src/rocky_man/processor/converter.py

@@ -40,11 +40,7 @@ class ManPageConverter:
         try:
             # Run mandoc with no arguments - it will show usage and exit
             # We just want to verify the command exists, not that it succeeds
-            subprocess.run(
-                ['mandoc'],
-                capture_output=True,
-                timeout=5
-            )
+            subprocess.run(["mandoc"], capture_output=True, timeout=5)
             return True
         except FileNotFoundError:
             # mandoc command not found
@@ -76,16 +72,27 @@ class ManPageConverter:
             # Check if mandoc output indicates this is a symlink/redirect
             # Pattern: <div class="manual-text">/usr/share/man/man8/target.8.gz</div>
             # or: <div class="manual-text">See the file /usr/share/man/man8/target.8.</div>
-            symlink_match = re.search(r'<div class="manual-text">(?:See the file )?(/usr/share/man/man\d+[a-z]*/([^/]+)\.(\d+[a-z]*)(?:\.gz)?)\.</div>', html)
+            # or: <div class="manual-text">See the file man1/builtin.1.</div>
+            symlink_match = re.search(
+                r'<div class="manual-text">.*?(?:See the file )?((?:/usr/share/man/)?man\d+[a-z]*/([^/]+)\.(\d+[a-z]*)(?:\.gz)?)\..*?</div>',
+                html,
+                re.DOTALL,
+            )
             if not symlink_match:
                 # Try simpler pattern without "See the file" or period
-                symlink_match = re.search(r'<div class="manual-text">(/usr/share/man/man\d+[a-z]*/([^/<]+)\.(\d+[a-z]*)(?:\.gz)?)</div>', html)
+                symlink_match = re.search(
+                    r'<div class="manual-text">.*?((?:/usr/share/man/)?man\d+[a-z]*/([^/<]+)\.(\d+[a-z]*)(?:\.gz)?).*?</div>',
+                    html,
+                    re.DOTALL,
+                )
 
             if symlink_match:
                 name = symlink_match.group(2)
                 section = symlink_match.group(3)
-                logger.info(f"{man_file.display_name} detected as symlink to {name}({section})")
-                html = self._generate_redirect_html({'name': name, 'section': section})
+                logger.info(
+                    f"{man_file.display_name} detected as symlink to {name}({section})"
+                )
+                html = self._generate_redirect_html({"name": name, "section": section})
 
             # Store in ManFile object
             man_file.html_content = html
@@ -96,7 +103,7 @@ class ManPageConverter:
 
             # Save HTML file
             output_path.parent.mkdir(parents=True, exist_ok=True)
-            with open(output_path, 'w', encoding='utf-8') as f:
+            with open(output_path, "w", encoding="utf-8") as f:
                 f.write(html)
 
             logger.debug(f"Converted {man_file.display_name} -> {output_path}")
@@ -107,9 +114,7 @@ class ManPageConverter:
             return False
 
     def convert_many(
-        self,
-        man_files: List[tuple],
-        max_workers: int = 10
+        self, man_files: List[tuple], max_workers: int = 10
     ) -> List[ManFile]:
         """Convert multiple man pages in parallel.
 
@@ -152,21 +157,21 @@ class ManPageConverter:
         """
         try:
             result = subprocess.run(
-                ['mandoc', '-T', 'html', '-O', 'fragment,toc'],
-                input=content.encode('utf-8'),
+                ["mandoc", "-T", "html", "-O", "fragment,toc"],
+                input=content.encode("utf-8"),
                 capture_output=True,
-                timeout=30
+                timeout=30,
             )
 
             if result.returncode != 0:
-                stderr = result.stderr.decode('utf-8', errors='replace')
+                stderr = result.stderr.decode("utf-8", errors="replace")
                 logger.warning(f"mandoc returned error: {stderr}")
                 # Sometimes mandoc returns non-zero but still produces output
                 if result.stdout:
-                    return result.stdout.decode('utf-8', errors='replace')
+                    return result.stdout.decode("utf-8", errors="replace")
                 return None
 
-            return result.stdout.decode('utf-8', errors='replace')
+            return result.stdout.decode("utf-8", errors="replace")
 
         except subprocess.TimeoutExpired:
             logger.error("mandoc conversion timed out")
@@ -186,14 +191,10 @@ class ManPageConverter:
         """
         # Remove empty parentheses in header cells
         html = re.sub(
-            r'<td class="head-ltitle">\(\)</td>',
-            '<td class="head-ltitle"></td>',
-            html
+            r'<td class="head-ltitle">\(\)</td>', '<td class="head-ltitle"></td>', html
         )
         html = re.sub(
-            r'<td class="head-rtitle">\(\)</td>',
-            '<td class="head-rtitle"></td>',
-            html
+            r'<td class="head-rtitle">\(\)</td>', '<td class="head-rtitle"></td>', html
         )
 
         # Strip leading/trailing whitespace
@@ -210,8 +211,8 @@ class ManPageConverter:
         Returns:
             HTML fragment for redirect page
         """
-        name = target_info['name']
-        section = target_info['section']
+        name = target_info["name"]
+        section = target_info["section"]
 
         # Generate the relative path to the target man page
         # Symlinks are in the same package, just different file names
@@ -229,7 +230,6 @@ class ManPageConverter:
 
         return html
 
-
     def _get_output_path(self, man_file: ManFile) -> Path:
         """Determine output path for HTML file.