Strange behaviour from XmlReader

[Jan Obrestad]

Hello,

I've been using the XmlReader class to read xml files.
Lately it has had some strange quirks. It seems to ignore elements when
there is no whitespace between them.

ex
<main><textline ><textelement>A </textelement>
<textelement> B</textelement<tex telement>C</textelement></textline></main>
(really on one line)

works fine, but with

<main><textline ><textelement>A </textelement><te xtelement>B</textelement><te xtelement>C</textelement></textline></main>

it skips element B.

and with
<main><textline ><textelement>A </textelement>
<textelement> B</textelement><te xtelement>C</textelement></textline></main>

it skips element C

A short program to demonstrate the problem:

using System;
using System.Xml;
using System.IO;

namespace XmlTest
{
class TestXml
{
public static void Main()
{
string xml =
"<main><textlin e><textelement> A</textelement><te xtelement>B</textelement><te xtelement>C</textelement></textline></main>";
StringReader strReader = new StringReader(xm l);
XmlReader reader = XmlReader.Creat e(strReader);
reader.MoveToCo ntent();
int i = 0;
while (!(reader.NodeT ype == XmlNodeType.End Element &&
reader.Name.Equ als("textline") ))
{
reader.Read();
if (reader.NodeTyp e == XmlNodeType.Ele ment &&
reader.Name.Equ als("textelemen t"))
{
string text = reader.ReadElem entString();
Console.WriteLi ne("{0} {1}", i, text);
i++;
}
}
Console.ReadLin e();
}
}
}

This now prints
0 A
1 C

It should have printed
0 A
1 B
2 C

Does anyone have any idea what I might be doing wrong?
I had always thought that whitespace was irrelevant in xml.


Jan Obrestad

[Martin Honnen]

Re: Strange behaviour from XmlReader

Jan Obrestad wrote:

string xml =
"<main><textlin e><textelement> A</textelement><te xtelement>B</textelement><te xtelement>C</textelement></textline></main>";
>
StringReader strReader = new StringReader(xm l);
XmlReader reader = XmlReader.Creat e(strReader);
reader.MoveToCo ntent();

Now the reader is positioned on the 'main' start tag.

int i = 0;
while (!(reader.NodeT ype == XmlNodeType.End Element &&
reader.Name.Equ als("textline") ))
{
reader.Read();
if (reader.NodeTyp e == XmlNodeType.Ele ment &&
reader.Name.Equ als("textelemen t"))
{
string text = reader.ReadElem entString();
Console.WriteLi ne("{0} {1}", i, text);
i++;
}
}

For the first run of the loop:
reader.Read() moves reader to the 'textline' start tag.
On the second run of the loop:
reader.Read() moves reader to the 'textelement' start tag.
reader.ReadElem entString() moves reader past the 'textelement' end
tag, meaning it is now positioned on the second 'textelement' start tag.
Output: i as 0, text as 'A'.
i set to 1
On the third run of the loop:
reader.Read() moves reader to the text node with contents 'B'.
So there you have the problem, your combination of
Read/ReadElementStri ng and your conditions fail to find the second
'textelement' start tag.

As a solution you might want to use ReadString() instead of
ReadElementStri ng().


--

Martin Honnen --- MVP XML

http://JavaScript.FAQTs.com/

[Jan Obrestad]

Re: Strange behaviour from XmlReader

Martin Honnen wrote:

As a solution you might want to use ReadString() instead of
ReadElementStri ng().

That fixed it.
Thank you!

Jan