"""Scrape AWS Security Hub FSBP controls index page into structured rows.

The index page (https://docs.aws.amazon.com/securityhub/latest/userguide/fsbp-standard.html)
lists controls as <p><a href="./<svc>-controls.html#<id>">[<Service>.<Number>] <Title></a></p>.
This module only parses the index — severity defaults to "medium" because the
index doesn't surface severity. Detail-page enrichment is a future task.
"""
from __future__ import annotations

import re
from dataclasses import dataclass, field
from urllib.parse import urljoin

from bs4 import BeautifulSoup

from scripts.control_resource_map import control_id_to_types


FSBP_INDEX_URL = (
    "https://docs.aws.amazon.com/securityhub/latest/userguide/"
    "fsbp-standard.html"
)


@dataclass
class FSBPRow:
    control_id: str
    title: str
    severity: str
    detail_url: str
    resource_types: list[str] = field(default_factory=list)
    requirement: str = ""


_ANCHOR_RE = re.compile(r"^\s*\[(?P<id>[A-Za-z][A-Za-z0-9]*\.\d+)\]\s*(?P<title>.+)$")


def parse_fsbp_index(html: str, base_url: str) -> list[FSBPRow]:
    soup = BeautifulSoup(html, "html.parser")
    rows: list[FSBPRow] = []
    seen: set[str] = set()
    for a in soup.find_all("a"):
        text = a.get_text(strip=True)
        m = _ANCHOR_RE.match(text)
        if not m:
            continue
        control_id = m.group("id")
        if control_id in seen:
            continue
        seen.add(control_id)
        title = m.group("title").strip()
        href = a.get("href", "")
        detail_url = urljoin(base_url, href) if href else ""
        rows.append(FSBPRow(
            control_id=control_id,
            title=title,
            severity="medium",
            detail_url=detail_url,
            resource_types=control_id_to_types(control_id),
        ))
    return rows