commons-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Stephen Colebourne <scolebou...@joda.org>
Subject Re: [VOTE] Release Commons Lang 3.0 (RC1)
Date Thu, 03 Mar 2011 11:59:27 GMT
I'm not overly enthused about some of the changes, but since I've not
been paying attention its difficult for me to vote/block. Anyway here
is my review:

ArrayUtils.hashCode() has been removed, but it had different
functionality to Arrays.hashCode wrt nested arrays.

        Object[] arrayA = new Object[] {new boolean[] {true, false},
new int[] {6, 7}};
        Object[] arrayB = new Object[] {new boolean[] {true, false},
new int[] {6, 7}};
        assertEquals(true, Arrays.hashCode(arrayB) == Arrays.hashCode(arrayA));

I don't love the new Pair class. We have an interface based version
here at OpenGamma to allow primitive implementations for performance.
I might be able to get our code released if there was interest.

ArrayUtils.toArray() javadoc has example code that won't compile
(missing "new") and also misses &lt; &gt; in places.

Some new methods use a different brace position from the existing
code, such as ArrayUtils.toArray(), Validate.matchesPattern(),
Validate.exclusiveBetween()...

Some new code isn't explicit about null handling, such as
AnnotationUtils, CharSequenceUtils, Validate.exclusiveBetween()...

Some new methods don't have an @since, such as Validate.notBlank(),
Validate.exclusiveBetween()...

StringUtils is now an odd mixture of methods that accept a
CharSequence and ones that don't. Looking at it, I'd prefer to see
CharSequenceUtils added to, and StringUtils methods just deal with
Strings (A CharSequence might be mutable, so it has a different set of
implications when writing code using it). But if that isn't OK, then
it would be better to see everything in StringUtils take a
CharSequence.

ToStringStyle doesn't have a serialization version ID.

While we've moved away from NullArgumentException, I suspect it may be
reasonably widely used.

DateUtils has added a new MODIFY int enum, rather than using a real
enum. Nor has the existing RANGE int enum been converted

The JavaVersion name field is not used.

ObjectUtils.firstNonNull() differs from Google Guava in that it can
return null. This is probably OK, but should be checked.

Range is only thread safe if the objects held inside are thread safe (javadoc).

Range.containsRange() might be better named containsAll()
Range.overlapsRange might be better named overlaps()

Public constants on StringEscapeUtils do not have javadoc.

The StringUtils.concat methods duplicate the existing join methods.

There are a number of TODOs in the code that might need addressing.

The migration guide exceptions section is missing a header.


Hope some of this helps (!)
Stephen


On 3 March 2011 07:39, Henri Yandell <flamefew@gmail.com> wrote:
> Looking to release 3.0; there's not been a lot of JIRA activity and
> it's 9 months or so since we released the beta.
>
> Downloads:
>
>  http://people.apache.org/~bayard/commons-lang3-RC1/
>
> Maven repo entry:
>
>  http://people.apache.org/~bayard/commons-lang3-RC1/maven/
>
> Website:
>
>  http://people.apache.org/~bayard/commons-lang3-RC1/site/
>
> Tag:
>
>  https://svn.apache.org/repos/asf/commons/proper/lang/tags/LANG_3_0_RC1
>
> [ ] +1
> [ ] -1, because:
>
>
> Thanks,
>
> Hen
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: dev-unsubscribe@commons.apache.org
> For additional commands, e-mail: dev-help@commons.apache.org
>
>

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@commons.apache.org
For additional commands, e-mail: dev-help@commons.apache.org


Mime
View raw message