (get-arg-var 1)] ; [arg2p (get-arg-var 2)]) ; (and (not (eq? arg3 arg2)) ; (not (eq? arg3 arg1)) ; (not (eq? arg3 arg2p)) ; (not (eq? arg2 arg1)) ; (eq? arg2 arg2p) ; (not (eq? arg1 arg2p))))) ; get-lifted-var maintains the mapping between let-bindings and the syntax object ; which is used to capture its index at runtime. ; unfortunately, it can't use "make-binding-source" because you need to compare the items ; with module-variable=?, which means that hash tables won't work. ; my weak-assoc lists are lists of two-element lists, where the first one is in a weak box. ; furthermore, the whole thing is in a box, to allow it to be banged when needed. (define (weak-assoc-add boxed-lst key value) (set-box! boxed-lst (cons (list (make-weak-box key) value) (unbox boxed-lst)))) (define (weak-assoc-search boxed-lst key eq-fun) (let* ([lst (unbox boxed-lst)] [found-val #f] [stripped (let loop ([remaining lst]) (if (null? remaining) null (let* ([first (car remaining)] [first-key (weak-box-value (car first))]) (if first-key (if (eq-fun key first-key) (begin (set! found-val (cadr first)) remaining) (cons first (loop (cdr remaining)))) (loop (cdr remaining))))))]) (set-box! boxed-lst stripped) found-val)) ; test cases: ; (define wa (box null)) ; (define-struct test ()) ; (weak-assoc-add wa 3 4) ; (weak-assoc-add wa 9 10) ; (= (weak-assoc-search wa 3 =) 4) ; (= (weak-assoc-search wa 9 =) 10) ; (= (weak-assoc-search wa 3 =) 4) ; (= (length (unbox wa)) 2) ; (define my-struct (make-test)) ; (weak-assoc-add wa my-struct 14) ; (= (length (unbox wa)) 3) ; (= (weak-assoc-search wa my-struct eq?) 14) ; (set! my-struct #f) ; (collect-garbage) ; (= (length (unbox wa)) 3) ; (= (weak-assoc-search wa 3 =) 4) ; (= (length (unbox wa)) 2) (define lifted-index 0) (define (next-lifted-symbol str) (let ([index lifted-index]) (set! lifted-index (+ lifted-index 1)) (datum->syntax-object #'here (string->symbol (string-append str (number->string index)))))) (define get-lifted-var (let ([assoc-table (box null)]) (lambda (stx) (let ([maybe-fetch (weak-assoc-search assoc-table stx module-identifier=?)]) (or maybe-fetch (begin (let* ([new-binding (next-lifted-symbol (string-append "lifter-" (format "~a" (syntax-object->datum stx)) "-"))]) (weak-assoc-add assoc-table stx new-binding) new-binding))))))) (define if-temp (syntax-property (datum->syntax-object #'here 'if-temp) 'stepper-binding-type 'stepper-temp)) ; gensyms needed by many modules: ; no-sexp is used to indicate no sexpression for display. ; e.g., on an error message, there's no sexp. (define no-sexp (gensym "no-sexp-")) ; skipped-step is used to indicate that the "before" step was skipped. (define skipped-step (gensym "skipped-step-")) ; multiple-highlight is used to indicate multiple highlighted expressions (define multiple-highlight (gensym "multiple-highlight-")) ; *unevaluated* is the value assigned to temps before they are evaluated. It's not a symbol so ; it won't need quoting in the source. Bit of a hack, I know. (define-struct *unevaluated-struct* ()) (define *unevaluated* (make-*unevaluated-struct*)) ; struct-flag : uninterned symbol (define struct-flag (gensym "struct-flag-")) ; list-partition takes a list and a number, and returns a 2vals containing 2 lists; the first one contains the ; first n elements of the list, and the second contains the remainder. If n is greater than ; the length of the list, the exn:application:mismatch exception is raised. (define (list-partition lst n) (if (= n 0) (2vals null lst) (if (null? lst) (list-ref lst 0) ; cheap way to generate exception (let*-2vals ([(first rest) (list-partition (cdr