URL decoding: UnsupportedEncodingException in Java

Question

What I understand from the documentation is that UnsupportedEncodingException can only be thrown if I specify a wrong encoding as the second parameter to URLDecoder.decode(String, String) method. Is it so? I need to know cases where this exception can be thrown.

Basically, I have this code segment in one of my functions:

if (keyVal.length == 2) {
    try {
        value = URLDecoder.decode(
            keyVal[1],
            "UTF-8");
    } catch (UnsupportedEncodingException e) {
          // Will it ever be thrown?
    }
}

Since I am explicitly mentioning "UTF-8", is there any way this exception can be thrown? Do I need to do anything in the catch block? Or, if my understanding is completely wrong, please let me know.

Stephen C · Accepted Answer · 2012-05-25 05:54:40Z

56

It cannot happen, unless there is something fundamentally broken in your JVM. But I think you should write this as:

try {
    value = URLDecoder.decode(keyVal[1], "UTF-8");
} catch (UnsupportedEncodingException e) {
    throw new AssertionError("UTF-8 is unknown");
    // or 'throw new AssertionError("Impossible things are happening today. " +
    //                              "Consider buying a lottery ticket!!");'
}

The cost of doing this is a few bytes of code that will "never" be executed, and one String literal that will never be used. That a small price for the protecting against the possibility that you may have misread / misunderstood the javadocs (you haven't in this case ...) or that the specs might change (they won't in this case ...)

edited May 25, 2012 at 5:54

answered May 17, 2011 at 11:37

Stephen C

723k95 gold badges849 silver badges1.3k bronze badges

Sign up to request clarification or add additional context in comments.

5 Comments

Pacerier Over a year ago

What are some practical likely examples that a JVM will be broken?

Stephen C Over a year ago

@Pacerier - I cannot think of any. You'd probably need to modify the "rt.jar" file (or worse) to break this. So it is theoretically possible ... but not something you are likely to ever encounter unless you are creating / using an experimental JVM.

Tomasz Nocoń Over a year ago

Is there any reason why this excpetion is checked?

Stephen C Over a year ago

Presumably because the designers thought that this was an error condition that a typical application should handle. (In hindsight, that was probably a mistake, but it can't be fixed without breaking binary compatibility.)

Stephen C Over a year ago

The other issue is that it is only "UTF-8" where there is a guarantee that the exception cannot be thrown. For other encoding names ... the exception is possible, and making the exception checked is a reasonable decision. Maybe the solution would be to add a URLDecoder.decodeUTF8( method that didn't throw the exception at all.

zergius · Accepted Answer · 2016-07-01 10:08:10Z

16

That's because of the odd choice to make UnsupportedEncodingException checked. No, it won't be thrown.

I usually do as follows:

} catch (UnsupportedEncodingException e) {
  throw new AssertionError("UTF-8 not supported");
}

edited Jul 1, 2016 at 10:08

zergius

4341 gold badge7 silver badges14 bronze badges

answered May 17, 2011 at 11:37

Bozho

599k147 gold badges1.1k silver badges1.2k bronze badges

Comments

Andreas Dolk · Accepted Answer · 2011-05-17 11:30:43Z

6

In your special case - no, it won't be thrown. Unless you execute your code in a Java runtime that does not support "UTF-8".

answered May 17, 2011 at 11:30

Andreas Dolk

115k20 gold badges185 silver badges275 bronze badges

6 Comments

Mat Over a year ago

Such a runtime is not supposed to exist - at least in JDK 1.6, UTF-8 is a standard charset. download.oracle.com/javase/6/docs/api/java/nio/charset/…

Andreas Dolk Over a year ago

@Mat - I believe this too but .. Oracle is not the only vendor of Java runtimes. The java language spec does not mention any encodings that have to be supported (except: UTF-16 which is the specified encoding for char)

Mat Over a year ago

Wow, thanks. I thought it did, and that javadoc seems to imply it but you're correct, that's not in the JLS.

sleske Over a year ago

@Andreas_D: While the JLS does not specify encodings, the javadocs linked by Mat are a specification as well, and are normative. So any certified Java implementions must support UTF-8 in the provided implementation of the Java core API.

Andreas Dolk Over a year ago

javadocs are normative? Do you have a reference for that statement? Honestly, I do not believe that one has to implement the JLS and all of the javaDoc details (of which version, btw?) in order to get a certificate.

|

k314159 · Accepted Answer · 2021-02-18 11:09:10Z

5

To answer an old question for newer readers:

Java 11 now has URLDecoder.decode(String, Charset); which does not throw. So you don't have to use a try-catch block at all. Just do:

URLDecoder.decode(keyVal[1], StandardCharsets.UTF_8);

answered Feb 18, 2021 at 11:09

k314159

12.4k2 gold badges29 silver badges72 bronze badges

Collectives™ on Stack Overflow

URL decoding: UnsupportedEncodingException in Java

4 Answers 4

5 Comments

Comments

6 Comments

Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

4 Answers 4

5 Comments

Comments

6 Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related