import base64
import json
import re

import requests
from bs4 import BeautifulSoup
from fake_useragent import UserAgent
from requests.adapters import HTTPAdapter
from urllib3.util.retry import Retry

from .Provider import Provider

# Precompile the different pattern matchers used during extraction:
# - REDIRECT_PATTERN pulls the intermediate redirect URL from the bootstrap
#   script so we can follow the provider's hand-off.
# - B64_PATTERN isolates the base64 encoded payload containing the ``source``
#   field once decoded.
# - HLS_PATTERN captures the base64 encoded HLS manifest for fallback when
#   no direct MP4 link is present.
REDIRECT_PATTERN = re.compile(r"https?://[^'\"<>]+")
B64_PATTERN = re.compile(r"var a168c='([^']+)'")
HLS_PATTERN = re.compile(r"'hls': '(?P<hls>[^']+)'")


class VOE(Provider):
    """VOE video provider implementation."""

    def __init__(self):
        self.RANDOM_USER_AGENT = UserAgent().random
        self.Header = {"User-Agent": self.RANDOM_USER_AGENT}

    def get_link(
        self, embedded_link: str, timeout: int
    ) -> tuple[str, dict]:
        """
        Extract direct download link from VOE embedded player.

        Args:
            embedded_link: URL of the embedded VOE player
            timeout: Request timeout in seconds

        Returns:
            Tuple of (direct_link, headers)
        """
        self.session = requests.Session()

        # Configure retries with backoff
        retries = Retry(
            total=5,  # Number of retries
            backoff_factor=1,  # Delay multiplier (1s, 2s, 4s, ...)
            status_forcelist=[500, 502, 503, 504],
            allowed_methods=["GET"],
        )

        adapter = HTTPAdapter(max_retries=retries)
        self.session.mount("https://", adapter)
        timeout = 30

        response = self.session.get(
            embedded_link,
            headers={"User-Agent": self.RANDOM_USER_AGENT},
            timeout=timeout,
        )

        redirect = re.search(r"https?://[^'\"<>]+", response.text)
        if not redirect:
            raise ValueError("No redirect found.")

        redirect_url = redirect.group(0)
        parts = redirect_url.strip().split("/")
        self.Header["Referer"] = f"{parts[0]}//{parts[2]}/"

        response = self.session.get(
            redirect_url, headers={"User-Agent": self.RANDOM_USER_AGENT}
        )
        html = response.content

        # Method 1: Extract from script tag
        extracted = self.extract_voe_from_script(html)
        if extracted:
            return extracted, self.Header

        # Method 2: Extract from base64 encoded variable
        htmlText = html.decode("utf-8")
        b64_match = B64_PATTERN.search(htmlText)
        if b64_match:
            decoded = base64.b64decode(b64_match.group(1)).decode()[::-1]
            source = json.loads(decoded).get("source")
            if source:
                return source, self.Header

        # Method 3: Extract HLS source
        hls_match = HLS_PATTERN.search(htmlText)
        if hls_match:
            decoded_hls = base64.b64decode(hls_match.group("hls")).decode()
            return decoded_hls, self.Header

        raise ValueError("Could not extract download link from VOE")

    def shift_letters(self, input_str: str) -> str:
        """Apply ROT13 shift to letters."""
        result = ""
        for c in input_str:
            code = ord(c)
            if 65 <= code <= 90:
                code = (code - 65 + 13) % 26 + 65
            elif 97 <= code <= 122:
                code = (code - 97 + 13) % 26 + 97
            result += chr(code)
        return result

    def replace_junk(self, input_str: str) -> str:
        """Replace junk character sequences."""
        junk_parts = ["@$", "^^", "~@", "%?", "*~", "!!", "#&"]
        for part in junk_parts:
            input_str = re.sub(re.escape(part), "_", input_str)
        return input_str

    def shift_back(self, s: str, n: int) -> str:
        """Shift characters back by n positions."""
        return "".join(chr(ord(c) - n) for c in s)

    def decode_voe_string(self, encoded: str) -> dict:
        """Decode VOE-encoded string to extract video source."""
        step1 = self.shift_letters(encoded)
        step2 = self.replace_junk(step1).replace("_", "")
        step3 = base64.b64decode(step2).decode()
        step4 = self.shift_back(step3, 3)
        step5 = base64.b64decode(step4[::-1]).decode()
        return json.loads(step5)

    def extract_voe_from_script(self, html: bytes) -> str:
        """Extract download link from VOE script tag."""
        soup = BeautifulSoup(html, "html.parser")
        script = soup.find("script", type="application/json")
        return self.decode_voe_string(script.text[2:-2])["source"]