Few Questions on Progressive parsing

[Susantha Kumara <su...@opensource.lk>]

Hi
 
I am using Xerces C++ SAX parser (version 2.2.0) in progressive parsing
mode in Axis C++. I have following 3 questions,
 
1.	If I call SAX2XMLReader's parseFirst or parseNext at an element
with namespace declarations (example :  <myElem xmlns:xx="yy" . >) the
parser fires startPrefixMapping events followed by startElement event.
But what I expect is parseNext (or parseFirst) to return after firing
only one event (say only a startPrefixMapping event). Then I will call
parseNext again to get the next event. Is this behavior intentional ?.
If so why ?
2.	I provide the parser my InputSource and BinInputStream
instances. And I found that the size of the buffer that is passed to
readBytes function is always some 48KB (given by iMaxToRead). So my
input stream waits to get that much of bytes. Is there any way to reduce
the size of this buffer ?.
3.	I found that when my input stream has more than 48K (size of the
buffer) of bytes the parser does not call the readBytes function again
to read the next part of the stream. What is the reason for this ?. Also
I haven't seen the "curPos" being called. What is the purpose of curPos
function ?
 
Your help for me to get a clear understanding of these issues is very
much appreciated.
 
Thanks in advance,
 
Susantha
Lanka Software Foundation
Office : +94112591785
Mobile : +94777420453