def process_md_file()

in check_all_deadlink.py [0:0]

• 37 lines of code
• 14 McCabe index (conditional complexity)

def process_md_file(file_path):
    link_pattern = re.compile(r"\[.*?\]\((.*?)\)")
    code_block_pattern = re.compile(r"^```.*$")

    with open(file_path, "r", encoding="utf-8") as f:
        content = f.read()

    lines = content.splitlines()
    in_code_block = False

    for line_number, line in enumerate(lines, start=1):
        # Skip codeblocks
        if code_block_pattern.match(line):
            in_code_block = not in_code_block
            continue

        if in_code_block:
            continue

        links = link_pattern.findall(line)

        for link in links:
            # Skip urls
            if (
                not urlparse(link).scheme
                and not os.path.isabs(link)
                and not (link[0] == "#")
            ):
                full_path = os.path.normpath(
                    os.path.join(os.path.dirname(file_path), link)
                )

                # Skip section headers
                if "#" in full_path:
                    full_path = full_path.split("#", 1)[0]

                if not full_path.endswith(".md") and not full_path.endswith(".mdx"):
                    full_path += ".md"
                md_exists = os.path.exists(full_path)
                mdx_exists = (
                    os.path.exists(full_path[:-3] + ".mdx")
                    if full_path.endswith(".md")
                    else False
                )

                if not md_exists and not mdx_exists:
                    print(
                        f"Error: File not found for link '{link}' in file '{file_path}:{line_number}'"
                    )