import os
import glob
import re

def check_links():
    docs_dir = "fluig_rag_docs"
    files = glob.glob(os.path.join(docs_dir, "**", "*.md"), recursive=True)
    broken_links = []
    
    # Regex para encontrar links markdown: [texto](link.md)
    # Suporta parênteses aninhados (até 1 nível)
    link_pattern = re.compile(r"\[.*?\]\(((?:[^()]+|\([^()]*\))*)\)")
    
    print(f"Validando links em {len(files)} arquivos...")
    
    for file_path in files:
        with open(file_path, "r", encoding="utf-8") as f:
            content = f.read()
            
        current_dir = os.path.dirname(file_path)
        links = link_pattern.findall(content)
        
        for link in links:
            # Ignorar links externos e âncoras puras
            if link.startswith("http") or link.startswith("#") or ":" in link:
                continue
            
            # Limpar âncoras do link local
            clean_link = link.split("#")[0]
            if not clean_link: continue
            
            target_path = os.path.abspath(os.path.join(current_dir, clean_link))
            
            if not os.path.exists(target_path):
                broken_links.append({
                    "file": file_path,
                    "link": link,
                    "target": target_path
                })

    if broken_links:
        print(f"\nEncontrados {len(broken_links)} links quebrados:")
        for bl in broken_links:
            print(f"Arquivo: {bl['file']} -> Link: {bl['link']}")
    else:
        print("\nNenhum link quebrado encontrado! 🎉")

if __name__ == "__main__":
    check_links()