Fix duplicated component identifier for names with non-ascii chars

Editor assessment

The change itself is well described — duplicated component identifiers for non-ASCII names is a precise, searchable symptom — but the header is roughly 95 characters long, well past the 72-character comfort zone, so it will be truncated on GitHub and in many changelog renderers. The leading 'fix' duplicates the type and could be dropped to reclaim characters. Structure is otherwise correct: compiler-core is the right scope, the PR reference sits in the header parentheses, and the 'fix #4422' footer uses valid token syntax after a blank line. There is no explanatory body, so release-note readers must click through for the underlying cause. As a fix, it maps to a patch-level version increment.