On Sat, 19 Jul 2003, Stephen Colebourne wrote:

> From: "Henri Yandell" <[EMAIL PROTECTED]>
> > > Having just updated the whitespace processing, I now think that
> isBlank()
> > > should be fully Unicode compliant and trim using
> Character.isWhitespace()
> > > not String.trim().
> >
> > So how is it different from StringUtils.isWhitespace?
>
> It isn't. My preference is now for
>
> isEmpty() - "" or null
> isBlank() - whitespace only, "" or null

Could just tell people to do isWhitespace && isEmpty  ? :)

> and maybe:
> isEmptyTrimmed() - trim() then "" or null

I'm not sure there's any need for this. I wonder how much of the chars
less than 32 count as whitespace? But it seems that isWhitespace is for
most people's usage a superset of isEmptyTrimmed.

> (plus isNotEmpty, isNotBlank, isEmptyNN, isBlankNN)

Still hard to decide how far to go with all these. isEmptyNN is a
single-atom method. isNotEmpty just replaces a ! sign etc etc.

Does anyone actually want isEmptyNN, or is it just that people are unhappy
with the null-handling in StringUtils? In which case I think the NN is the
wrong solution, we need to be thinking about an ability to create a
StringUtils with a strategy or having an underlying hidden class and 3
facade's for the different strategies. All 3.0.

I think removing the trim from isEmpty is good, it's not expected by
users. I think the isEmpty-w/ trim people can just use isWhitespace [we
can javadoc them over].

Sorry to take so while to wake up on this,

Hen


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

Reply via email to