lst) (- n 1))]) (2vals (cons (car lst) first) rest))))) ; (define expr-read read-getter) ; (define set-expr-read! read-setter) (define (list-take n a-list) (if (= n 0) null (cons (car a-list) (list-take (- n 1) (cdr a-list))))) (define (flatten-take n a-list) (apply append (list-take n a-list))) (define-values (closure-table-put! closure-table-lookup in-closure-table) (let ([closure-table (make-hash-table 'weak)]) (values (lambda (key value) (hash-table-put! closure-table key value) key) ; this return allows a run-time-optimization (lambda args ; key or key & failure-thunk (apply hash-table-get closure-table args)) (lambda (key) (let/ec k (hash-table-get closure-table key (lambda () (k #f))) #t))))) ;(begin (closure-table-put! 'foo 'bar) ; (and (eq? (in-closure-table 'blatz) #f) ; (eq? (in-closure-table 'foo) #t))) ;; arglist : for our puposes, an ilist is defined like this: ;; arglist : (union identifier? null? (cons identifier? arglist?) (syntax (cons identifier? arglist?)) ;; ... where an ilist val can be anything _except_ a pair or null ;; arglist->ilist : turns an (possibly improper) arglist into a (possibly improper) list of syntax objects (define (arglist->ilist arglist) (let loop ([ilist arglist]) (cond [(identifier? ilist) ilist] [(pair? ilist) (cons (car ilist) (loop (cdr ilist)))] [(and (syntax? ilist) (pair? (syntax-e ilist))) (loop (syntax-e ilist))] [(null? ilist) null]))) ; arglist-flatten : produces a list containing the elements of the ilist (define (arglist-flatten arglist) (let loop ([ilist arglist]) (cond [(identifier? ilist) (cons ilist null)] [(or (null? ilist) (and (syntax? ilist) (null? (syntax-e ilist)))) null] [(pair? ilist) (cons (car ilist) (loop (cdr ilist)))] [(and (syntax? ilist) (pair? (syntax-e ilist))) (loop (syntax-e ilist))]))) ; zip : (listof 'a) (listof 'b) (listof 'c) ... -> (listof (list 'a 'b 'c ...)) ; zip reshuffles lists of items into a list of item-lists. Look at the contract, okay? (define zip (lambda args (apply map list args))) (define let-counter (syntax-property #'let-counter 'stepper-binding-type 'stepper-temp)) ; syntax-pair-map (using the def'ns of the MzScheme docs): (define (syntax-pair-map pair fn) (cons (fn (car pair)) (cond [(syntax? (cdr pair)) (fn (cdr pair))] [(pair? (cdr pair)) (syntax-pair-map (cdr pair) fn)] [(null? (cdr pair)) null]))) (define (make-queue) (box null)) (define (queue-push queue new) (set-box! queue (append (unbox queue) (list new)))) (define (queue-pop queue) (if (null? (unbox queue)) (error 'queue-pop "no elements in queue") (let ([first (car (unbox queue))]) (set-box! queue (cdr (unbox queue))) first))) (define (queue-length queue) (length (unbox queue))) (define (rebuild-stx new old) (syntax-recertify (datum->syntax-object old new old old) old (current-code-inspector) #f)) (define break-kind? (symbols 'normal-break 'result-exp-break 'result-value-break 'double-break 'late-let-break 'expr-finished-break 'define-struct-break)) ; functional update package (define (swap-args 2-arg-fun) (lambda (x y) (2-arg-fun y x))) (define second-arg (lambda (dc y) y)) (define up-mappings `((rebuild ((,car ,(lambda (stx new) (cons new (cdr stx)))) (,cdr ,(lambda (stx new) (cons (car stx) new))) (,syntax-e ,(swap-args rebuild-stx)))) (discard ((,car ,second-arg) (,cdr ,second-arg) (,syntax-e ,second-arg))))) (define (update fn-list val fn traversal) (if (null? fn-list) (fn val) (let* ([down (car fn-list)] [up (cadr (assq down (cadr (assq traversal up-mappings))))]) (up val (update (cdr fn-list) (down val) fn traversal))))) (define (skipto-annotate fn-list stx annotater) (update fn-list stx