path: root/tests/integration/harness.py

"""Starts a real opentracker and talks to it over the wire.

Everything here drives the binary rather than the code, so these tests cover
what the unit suite cannot reach: the event loop, the HTTP parser and the
torrent store working together. Run against a sanitizer build and the tracker
reports its own memory errors, which the teardown turns into test failures.
"""

import os
import shutil
import signal
import socket
import struct
import subprocess
import sys
import tempfile
import time
import unittest
import urllib.error
import urllib.parse
import urllib.request

TRACKER = os.environ.get("OPENTRACKER", os.path.join(os.path.dirname(__file__), "..", "..", "opentracker.debug"))


def bdecode(data):
    """Enough bencode for what a tracker answers."""

    def value(i):
        if data[i : i + 1] == b"d":
            i, out = i + 1, {}
            while data[i : i + 1] != b"e":
                i, key = value(i)
                i, out[key] = value(i)
            return i + 1, out
        if data[i : i + 1] == b"l":
            i, out = i + 1, []
            while data[i : i + 1] != b"e":
                i, item = value(i)
                out.append(item)
            return i + 1, out
        if data[i : i + 1] == b"i":
            end = data.index(b"e", i)
            return end + 1, int(data[i + 1 : end])
        colon = data.index(b":", i)
        length = int(data[i:colon])
        return colon + 1 + length, data[colon + 1 : colon + 1 + length]

    return value(0)[1]


def free_port():
    probe = socket.socket()
    probe.bind(("127.0.0.1", 0))
    port = probe.getsockname()[1]
    probe.close()
    return port


class Tracker:
    """One tracker process, listening on loopback."""

    def __init__(self, extra_config="", files=None):
        self.port = free_port()
        self.dir = tempfile.mkdtemp(prefix="ot-integration-")
        self.log = None
        self.process = None
        try:
            for name, content in (files or {}).items():
                with open(os.path.join(self.dir, name), "wb") as handle:
                    handle.write(content)
            self.config = os.path.join(self.dir, "opentracker.conf")
            with open(self.config, "w") as handle:
                handle.write("listen.tcp 127.0.0.1:%d\n" % self.port)
                handle.write("listen.udp 127.0.0.1:%d\n" % self.port)
                handle.write(extra_config.replace("{dir}", self.dir))
            self.log = open(os.path.join(self.dir, "tracker.log"), "w+b")
            # Leak detection is left at the sanitizer's own default, which is on
            # where LeakSanitizer exists and off where it does not. Asking for it
            # by name makes the tracker refuse to start on macOS and FreeBSD.
            environment = dict(os.environ, ASAN_OPTIONS="abort_on_error=0")
            self.process = subprocess.Popen([TRACKER, "-f", self.config], stdout=self.log, stderr=self.log, env=environment)
            self._wait_for_port()
        except BaseException:
            self._cleanup()
            raise

    def _cleanup(self):
        """Close what we opened and take the scratch directory with it."""
        if self.process is not None and self.process.poll() is None:
            self.process.kill()
            self.process.wait()
        if self.log is not None:
            self.log.close()
            self.log = None
        shutil.rmtree(self.dir, ignore_errors=True)

    def _wait_for_port(self):
        for _ in range(200):
            if self.process.poll() is not None:
                raise AssertionError("tracker exited at once:\n" + self.output())
            try:
                probe = socket.create_connection(("127.0.0.1", self.port), timeout=0.2)
                probe.close()
                return
            except OSError:
                time.sleep(0.05)
        raise AssertionError("tracker never listened on %d:\n%s" % (self.port, self.output()))

    def output(self):
        if self.log is None:
            return ""
        self.log.flush()
        self.log.seek(0)
        return self.log.read().decode("utf-8", "replace")

    def stop(self):
        """Returns anything the tracker complained about on the way out.

        The interrupt rather than a kill is what makes the leak check work at
        all: LeakSanitizer reports from an exit handler, so a killed tracker
        reports nothing and the suite passes over leaks it never looked for.
        """
        if self.process is not None and self.process.poll() is None:
            self.process.send_signal(signal.SIGINT)
            try:
                self.process.wait(timeout=5)
            except subprocess.TimeoutExpired:
                self.process.kill()
                self.process.wait()
        complaints = [
            line
            for line in self.output().splitlines()
            if "Sanitizer" in line or "runtime error" in line or line.startswith("ERROR")
        ]
        self._cleanup()
        return complaints

    # ----------------------------------------------------------- over http

    def get(self, path, timeout=10):
        return urllib.request.urlopen("http://127.0.0.1:%d%s" % (self.port, path), timeout=timeout).read()

    def announce(self, info_hash, peer_id=b"-OT0001-aaaaaaaaaaaa", port=6881, left=0, **extra):
        query = {
            "info_hash": urllib.parse.quote_from_bytes(info_hash, safe=""),
            "peer_id": urllib.parse.quote_from_bytes(peer_id, safe=""),
            "port": port,
            "left": left,
        }
        query.update(extra)
        return bdecode(self.get("/announce?" + "&".join("%s=%s" % item for item in query.items())))

    def scrape(self, info_hash):
        return bdecode(self.get("/scrape?info_hash=" + urllib.parse.quote_from_bytes(info_hash, safe="")))

    def stats(self, mode):
        return self.get("/stats?mode=" + mode)

    def get_with_headers(self, path, headers=None, timeout=10):
        """The body and the response headers, for the compressed answers."""
        request = urllib.request.Request("http://127.0.0.1:%d%s" % (self.port, path), headers=headers or {})
        with urllib.request.urlopen(request, timeout=timeout) as response:
            return response.read(), dict(response.headers)

    def torrent_count(self):
        return int(self.stats("torr").split()[0])

