Re: XML-SAX not quite right II

[Adam Glauser <adamglauser@xxxxxxxxxxxx>]

Sharon wrote:
> when event = *XML_CHARS;                                     
>        eventval = event + 1;                                 
>        if %subst(chars : eventval : stringLen) = 'DataArea'; 
>        pathvalue = pathvalue;                                
>        endif;                                                
>                                                                           
> > EVAL chars                                                              
>   CHARS =                                                                 
>             ....5...10...15...20...25...30...35...40...45...50...55...60  
>        1   '  <DataArea>    <storeNumber>7</storeNumber>    <trans' 

DataArea appears to be an element in your XML file (your DTD from the 
previous message* also shows this).  You should be processing the 
DataArea start tag using the *XML_START_ELEMENT event.

I haven't written code using XML-SAX, but based on my experience with 
other SAX parsers, I think the *XML_CHARS event is meant to return the 
characters _between_ elements.  Based on the XML document you posted in 
your other message*, I think that '%subst(chars : 1 : stringLen)' should 
be equal to the following values for the first few *XML_CHARS events:
1:  x'0D25' - this is the END_OF_LINE marker for the first line
2:  x'0D25' - this is the END_OF_LINE marker for the second
3:  x'0D254040' - this is the END_OF_LINE marker for the third line,
      plus the two spaces preceding the <DataArea> tag.
4:  x'0D2540404040' - this is the END_OF_LINE marker for the fourth
      line, plus the leading spaces of the sixth line
5:  '7' or x'F7' - this is the value contained in the <storeNumber>
      element.

As you can see, *XML_CHARS should fire for _any_ data that is between 
elements, including whitespace characters.  The XML-SAX opcode has an 
option that allows you to trim whitespace, so you'd end up with empty 
strings for 1-4.  That may help you simplify the *XML_CHARS processing. 
 According to Barbara Morris, there is even an example of this in the 
XML-SAX documentation**.

I've found Jon Paris and Susan Gantner's articles in IBM Systems 
magazine helpful for these sorts of topics.  This one*** covers XML-SAX, 
and should help you get a better handle on how your SaxHandler routine 
ought to work.

Hope this helps,
Adam

*   http://archive.midrange.com/rpg400-l/200811/msg00144.html
**  http://archive.midrange.com/rpg400-l/200704/msg00184.html
*** http://www.ibmsystemsmag.com/i5/march06/coverstory/9021p5.aspx