Hi Antonio,

I am trying to apply your suggested code to my case. In my case I get the substring (between, for example, NV and NV (with brackets). And then I have to insert this substring back into the long string, using your CDATA tags. Here is my example:

cOvString = STREXTRACT( C_WOAUDIT.AUDIT_DESC, "<Fld><Name>WODESCR</Name><OV>","</OV>")
*-- So the string cOvString has bad characters.  
*-- Now I need to put this back into the field AUDIT_DESC and add the CDATA tags.  My attempt:
REPLACE AUDIT_DESC with 
STRTRAN( AUDIT_DESC, "<Fld><Name>WODESCR</Name><OV>" + cOvString + "</OV>", 
"<Fld><Name>WODESCR</Name><OV>![CDATA[" + cOvString + "![CDATA]</OV>" ) IN C_WOAUDIT

Did I add the exclamation sign and the CDATA correctly? I don't know if you had the exclamation sign as part of the sample code or it has to be in the string.

TIA



>Dmitry, you may encapsulate the user provided data inside CDATA sections, and then build the XML elements using them.
>
>

>CREATE CURSOR CUR_TEMP (NAME C(20), OV M, NV M)
>
>LOCAL UserProvidedName AS String
>LOCAL UserProvidedOV AS String
>LOCAL UserProvidedNV AS String
>
>LOCAL cAuditDescr AS String
>
>m.UserProvidedName = "Benny & Sons, LLC"
>m.UserProvidedOV = "This tests ]]> the end of CDATA section."
>m.UserProvidedNV = "Some long description with <, >, & and so forth..."
>
>TEXT TO m.cAuditDescr TEXTMERGE NOSHOW
><?xml version="1.0"?>
><VFPDATA >
>	<row>
>		<name><![CDATA[<<STRTRAN(m.UserProvidedName, "]]>", "]]>]]&" + "gt;<![CDATA[")>>]]></name>
>		<ov><![CDATA[<<STRTRAN(m.UserProvidedOV, "]]>", "]]>]]&" + "gt;<![CDATA[")>>]]></ov>
>		<nv><![CDATA[<<STRTRAN(m.UserProvidedNV, "]]>", "]]>]]&" + "gt;<![CDATA[")>>]]></nv>
>	</row>
></VFPDATA>
>ENDTEXT
>
>MESSAGEBOX(m.cAuditDescr)
>
>XMLTOCURSOR(m.cAuditDescr, "Cur_Temp", 8192)
>
>BROWSE
>

>
>
>>Antonio,
>>
>>The XML is based on a content of a memo field (editbox) on one of the forms.
>>What I am thinking to do is, before converting the memo into the string and into the XML, search for any of these "bad" characters and convert them to "good" characters. For example, convert

<

to LT,

>

to GT and

&

to Ampersand.
>>I don't know how often users actually use these characters. I came upon them when testing and using my sample DB where I use all kind of weird characters.
>>Thanks
>>
>>>Dmitry,
>>>
>>>Since you're allowing the user to edit or prepare the XML document, there is not much you can do besides error trapping. Using current parsers like MSXML you cannot turn a not well-formed XML document into a well-formed XML document. You would have to write a parser for the effect.
>>>
>>>Does the user has to prepare the data for your system in XML?
>>>
>>>>Hi,
>>>>
>>>>I am converting a string into a cursor. The string has values entered by the user (so, I have no control). Here is the simple example:
>>>>

>>>>CREATE CURSOR CUR_TEMP (NAME C(20), OV M, NV M)
>>>>XMLTOCURSOR( cAuditDescr, 'CUR_TEMP', 8192 )
>>>>

>>>>
>>>>What happens is the string cAuditDescr could have characters such as a Less-Than (

<

, Ampersand

&

, Greater-Than (

>

). Any of these characters cause a run-time error.
>>>>
>>>>How do you suggest I change the code above to allow the values to be in the cursor ("CUR_TEMP") without a run-time error?
>>>>
>>>>TIA