errata.txt

Errata: 1st Printing EOPL3
Updated Wed Dec 12 14:44:27 2018

Page 5:
In the deduction tree 2/3 of the way down the page, all the hypotheses that say "xxx in N" should say "xxx in Int", 
as described in Definition 1.1.5 (on p. 4)
h/t: Kyle Stansfield (stanfield@ucsb.edu)

Page 44:
Exercise 2.19:
move-to-left => move-to-left-son
move-to-right => move-to-right-son

Page 45:
Exercise 2.19: The definition of t1 is missing a right paren.

Page 45:
Exercise 2.20
at-leaf? is already in exercise 2.19

Page 84:
Exercise 3.32
mututally => mutually

Page 101:
Exercise 3.39
Delete "and 'unpack'"

Page 116:
Figure 4.6
The return type of value-of should be Answer.

Page 124:
Section 4.4:
Line 5:
newpair ==> make-pair

Page 180:
205 is printed by the producer. So the text should say:

So the first two items, 300 and 205, are printed by the main thread and the subthread respectively.

Page 195:
Figure 6.1, in apply-cont, the case for fact1-cont:
(begin
  (set! cont saved-cont)
  (set! n saved-n)
  (apply-cont))))))
==>
(begin
  (set! cont saved-cont)
  (set! val (* val saved-n))  ;; changed here
  (apply-cont)))
  
Page 196, Figure 6.2, next-to-last line of apply-cont:
(set! n saved-n)
=>
(set! val (* val saved-n)) 
 
Page 208:
a-program ==> cps-a-program
const-exp ==> cps-const-exp
var-exp ==> cps-var-exp

Page: 209:
The contract should be
apply-procedure/k : Proc * ExpVal * Cont -> ExpVal.
Also, "cps-" should precede "let-exp", "letrec-exp",
"if-exp", and "call-exp".

Page: 216:
Line 10 (line 4 of the second display):
The last paren should be preceded by ">>", leading to "exp_n>> K) >>)"

Lines 9 and 14 of the final display:
The two closing parentheses should each be preceded by ">>" leading to ">>) >>)"

Page 217:
Line 7 of last display
There is an unmatched set of <<'s and >>'s.
Need to move the '>>' to the next term following an equal sign
and have it follow the "K)"

Page 219:
The type and name say it is for cps-of-diff-exp, but
the code is for cps-of-sum-exp.  So, we need the type:
cps-of-sum-exp : Listof (InpExp) * SimpleExp -> TfExp

Page 223-224
The code for cps-of-let-exp is incorrect, because it puts k-exp inside
the scope of id.  The solution is to translate the 'let' into the immediate
application of a lambda expression.  Then cps-of-exps will make the
needed fresh variables. 

  (define cps-of-let-exp
    (lambda (id rhs body k-exp)
      (cps-of-exp
       (call-exp
        (proc-exp (list id) body)
        (list rhs))
       k-exp)))

The proposed code in Exercise 6.27 suffers from the same error.


Page 223:
The type left off the type of the letrec-body.
It is, of course, just InpExp.  Here is what the type should say.
  ;; cps-of-letrec-exp :
  Listof(Listof(Var)) * Listof(InpExp) * InpExp * SimpleExp -> TfExp

Page 265
Section 7.4.3
optype ==> otype

Page 230
Line -4
apply-procedure ==> apply-procedure/k

Page 231
Line 2
apply-procedure ==> apply-procedure/k

Page 273
Exercise 7.28
In the example,
The first 'let' should be 'letrec', to fit into the syntax of INFERRED
Also
((foo f) cdr(x)))  => (foo cdr(x)))

Page 290
Exercise 8.5
 interface
        [even : int -> bool
         odd  : int -> bool]
=>
 interface
        [even : (int -> bool)
         odd  : (int -> bool)]

Page 303:
(name type?) => (name symbol?)
Note: in the repo, this field is called t-name.
See chapter8/abstract-types-lang/static-data-structures.scm

Page 337 et seq:
The code in
https://github.com/mwand/eopl3/blob/master/chapter9/classes is
inconsistent .  data-structures.scm defines ExpVal to include objects,
with the injector obj-val.  However, the rest of the interpreter
doesn't use obj-val.  Objects are passed around without the wrapper.
The code still works, because method-call-exp doesn't check to see
whether its argument is an object.  method-call-exp should call a
function like expval->obj, in the same way that diff-exp calls
expval->num. 

Page 341:
extend-env* => extend-env

Page 362:
2 lines above the display
instance => interface

Page 378: Line 2:
Delete ", and"