Fix session variable search discoverability in docs

Enhance session variable descriptions with variable names to improve
search discoverability. Session variables like default_transaction_use_follower_reads
were not findable via search because variable names appeared only in
filtered table cells while descriptions lacked the variable names.

The fix prepends session variable names to their descriptions for
session-variables.html, making them discoverable without creating
additional index records.

Fixes issue where customers couldn't find documented session variables
through main docs search functionality.

Author: ebembi-crdb

src/current/algolia_index_intelligent_bloat_removal.py

@@ -672,6 +672,30 @@ def get_git_last_modified(file_path: pathlib.Path) -> str:
         GIT_DATE_CACHE[cache_key] = date
         return date
 
+def enhance_session_variable_content(content: str, element, context: Dict[str, str]) -> str:
+    """Add session variable name to description records where missing for better discoverability."""
+    # Only for session-variables.html page
+    if 'session-variables.html' not in context.get('url', ''):
+        return content
+        
+    # Check if this is a description cell adjacent to a variable name cell
+    if element.name == 'td':
+        prev_sibling = element.find_previous_sibling('td')
+        if prev_sibling:
+            prev_text = extract_text_with_spaces(prev_sibling).strip()
+            
+            # If previous cell contains a session variable name pattern
+            if (re.match(r'^\w+(_\w+)+$', prev_text) and 
+                '_' in prev_text and 
+                len(prev_text) > 5 and
+                len(prev_text) < 50 and
+                prev_text not in content):
+                
+                # Prepend variable name to description for discoverability
+                return f"{prev_text}: {content}"
+    
+    return content
+
 def extract_records_from_html(html_path: pathlib.Path, versions: Dict[str, str] = None) -> List[Dict[str, Any]]:
     """Proven extraction + intelligent bloat removal."""
     if should_exclude_file(str(html_path), versions):
@@ -738,6 +762,9 @@ def extract_records_from_html(html_path: pathlib.Path, versions: Dict[str, str]
             continue
 
         text = extract_text_with_spaces(element)
+        
+        # Enhance session variable content for better discoverability
+        text = enhance_session_variable_content(text, element, filter_context)
 
         # INTELLIGENT BLOAT REMOVAL - context-aware filtering
         if bloat_filter.is_bloat_content(text, filter_context):