Add a script to lift CERT risk assessment tags from help files

lcartey · lcartey · commit a1f45ff00ee8 · 2025-05-01T23:39:39.000+01:00
diff --git a/scripts/add_risk_assessment_tags.py b/scripts/add_risk_assessment_tags.py
@@ -0,0 +1,160 @@
+#!/usr/bin/env python3
+"""
+Add risk assessment tags to rule package JSON files.
+
+This script:
+1. Iterates through each JSON file in rule_packages directory
+2. Looks for CERT-C or CERT-CPP sections
+3. For each rule, finds the corresponding markdown file
+4. Extracts risk assessment data from the markdown file
+5. Adds risk assessment data as tags to each query in the JSON file
+"""
+
+import os
+import json
+import re
+import glob
+from bs4 import BeautifulSoup
+import logging
+
+logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s')
+logger = logging.getLogger(__name__)
+
+def find_rule_packages():
+    """Find all JSON rule package files in the rule_packages directory."""
+    repo_root = os.path.dirname(os.path.dirname(os.path.abspath(__file__)))
+    rule_packages_dir = os.path.join(repo_root, "rule_packages")
+    return glob.glob(os.path.join(rule_packages_dir, "**", "*.json"), recursive=True)
+
+def extract_risk_assessment_from_md(md_file_path):
+    """Extract risk assessment data from the markdown file."""
+    risk_data = {}
+    
+    try:
+        with open(md_file_path, 'r', encoding='utf-8') as f:
+            content = f.read()
+        
+        # Find the Risk Assessment section
+        risk_section_match = re.search(r'## Risk Assessment(.*?)##', content, re.DOTALL)
+        if not risk_section_match:
+            # Try to find it as the last section
+            risk_section_match = re.search(r'## Risk Assessment(.*?)$', content, re.DOTALL)
+            if not risk_section_match:
+                logger.warning(f"No Risk Assessment section found in {md_file_path}")
+                return risk_data
+        
+        risk_section = risk_section_match.group(1)
+        
+        # Look for the table with risk assessment data
+        table_match = re.search(r'<table>(.*?)</table>', risk_section, re.DOTALL)
+        if not table_match:
+            logger.warning(f"No risk assessment table found in {md_file_path}")
+            return risk_data
+        
+        table_html = table_match.group(0)
+        soup = BeautifulSoup(table_html, 'html.parser')
+        
+        # Find all rows in the table
+        rows = soup.find_all('tr')
+        if len(rows) < 2:  # Need at least header and data row
+            logger.warning(f"Incomplete risk assessment table in {md_file_path}")
+            return risk_data
+        
+        # Extract headers and values
+        headers = [th.get_text().strip() for th in rows[0].find_all('th')]
+        values = [td.get_text().strip() for td in rows[1].find_all('td')]
+        
+        # Create a dictionary of headers and values
+        if len(headers) == len(values):
+            for i, header in enumerate(headers):
+                risk_data[header] = values[i]
+        else:
+            logger.warning(f"Header and value count mismatch in {md_file_path}")
+        
+    except Exception as e:
+        logger.error(f"Error extracting risk assessment from {md_file_path}: {e}")
+    
+    return risk_data
+
+def find_md_file(rule_id, short_name, language):
+    """Find the markdown file for the given rule ID and short name."""
+    repo_root = os.path.dirname(os.path.dirname(os.path.abspath(__file__)))
+    md_path = os.path.join(repo_root, language, "cert", "src", "rules", rule_id, f"{short_name}.md")
+    
+    if os.path.exists(md_path):
+        return md_path
+    else:
+        # Try without short name (sometimes the file is named after the rule ID)
+        md_path = os.path.join(repo_root, language, "cert", "src", "rules", rule_id, f"{rule_id}.md")
+        if os.path.exists(md_path):
+            return md_path
+        else:
+            logger.warning(f"Could not find markdown file for {language} rule {rule_id} ({short_name})")
+            return None
+
+def process_rule_package(rule_package_file):
+    """Process a single rule package JSON file."""
+    try:
+        with open(rule_package_file, 'r', encoding='utf-8') as f:
+            data = json.load(f)
+        
+        modified = False
+        
+        # Look for CERT-C and CERT-CPP sections
+        for cert_key in ["CERT-C", "CERT-C++"]:
+            if cert_key in data:
+                language = "c" if cert_key == "CERT-C" else "cpp"
+                
+                # Process each rule in the CERT section
+                for rule_id, rule_data in data[cert_key].items():
+                    if "queries" in rule_data:
+                        for query in rule_data["queries"]:
+                            if "short_name" in query:
+                                md_file = find_md_file(rule_id, query["short_name"], language)
+                                
+                                if md_file:
+                                    risk_data = extract_risk_assessment_from_md(md_file)
+                                    
+                                    if risk_data:
+                                        # Add risk assessment data as tags
+                                        if "tags" not in query:
+                                            query["tags"] = []
+                                        
+                                        # Add each risk assessment property as a tag
+                                        for key, value in risk_data.items():
+                                            key_sanitized = key.lower().replace(" ", "-")
+                                            if key_sanitized == "rule":
+                                                # skip rule, as that is already in the rule ID
+                                                continue
+                                            tag = f"external/cert/{key_sanitized}/{value.lower()}"
+                                            if tag not in query["tags"]:
+                                                query["tags"].append(tag)
+                                                modified = True
+                                                logger.info(f"Added tag {tag} to {rule_id} ({query['short_name']})")
+        
+        # Save the modified data back to the file if any changes were made
+        if modified:
+            with open(rule_package_file, 'w', encoding='utf-8') as f:
+                json.dump(data, f, indent=2)
+            logger.info(f"Updated {rule_package_file}")
+        else:
+            logger.info(f"No changes made to {rule_package_file}")
+            
+    except Exception as e:
+        logger.error(f"Error processing {rule_package_file}: {e}")
+
+def main():
+    """Main function to process all rule packages."""
+    logger.info("Starting risk assessment tag addition process")
+    
+    rule_packages = find_rule_packages()
+    logger.info(f"Found {len(rule_packages)} rule package files")
+    
+    for rule_package in rule_packages:
+        logger.info(f"Processing {rule_package}")
+        process_rule_package(rule_package)
+    
+    logger.info("Completed risk assessment tag addition process")
+
+if __name__ == "__main__":
+    main()