annotater 'rebuild)) ; test cases ; (equal? (syntax-object->datum (skipto-annotate (list syntax-e car syntax-e cdr cdr car) #'((a b c) (d e f) (g h i)) (lambda (dc) #'foo))) ; '((a b foo) (d e f) (g h i))) (define (skipto-reconstruct fn-list stx reconstructer) (update fn-list stx reconstructer 'discard)) ; BINDING-/VARREF-SET FUNCTIONS ; note: a BINDING-SET which is not 'all may be used as a VARREF-SET. ; this is because they both consist of syntax objects, and a binding ; answers true to bound-identifier=? with itself, just like a varref ; in the scope of that binding would. ; binding-set-union: (listof BINDING-SET) -> BINDING-SET ; varref-set-union: (listof VARREF-SET) -> VARREF-SET (define profiling-table (make-hash-table 'equal)) (define (reset-profiling-table) (set! profiling-table (make-hash-table 'equal))) (define (get-set-pair-union-stats) (hash-table-map profiling-table (lambda (k v) (list k (unbox v))))) ; test cases : ; (profiling-table-incr 1 2) ; (profiling-table-incr 2 3) ; (profiling-table-incr 2 1) ; (profiling-table-incr 1 2) ; (profiling-table-incr 2 1) ; ; (equal? (get-set-pair-union-stats) ; `(((2 . 3) 1) ((2 . 1) 2) ((1 . 2) 2))) ; until this remove* goes into list.ss? (define (set-pair-union a-set b-set comparator) (cond [(null? b-set) a-set] [(null? a-set) b-set] [else (append (remove* a-set b-set comparator) a-set)])) (define (varref-set-pair-union a-set b-set) (set-pair-union a-set b-set free-identifier=?)) (define (binding-set-pair-union a-set b-set) (cond [(eq? a-set 'all) 'all] [(eq? b-set 'all) 'all] [else (set-pair-union a-set b-set eq?)])) (define (pair-union->many-union fn) (lambda (args) (foldl fn null args))) (define binding-set-union (pair-union->many-union binding-set-pair-union)) (define varref-set-union (pair-union->many-union varref-set-pair-union)) ; binding-set-varref-set-intersect : BINDING-SET VARREF-SET -> BINDING-SET ; return the subset of varrefs that appear in the bindings (define (binding-set-varref-set-intersect bindings varrefs) (cond [(eq? bindings 'all) varrefs] [else (filter (lambda (varref) (ormap (lambda (binding) (bound-identifier=? binding varref)) bindings)) varrefs)])) ; varref-set-remove-bindings : VARREF-SET (BINDING-SET - 'all) -> VARREF-SET ; remove bindings from varrefs (define (varref-set-remove-bindings varrefs bindings) (cond [(eq? bindings 'all) (error 'varref-set-remove-bindings "binding-set 'all passed as second argument, first argument was: ~s" varrefs)] [else (remove* bindings varrefs bound-identifier=?)])) ; sublist returns the list beginning with element and ending just before element . ; (-> number? number? list? list?) (define (sublist begin end lst) (if (= end 0) null (if (= begin 0) (cons (car lst) (sublist 0 (- end 1) (cdr lst))) (sublist (- begin 1) (- end 1) (cdr lst))))) ; attach-info : SYNTAX-OBJECT SYNTAX-OBJECT -> SYNTAX-OBJECT ; attach-info attaches to a generated piece of syntax the origin & source information of another. ; we do this so that macro unwinding can tell what reconstructed syntax came from what original syntax (define (attach-info stx expr) (let* ([it (syntax-property stx 'user-origin (syntax-property expr 'origin))] [it (syntax-property it 'user-stepper-hint (syntax-property expr 'stepper-hint))] [it (syntax-property it 'user-stepper-else (syntax-property expr 'stepper-else))] [it (syntax-property it 'user-stepper-define-type (syntax-property expr 'stepper-define-type))] [it (syntax-property it 'user-stepper-proc-define-name (syntax-property expr 'stepper-proc-define-name))] [it (syntax-property it 'user-stepper-and/or-clauses-consumed (syntax-property expr 