Mailing-List: contact commons-dev-help@jakarta.apache.org; run by ezmlm
Precedence: bulk
Reply-To: "Jakarta Commons Developers List" <commons-dev@jakarta.apache.org>
Received-SPF: pass (asf.osuosl.org: domain of sandymac@gmail.com designates
 64.233.162.202 as permitted sender)
DomainKey-Signature: a=rsa-sha1; q=dns; c=nofws;
        s=beta; d=gmail.com;
        h=received:message-id:date:from:sender:to:subject:in-reply-to:mime-version:content-type:references;
        b=mm/jrIRZHsrJSueEbtUHZ72HUORLorNccEMEJ01sDmgThuEY+B+2S4Lh3DlXMUrOG2KhJZOfPhgBf1+O2xmQz6InUp2+KgAcIaMPwYbqrcw15U3h4pjrU5oZymqujaUn24cgaFP06DP4ArA0eltULK9ytcTZT0dsKCQBSSvDeKY=
Message-ID: <6bde122b0603051845v1116f082ve928171a8b2c8451@mail.gmail.com>
Date: Sun, 5 Mar 2006 21:45:48 -0500
From: "Sandy McArthur" <sandymac@apache.org>
Sender: sandymac@gmail.com
To: "Jakarta Commons Developers List" <commons-dev@jakarta.apache.org>
Subject: Re: [io] LineIterator suggestions [was: LineIterator finalize]
In-Reply-To: <6bde122b0603051640t187616f2w1803ea8b40d59316@mail.gmail.com>
MIME-Version: 1.0
Content-Type: multipart/mixed;
	boundary="----=_Part_6800_20511998.1141613148002"
References: <6bde122b0603051640t187616f2w1803ea8b40d59316@mail.gmail.com>

------=_Part_6800_20511998.1141613148002
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable
Content-Disposition: inline

Attached is the changed I'd make. If no one objects to those changes I
can commit it myself.

On 3/5/06, Sandy McArthur <sandymac@apache.org> wrote:
> On 3/5/06, Stephen Colebourne <scolebourne@btopenworld.com> wrote:
> > Sandy McArthur wrote:
> >  >>>I don't think LineIterator should have a finalizer method and I
> >  >>>believe the JavaDocs in that class about resource leaks are wrong a=
nd
> >  >>>unnecessarily alarming.
> > How is the javadoc over the top? I'll happily make changes, or go ahead
> > yourself.
>
> I checked out the IO trunk and here is what I'd change relating to the
> current LineIterator:
>
> * I think IOIterator should be removed. It's based on the premise that
> an Iterator needs special action else it will leak resources. Also
> there is only one implementation of this interface, which doesn't
> actually leak anything. I don't believe it's utility justify it's
> existence. Maybe in a later release if you find yourself adding a
> number of Iterators with a close() method you can add it back in and
> retro fit LineIterator to implement it.
>
> * Don't automatically closing the Reader when the last line is read.
> The LineIterator potentially breaks being used with the
> java.io.PushbackReader. PushbackReader lets the Reader backup so to
> speak but it cannot do that if the Reader is closed.
>
> * Either make LineIterator final or change the way hasNext works to
> allow meaningful subclassing. As hasNext() is currently implemented
> there is no way for a sub-class that filters lines that only match a
> regex without reimplementing hasNext() completely.
>
> * Remove the static method closeQuietly(LineIterator). I don't think
> it's useful enough to justify itself.
>
> * Change the constructor to throw an IllegalArguementException not a
> NullPointerException. I personally view an NPE as the result of trying
> to dereference a field or method of null. The constructor doesn't
> actually dereference reader, we are testing that the argument reader
> is a legal and meaningful value to preemptively prevent a future NPE.
>
> --
> Sandy McArthur
>
> "He who dares not offend cannot be honest."
> - Thomas Paine
>


--
Sandy McArthur

"He who dares not offend cannot be honest."
- Thomas Paine


------=_Part_6800_20511998.1141613148002
Content-Type: text/plain; charset=us-ascii

---------------------------------------------------------------------
To unsubscribe, e-mail: commons-dev-unsubscribe@jakarta.apache.org
For additional commands, e-mail: commons-dev-help@jakarta.apache.org
------=_Part_6800_20511998.1141613148002--