FinDataHub/Scripts/bin/tests/test_sync_bridge.py

"""sync_bridge.py 测试 — _pump_stream 流转发 + 子进程管理 + 辅助函数."""

from __future__ import annotations

import io
import os
import subprocess
import sys
import threading
from unittest.mock import patch

import pytest

# 将 Scripts/bin 加入 sys.path 以导入 sync_bridge
_script_dir = os.path.dirname(os.path.dirname(os.path.abspath(__file__)))
if _script_dir not in sys.path:
    sys.path.insert(0, _script_dir)

import sync_bridge


# ── _pump_stream ─────────────────────────────────────────────

class _FakeSink:
    """记录写入的 sink，模拟 sys.stdout.buffer / sys.stderr.buffer."""

    def __init__(self):
        self.written: list[bytes] = []
        self.flush_calls: list[int] = []

    def write(self, data: bytes):
        self.written.append(data)

    def flush(self):
        self.flush_calls.append(len(self.written))


class TestPumpStream:
    def test_forwards_all_chunks(self):
        """_pump_stream 将 fd 中所有数据块转发到 sink."""
        data = [b"hello\n", b"world\n", b""]
        fd = _FakeReadFD(data)
        sink = _FakeSink()
        sync_bridge._pump_stream(fd, sink)
        assert sink.written == [b"hello\n", b"world\n"]

    def test_empty_fd_no_write(self):
        """fd 立即返回空字节时不写入任何内容."""
        sink = _FakeSink()
        sync_bridge._pump_stream(_FakeReadFD([b""]), sink)
        assert sink.written == []

    def test_os_error_breaks_loop(self):
        """OSError 时优雅退出（不传播异常）."""
        sink = _FakeSink()
        # os.read 在无效 fd 上抛出 OSError
        sync_bridge._pump_stream(-1, sink)
        assert sink.written == []

    def test_large_chunk_forwarded(self):
        """大块数据（>8192 字节）正确转发."""
        data = b"x" * 10000
        fd = _FakeReadFD([data, b""])
        sink = _FakeSink()
        sync_bridge._pump_stream(fd, sink)
        assert len(sink.written) == 1
        assert sink.written[0] == data

    def test_flush_called_after_each_write(self):
        """每次 write 后调用 flush，确保实时转发不被缓冲."""
        data = [b"chunk1\n", b"chunk2\n", b""]
        fd = _FakeReadFD(data)
        sink = _FakeSink()
        sync_bridge._pump_stream(fd, sink)
        # 每次 write 后应紧跟一次 flush：written[0] 后 flush 时 written 长度为 1
        assert len(sink.flush_calls) == 2
        assert sink.flush_calls == [1, 2]

    def test_multiple_chunks_each_flushed(self):
        """多个数据块时每个块后都 flush."""
        data = [b"a\n", b"b\n", b"c\n", b"d\n", b""]
        fd = _FakeReadFD(data)
        sink = _FakeSink()
        sync_bridge._pump_stream(fd, sink)
        assert len(sink.flush_calls) == 4
        assert sink.flush_calls == [1, 2, 3, 4]


# ── _find_project ────────────────────────────────────────────

class TestFindProject:
    def test_returns_first_candidate_with_pyproject(self, tmp_path, monkeypatch):
        """返回第一个存在 pyproject.toml 的候选路径."""
        # 创建包含 pyproject.toml 的目录
        proj_dir = tmp_path / "test_proj"
        proj_dir.mkdir()
        (proj_dir / "pyproject.toml").write_text("")

        # mock HOME 和 os.environ
        monkeypatch.setattr(sync_bridge, "HOME", tmp_path)
        monkeypatch.delenv("APP_RESOURCES", raising=False)

        result = sync_bridge._find_project("test_proj")
        assert result == proj_dir

    def test_falls_back_to_first_candidate(self, tmp_path, monkeypatch):
        """所有候选都不存在时返回第一个候选（不崩溃）."""
        monkeypatch.setattr(sync_bridge, "HOME", tmp_path)
        monkeypatch.delenv("APP_RESOURCES", raising=False)

        result = sync_bridge._find_project("nonexistent_proj")
        # 返回第一个候选路径
        assert result is not None


