tdxalert 0.1.0__tar.gz

tdxalert-0.1.0/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 mengyiheng
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

tdxalert-0.1.0/PKG-INFO

@@ -0,0 +1,121 @@
+Metadata-Version: 2.4
+Name: tdxalert
+Version: 0.1.0
+Summary: 盘中预警推送引擎：可插拔数据源 + 基础规则集 + 微信推送，零依赖
+Author: mengyiheng
+License: MIT
+Keywords: stock,alert,quant,wechat,monitor
+Classifier: Development Status :: 4 - Beta
+Classifier: Intended Audience :: Developers
+Classifier: Intended Audience :: Financial and Insurance Industry
+Classifier: License :: OSI Approved :: MIT License
+Classifier: Programming Language :: Python :: 3
+Classifier: Topic :: Office/Business :: Financial :: Investment
+Requires-Python: >=3.9
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Dynamic: license-file
+
+<p align="center"><img src="assets/cover.png" alt="tdxalert" width="720"/></p>
+
+# tdxalert — 盘中预警推送引擎
+
+把你的**全市场实时快照**接进来，按规则监控，触发即推送**微信**。
+
+- 🔌 **数据源可插拔**：引擎只吃标准化的快照帧——你自己的采集器、pytdx、
+  [tdxreplay](https://gitee.com/mengyiheng/tdxreplay) 回放流，都能接
+- 📏 **基础规则集**：价格上/下破、涨跌幅超阈、成交量异常放大、参考线上穿/下穿，
+  每条规则独立冷却时间
+- 💬 **微信推送**：Server酱 / PushPlus，另附钉钉/企微 webhook 与控制台渠道
+- 🪶 **零第三方依赖**，纯 Python 标准库
+- 🧪 规则引擎与推送解耦，方便先接回放源**离线验证预警逻辑**再上实盘
+
+## 安装
+
+```bash
+pip install tdxalert
+```
+
+## 快速开始
+
+### 1. 写配置 `alert.json`
+
+```json
+{
+  "notifiers": { "serverchan": { "key": "SCTxxxxxxxx 你的SendKey" } },
+  "rules": [
+    { "type": "price_above", "code": "600519", "value": 1500, "cooldown": 600 },
+    { "type": "pct_below",   "code": "*",      "value": -3 },
+    { "type": "vol_spike",   "code": "*",      "window": 20, "ratio": 3 },
+    { "type": "change_dir",  "code": "000001", "value": 10.5 }
+  ]
+}
+```
+
+### 2. 把你的快照流接进来（每行一帧 JSON）
+
+```
+{"600519": {"price": 1501.2, "pct": 1.2, "vol": 3200, "name": "贵州茅台"}, "000001": {...}}
+```
+
+```bash
+your_snapshot_source | tdxalert run --config alert.json
+```
+
+### 3. 或者只用 Python API
+
+```python
+from tdxalert import AlertEngine
+
+engine = AlertEngine.from_config(cfg)
+for frame in my_snapshot_loop():   # 你的 REA / 采集器
+    engine.feed(frame)             # 触发即推微信
+```
+
+## 规则类型
+
+| type | 含义 | 专属参数 |
+|---|---|---|
+| `price_above` / `price_below` | 价格上破 / 下破 | `value` 阈值 |
+| `pct_above` / `pct_below` | 涨跌幅超阈（%） | `value` |
+| `vol_spike` | 成交量较滚动均值放大 | `window` 窗口根数、`ratio` 倍数 |
+| `change_dir` | 上穿 / 下穿参考线 | `value` 参考价 |
+
+通用参数：`code`（`"*"`=全市场）、`cooldown` 冷却秒数（默认 300）。
+
+## 快照帧字段约定
+
+| 字段 | 必需 | 说明 |
+|---|---|---|
+| `price` | ✅ | 最新价 |
+| `pct` | 建议 | 涨跌幅 %（正涨负跌） |
+| `vol` | 建议 | 成交量（vol_spike 需要） |
+| `name` | 可选 | 证券名称（出现在告警文案里） |
+
+## 提示
+
+- **先离线验证再上实盘**：把 [tdxreplay](https://gitee.com/mengyiheng/tdxreplay)
+  的回放流转成快照帧喂进来，历史数据上确认规则触发符合预期后再接实时源；
+- Server酱 SendKey 获取：<https://sct.ftqq.com/>（微信扫码即用）；
+- 规则误报多就加大 `cooldown`；放量告警灵敏度高，`window` 建议不低于 10。
+
+## 免责声明
+
+本项目仅为个人研究工具，不构成任何投资建议；告警可能延迟、丢失或误报。
+使用前请[阅读完整免责声明](DISCLAIMER.md)。
+
+## License
+
+MIT
+
+## 量化相关定制需求
+
+预警引擎恰恰是最值得按个人习惯定制的模块：
+
+| 方向 | 示例 |
+|---|---|
+| 📏 规则扩展 | 接入你的策略条件、多规则组合、板块/概念联动预警 |
+| 💬 渠道扩展 | Telegram/邮件/短信、按告警级别分流推送 |
+| 🔌 源适配 | 把你现有的采集器/行情源接成标准快照帧 |
+
+**联系合作**：[Gitee Issues](https://gitee.com/mengyiheng/tdxalert/issues) 或 [Gitee 私信](https://gitee.com/mengyiheng)。规则类需求请附：监控对象、触发条件、期望推送方式。

tdxalert-0.1.0/README.md

@@ -0,0 +1,103 @@
+<p align="center"><img src="assets/cover.png" alt="tdxalert" width="720"/></p>
+
+# tdxalert — 盘中预警推送引擎
+
+把你的**全市场实时快照**接进来，按规则监控，触发即推送**微信**。
+
+- 🔌 **数据源可插拔**：引擎只吃标准化的快照帧——你自己的采集器、pytdx、
+  [tdxreplay](https://gitee.com/mengyiheng/tdxreplay) 回放流，都能接
+- 📏 **基础规则集**：价格上/下破、涨跌幅超阈、成交量异常放大、参考线上穿/下穿，
+  每条规则独立冷却时间
+- 💬 **微信推送**：Server酱 / PushPlus，另附钉钉/企微 webhook 与控制台渠道
+- 🪶 **零第三方依赖**，纯 Python 标准库
+- 🧪 规则引擎与推送解耦，方便先接回放源**离线验证预警逻辑**再上实盘
+
+## 安装
+
+```bash
+pip install tdxalert
+```
+
+## 快速开始
+
+### 1. 写配置 `alert.json`
+
+```json
+{
+  "notifiers": { "serverchan": { "key": "SCTxxxxxxxx 你的SendKey" } },
+  "rules": [
+    { "type": "price_above", "code": "600519", "value": 1500, "cooldown": 600 },
+    { "type": "pct_below",   "code": "*",      "value": -3 },
+    { "type": "vol_spike",   "code": "*",      "window": 20, "ratio": 3 },
+    { "type": "change_dir",  "code": "000001", "value": 10.5 }
+  ]
+}
+```
+
+### 2. 把你的快照流接进来（每行一帧 JSON）
+
+```
+{"600519": {"price": 1501.2, "pct": 1.2, "vol": 3200, "name": "贵州茅台"}, "000001": {...}}
+```
+
+```bash
+your_snapshot_source | tdxalert run --config alert.json
+```
+
+### 3. 或者只用 Python API
+
+```python
+from tdxalert import AlertEngine
+
+engine = AlertEngine.from_config(cfg)
+for frame in my_snapshot_loop():   # 你的 REA / 采集器
+    engine.feed(frame)             # 触发即推微信
+```
+
+## 规则类型
+
+| type | 含义 | 专属参数 |
+|---|---|---|
+| `price_above` / `price_below` | 价格上破 / 下破 | `value` 阈值 |
+| `pct_above` / `pct_below` | 涨跌幅超阈（%） | `value` |
+| `vol_spike` | 成交量较滚动均值放大 | `window` 窗口根数、`ratio` 倍数 |
+| `change_dir` | 上穿 / 下穿参考线 | `value` 参考价 |
+
+通用参数：`code`（`"*"`=全市场）、`cooldown` 冷却秒数（默认 300）。
+
+## 快照帧字段约定
+
+| 字段 | 必需 | 说明 |
+|---|---|---|
+| `price` | ✅ | 最新价 |
+| `pct` | 建议 | 涨跌幅 %（正涨负跌） |
+| `vol` | 建议 | 成交量（vol_spike 需要） |
+| `name` | 可选 | 证券名称（出现在告警文案里） |
+
+## 提示
+
+- **先离线验证再上实盘**：把 [tdxreplay](https://gitee.com/mengyiheng/tdxreplay)
+  的回放流转成快照帧喂进来，历史数据上确认规则触发符合预期后再接实时源；
+- Server酱 SendKey 获取：<https://sct.ftqq.com/>（微信扫码即用）；
+- 规则误报多就加大 `cooldown`；放量告警灵敏度高，`window` 建议不低于 10。
+
+## 免责声明
+
+本项目仅为个人研究工具，不构成任何投资建议；告警可能延迟、丢失或误报。
+使用前请[阅读完整免责声明](DISCLAIMER.md)。
+
+## License
+
+MIT
+
+## 量化相关定制需求
+
+预警引擎恰恰是最值得按个人习惯定制的模块：
+
+| 方向 | 示例 |
+|---|---|
+| 📏 规则扩展 | 接入你的策略条件、多规则组合、板块/概念联动预警 |
+| 💬 渠道扩展 | Telegram/邮件/短信、按告警级别分流推送 |
+| 🔌 源适配 | 把你现有的采集器/行情源接成标准快照帧 |
+
+**联系合作**：[Gitee Issues](https://gitee.com/mengyiheng/tdxalert/issues) 或 [Gitee 私信](https://gitee.com/mengyiheng)。规则类需求请附：监控对象、触发条件、期望推送方式。

