harmony-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Stepan Mishura" <stepan.mish...@gmail.com>
Subject Re: [classlib][swing] compatibility: j.s.text.GapContent.replace() behaviour
Date Fri, 03 Nov 2006 10:49:17 GMT
On 11/3/06, Ivanov, Alexey A wrote:
> >-----Original Message-----
> >From: Stepan Mishura
> >Sent: Friday, November 03, 2006 9:43 AM
> >To: harmony-dev@incubator.apache.org
> >Subject: Re: [classlib][swing] compatibility:
> j.s.text.GapContent.replace()
> >behaviour
> >
> >On 11/2/06, Ivanov, Alexey A wrote:
> >>
> >> Hi all,
> >>
> >> I've started fixing HARMONY-1809. To remove throws clause from the
> >> declaration of replace method, as it was proposed by Oleg in
> >> HARMONY-1975, I placed removeItems() and insertItems() calls into
> >> try-catch block. This would work OK for any valid arguments.
> >>
> <SNIP>
> >>
> >>
> >> Any objections, comments, opinions?
> >
> >
> >
> >Hi Alexey,
> Hi Stepan,
> >I'm not quite convinced by your evaluation (I'm not an expert in Swing
> >so I may be wrong). My experiments with GapContent showed that Harmony
> Let me give a quick introduction what GapContent is then. This class
> serves as the default storage for all Document implementations within
> javax.swing.text. It stores text in char[] array... but with a gap in
> it.
> Using the default constructor the array will have length 10. And one
> character will be placed into the buffer; the other 9 characters in the
> array will be the gap, and they are not considered to be portion of the
> content. This prevents frequent re-allocations of the underlying storage
> array where text is inserted or removed. Moving the gap is cheaper than
> re-allocating the array.
> When an insert or remove occurs, the gap is moved so that it starts at
> the position of insert or remove.

Thanks for the explanation.

>initially created different object then RI, for example, if you create
> an
> >object with GapContent() constructor, Harmony will return start==0 and
> >end==9 while RI will return start==1 and end==10. The next point that
> It doesn't really matter where the gap is. This difference shows only
> that the gap in the text buffer is located at different location. Using
> content.shiftGap(0) on RI, you'll get start = 0, end = 9. Accordingly
> using content.shiftGap(1) on Harmony, you'll get start = 1, end = 10.

IMO it might be compatibly issue. I can extend GapContent class and it may
depend on what getGapStart() and getGapEnd() return.

What matters is that content.length() = 1 in both cases, and
> content.getString(0, content.length()) returns "\n".
> Actually, setting the gap to be at 0 is more efficient because the next
> (or first) insert is likely to happen at position 0 rather than 1.
> Therefore to insert text at position 0, the gap will be moved to 0 on RI
> (which involves array copying and some other operations). This won't be
> the case with Harmony because the gap is already at 0.

If you do:
GapContent gc = new GapContent();
gc.insertString(0, "text");

then gc.getGapStart() returns 4 on RI not 0 as you wrote.

>confuses me that the spec. says about position as "logical position in
> the
> >storage"...and... "This is not the location in the underlying storage
> >array". So negative value for position may be considered as valid.
> It's all about how the text is stored in GapContent. Because there's a
> gap modeled "the location in the underlying storage array" differs from
> "the logical position in storage". If you look at the spec of any
> methods which throw BadLocationException, you'll see that position (it's
> called 'where') is to be >= 0. Since this method is used to perform
> operations of insertString() and remove() in RI, negative positions are
> invalid. I believe they put no checks here because validation of
> parameters is performed in those methods which call replace().

As I understood the spec. it is true for methods that change content but
there is no such restriction for methods that work with gap (see
shiftGapEndUp, shiftGapStartDown, shiftEnd, shiftGap)

Another point is that replace() in never used in Harmony implementation
> - it's called from tests only.

But it can be used by some app. so it should be compatible.


>> Alexey.
>> P.S. The related JIRA issues:
>> https://issues.apache.org/jira/browse/HARMONY-1809
>> https://issues.apache.org/jira/browse/HARMONY-1975
>> GapContent Javadoc:
>> Description of GapContent.replace:
>> #replace(int,%20int,%20java.lang.Object,%20int)
>> --
>> Alexey A. Ivanov
>> Intel Middleware Product Division
>Stepan Mishura
>Intel Middleware Products Division
>Terms of use : http://incubator.apache.org/harmony/mailing.html
>To unsubscribe, e-mail: harmony-dev-unsubscribe@incubator.apache.org
>For additional commands, e-mail: harmony-dev-help@incubator.apache.org

Alexey A. Ivanov
Intel Middleware Product Division

Stepan Mishura
Intel Middleware Products Division

Terms of use : http://incubator.apache.org/harmony/mailing.html
To unsubscribe, e-mail: harmony-dev-unsubscribe@incubator.apache.org
For additional commands, e-mail: harmony-dev-help@incubator.apache.org

  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message