HTML Import: Pages render blank when source contains non-standard HTML elements (e.g. <tabset>)

XMLWordPrintable

    • 1
    • Severity 3 - Minor

      Issue Summary

      When performing an "Import from other tools > HTML import" (Beta), the presence of non-standard HTML tags such as <tabset> (commonly used by AngularJS, Angular Material, or similar JS frameworks) causes the HTML-to-ADF conversion to fail silently, resulting in an empty page once imported to Confluence.

      The blank page's Version History shows "This comparison involves a version from the legacy editor" on v1 and the page header displays "Cloud editor," suggesting the page was created with an inconsistent editor state.

      Steps to Reproduce

      1. Create or open an HTML file containing non-standard elements such as <tabset> (example below):  
        • <!DOCTYPE html><html><head><title>Testing HTML Import</title></head><body><p>Hello World!</p><tabset> <template v-slot:tabs=""><li>Schema</li><li>Example</li></template> <template v-slot:content=""> <div class="tabset-panel"><pre v-pre="true" class="prismjs line-numbers"><code class="language-json">{}</code></pre></div> <div class="tabset-panel"><p>some text within the TABSET</p><pre v-pre="true" class="prismjs line-numbers"><code class="language-json">{}</code></pre> </div> </template></tabset><p>End of page</p></body></html>
      1. Place the HTML file inside a folder and compress it into a .zip file (standard HTML import structure).
      2. Navigate to Spaces → Import work from other tools → HTML import.
      3. Upload the .zip and complete the import.
      4. Once the import finishes, check the resulting page.

      Expected Results

      The HTML file should be imported with its full content, without any loss of data.
      Non-standard elements that cannot be converted should either be stripped gracefully (preserving inner text content) or the importer should surface a clear error identifying which elements caused the failure.

      Actual Results

      The HTML file is imported empty, with no content visible. If it had any attachments, those will be linked accordingly, but no visible content appears on the page itself.

      The importer email notification reports "X pages couldn't be imported, or were imported incorrectly" but does not identify which pages failed or why. The issue likely affects any non-standard/custom HTML element, not just <tabset>.

      Workaround

      Once any blank pages are identified after a test import, perform a manual cleanup on all affected pages' source files (for example, replacing <tabset> blocks with standard HTML such as <div> containers or <table> elements), and then perform a new HTML import with the corrected files.

              Assignee:
              Unassigned
              Reporter:
              Rodrigo Bozza (Atlassian)
              Votes:
              0 Vote for this issue
              Watchers:
              2 Start watching this issue

                Created:
                Updated: