processed_files.txt

c2dc67f3bfa73d31817d7b30d37d5d5535038230
b33e5713ae0bfe6b3893dad3cfbb7ab0355c5f5d
2563bef0a467e0b6c9abefac1adcdc96af523005
e33cf5eddce94aa9ab81e0fdd78c9c9b4ddef17b
f1922ef8e28d064415d52f1c1a2cdde69e00ece3
b657b93f2006b6c9833e42fdaab0446f4de830fe
b657b93f2006b6c9833e42fdaab0446f4de830fe
8daee5ed6b3d55195a18ac918ca8d794972c7022
2a0f08a595b51cbb447c59a331fc0d9f3efb88c9
85c64476c3d25a37652bcd49ebf81251ceff1a2e
b6d1d6e58c6c7b471816dfdc2c2841e838ea1841
5151768eb7d99e1de53b8ab24d872a305e66d016
90a14fb353d17b0492fde037700a2e5f4c1d8810
69d4c209ca90b7e5d02ffe797093ddc0bc19d30b
3a5d508188c45e6ada22c4622a5cd798ebc30cb1
97d676617b489b56ed69ebf49da8bd90a11544d1
cdc8af65ae6ebc98ff6b0d79874da9e5a6e229ed
6b18e7862e1d6c778ecb459005b26da555278df5
597f04eddd8520ca1278d9741bed12b7c8fe9264
0b466dacc6a090ed252805a81ce3da896ffc2c00
35f8bb75855efeb215451cd3faf3c312897a4819
319afb98536ee0e336e6a122aa48324bda516cda
066f26577dc2d7642f8f4600d42311665f0ae32c
c5e711a76bf345c3ed98a9a847399601c071aeab
3a5d508188c45e6ada22c4622a5cd798ebc30cb1
da9af35e64a9ac05e5956d4b0646cdfd03074d11
97d676617b489b56ed69ebf49da8bd90a11544d1
d1f9a0477fef164bde91f80cc5a3eea1b472f96b
e2dd433d9e959b2bea90e7341c3b8db32ffef860
92f83e6b481fec1a72bdab2a06266d6a8f421355
218bec77d9254f7a93f3a2bbe01c995fb0eb2428
3a5d508188c45e6ada22c4622a5cd798ebc30cb1
827406678a8c289ba9a4de1470ccd18a36eba460
e33cf5eddce94aa9ab81e0fdd78c9c9b4ddef17b
6633b122a2d3fea5448bb199fbd9cf9dfeb4101a
b6d1d6e58c6c7b471816dfdc2c2841e838ea1841
3a5d508188c45e6ada22c4622a5cd798ebc30cb1
b9bf879d7e1118088197e89c8270d7f875b08511
5af1e7e3aef2704645dbb96d3820c1e3a8d0a353

script.py

import requests
import base64
import re
from datetime import datetime
import os

GITHUB_TOKEN = "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"

def get_processed_files():
    if os.path.exists('processed_files.txt'):
        with open('processed_files.txt', 'r') as f:
            return set(f.read().splitlines())
    return set()

def update_processed_files(commit_hash):
    with open('processed_files.txt', 'a') as f:
        f.write(f"{commit_hash}\n")

def get_file_content(file_url):
    headers = {'Authorization': f'token {GITHUB_TOKEN}'}
    response = requests.get(file_url, headers=headers)
    if response.status_code == 200:
        try:
            return base64.b64decode(response.json()['content']).decode('utf-8')
        except requests.exceptions.JSONDecodeError:
            return response.text
    else:
        print(f"Failed to fetch file: {file_url}")
        return None

def format_file(content):
    parts = content.split("# Findings", 1)
    if len(parts) > 1:
        content = parts[1]
    else:
        content = parts[0]
    
    formatted = "**Auditor**\n\n[Shieldify Security](https://x.com/ShieldifySec)\n\n# Findings\n\n"
    
    lines = content.split('\n')
    current_risk = None
    
    for i, line in enumerate(lines):
        if re.match(r'^\s*#\s*\[(C|H|M|L)-\d+\]', line):
            finding_type = re.search(r'\[(C|H|M|L)-', line).group(1)
            if finding_type in ['C', 'H']:
                current_risk = "High Risk"
            elif finding_type == 'M':
                current_risk = "Medium Risk"
            elif finding_type == 'L':
                current_risk = "Low Risk"
            
            formatted += f"## {current_risk}\n"
            formatted += "### " + re.sub(r'^#+\s*', '', line.strip()) + "\n"

        
        elif line.strip().startswith(('#', '##', '###', '####')):
            formatted += f"**{line.lstrip('#').strip()}**\n"
        
        else:
            formatted += line + '\n'
    
    formatted = re.sub(r'\n{3,}', '\n\n', formatted)
    
    return formatted.strip() + '\n'

def get_file_info(repo_owner, repo_name, file_path):
    api_url = f"https://api.github.com/repos/{repo_owner}/{repo_name}/commits"
    params = {
        'path': file_path,
        'per_page': 1
    }
    headers = {'Authorization': f'token {GITHUB_TOKEN}'}
    response = requests.get(api_url, params=params, headers=headers)
    if response.status_code == 200:
        commits = response.json()
        if commits:
            creation_date = commits[0]['commit']['author']['date']
            commit_hash = commits[0]['sha']
            return datetime.strptime(creation_date, "%Y-%m-%dT%H:%M:%SZ").strftime('%Y-%m-%d'), commit_hash
    return None, None

def get_protocol_name(filename):
    return filename.replace('-Security-Review', '').replace('.md', '')

api_url = "https://api.github.com/repos/shieldify-security/audits-portfolio-md/contents"
repo_owner = "shieldify-security"
repo_name = "audits-portfolio-md"

processed_files = get_processed_files()

headers = {'Authorization': f'token {GITHUB_TOKEN}'}
response = requests.get(api_url, headers=headers)
if response.status_code == 200:
    files = response.json()
    for file in files:
        if file['name'].endswith('.md'):
            print(f"Processing {file['name']}")
            
            try:
                file_path = f"/{file['name']}"
                date, commit_hash = get_file_info(repo_owner, repo_name, file_path)
                
                if not date or not commit_hash:
                    print(f"Failed to get file info for {file['name']}, skipping...")
                    continue
                
                if commit_hash in processed_files:
                    print(f"File {file['name']} already processed, skipping...")
                    continue
                
                file_content = get_file_content(file['download_url'])
                
                if file_content:
                    formatted_content = format_file(file_content)
                    
                    protocol_name = get_protocol_name(file['name'])
                    
                    new_filename = f"{date}-{protocol_name}.md"
                    
                    with open(new_filename, 'w', encoding='utf-8') as f:
                        f.write(formatted_content)
                    
                    update_processed_files(commit_hash)
            except Exception as e:
                print(f"Error processing {file['name']}: {str(e)}")

    print("Formatting complete!")
else:
    print(f"Failed to fetch directory contents. Status code: {response.status_code}")