# ── load_token ───────────────────────────────────────────────

class TestLoadToken:
    def test_reads_from_env_file(self, tmp_path, monkeypatch):
        """从 .env 文件读取 TUSHARE_TOKEN."""
        env_file = tmp_path / ".env"
        env_file.write_text("TUSHARE_TOKEN=test-token-123\nOTHER=value\n")

        result = sync_bridge.load_token(tmp_path)
        assert result == "test-token-123"

    def test_quoted_token_stripped(self, tmp_path):
        """带引号的 token 值正确去除引号."""
        env_file = tmp_path / ".env"
        env_file.write_text('TUSHARE_TOKEN="quoted-token"\n')

        result = sync_bridge.load_token(tmp_path)
        assert result == "quoted-token"

    def test_no_env_file_returns_env_var(self, tmp_path, monkeypatch):
        """无 .env 文件时回退到环境变量."""
        monkeypatch.setenv("TUSHARE_TOKEN", "env-token")
        result = sync_bridge.load_token(tmp_path)
        assert result == "env-token"

    def test_no_env_file_no_env_var_returns_empty(self, tmp_path, monkeypatch):
        """无 .env 且无环境变量时返回空字符串."""
        monkeypatch.delenv("TUSHARE_TOKEN", raising=False)
        result = sync_bridge.load_token(tmp_path)
        assert result == ""


# ── run_akshare ───────────────────────────────────────────────

class TestRunAkshare:
    def test_returns_exit_code(self, tmp_path, monkeypatch):
        """子进程正常退出时返回退出码."""
        # 使用 echo 命令模拟子进程
        monkeypatch.setattr(subprocess, "Popen", _make_fake_popen(returncode=0))
        monkeypatch.setattr(sync_bridge, "find_python", lambda p: "/usr/bin/python3")

        rc = sync_bridge.run_akshare(tmp_path, "daily", 4, "/usr/bin/python3")
        assert rc == 0

    def test_returns_nonzero_exit_code(self, tmp_path, monkeypatch):
        """子进程异常退出时返回非零退出码."""
        monkeypatch.setattr(subprocess, "Popen", _make_fake_popen(returncode=1))
        monkeypatch.setattr(sync_bridge, "find_python", lambda p: "/usr/bin/python3")

        rc = sync_bridge.run_akshare(tmp_path, "daily", 4, "/usr/bin/python3")
        assert rc == 1

    def test_stdout_stderr_threads_started(self, tmp_path, monkeypatch):
        """stdout/stderr 转发线程被启动."""
        threads_started = []

        class _FakeThread:
            def __init__(self, target, args, daemon=False):
                self.target = target
                self.args = args
                self.daemon = daemon
                threads_started.append(self)

            def start(self):
                pass

            def join(self):
                pass

        monkeypatch.setattr(threading, "Thread", _FakeThread)
        monkeypatch.setattr(subprocess, "Popen", _make_fake_popen(returncode=0))
        monkeypatch.setattr(sync_bridge, "find_python", lambda p: "/usr/bin/python3")

        sync_bridge.run_akshare(tmp_path, "daily", 4, "/usr/bin/python3")
        assert len(threads_started) == 2
        assert threads_started[0].daemon is True
        assert threads_started[1].daemon is True

    def test_popen_cmd_constructed_correctly(self, tmp_path, monkeypatch):
        """Popen cmd 参数正确构造：python -m akshare_fetcher.cli sync --type X --workers N."""
        popen_calls = []

        class _CapturePopen:
            def __init__(self, cmd, **kwargs):
                popen_calls.append((cmd, kwargs))
                self._stdout = _FakeReadFD([b""])
                self._stderr = _FakeReadFD([b""])

            @property
            def stdout(self):
                return self._stdout

            @property
            def stderr(self):
                return self._stderr

            def wait(self):
                return 0

        monkeypatch.setattr(subprocess, "Popen", _CapturePopen)
        monkeypatch.setattr(sync_bridge, "find_python", lambda p: "/usr/bin/python3")

        sync_bridge.run_akshare(tmp_path, "weekly", 8, "/usr/bin/python3")
        assert len(popen_calls) == 1
        cmd = popen_calls[0][0]
        assert cmd[0] == "/usr/bin/python3"
        assert cmd[1] == "-m"
        assert cmd[2] == "akshare_fetcher.cli"
        assert cmd[3] == "sync"
        assert "--type" in cmd
        assert "weekly" in cmd
        assert "--workers" in cmd
        assert "8" in cmd

    def test_popen_env_contains_required_vars(self, tmp_path, monkeypatch):
        """Popen env 包含 PYTHONPATH、AKSHARE_HOME、PYTHONUNBUFFERED."""
        popen_calls = []

        class _CapturePopen:
            def __init__(self, cmd, **kwargs):
                popen_calls.append((cmd, kwargs))
                self._stdout = _FakeReadFD([b""])
                self._stderr = _FakeReadFD([b""])

            @property
            def stdout(self):
                return self._stdout

            @property
            def stderr(self):
                return self._stderr

            def wait(self):
                return 0

        monkeypatch.setattr(subprocess, "Popen", _CapturePopen)
        monkeypatch.setattr(sync_bridge, "find_python", lambda p: "/usr/bin/python3")

        sync_bridge.run_akshare(tmp_path, "daily", 4, "/usr/bin/python3")
        env = popen_calls[0][1].get("env", {})
        assert env["PYTHONPATH"] == f"{tmp_path}/src"
        assert env["AKSHARE_HOME"] == str(tmp_path)
        assert env["PYTHONUNBUFFERED"] == "1"

    def test_popen_uses_pipes_for_streams(self, tmp_path, monkeypatch):
        """Popen 使用 PIPE 捕获 stdout 和 stderr."""
        popen_calls = []

        class _CapturePopen:
            def __init__(self, cmd, **kwargs):
                popen_calls.append((cmd, kwargs))
                self._stdout = _FakeReadFD([b""])
                self._stderr = _FakeReadFD([b""])

            @property
            def stdout(self):
                return self._stdout

            @property
            def stderr(self):
                return self._stderr

            def wait(self):
                return 0

        monkeypatch.setattr(subprocess, "Popen", _CapturePopen)
        monkeypatch.setattr(sync_bridge, "find_python", lambda p: "/usr/bin/python3")

        sync_bridge.run_akshare(tmp_path, "monthly", 2, "/usr/bin/python3")
        kwargs = popen_calls[0][1]
        assert kwargs["stdout"] == subprocess.PIPE
        assert kwargs["stderr"] == subprocess.PIPE
        assert kwargs["cwd"] == str(tmp_path)


# ── run_tushare ──────────────────────────────────────────────

class TestRunTushare:
    def test_no_token_returns_1(self, tmp_path, monkeypatch):
        """无 token 时立即返回 1."""
        monkeypatch.setattr(sync_bridge, "load_token", lambda p: "")
        rc = sync_bridge.run_tushare(tmp_path, "daily", 4, "/usr/bin/python3")
        assert rc == 1

    def test_batch_success_returns_0(self, tmp_path, monkeypatch):
        """所有 fetch type 成功时返回 0."""
        monkeypatch.setattr(sync_bridge, "load_token", lambda p: "token")
        monkeypatch.setattr(subprocess, "Popen", _make_fake_popen(returncode=0))
        monkeypatch.setattr(sync_bridge, "find_python", lambda p: "/usr/bin/python3")

        # daily batch 包含 ["daily", "market_sentiment"]
        rc = sync_bridge.run_tushare(tmp_path, "daily", 4, "/usr/bin/python3")
        assert rc == 0

    def test_batch_partial_failure_returns_nonzero(self, tmp_path, monkeypatch):
        """任一 fetch type 失败时返回非零."""
        monkeypatch.setattr(sync_bridge, "load_token", lambda p: "token")

        class _FailThenOkPopen:
            _call_count = 0

            def __init__(self, *args, **kwargs):
                _FailThenOkPopen._call_count += 1

            def wait(self):
                if _FailThenOkPopen._call_count == 1:
                    return 1  # 第一个 fetch type 失败
                return 0

            @property
            def stdout(self):
                return _FakeReadFD([])

