Consider 1-4 done of these.

Main task is documenting WordUtils and checking Duration/StopWatch.

Stephen

----- Original Message ----- From: "Henri Yandell" <[EMAIL PROTECTED]>
To: "Jakarta Commons Developers List" <[EMAIL PROTECTED]>
Sent: Saturday, December 18, 2004 3:21 AM
Subject: [lang] 2.1...



Sorry for the silence from me. As I've probably said a million times,
baby happened.

I'm full of coding energy, or perhaps the need to 'have coded' that
leads to coding happening and looking to direct some of it to Lang
2.1.

Looking at the wiki and bugzilla, we were nearly there as far as I
recall. Not a lot has cropped up in the 2 months or so that have
passed, couple of bugs I've easily closed out and a couple of trickier
ones.

Gary seems to have taken care of
http://issues.apache.org/bugzilla/show_bug.cgi?id=32625. Is it ready
for closure Gary? Despite Matt Blum's comment on a work-around, it
still sounds like a fine change to the library.

The only other new bug is a note on incorrect javadoc and should be an
easy one: http://issues.apache.org/bugzilla/show_bug.cgi?id=32619.

Stephen listed the following as smelling:

-------
1)
- ArrayUtils.lastIndex()
Gets the last valid index of an array. Surely users would just call
getLength() - 1?

2)
- ClassUtils.CLASS_NAME_COMPARATOR
- ClassUtils.PACKAGE_NAME_COMPARATOR
Although potentially useful, they seem quite specific. What about comparing
packages using Class objects? (I've never used a Package object in my work)
Or comparing just the short name of a class?


3)
- NotImplementedException
This now implements Nestable, with a lot of extra methods. None of the other
exception classes in the main package have been changed. We should revert
this change, or find a lighter weight solution that works with
ExceptionUtils.


4)
- Validate.allElementsOfClass
Should rename to allElementsOfType, and use instanceof style check, not
class equals style check

Still TODO:
5)
- WordUtils
Capitalize with separator methods need to define null handling for delimiter
list, and better javadoc for two of the three methods


6)
- time.DurationFormatUtils (complete?)
7)
- time.StopWatch (complete?)
8)
- text subpackage (not in 2.1)
---------

I think we were all agreed on 8. 6+7 I'm open to ideas on
improvements, I think they were complete but need to look again to
convince myself.

3, 4 and 5 all look to be quite agreeable and so just need to be done.

1 and 2 look to be "justify thy existence" questions to the relevant
features. So we need to decide whether we think they should go or not.

1 seems to be a classic "it's semantically better" argument, whereas 2
seems to be a "it's not generic enough". Usually we keep code with
semantically better arguments and drop not-generic enough code I
think, so 1 would stay and 2 would be rm'd. I'm not tied to either
though, so ymmv.

I've clovered and uploaded to:
http://www.apache.org/~bayard/commons-lang-2.1/clover/

Thoughts?

Hen

---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]





--------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]



Reply via email to