Re: [Laszlo-user] escape handling in xpathQuery()

[P T Withington]

When your .lzx file is parsed:

    <data>&lt;ABCD&gt;</data>
will be stored internally as a data node with the text `<ABCD>`.  You  
have to use &lt; to get past the XML parser:

    <data><ABCD></data>

would parse ABCD as an empty child node (and give an error because  
there is no matching end tag).  `&lt;` escapes the `<` past the XML  
parser, so that it is treated as text (or character data), rather  
than the beginning of an XML tag.  You could also have said:

    <data><![CDATA[<ABCD>]]></data>

`<![CDATA[` tells the XML parser to not treat `<` and `&` as special  
characters until the matching `]]>` is hit.

Serialize will unescape the data contents because its contract is to  
create a representation that can be parsed (unserialized) back to the  
original data.  So it shows `&lt;ABCD&gt;`.  xpathQuery returns the  
raw data `<ABCD>`.  If you want the result of the query to be  
escaped, you would have to do it yourself:

    Debug.write( "escaped dp xpathQuery: " + LzBrowser.xmlEscape 
(dp.xpathQuery("data/text()")) );


On 2007-03-12, at 22:25 EDT, Daniel Williams wrote:

mt1,
I would think that Laszlo would call a 3rd party library for  
parsing the XML.
You may want to ask the developers how they parse the XML for a  
dataset.

Happy hunting...

-Daniel

mt1 wrote:
Daniel,

 Thank you for your point.
 And i am sorry cause of my dump, but i do not understand what you  
said.
 I know i am using a strict character.
 My question is
     1. described in the dataset tag     &#38;#60;ABCD&#38;#62;
     2. when using serialize()              &amp;#60;ABCD&amp;#62;
     3. when using xpathQuery()         &#60;ABCD&#62;

 In the case of 2, transferred '&#38' to '&amp'.
 In the case of 3, transferred '&#38' to '&'.
 And in the example code of setText() transferred '&#60' to '<'  
when it shown in then view.

 where those are converted to ?  *.as, *.lzx or LPS ?
 I could not find out the converting mechanism of strict characters.

 Give me some hints.

 mt1

Daniel Williams wrote:
mt1,
I missed the #38 before the ">", that's way it was displayed in the
debug window.
So the data should read &#38;#60;ABCD&#38;#62;

My point is this http://www.w3.org/TR/REC-xml/#sec-external-ent,  
your
working with XML which has a strict character set.

-Daniel


mt1 wrote:

Thank you Daniel, but I am getting confuse more.
Your code has shown "&#60;ABCD> " in the debug window and  
"<ABCD>" in
the text area.
That mean xpathQuery() could not handle '&#38;#60;' but <text>  
tag could
do it, right?
Where your point is ?

Thx
mt1

Daniel Williams wrote:


mt1,
I don't think this is a Laszlo code issue. You may need to use the
numeric reference , here is a list.

Ampersand &amp; & &#38;#38;
Left angle bracket &lt; < &#38;#60;
Right angle bracket &gt; > &#62;
Straight quotation mark &quot; " &#39;
Apostrophe &apos; ' &#34;


Try this code out if you want to display "<ABCD>" to the user,  
FYI I'm
running 3.3.3 .

<canvas debug="true">
<dataset name="ds">
<data>&#38;#60;ABCD&#62;</data>
</dataset>

<handler name="oninit">
var dp = new LzDatapointer();
dp.setPointer(ds);

Debug.write( "dp serialize : " + dp.serialize() );
Debug.write( "dp xpathQuery: " + dp.xpathQuery("data/text()") );

var text = dp.xpathQuery('data/text()');
testText.setAttribute('text', text);

</handler>

<view bgcolor="blue" width="100" height="100">
<text id="testText" resize="true" text="Hello"/>
</view>
</canvas>

Just my 2 cents.
-Daniel

mt1 wrote:


Any hints?

mt1 wrote:


Yes i thought so, but there were not it in xpathQuery()  
method or
*.as files.
So i confused where it has been done.
You can see that where is ?

Thanks
mt1

Not Zippy wrote:


I believe the data is being unescaped.
Debug.write(unescape("&lt;ABCD&gt;") )


On 3/5/07, *mt1* < [EMAIL PROTECTED]  
<mailto:[EMAIL PROTECTED]>>
wrote:

Hi forks,

I am wondering how to handle a escape characters in
LzDatapointer.xpathQuery().
This is a example code.

<canvas debug="true">
<dataset name="ds">
<data>&lt;ABCD&gt;</data>
</dataset>

<handler name="oninit">
var dp = new LzDatapointer();
dp.setPointer(ds);

Debug.write( "dp serialize : " + dp.serialize() );
Debug.write( "dp xpathQuery: " + dp.xpathQuery("data/text 
()") );
</handler>
</canvas>

You can see the data with escaped in <dataset> tag.
And the first debug line ( using serialize() ) shows the  
strings
without
modification,
but the second debug line ( using xpathQuery() ) shows "<ABCD>"
out of
the way by
operating unescaped.
I took a look around LzDatapointer.as or LzDataElement.as, but
xpathQuery() do not
translate from '&gt' to '<' anywhere.

Please let me know why xpathQuery() can do such like that ?

mt1