            @property
            def stderr(self):
                return _FakeReadFD([])

        monkeypatch.setattr(subprocess, "Popen", _FailThenOkPopen)
        monkeypatch.setattr(sync_bridge, "find_python", lambda p: "/usr/bin/python3")

        # 重置计数器
        _FailThenOkPopen._call_count = 0
        rc = sync_bridge.run_tushare(tmp_path, "daily", 4, "/usr/bin/python3")
        assert rc == 1

    def test_popen_env_contains_token(self, tmp_path, monkeypatch):
        """tushare Popen env 包含 TUSHARE_TOKEN."""
        popen_calls = []

        class _CapturePopen:
            def __init__(self, cmd, **kwargs):
                popen_calls.append((cmd, kwargs))
                self._stdout = _FakeReadFD([b""])
                self._stderr = _FakeReadFD([b""])

            @property
            def stdout(self):
                return self._stdout

            @property
            def stderr(self):
                return self._stderr

            def wait(self):
                return 0

        monkeypatch.setattr(subprocess, "Popen", _CapturePopen)
        monkeypatch.setattr(sync_bridge, "load_token", lambda p: "test-token-abc")
        monkeypatch.setattr(sync_bridge, "find_python", lambda p: "/usr/bin/python3")

        sync_bridge.run_tushare(tmp_path, "daily", 4, "/usr/bin/python3")
        assert len(popen_calls) >= 1
        env = popen_calls[0][1].get("env", {})
        assert env["TUSHARE_TOKEN"] == "test-token-abc"
        assert env["TUSHARE_HOME"] == str(tmp_path)
        assert env["PYTHONUNBUFFERED"] == "1"

    def test_weekly_batch_uses_correct_fetch_types(self, tmp_path, monkeypatch):
        """weekly batch 使用 [fundamental, static, event]."""
        popen_calls = []

        class _CapturePopen:
            def __init__(self, cmd, **kwargs):
                popen_calls.append((cmd, kwargs))
                self._stdout = _FakeReadFD([b""])
                self._stderr = _FakeReadFD([b""])

            @property
            def stdout(self):
                return self._stdout

            @property
            def stderr(self):
                return self._stderr

            def wait(self):
                return 0

        monkeypatch.setattr(subprocess, "Popen", _CapturePopen)
        monkeypatch.setattr(sync_bridge, "load_token", lambda p: "token")
        monkeypatch.setattr(sync_bridge, "find_python", lambda p: "/usr/bin/python3")

        sync_bridge.run_tushare(tmp_path, "weekly", 4, "/usr/bin/python3")
        # weekly batch: fundamental, static, event = 3 Popen 调用
        assert len(popen_calls) == 3

    def test_unknown_sync_type_falls_back_to_daily(self, tmp_path, monkeypatch):
        """未知 sync_type 回退到 ["daily"]."""
        popen_calls = []

        class _CapturePopen:
            def __init__(self, cmd, **kwargs):
                popen_calls.append((cmd, kwargs))
                self._stdout = _FakeReadFD([b""])
                self._stderr = _FakeReadFD([b""])

            @property
            def stdout(self):
                return self._stdout

            @property
            def stderr(self):
                return self._stderr

            def wait(self):
                return 0

        monkeypatch.setattr(subprocess, "Popen", _CapturePopen)
        monkeypatch.setattr(sync_bridge, "load_token", lambda p: "token")
        monkeypatch.setattr(sync_bridge, "find_python", lambda p: "/usr/bin/python3")

        sync_bridge.run_tushare(tmp_path, "unknown_type", 4, "/usr/bin/python3")
        assert len(popen_calls) == 1  # 仅 daily

    def test_threads_started_per_fetch_type(self, tmp_path, monkeypatch):
        """每个 fetch type 启动 2 个 daemon 线程（stdout/stderr 转发）.

