Issue Bug Resolved

#111455Import turns rich text page links into a bare page uid

Oliver Bartsch 11/14/26, 17:41 updated 11/17/26, 15:50

Description

Problem

After importing a T3D file, links to pages in rich text fields lose the
@t3://@ scheme, their additional parameters and their fragment:
the href t3://page?uid=42&L=1#top is imported as 57#top, the href
@t3://page?uid=42&L=1@ as 57.

Steps to reproduce

  1. Create a text element with a link to page 42 with additional parameters
    &L=1 and anchor top.
  2. Export the page tree including page 42 with EXT:impexp and import it into
    another tree.
  3. The imported link’s href is the new page uid plus fragment, without
    t3://page?uid= and without &L=1.

Cause

TypolinkTagSoftReferenceParser::parse() replaces the whole href of a page
link by a soft reference token (and a second token for a numeric fragment),
while the token value is only the page uid. the import of EXT:impexp
substitutes the uid for the token, so everything else in the href is gone. The
exclude mode of the import is affected too: it puts back the bare uid.

Suggested fix

Tokenize only the uid inside the href, as the parser already does for record
links: anchor the match on (?:\?|&(?:amp;)?)uid=(\d+) and replace that span
with the token, keeping the rest of the href.

Reviews 3

History 812

  1. Oliver Bartsch commented and changed status Under Review Resolved done 0 % 100 % <span data-processor="relativetime" data-value="2026-11-17T14:15:16+00:00" > 11/17/26, 14:15 </span>

    Applied in changeset commit:a6998a353cc40439e2e349b73843d7b60b265462.

  2. Gerrit Code Review pushed patch set 2 for 14.3 to change 97750 status Resolved Under Review <span data-processor="relativetime" data-value="2026-11-17T15:36:35+00:00" > 11/17/26, 15:36 </span>

  3. Oliver Bartsch commented and changed status Under Review Resolved <span data-processor="relativetime" data-value="2026-11-17T15:50:16+00:00" > 11/17/26, 15:50 </span>

    Applied in changeset commit:d36f1aac76747529aa1374507f5829cd73fec445.