import shlex


def tokenize(text) -> list:
    lexer = shlex.shlex(text, posix=True)
    lexer.whitespace_split = True
    lexer.commenters = "#"
    lexer.wordchars += ".:/-@(),"
    return list(lexer)


def parse_sites_imports(text):
    tokens = tokenize(text)

    result = {}
    current_site = None
    depth = 0
    i = 0

    while i < len(tokens):
        tok = tokens[i]

        # detect site start
        if i + 1 < len(tokens) and tokens[i + 1] == "{":
            if depth == 0:
                sites = [s.strip() for s in tok.split(",")]
                for s in sites:
                    result.setdefault(s, [])
                current_site = sites
            depth += 1
            i += 2
            continue

        if tok == "{":
            depth += 1

        elif tok == "}":
            depth -= 1
            if depth == 0:
                current_site = None

        elif tok == "import" and i + 1 < len(tokens) and current_site:
            name = tokens[i + 1]
            for site in current_site:
                result[site].append(name)
            i += 2
            continue
        i += 1
    return result


def filter_globals_out(data) -> dict:
    import re

    ret = dict()
    for key, values in data.items():
        if re.match(r"\(\w+\)", key):  # exclude import definitions e.g (auth)
            continue
        ret[key] = values
    return ret


def parse(caddy_string) -> dict[str, list[str]]:
    return filter_globals_out(parse_sites_imports(caddy_string))

if __name__ == "__main__":
    with open("Caddyfile") as f:
        data = parse(f.read())
    print(data)