tdxalert-0.1.0/pyproject.toml

@@ -0,0 +1,27 @@
+[build-system]
+requires = ["setuptools>=68"]
+build-backend = "setuptools.build_meta"
+
+[project]
+name = "tdxalert"
+version = "0.1.0"
+description = "盘中预警推送引擎：可插拔数据源 + 基础规则集 + 微信推送，零依赖"
+readme = "README.md"
+requires-python = ">=3.9"
+license = { text = "MIT" }
+authors = [{ name = "mengyiheng" }]
+keywords = ["stock", "alert", "quant", "wechat", "monitor"]
+classifiers = [
+    "Development Status :: 4 - Beta",
+    "Intended Audience :: Developers",
+    "Intended Audience :: Financial and Insurance Industry",
+    "License :: OSI Approved :: MIT License",
+    "Programming Language :: Python :: 3",
+    "Topic :: Office/Business :: Financial :: Investment",
+]
+
+[project.scripts]
+tdxalert = "tdxalert.cli:main"
+
+[tool.setuptools.packages.find]
+where = ["src"]

tdxalert-0.1.0/setup.cfg

@@ -0,0 +1,4 @@
+[egg_info]
+tag_build = 
+tag_date = 0
+

tdxalert-0.1.0/src/tdxalert/__init__.py

@@ -0,0 +1,23 @@
+"""tdxalert — 盘中预警推送引擎（数据源无关）。
+
+数据源可插拔：任何能产出 ``{code: {"price": .., "pct": .., "vol": ..}}``
+快照帧的代码都能接入，包括你自己的全市场采集器、pytdx、或 tdxreplay 回放流。
+
+    from tdxalert import AlertEngine
+
+    engine = AlertEngine.from_config({
+        "notifiers": {"serverchan": {"key": "SCTxxx"}},
+        "rules": [
+            {"type": "price_above", "code": "600519", "value": 1500},
+            {"type": "vol_spike", "code": "*", "window": 20, "ratio": 3},
+        ],
+    })
+    for frame in your_snapshot_loop():     # 你的 REA 全市场快照
+        engine.feed(frame)                 # 触发即推送微信
+"""
+
+from .engine import AlertEngine
+from .rules import Rule, evaluate, parse
+
+__version__ = "0.1.0"
+__all__ = ["AlertEngine", "Rule", "evaluate", "parse", "__version__"]

