xliff export import problems

[antton]

Hi,

The xliff export-import function has some problems when we use javascript code in any idevice.

For example if we write the following easy function.

<script type="text/javascript">// <![CDATA[
function programa(sw) {
var pic;
var testua;
switch (sw) {
    case 0:
        pic = "MCN04_AUX_R09_00_SolucionModales.png";
	testua = "<h1>hola</h1>";
        break;
    case 1:
        pic = "MCN04_AUX_R09_N20_SolucionModales.png";
	testua = "<h1>hola2</h1>";
        break;
    case 2:
        pic = "MCN04_AUX_R09_N30_SolucionModales.png";
	testua = "<h1>hola3</h1>";
        break;
   
}
document.getElementById('myImage').src = pic;
document.getElementById('codea').innerHTML = testua;

}
// ]]></script>

It return the following function to the new import idevice.

<script type="text/javascript">// &lt;![CDATA[
function programa(sw) {
var pic;
var testua;
switch (sw) {
    case 0:
        pic = &quot;MCN04_AUX_R09_00_SolucionModales.png&quot;;
	testua = &quot;&lt;h1&gt;hola&lt;/h1&gt;&quot;;
        break;
    case 1:
        pic = &quot;MCN04_AUX_R09_N20_SolucionModales.png&quot;;
	testua = &quot;&lt;h1&gt;hola2&lt;/h1&gt;&quot;;
        break;
    case 2:
        pic = &quot;MCN04_AUX_R09_N30_SolucionModales.png&quot;;
	testua = &quot;&lt;h1&gt;hola3&lt;/h1&gt;&quot;;
        break;

}
document.getElementById('myImage').src = pic;
document.getElementById('codea').innerHTML = testua;

}
// ]]&gt;</script>

As we can see it escape all the html entities with something like (htmlentities in PHP) but it doesn't reformat again with something like (html_entity_decode in PHP) to make it work properly then.

Would be nice to fix it ;)

[antton]

It looks like the problem could be when you export the content wrapping on CDATA and you try to import it again, the CDATA of the javascript code is inside the content general CDATA and the parses never looks for that part so it let it "broken".

I hope someone could check it!

[ignaciogros]

@JuanCasanovaSdweb:

I think it's just as @antton says: eXe should replace those entities (<...) with the original code (<...).

They're using XLIFF to translate many contents, and TinyMCE adds CDATA automatically, so solving this might be important for some projects.