        安全审计修复（4b62693）：run_tushare 从 capture_output=True（可能
        缓冲满后死锁）改为 Popen + _pump_stream 线程转发。此测试验证
        线程确实被创建且为 daemon，确保子进程输出不会被丢弃。
        """
        threads_started = []

        class _FakeThread:
            def __init__(self, target, args, daemon=False):
                self.target = target
                self.args = args
                self.daemon = daemon
                threads_started.append(self)

            def start(self):
                pass

            def join(self):
                pass

        monkeypatch.setattr(threading, "Thread", _FakeThread)
        monkeypatch.setattr(sync_bridge, "load_token", lambda p: "token")
        monkeypatch.setattr(subprocess, "Popen", _make_fake_popen(returncode=0))
        monkeypatch.setattr(sync_bridge, "find_python", lambda p: "/usr/bin/python3")

        # daily batch = ["daily", "market_sentiment"] → 2 fetch types × 2 threads = 4
        sync_bridge.run_tushare(tmp_path, "daily", 4, "/usr/bin/python3")
        assert len(threads_started) == 4
        assert all(t.daemon is True for t in threads_started)

    def test_popen_uses_pipes_not_capture_output(self, tmp_path, monkeypatch):
        """Popen 使用 stdout=PIPE/stderr=PIPE，不使用 capture_output.

        capture_output=True 会缓冲子进程 stderr（tqdm 进度），
        缓冲满后阻塞子进程写入导致死锁。必须使用 PIPE + 线程转发。
        """
        popen_calls = []

        class _CapturePopen:
            def __init__(self, cmd, **kwargs):
                popen_calls.append((cmd, kwargs))
                self._stdout = _FakeReadFD([b""])
                self._stderr = _FakeReadFD([b""])

            @property
            def stdout(self):
                return self._stdout

            @property
            def stderr(self):
                return self._stderr

            def wait(self):
                return 0

        monkeypatch.setattr(subprocess, "Popen", _CapturePopen)
        monkeypatch.setattr(sync_bridge, "load_token", lambda p: "token")
        monkeypatch.setattr(sync_bridge, "find_python", lambda p: "/usr/bin/python3")

        sync_bridge.run_tushare(tmp_path, "daily", 4, "/usr/bin/python3")
        assert len(popen_calls) >= 1
        for _, kwargs in popen_calls:
            assert kwargs.get("stdout") == subprocess.PIPE
            assert kwargs.get("stderr") == subprocess.PIPE
            assert "capture_output" not in kwargs or not kwargs.get("capture_output")


# ── helpers ──────────────────────────────────────────────────

class _FakeReadFD:
    """模拟 os.read 可读取的文件描述符."""

    def __init__(self, chunks: list[bytes]):
        self._chunks = list(chunks)
        self._idx = 0

    def fileno(self):
        return 999  # 虚拟 fd 号


def _mock_os_read_factory(chunks: list[bytes]):
    """创建 os.read mock，按顺序返回 chunks."""

    def fake_read(fd, n):
        # 仅在 fd 是我们创建的 FakeReadFD 时才拦截
        if isinstance(fd, _FakeReadFD):
            if fd._idx >= len(fd._chunks):
                return b""
            chunk = fd._chunks[fd._idx]
            fd._idx += 1
            return chunk
        return os.read(fd, n)

    return fake_read


def _make_fake_popen(returncode=0):
    """创建 fake subprocess.Popen 类."""

    class FakePopen:
        def __init__(self, cmd, **kwargs):
            self.cmd = cmd
            self.kwargs = kwargs
            self._stdout = _FakeReadFD([b"ok\n", b""])
            self._stderr = _FakeReadFD([b""])

        @property
        def stdout(self):
            return self._stdout

        @property
        def stderr(self):
            return self._stderr

        def wait(self):
            return returncode

    return FakePopen


# 全局 fixture：用 mock os.read 替换 _pump_stream 中的 os.read 调用
@pytest.fixture(autouse=True)
def _mock_os_read(monkeypatch):
    """注入 fake os.read，使 _pump_stream 使用 FakeReadFD."""
    orig_read = os.read

    def fake_read(fd, n):
        if isinstance(fd, _FakeReadFD):
            if fd._idx >= len(fd._chunks):
                return b""
            chunk = fd._chunks[fd._idx]
            fd._idx += 1
            return chunk
        try:
            return orig_read(fd, n)
        except OSError:
            raise

    monkeypatch.setattr(os, "read", fake_read)