I think rather than a few patches here and there that solve relative
problems, somebody on the Xerces team should look into making serialization
more extensible.
Looking at the source code Assaf Arkin([EMAIL PROTECTED]) is the author of
much of serialization, maybe if he could spare some time and make the
architecture a little more extensible, that would be a good idea.
So if he's listening, could you please do a revisit on serialization.
Sandeep Randhawa
-----Original Message-----
From: Gunnlaugur Thor Briem <[EMAIL PROTECTED]>
To: [EMAIL PROTECTED] <[EMAIL PROTECTED]>
Date: Tuesday, July 24, 2001 12:28 AM
Subject: RE: [PATCH] org.apache.xml.serialize extensibility changes
Hi again,
Right you are Sandeep ... at the risk of overtaxing committer
patience, here's another (current) patch that clips off a few
more finals and package visibilities. I'm leaving two printText
methods in BaseMarkupSerializer final though -- can't see that
they hurt much, and they may help performance a tiny bit.
I also included a little tweak in XMLSerializer -- it's a waste to
construct a new AttributesImpl on each startElement, since most of
the time no change is made. So I changed extractNamespaces to copy
the Attributes lazily, once we know that we need to remove an
attribute.
Agreed?
Cheers,
- Gulli
-----Original Message-----
From: Sandeep Randhawa [mailto:[EMAIL PROTECTED]]
Sent: 23. j�l� 2001 17:20
To: [EMAIL PROTECTED]
Subject: Fw: [PATCH] org.apache.xml.serialize extensibility changes
------------------------------
I Second the request.
------------------------------
Also, for Encodings to be of any use publically, static EncodingInfo
getEncodingInfo(String encoding) should also be public.
For ElementState similarly, to be of any use the fields should be
accessible.
For Neil Graham,
Shoudn't xxxSerializer in general be non final?
Sandeep Randhawa
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
