Replace with whitespace

问题:

I try to parse this xml-information:

<text:p >Lorem<text:s/>ipsum.</text:p>

Therefore I'm using XMLReader. Nearly everything is working as I need it. But the <text:s/>-element makes some trouble for me. As I want to remove any formatting tags (i.e. bold) I'm using expand()->textContent to get just the text:

$reader = new XMLReader();
if (!$reader->open("content.xml");
while ($reader->read()) {
    if ($reader->nodeType == XMLREADER::ELEMENT && $reader->name === 'text:p') echo utf8_decode($reader->expand()->textContent);
}

In this case I would get 'Loremipsum.' instead of 'Lorem ipsum.'. How can I replace every <text:s/> with a whitespace.

Update: I did it this way: preg_replace("/<\\/?text:s(\\s+.*?>|>)/", " ", utf8_decode($reader->readInnerXML()))

Update:

If I'm using DOMDocument for parsing, how do I have to change the syntax?

$reader = new DOMDocument();
$reader->load("zip://folder/".$file.".odt#content.xml");

while ($reader->read()){ 
    if ($reader->nodeType == XMLREADER::ELEMENT && $reader->name === 'text:h') { 
        if ($reader->getAttribute('text:outline-level')=="2") $html .= '<h2>'.$reader->expand()->textContent.'</h2>';
    }
    elseif ($reader->nodeType == XMLREADER::ELEMENT && $reader->name === 'text:p') { 
        if ($reader->getAttribute('text:style-name')=="Standard") {
            $str = $reader->readInnerXML(); 
            // replace text:s-elements with " " at this point
        }
    }
}

回答1:

You don't want to output the <text:p> elements, but you want to output just text-nodes and the <text:s> element as just a space:

 $reader = new XMLReader();
 $result = $reader->open("content.xml");
 if (!$result) {
     throw new UnexpectedValueException('Could not open XML file for reading.');
 }

while ($reader->read()) {
    if ($reader->nodeType == XMLREADER::ELEMENT && $reader->name === 'text:s') {
        echo " "; // SPACE
    }
    if ($reader->nodeType == XMLREADER::TEXT) {
        echo $reader->textContent;
    }
}

So it's more a problem with the processing logic and less a technical issue with XMLReader.

Some note on the character encoding I've left out in my example:

The conversion to Latin-1 you do (utf8_decode) should normally not be necessary if you deliver the output as UTF-8. See Character encodings.

If it's necessary for your target output, it's most likely not necessary to take care of it at that place, see ob_iconv_handler.