[DIGESTER-96] Multiple element body parts problem

• Type: Improvement
• Status: Closed
• Priority: Minor
• Resolution: Fixed
• Affects Version/s: 1.1

Environment

Operating System: All
Platform: All

• Bugzilla Id: 3893
• Assignee: Unassigned
• Reporter: Teodor Danciu
• Created: 01/Oct/01 08:51
• Updated: 09/Mar/07 20:39
• Resolved: 09/Mar/07 20:39

Description

Hi,

I have this problem when using Digester to parse my XML file.
Suppose we have the following XML structure:

<parent>
firstPart
<child>
child1
</child>
secondPart
<child>
child2
</child>
lastPart
</parent>

The corresponding DTD structure would be:

<!ELEMENT parent (#PCDATA | child)*>
<!ELEMENT child (#PCDATA)>

I saw that the body() method of a Rule object is called only when the end
of the matching pattern is encountered, and that's why we can only retrieve
the "lastPart" portion of the <parent> element body, presented in the
previous example.
Even more, in such a case, the body() method would receive "child2lastPart"
as parameter.
I think this is because we assumed that an element would never have split
body content like the one I constructed.

I propose a solution in witch to call the body() method of a rule
on characters() event of the matching element pattern,
rather than on endElement(), in order to be able to treat all the body chunks.

What do you think about this ?

Thanks!

Best regards,
Teodor Danciu