Fix sanitize-config.py to properly remove only admin and dev sections

.github/scripts/sanitize-config.py

@@ -7,11 +7,34 @@
 """
 
 import sys
-import re
+
+
+def find_section_end(lines, start_idx, base_indent):
+    """Find the end of a YAML section based on indentation."""
+    idx = start_idx + 1
+    while idx < len(lines):
+        line = lines[idx]
+        # Skip empty lines and comments
+        if line.strip() == '' or line.strip().startswith('#'):
+            idx += 1
+            continue
+
+        # Get the indentation of the current line
+        current_indent = len(line) - len(line.lstrip())
+
+        # If we find a line with same or less indentation than the section header,
+        # the section has ended
+        if current_indent <= base_indent:
+            return idx
+
+        idx += 1
+
+    # If we reach the end of file, return the length
+    return len(lines)
 
 
 def sanitize_config(config_file_path):
-    """Remove node.dev and node.admin sections from config file using regex."""
+    """Remove node.dev and node.admin sections from config file."""
     print(f"Sanitizing config file: {config_file_path}")
 
     # Read the YAML file
@@ -20,25 +43,61 @@ def sanitize_config(config_file_path):
 
     print(f"Original file size: {len(content)} bytes")
 
-    # Remove node.admin section
-    # This regex matches the admin: line and all indented content that follows
-    admin_pattern = r'(\n\s+admin:\s*\n(?:\s+(?:port:\s*\d+|\S.*)\s*\n)*)'
-    if re.search(admin_pattern, content):
-        content = re.sub(admin_pattern, '\n', content)
-        print("Removed node.admin section")
+    # Split into lines for easier processing
+    lines = content.splitlines(keepends=True)
+
+    # Track lines to remove
+    lines_to_remove = set()
+
+    # Find and mark node.admin and node.dev sections
+    i = 0
+    while i < len(lines):
+        line = lines[i]
+        stripped = line.strip()
+
+        # Check if this is an admin: or dev: line under node:
+        if stripped in ['admin:', 'dev:']:
+            # Get the indentation of this line
+            indent = len(line) - len(line.lstrip())
+
+            # Make sure this is under a node: section by checking previous lines
+            # Look for 'node:' at a lower indentation level
+            is_under_node = False
+            for j in range(i-1, -1, -1):
+                prev_line = lines[j].strip()
+                if prev_line == '':
+                    continue
+                prev_indent = len(lines[j]) - len(lines[j].lstrip())
+                if prev_indent < indent and prev_line.startswith('node:'):
+                    is_under_node = True
+                    break
+                elif prev_indent < indent:
+                    # Found a different section at lower indent
+                    break
+
+            if is_under_node:
+                # Find the end of this section
+                section_end = find_section_end(lines, i, indent)
+
+                # Mark all lines in this section for removal
+                for idx in range(i, section_end):
+                    lines_to_remove.add(idx)
+
+                print(f"Removed node.{stripped[:-1]} section (lines {i+1}-{section_end})")
+
+                # Skip to the end of this section
+                i = section_end - 1
+
+        i += 1
 
-    # Remove node.dev section
-    # This regex matches the dev: line and all indented content that follows
-    dev_pattern = r'(\n\s+dev:\s*\n(?:\s+\S.*\s*\n)*)'
-    if re.search(dev_pattern, content):
-        content = re.sub(dev_pattern, '\n', content)
-        print("Removed node.dev section")
+    # Remove marked lines
+    new_lines = [line for idx, line in enumerate(lines) if idx not in lines_to_remove]
 
     # Write back to file
     with open(config_file_path, 'w') as f:
-        f.write(content)
+        f.writelines(new_lines)
 
-    print(f"Sanitized file size: {len(content)} bytes")
+    print(f"Sanitized file size: {sum(len(line) for line in new_lines)} bytes")
     print("Config sanitization completed")