fix linking of identifiers that contain unicode characters (#831)

Author: iFreilicht

CHANGELOG.md

@@ -4,6 +4,8 @@
 
 ## Unreleased: pdoc next
 
+- Fix linking of identifiers that contain unicode characters.
+  ([#831](https://github.com/mitmproxy/pdoc/issues/831), @iFreilicht)
 
 ## 2025-06-04: pdoc 15.0.4
 

pdoc/render_helpers.py

@@ -408,23 +408,23 @@ def linkify_repl(m: re.Match):
             # First part of the identifier (e.g. "foo") - this is optional for relative references.
             (?:
                 \b
-                (?!\d)[a-zA-Z0-9_]+
+                (?!\d)\w+
                 |
                 \.*  # We may also start with multiple dots.
             )
             # Rest of the identifier (e.g. ".bar" or "..bar")
             (?:
                 # A single dot or a dot surrounded with pygments highlighting.
                 (?:\.|</span><span\ class="o">\.</span><span\ class="n">)
-                (?!\d)[a-zA-Z0-9_]+
+                (?!\d)\w+
             )+
             (?:\(\)|\b(?!\(\)))  # we either end on () or on a word boundary.
             (?!</a>)  # not an existing link
             (?![/#])  # heuristic: not part of a URL
 
             | # Part 2: `foo` or `foo()`. `foo.bar` is already covered with part 1.
             (?<=<code>)
-                 (?!\d)[a-zA-Z0-9_]+
+                 (?!\d)\w+
             (?:\(\))?
             (?=</code>(?!</a>))
             """,

test/testdata/demo_long.html

@@ -1697,7 +1697,7 @@ <h6 id="be-careful">Be Careful!</h6>
             <div class="docstring"><p>The following link should be created properly:</p>
 
 <ul>
-<li><code>MötörCrüe.créer_naïveté</code></li>
+<li><code><a href="#MötörCrüe.créer_naïveté">MötörCrüe.créer_naïveté</a></code></li>
 </ul>
 </div>