Errata for Programming Clojure

#41419: Method isBlank of StringUtils class contains an error -
"The method must return of type boolean" .

One of variants of correction of an error:

public class StringUtils {
public static boolean isBlank(String str) {
int strLen;
if (str == null || (strLen = str.length()) == 0) {
return true;
}
for (int i = 0; i < strLen; i++) {
if ((Character.isWhitespace(str.charAt(i)) == false)) {
return false;
}
}
return true; // INSERT FOR CORRECTION
}
} --Serg

#41857: The remark about "following the *nix instructions" for Cygwin is not accurate. From a Cygwin shell, the repl.sh script does not work. You must change the -cp argument (quote it and change colons to semicolons). The reason is that java for Windows uses a different path separator.

#40813: In the example of redefining the hello function to be hello-with-memory, the text says you can verify the new functionality by typing (hello "Rich") and then typing (hello "Rich") again. If you use (load-file "examples/introduction.clj") to automatically enter the example code instead of typing it in by hand, however, the example fails, because the example file defines everything in the "examples.introduction" namespace. When you type just (hello "Name"), you're really getting (user/hello "Name"), which is the simpler function you typed in before. The new function lives in "examples.introduction/hello".

Typing (examples.introduction/hello "Rich") twice in a row does display the new functionality.--mark nutter

#41655: (require 'clojure.contrib.str-utils)
and
(require 'examples.introduction) result in a java.io.FileNotFoundException - could not locate on classpath. I downloaded the zip to c:\code but the book has not told me to amend classpath? I used the repl.bat under C:\code\bin.

Thanks.--andy

#41693: This may be a misunderstanding on my part, but for the code example:
[code]
(defn index-filter [pred coll]
(when pred
(for [[idx elt] (indexed coll) :when (pred elt)] idx)))
[/code]

the book states "The index/element pairs of (indexed coll) are bound to the names idx and elt but only when (pred elt) is true."

It seems that elt, at the very least, must get bound regardless in order to evaluate whether (pred elt) is true; it does not make sense to me that it could base its decision to bind or not based on whether an as-yet-unbound name satisfies a predicate.

Alternately, if it is clever enough to somehow recognize "I only want to bind values to [idx elt] if the value that would be bound to elt satisfies pred, so I will test pred with that value." then that should probably be explicitly stated, though I am not sure that that is actually different than binding.--Joel Jorgensen

#41026: Using the "binding" form does not work properly on Mac OS Snow Leopard using the code downloaded from the Programming Clojure website.

(def foo 10)
=> #'user/form

(binding [foo 42] foo)
=> 10 (should be 42)

The Clojure version reported by the REPL is 1.1.0-alpha-SNAPSHOT.--Cameron McKay

#41568: The example shows what would happen if Clojure allowed unqualified symbols "start" and "result" in a macro. The example continues with "start" being bound to the user's value, 1. This gives an erroneous result on the time it takes to add the numbers, NOT the result of adding the numbers.

So, in the middle of the page, "All of
a sudden, one plus two seems to equal 1228277342451783002." should be "All of
a sudden, adding one plus two seems to take 1228277342451783002 nanoseconds."--Fred Daoud