import os
import nbformat as nbf
import re
import requests
# Code to populate file_list with relevant .ipynb files
file_list = []

for root, dirs, files in os.walk("qiskit-tutorials/tutorials/"):
    for file in files:
        if file.endswith(".ipynb"):
            file_list.append(os.path.join(root, file))  # Including the relative path of the file

INLINE_LINK_RE = re.compile(r'\[([^\]]+)\]\(([^)]+)\)')
for file_path in file_list:
    try:
        with open(file_path) as file:
            nb = nbf.read(file, 4)
    except OSError as e:
        print(f"Error occurred while reading file: {file_path}")
        print("Error message:", str(e))
        print()
        continue

    for cell in nb["cells"]:
        if cell["cell_type"] != "markdown":
            continue
        links = list(INLINE_LINK_RE.findall(cell["source"]))
        for link in links:
            if not link:
                continue
            link_text, link_url = link
            if link_url.lower().endswith(('.png', '.jpg', '.jpeg', '.gif')):
                continue  # Skip image links
            try:
                response = requests.get(link_url)
                if response.status_code != 200:
                    print(f"Broken link found in file: {file_path}")
                    print(f"Link: {link_text}")
                    print(f"URL: {link_url}")
                    print("HTTP status code:", response.status_code)
                    print()
            except requests.RequestException as e:
                print(f"Error occurred while checking link in file: {file_path}")
                print(f"Link: {link_text}")
                print(f"URL: {link_url}")
                print("Error message:", str(e))
                print()