Forum Stats

  • 3,853,749 Users
  • 2,264,264 Discussions


Export of app with embedded unicode characters

Scott Wesley
Scott Wesley Member Posts: 6,260 Gold Crown

We have an old application that embedded some HTML as part of an image gallery lightbox implementation.

This HTML included some 'higher-plane' unicode characters.

    <a class="prev">‹</a>    <a class="next">›</a>    <a class="close">×</a>

during regression testing I noticed the following for the "previous" button


instead of


I found that when I exported from 5.1, these values encoded to

'    <a class="prev">‹</a>','    <a class="next">›</a>','    <a class="close">×</a>',

When exporting from 18.2, they looked like

unistr('    <a class="prev">\2039</a>'),unistr('    <a class="next">\203A</a>'),unistr('    <a class="close">\00D7</a>'),

Exporting/Importing from the same APEX version is fine, but when importing a 5.1 export into 18.2, they don't transform as expected.

It looks like there has been some intended improvement to the export, I was just curious on details.

I'm working around the problem by fixing the page in the new environment, exporting that page, and including it in the 18.2 rollout.

Alternatively, I could transfer that code into a PL/SQL Dynamic Content region.