'stepper-and/or-clauses-consumed))] [it (syntax-property it 'stepper-xml-hint (syntax-property expr 'stepper-xml-hint))] [it (syntax-property it 'user-source (syntax-source expr))] [it (syntax-property it 'user-position (syntax-position expr))]) it)) (define (transfer-info stx expr) (let* ([it (syntax-property stx 'user-origin (syntax-property expr 'user-origin))] [it (syntax-property it 'user-stepper-hint (syntax-property stx 'user-stepper-hint))] [it (syntax-property it 'user-stepper-else (syntax-property expr 'user-stepper-else))] [it (syntax-property it 'user-stepper-define-type (syntax-property expr 'user-stepper-define-type))] [it (syntax-property it 'user-stepper-proc-define-name (syntax-property expr 'user-stepper-proc-define-name))] [it (syntax-property it 'user-stepper-and/or-clauses-consumed (syntax-property expr 'user-stepper-and/or-clauses-consumed))] [it (syntax-property it 'stepper-xml-hint (syntax-property expr 'stepper-xml-hint))] [it (syntax-property it 'user-source (syntax-property expr 'user-source))] [it (syntax-property it 'user-position (syntax-property expr 'user-position))] [it (syntax-property it 'stepper-highlight (or (syntax-property expr 'stepper-highlight) (syntax-property it 'stepper-highlight)))]) it)) (define (values-map fn . lsts) (apply values (apply map list (apply map (lambda args (call-with-values (lambda () (apply fn args)) list)) lsts)))) ; produces the list of numbers from a to b (inclusive) (define (a...b a b) (if (= a b) (list a) (cons a (a...b (+ a 1) b)))) ;; re-intern-identifier : (identifier? -> identifier?) ;; re-intern-identifier : some identifiers are uninterned, which breaks ;; test cases. re-intern-identifier takes an identifier to a string ;; and back again to make in into an interned identifier. (define (re-intern-identifier identifier) #`#,(string->symbol (symbol->string (syntax-e identifier)))) (provide/contract [syntax-object->hilite-datum ((syntax?) ; input (boolean?) ; ignore-highlight? . opt-> . any/c)]) ; sexp with explicit tags ;; syntax-object->hilite-datum : takes a syntax object with zero or more ;; subexpressions tagged with the 'stepper-highlight', 'stepper-xml-hint', and 'stepper-xml-value-hint' syntax-properties ;; and turns it into a datum, where expressions with the named ;; properties result in (hilite ), (xml-box ), (scheme-box ) and (splice-box ) rather than . It also ;; re-interns all identifiers. In cases where a given expression has more than one of these, they appear in the order ;; listed. That is, an expression with both highlight and xml-box annotations will result it (hilite (xml-box )) ;; ;; this procedure is useful in checking the output of the stepper. (define syntax-object->hilite-datum (opt-lambda (stx [ignore-highlight? #f]) (let ([datum (syntax-case stx () [(a . rest) (cons (syntax-object->hilite-datum #`a) (syntax-object->hilite-datum #`rest))] [id (identifier? stx) (string->symbol (symbol->string (syntax-e stx)))] [else (if (syntax? stx) (syntax-object->datum stx) stx)])]) (let* ([it (case (syntax-property stx 'stepper-xml-hint) [(from-xml-box) `(xml-box ,datum)] [(from-scheme-box) `(scheme-box ,datum)] [(from-splice-box) `(splice-box ,datum)] [else datum])] [it (case (syntax-property stx 'stepper-xml-value-hint) [(from-xml-box) `(xml-box-value ,it)] [else it])] [it (if (and (not ignore-highlight?) (syntax-property stx 'stepper-highlight)) `(hilite ,it) it)]) it)))) ;; finished-xml-box-table : this table tracks values that are the result ;; of evaluating xml boxes. These values should be rendered as xml boxes, ;; and not as simple lists. (define finished-xml-box-table (make-hash-table 'weak)) (provide/contract [syntax-object->interned-datum (syntax? ; input . -> . any)]) ; sexp ;; syntax-object->interned-datum : like syntax-object->datum, except ;; that it re-interns all identifiers. Useful in checking whether ;; two sexps will have the same printed representation. (define (syntax-object->interned-datum stx) (syntax-case stx () [(a . rest) (cons (syntax-object->interned-datum #`a) (syntax-object->interned-datum #`rest))] [id (identifier? stx) (string->symbol (symbol->string (syntax-e stx)))] [else (if (syntax? stx) (syntax-object->datum stx) stx)])) ;; the xml-snip-creation@ unit accepts the xml-snip% and scheme-snip% classes and provides ;; functions which map a "spec" to an xml-snip. ;; An xml-spec is (listof xml-spec-elt) ;; An xml-spec-elt is either ;; - a string, ;; - (cons/c 'scheme-box scheme-spec), or ;; - (cons/c 'splice-box scheme-spec) ;; ;; A scheme-spec is (listof scheme-spec-elt) ;; A scheme-spec-elt is either ;; - a string, or ;; - (cons ... oh crud. #;(define xml-snip-creation@ (unit/sig (create-xml-snip create-scheme-snip create-splice-snip) (import (xml-snip% scheme-snip%)) (define (construct-xml-box spec) (let* ([new-xml-box (instantiate xml-snip% () [eliminate-whitespace-in-empty-tags? #t])] ; need to check what the languages themselves do here [xml-editor (send new-xml-box get-editor)]) (for-each (match-lambda [`(scheme-box ,@(schemeboxspec ...)) (send new-xml-box insert (construct-scheme-box #f schemeboxspec))] [`(splice-box ,@(spliceboxspec ...)) (send new-xml-box insert (construct-scheme-box #f spliceboxspec))] [(? string? text) (send xml-editor insert text)]) spec) new-xml-box)) (define (construct-scheme-box splice? spec) (let* ([new-scheme-box (instantiate scheme-snip% () [splice? splice?])] [scheme-editor (send new-scheme-box get-editor)]) (for-each (match-lambda [`(xml-box ,@(xmlspec ...)) (send scheme-editor insert (construct-xml-box xmlspec))] [(? string? text) (send scheme-editor insert text)]) spec))))) ) ; test cases ;(require shared) ;(load "/Users/clements/plt/tests/mzscheme/testing.ss") ; ;(define (a sym) ; (syntax-object->datum (get-lifted-var sym))) ;(define cd-stx ; (datum->syntax-object #f 'cd)) ;(test 'lifter-ab-0 a (datum->syntax-object #f 'ab)) ;(test 'lifter-cd-1 a cd-stx) ;(test 'lifter-ef-2 a (datum->syntax-object #f 'ef)) ;(test 'lifter-cd-1 a cd-stx) ; ;(test '(a b c) map syntax-e (arglist->ilist #'(a b c))) ;(test '(a b c) map syntax-e (arglist->ilist #'(a . (b c)))) ;(test 'a syntax-e (arglist->ilist #'a)) ;(let ([result (arglist->ilist #' (a b . c))]) ; (test 'a syntax-e (car result)) ; (test 'b syntax-e (cadr result)) ; (test 'c syntax-e (cddr result))) ;(test '(a b c) map syntax-e (arglist-flatten #'(a b c))) ;(test '(a b c) map syntax-e (arglist-flatten #'(a . (b c)))) ;(test '(a b c) map syntax-e (arglist-flatten #'(a b . c))) ;(test '(a) map syntax-e (arglist-flatten #'a)) ; ;(define (add1 x) (+ x 1)) ;(test '(3 4 5) ilist-map add1 '(2 3 4)) ;(test '(3 4 . 5) ilist-map add1 '(2 3 . 4)) ; ;(test '(2 3 4) ilist-flatten '(2 3 4)) ;(test '(2 3 4) ilist-flatten '(2 3 . 4)) ; ;(define new-queue (make-queue)) ;(test (void) queue-push new-queue 1) ;(test (void) queue-push new-queue 2) ;(test 1 queue-pop new-queue) ;(test (void) queue-push new-queue 3) ;(test 2 queue-pop new-queue) ;(test 3 queue-pop new-queue) ;(err/rt-test (queue-pop new-queue) exn:user?) ;(equal? ; (call-with-values (lambda () ; (values-map (lambda (a b) (values (+ a b) (- a b))) ; `(1 2 3 4 5) ; `(9 8 7 6 5))) ; (lambda (sums diffs) ; (list sums diffs))) ; `((10 10 10 10 10) ; (-8 -6 -4 -2 0)))