camoufox/pythonlib/camoufox/async_api.py

import asyncio
import json as _json
import urllib.request
from functools import partial
from typing import Any, Dict, List, Optional, Union, overload
from urllib.parse import urlparse

from playwright.async_api import (
    Browser,
    BrowserContext,
    Playwright,
    PlaywrightContextManager,
)
from typing_extensions import Literal

from camoufox.virtdisplay import VirtualDisplay

from .fingerprints import generate_context_fingerprint
from .utils import (
    async_attach_vd,
    attach_no_viewport_default,
    launch_options,
    spoofs_window_dimensions,
)


class AsyncCamoufox(PlaywrightContextManager):
    """
    Wrapper around playwright.async_api.PlaywrightContextManager that automatically
    launches a browser and closes it when the context manager is exited.
    """

    def __init__(self, **launch_options):
        super().__init__()
        self.launch_options = launch_options
        self.browser: Optional[Union[Browser, BrowserContext]] = None

    async def __aenter__(self) -> Union[Browser, BrowserContext]:
        _playwright = await super().__aenter__()
        try:
            self.browser = await AsyncNewBrowser(_playwright, **self.launch_options)
        except BaseException as e:
            # Any launch failure (InvalidProxy, missing browser, bad options, ...)
            # must tear down the playwright session started above so the driver
            # process/connection is not leaked.
            await super().__aexit__(type(e), e, e.__traceback__)
            raise
        return self.browser

    async def __aexit__(self, *args: Any):
        # Run the base teardown even if browser.close() raises (e.g. the browser
        # process already crashed), so the playwright driver/connection is not leaked.
        try:
            if self.browser:
                await self.browser.close()
        finally:
            await super().__aexit__(*args)


@overload
async def AsyncNewBrowser(
    playwright: Playwright,
    *,
    from_options: Optional[Dict[str, Any]] = None,
    persistent_context: Literal[False] = False,
    **kwargs,
) -> Browser: ...


@overload
async def AsyncNewBrowser(
    playwright: Playwright,
    *,
    from_options: Optional[Dict[str, Any]] = None,
    persistent_context: Literal[True],
    **kwargs,
) -> BrowserContext: ...


async def AsyncNewBrowser(
    playwright: Playwright,
    *,
    headless: Optional[Union[bool, Literal['virtual']]] = None,
    from_options: Optional[Dict[str, Any]] = None,
    persistent_context: bool = False,
    debug: Optional[bool] = None,
    **kwargs,
) -> Union[Browser, BrowserContext]:
    """
    Launches a new browser instance for Camoufox given a set of launch options.

    Parameters:
        from_options (Dict[str, Any]):
            A set of launch options generated by `launch_options()` to use
        persistent_context (bool):
            Whether to use a persistent context.
        **kwargs:
            All other keyword arugments passed to `launch_options()`.
    """
    if headless == 'virtual':
        virtual_display = VirtualDisplay(debug=debug)
        kwargs['virtual_display'] = virtual_display.get()
        headless = False
    else:
        virtual_display = None

    if not from_options:
        from_options = await asyncio.get_event_loop().run_in_executor(
            None,
            partial(launch_options, headless=headless, debug=debug, **kwargs),
        )

    # Playwright's default viewport deadlocks Juggler when the window is spoofed
    # to a different size (daijro/camoufox#666), so default to no_viewport.
    no_viewport_default = spoofs_window_dimensions(from_options)

    # Persistent context
    if persistent_context:
        if no_viewport_default and not ('viewport' in from_options or 'no_viewport' in from_options):
            from_options = {**from_options, 'no_viewport': True}
        context = await playwright.firefox.launch_persistent_context(**from_options)
        return await async_attach_vd(context, virtual_display)

    # Browser
    browser = await playwright.firefox.launch(**from_options)
    if no_viewport_default:
        attach_no_viewport_default(browser)
    return await async_attach_vd(browser, virtual_display)


def _proxy_url_with_creds(proxy: Dict[str, str]) -> str:
    """Builds a proxy URL string with embedded credentials."""
    parsed = urlparse(proxy.get("server", ""))
    user = proxy.get("username", "")
    pwd = proxy.get("password", "")
    if user and pwd:
        return f"{parsed.scheme}://{user}:{pwd}@{parsed.netloc}"
    return proxy.get("server", "")


async def _resolve_proxy_geo(proxy: Dict[str, str]) -> Dict[str, Optional[str]]:
    """Queries ip-api.com through the proxy for the exit IP and timezone."""
    proxy_url = _proxy_url_with_creds(proxy)

    def _fetch() -> Dict[str, Optional[str]]:
        handler = urllib.request.ProxyHandler({"http": proxy_url, "https": proxy_url})
        opener = urllib.request.build_opener(handler)
        try:
            with opener.open("http://ip-api.com/json?fields=query,timezone", timeout=10) as resp:
                data = _json.loads(resp.read())
                return {"ip": data.get("query") or None, "timezone": data.get("timezone") or None}
        except Exception:
            return {"ip": None, "timezone": None}

    return await asyncio.get_event_loop().run_in_executor(None, _fetch)


async def AsyncNewContext(
    browser: Browser,
    *,
    preset: Optional[Dict[str, Any]] = None,
    os: Optional[str] = None,
    ff_version: Optional[str] = None,
    webrtc_ip: Optional[str] = None,
    proxy: Optional[Dict[str, str]] = None,
    geolocation: Optional[Dict[str, float]] = None,
    **context_kwargs: Any,
) -> BrowserContext:
    """
    Creates a new browser context with a unique fingerprint identity.

    Each context gets its own real fingerprint preset (navigator, screen, WebGL, fonts, etc.)
    with unique seeds for audio, canvas, and font spacing noise. All values are applied
    via addInitScript so they self-destruct before page scripts can detect them.

    Parameters:
        browser: A Browser instance from AsyncNewBrowser or AsyncCamoufox.
        preset: A specific fingerprint preset dict to use. If None, picks randomly.
        os: Target OS for preset selection ("windows", "macos", "linux").
        ff_version: Firefox version string for UA patching.
        webrtc_ip: IPv4 address to spoof for WebRTC ICE candidates.
        proxy: Per-context proxy (Playwright format: {"server": "...", "username": "...", "password": "..."}).
        geolocation: Per-context geolocation ({"latitude": float, "longitude": float}).
        **context_kwargs: Additional Playwright new_context() options.
    """
    # Auto-derive WebRTC IP and timezone from proxy's exit IP when not explicitly provided
    if proxy and (not webrtc_ip or "timezone_id" not in context_kwargs):
        geo = await _resolve_proxy_geo(proxy)
        if not webrtc_ip:
            webrtc_ip = geo["ip"]
        if "timezone_id" not in context_kwargs and geo["timezone"]:
            context_kwargs["timezone_id"] = geo["timezone"]

    fp = await asyncio.get_event_loop().run_in_executor(
        None,
        lambda: generate_context_fingerprint(preset=preset, os=os, ff_version=ff_version, webrtc_ip=webrtc_ip),
    )

    # Merge generated context options with user overrides (user wins)
    opts: Dict[str, Any] = {**fp['context_options'], **context_kwargs}
    if proxy:
        opts['proxy'] = proxy
    if geolocation:
        opts['geolocation'] = geolocation
        opts.setdefault('permissions', ['geolocation'])

    context = await browser.new_context(**opts)
    await context.add_init_script(fp['init_script'])
    return context