OCPJP8 errata addition, page 213

[Peter Deak]

In the middle of page 213, the book asks which functional interface would we fill in the blank with to make the code compile. It concludes:

"The only functional interface meeting all three of those criteria is DoubleToIntFunction"

However, that is not the only functional interface that would allow the code to compile. You can also use ToIntFunction<Double> or ToIntFunction<Object>, and the code above will compile.

[Peter Deak]

Also, on the middle of page 306 there appears to be a word missing in the sentence:

"They made it legal to write Exception in the catch block but really only a limited set of exceptions"

I guess it's meant to be something like:

"They made it legal to write Exception in the catch block but really only mean a limited set of exceptions"

[Peter Deak]

Also, at the bottom of page 458, it claims that the following code throws an exception:

Path path1 = Paths.get(new URI("file://pandas/cuddly.png"));

This doesn't appear to be the case on Windows. When I run this code on java version "1.8.0_74":

Path path1 = Paths.get(new URI("file://pandas/cuddly.png"));
System.out.println(path1);

I get the output below, with no exception:

\\pandas\cuddly.png\

[Peter Deak]

Also, on page 573, answer to question 20. The answer and explanation are correct, but the code would fail to compile even before the reasons given; it contains a try-with-resources statement with no { } block. Instead it has an invalid try { .... }; form.

[Peter Deak]

Also, on page 563, question 10 explanation.

The answer C is correct, but based on the three argument overload for reduce, it is c1 that would be an int and c2 would be the String. The explanation claims the opposite (that c1 is the String and c2 is the int).

[Jeanne Boyarsky]

page 213 - You’re right! Have a cow for that really good catch. I wasn’t thinking about autoboxing! Have a cow for that great find!
page 306 - agreed
page 458 - confirmed
page 563 - confirmed
page 573 - yuck. I missed this in a number of places. (Because I create questions by copy/pasting code.)

[Frenkel Smeijers]

Peter Deak wrote:In the middle of page 213, the book asks which functional interface would we fill in the blank with to make the code compile. It concludes:

"The only functional interface meeting all three of those criteria is DoubleToIntFunction"

However, that is not the only functional interface that would allow the code to compile. You can also use ToIntFunction<Double> or ToIntFunction<Object>, and the code above will compile.

And ToIntFunction<Number>.

[Jeanne Boyarsky]

Right!

[Frenkel Smeijers]

There are even more possible solutions:
ToIntFunction (this gives a compile warning)
ToIntFunction<? super Double>
ToIntFunction<? super Number>
ToIntFunction<? super Object>

ToIntFunction<Serializable>
ToIntFunction<? super Serializable>    

ToIntFunction<Comparable>
ToIntFunction<Comparable<?>>
ToIntFunction<Comparable<Double>>
ToIntFunction<Comparable<? super Double>>
ToIntFunction<Comparable<? extends Double>>
ToIntFunction<Comparable<? extends Number>>
ToIntFunction<Comparable<? extends Object>>

ToIntFunction<? super Comparable>
ToIntFunction<? super Comparable<?>>
ToIntFunction<? super Comparable<Double>>
ToIntFunction<? super Comparable<? super Double>>
ToIntFunction<? super Comparable<? extends Double>>
ToIntFunction<? super Comparable<? extends Number>>
ToIntFunction<? super Comparable<? extends Object>>

[Jeanne Boyarsky]

The errata fix avoids that by saying

DoubleToIntFunction" is not true. It should say "DoubleToIntFunction and ToIntFunction meet all three of these criteria. As does "ToIntFunction"

It doesn't get into all the generic variations.