    # ------------------------------------------------------------ over udp

    def udp_connect(self):
        sock = socket.socket(socket.AF_INET, socket.SOCK_DGRAM)
        sock.settimeout(5)
        sock.sendto(struct.pack(">QII", 0x41727101980, 0, 1), ("127.0.0.1", self.port))
        answer = sock.recv(2048)
        action, _, connection_id = struct.unpack(">IIQ", answer[:16])
        assert action == 0, "udp connect answered action %d" % action
        return sock, connection_id

    def udp_announce(self, sock, connection_id, info_hash, left=0, event=0, port=6881, numwant=-1):
        sock.sendto(
            struct.pack(
                ">QII20s20sQQQIIIiH", connection_id, 1, 2, info_hash, b"p" * 20, 0, left, 0, event, 0, 0, numwant, port
            ),
            ("127.0.0.1", self.port),
        )
        answer = sock.recv(2048)
        action, _, interval, leechers, seeders = struct.unpack(">IIIII", answer[:20])
        assert action == 1, "udp announce answered action %d" % action
        return {"interval": interval, "leechers": leechers, "seeders": seeders, "peers": answer[20:]}

    def udp_scrape(self, sock, connection_id, hashes):
        """One (seeders, completed, leechers) triple per hash, in order."""
        sock.sendto(struct.pack(">QII", connection_id, 2, 3) + b"".join(hashes), ("127.0.0.1", self.port))
        answer = sock.recv(2048)
        action, _ = struct.unpack(">II", answer[:8])
        assert action == 2, "udp scrape answered action %d" % action
        return [struct.unpack(">III", answer[8 + 12 * i : 20 + 12 * i]) for i in range(len(hashes))]

    def connect(self):
        """A connection the caller drives and closes."""
        return Connection(self.port)

    def udp_raw(self, payload, timeout=1.0):
        """Sends a packet exactly as given. Returns the answer, or None if the
        tracker chose to say nothing, which is how it treats a broken one."""
        sock = socket.socket(socket.AF_INET, socket.SOCK_DGRAM)
        sock.settimeout(timeout)
        try:
            sock.sendto(payload, ("127.0.0.1", self.port))
            try:
                return sock.recv(2048)
            except socket.timeout:
                return None
        finally:
            sock.close()


class Connection:
    """One HTTP connection the caller keeps open, which urllib will not do.

    Reads exactly one reply at a time, so a test can tell "the tracker
    answered and kept the socket" from "the tracker answered and hung up"."""

    def __init__(self, port, timeout=5):
        self.sock = socket.create_connection(("127.0.0.1", port), timeout=timeout)
        self.sock.settimeout(timeout)
        self.buffer = b""

    def send(self, path, connection="Keep-Alive", version="1.1"):
        request = "GET %s HTTP/%s\r\nHost: tracker\r\n" % (path, version)
        if connection is not None:
            request += "Connection: %s\r\n" % connection
        self.sock.sendall((request + "\r\n").encode())

    def _fill(self):
        chunk = self.sock.recv(65536)
        if not chunk:
            raise EOFError("the tracker closed the connection")
        self.buffer += chunk

    def reply(self):
        """The one complete reply at the front of the stream, headers and body."""
        while b"\r\n\r\n" not in self.buffer:
            self._fill()
        head, rest = self.buffer.split(b"\r\n\r\n", 1)
        headers = {}
        for line in head.split(b"\r\n")[1:]:
            name, _, value = line.partition(b":")
            headers[name.strip().lower()] = value.strip()

        if headers.get(b"transfer-encoding") == b"chunked":
            while not rest.endswith(b"0\r\n\r\n"):
                self.buffer = head + b"\r\n\r\n" + rest
                self._fill()
                rest = self.buffer.split(b"\r\n\r\n", 1)[1]
            body, self.buffer = rest, b""
        else:
            length = int(headers.get(b"content-length", b"0"))
            while len(rest) < length:
                self.buffer = head + b"\r\n\r\n" + rest
                self._fill()
                rest = self.buffer.split(b"\r\n\r\n", 1)[1]
            body, self.buffer = rest[:length], rest[length:]
        return head.split(b"\r\n")[0], headers, body

    def is_open(self):
        """True if the tracker is still holding the connection."""
        try:
            self.sock.settimeout(0.4)
            return self.sock.recv(1) != b""
        except socket.timeout:
            return True
        except OSError:
            return False
        finally:
            try:
                self.sock.settimeout(5)
            except OSError:
                pass

    def close(self):
        self.sock.close()


class TrackerTestCase(unittest.TestCase):
    """Gives every test its own tracker and fails it on any sanitizer output."""

    config = ""
    files = None

    def setUp(self):
        self.tracker = Tracker(self.config, self.files)

    def tearDown(self):
        # A connection the test closed is only reaped when the event loop gets
        # round to it, and a cookie still held at exit reads as a leak. One
        # more request puts us behind that work in the loop's queue.
        try:
            self.tracker.torrent_count()
        except Exception:
            pass
        complaints = self.tracker.stop()
        self.assertEqual(complaints, [], "the tracker reported:\n" + "\n".join(complaints))

    def hash_for(self, name):
        """A distinct info hash per test, so tests cannot leak into each other."""
        return (name.encode() + b"\0" * 20)[:20]