Search and Replace strings in a multiple XML

[nitekram]

Are you looking just to delete this string ( <Eq Source="LATEX"><![CDATA[ $$begin{array}{*{20}l} Kw+Pa &= V, P^Tw &= 0, end{array} $$ ]]>></Eq> ) - if so you can use the same code above and then just use "" (a set of double quotes, or single quotes) for the replace parameter

If I am wrong - can you describe a little more?

[aguynamedray]

There are lots of CDATA sections in xml. But I need to delete the <Eq Source="LATEX"> as well as the sub element and the data in it. Can you show me how? Thanks.

[nitekram]

I do not know the structure of XML well enough to understand what you are looking for. Can you provide more details and examples of what you are trying to do?

[aguynamedray]

To make it clear, how do I replace special characters and numbers (e.g. <![CDATA[[ $$ or , ^, &, and soon) in xml?

[nitekram]

If you are just talking about removing special characters then you could use the same function that was written. Put the special character in for the search string and then search for it - if the string is going to be the same across the board, or make a list of your string matches. So using the check to make sure the string/line has the character in question and if it does - replace with NULL (if beta user) or '' "" (a single quote 2 times, or a double quote 2 times) - that will replace it with nothing - not even a space.

But I think you are looking for something else?

[aguynamedray]

Yup! Your right, I was looking for something else. I would like to delete the whole <Eq ID="DE1">...</Eq> node as well as the child node in all xml found in output directory.

Edited March 11, 2013 by aguynamedray

[nitekram]

<Eq ID="DE1">...</Eq>

Taking this example - I believe we are on the same page just reading a different language. If the string or NODE all start with <E then we search for that, and then find all characters from that point to the next point q>

Clear as mud? You will need to look at StringMid & StingLen in the help file - they should give you a idea of how to use it to find what you are looking for. There are some gurus on here that use RegEx that might have a way, but I do not use that - YET

[Nessie]

Try this:

#include <File.au3>

$xml_file = _FileListToArray(@ScriptDir, "*.xml", 1)
If @error Then
    MsgBox(16, "Error", "No xml file")
    Exit
EndIf

For $i = 1 To UBound($xml_file) - 1
    $xml_path = @ScriptDir & "\" & $xml_file[$i]
    _ReplaceStringInFile($xml_path, '<EqSource Format="TEX">', '<Source Format="LATEX">')
    _ReplaceStringInFile($xml_path, '</EqSource>', "</Source>")

    $Open = FileOpen($xml_path)
    $Read = FileRead($Open)

    $sRegex = StringRegExpReplace($Read, '(<Eq ID="DE1">.*?</Eq>)', "")

    If Not @error Then
        $Open = FileOpen($xml_path,2)
        FileWrite($Open, $sRegex)
    EndIf

    FileClose($Open)
Next

Hi!

Edited March 11, 2013 by Nessie

[jdelaney]

I've always found xpaths to be much easier to work with than regexp:

$stest = '<?xml version="1.0"?><items><item name="racecar" test="test2"><type title="value">value1</type><type title="value">value2</type></item></items>'
$oXML=ObjCreate("Microsoft.XMLDOM")
;~ $stest = @DesktopDir & "\xml2.xml"
$oXML.loadxml($stest)

$oNodes = $oXML.selectNodes( '//items/item' )
For $oNode In $oNodes
$oNode.text = "whatever value you need"
Next
ConsoleWrite ( $oXML.xml & @CRLF )

when loading a file to the xml object, use .load instead

Edited March 11, 2013 by jdelaney