Return-Path: 
 <dev-return-118283-apmail-commons-dev-archive=commons.apache.org@commons.apache.org>
Delivered-To: apmail-commons-dev-archive@www.apache.org
Received: (qmail 92467 invoked from network); 7 Mar 2010 02:07:13 -0000
Received: from unknown (HELO mail.apache.org) (140.211.11.3)
  by 140.211.11.9 with SMTP; 7 Mar 2010 02:07:13 -0000
Received: (qmail 18782 invoked by uid 500); 7 Mar 2010 02:06:54 -0000
Delivered-To: apmail-commons-dev-archive@commons.apache.org
Received: (qmail 18481 invoked by uid 500); 7 Mar 2010 02:06:54 -0000
Mailing-List: contact dev-help@commons.apache.org; run by ezmlm
Precedence: bulk
List-Help: <mailto:dev-help@commons.apache.org>
List-Unsubscribe: <mailto:dev-unsubscribe@commons.apache.org>
List-Post: <mailto:dev@commons.apache.org>
List-Id: <dev.commons.apache.org>
Reply-To: "Commons Developers List" <dev@commons.apache.org>
Delivered-To: mailing list dev@commons.apache.org
Received: (qmail 18472 invoked by uid 99); 7 Mar 2010 02:06:54 -0000
Received: from athena.apache.org (HELO athena.apache.org) (140.211.11.136)
    by apache.org (qpsmtpd/0.29) with ESMTP; Sun, 07 Mar 2010 02:06:54 +0000
X-ASF-Spam-Status: No, hits=-2.3 required=10.0
	tests=RCVD_IN_DNSWL_MED,SPF_PASS
X-Spam-Check-By: apache.org
Received-SPF: pass (athena.apache.org: domain of GGregory@seagullsoftware.com
 designates 216.82.249.243 as permitted sender)
Received: from [216.82.249.243] (HELO mail73.messagelabs.com) (216.82.249.243)
    by apache.org (qpsmtpd/0.29) with ESMTP; Sun, 07 Mar 2010 02:06:49 +0000
X-VirusChecked: Checked
X-Env-Sender: GGregory@seagullsoftware.com
X-Msg-Ref: server-10.tower-73.messagelabs.com!1267927587!40602682!1
X-StarScan-Version: 6.2.4; banners=-,-,-
X-Originating-IP: [137.134.240.188]
Received: (qmail 16979 invoked from network); 7 Mar 2010 02:06:28 -0000
Received: from unknown (HELO postal.rocketsoftware.com) (137.134.240.188)
  by server-10.tower-73.messagelabs.com with AES128-SHA encrypted SMTP;
 7 Mar 2010 02:06:28 -0000
Received: from NWT-S-MBX1.rocketsoftware.com ([fe80::34fc:6d7f:6837:62b]) by
 nwt-s-cas1.rocketsoftware.com ([::1]) with mapi; Sat, 6 Mar 2010 21:06:16
 -0500
From: Gary Gregory <GGregory@seagullsoftware.com>
To: Commons Developers List <dev@commons.apache.org>
Subject: RE: [lang] LANG-510
Thread-Topic: [lang] LANG-510
Thread-Index: AQHKvEbi8ulFHl8c4EqAJviJgphQzJHltEJw
Date: Sun, 7 Mar 2010 02:05:07 +0000
Message-ID: 
 <02AA127CD8DCDE48BC7D2DFB6C87083A0A8256@nwt-s-mbx1.rocketsoftware.com>
References: <31cc37361003050132m260425d2sf41a6441539e8653@mail.gmail.com>
In-Reply-To: <31cc37361003050132m260425d2sf41a6441539e8653@mail.gmail.com>
Accept-Language: en-US
Content-Language: en-US
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
MIME-Version: 1.0

Here is another issue that I find confusing at first glance with Generics. =
I redefined defaultIfEmpty in my sandbox as follows:

    public static <T extends CharSequence> T defaultIfEmpty(T str, T defaul=
tStr) {
        return StringUtils.isEmpty(str) ? defaultStr : str;
    }

This is simple enough but I was surprised that this compiled:

assertEquals("abc", StringUtils.defaultIfEmpty(new StringBuilder("abc"), "N=
ULL"));

I thought that all T's should match, but they do not, the compiler works it=
s type inference magic to match up the call site to the method, such that b=
oth arguments are of type CharSequence. The API as coded can return both a =
String and a StringBuilder depending on the value of the input. That's the =
confusing part.

The question is: Is it worth this potential confusion to allow any CharSequ=
ence to be used?

In this case, the "<T extends CharSequence>" version is the same as:

    public static CharSequence defaultIfEmpty(CharSequence str, CharSequenc=
e defaultStr) {
        return StringUtils.isEmpty(str) ? defaultStr : str;
    }

Are there cases where the "<T extends CharSequence>" version is needed?

Gary=20

> -----Original Message-----
> From: Henri Yandell [mailto:flamefew@gmail.com]
> Sent: Friday, March 05, 2010 01:33
> To: Commons Developers List
> Subject: [lang] LANG-510
>=20
> Thinking further on moving StringUtils to CharSequence, I'd like to
> take the String left(String, int) method as an example. It depends on
> substring(int, int), so is entirely possibly to move over to
> subSequence(int, int).
>=20
> Hypothetical new method:
>=20
>     CharSequence left(CharSequence, int)
>=20
> The downside here is that users currently storing this in a String are
> going to have to cast. Generics to the rescue.
>=20
>     <T extends CharSequence> T left(T, int)
>=20
> This hits two problems:
>=20
> 1) EMPTY is returned when the int is less than 0; EMPTY is a String and n=
ot T.
> 2) subSequence returns CharSequence and not T.
>=20
> I could add a wrapper method to make Strings nicer:
>=20
>     public static String left(String str, int len) {
>         if (str =3D=3D null) {
>             return null;
>         }
>         return left( (CharSequence) str, len).toString();
>     }
>=20
> But that doesn't help the StringBuffer/StrBuilder/StringBuilder user;
> they still get a sucky API.
>=20
> Am I missing anything obvious here, or should I give up the ghost on
> trying to take these methods to CharSequence APIs?
>=20
> Hen
>=20
> ---------------------------------------------------------------------
> 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