tdxalert-0.1.0/src/tdxalert/__main__.py

@@ -0,0 +1,3 @@
+from .cli import main
+
+raise SystemExit(main())

tdxalert-0.1.0/src/tdxalert/cli.py

@@ -0,0 +1,53 @@
+"""tdxalert.cli — 命令行：读 JSONL 快照流，跑规则，推微信。
+
+用法:
+  tdxalert run --config alert.json --input quotes.jsonl     # 快照文件/管道
+  your_snapshot_source | tdxalert run --config alert.jsonl  # 接你自己的采集器
+
+快照帧格式（每行一个 JSON）::
+  {"600519": {"price": 1501.2, "pct": 1.2, "vol": 3200, "name": "贵州茅台"}, ...}
+"""
+
+from __future__ import annotations
+
+import argparse
+import json
+import sys
+import time
+
+
+def main(argv: list[str] | None = None) -> int:
+    p = argparse.ArgumentParser(prog="tdxalert", description="盘中预警推送引擎")
+    sub = p.add_subparsers(dest="cmd", required=True)
+    run = sub.add_parser("run", help="消费 JSONL 快照流并推送告警")
+    run.add_argument("--config", required=True, help="配置 JSON（规则 + 推送渠道）")
+    run.add_argument("--input", default="-", help="JSONL 快照流文件，默认 stdin")
+    run.add_argument("--interval", type=float, default=0.0,
+                     help="每帧之间的模拟间隔秒数（回放测试用）")
+    a = p.parse_args(argv)
+
+    from .engine import AlertEngine
+
+    cfg = json.loads(open(a.config, encoding="utf-8-sig").read())
+    engine = AlertEngine.from_config(cfg)
+    print(f"[tdxalert] 规则 {len(engine.rules)} 条，渠道: {list(engine.notifiers)}")
+
+    stream = sys.stdin if a.input == "-" else open(a.input, encoding="utf-8")
+    n = 0
+    for line in stream:
+        line = line.strip().lstrip("\ufeff")
+        if not line:
+            continue
+        frame = json.loads(line)
+        alerts = engine.feed(frame)
+        for text in alerts:
+            print(f"[触发] {text}")
+        n += 1
+        if a.interval:
+            time.sleep(a.interval)
+    print(f"[tdxalert] 处理 {n} 帧，触发 {len(engine.sent)} 条告警")
+    return 0
+
+
+if __name__ == "__main__":
+    raise SystemExit(main())

tdxalert-0.1.0/src/tdxalert/engine.py

@@ -0,0 +1,45 @@
+"""tdxalert.engine — 预警引擎：吃快照，吐告警。
+
+数据源无关：任何能产出 ``{code: {"price": float, "pct": float, "vol": float,
+"name": str}}`` 快照的代码都能接入::
+
+    engine = AlertEngine.from_config(cfg)
+    for frame in your_snapshot_loop():      # 你的 REA 快照
+        for msg in engine.feed(frame):
+            ...                             # 已推送到配置的渠道
+"""
+
+from __future__ import annotations
+
+from typing import Any, Dict, List
+
+from .notify import build_notifiers
+from .rules import Rule, evaluate, parse
+
+
+class AlertEngine:
+    def __init__(self, rules: List[Rule], notifiers: Dict[str, Any] | None = None):
+        self.rules = rules
+        self.notifiers = build_notifiers(notifiers or {})
+        self.sent: List[str] = []
+
+    @classmethod
+    def from_config(cls, cfg: Dict[str, Any]) -> "AlertEngine":
+        rules = [parse(c) for c in cfg.get("rules", [])]
+        return cls(rules, cfg.get("notifiers", {}))
+
+    def feed(self, quotes: Dict[str, Dict[str, Any]]) -> List[str]:
+        """处理一帧全市场快照，返回触发的告警文本（同时推送渠道）。"""
+        alerts: List[str] = []
+        for rule in self.rules:
+            for code, q in quotes.items():
+                msg = evaluate(rule, code, q)
+                if msg:
+                    name = q.get("name", "")
+                    alerts.append(f"{name} {msg}" if name else msg)
+        if alerts:
+            for text in alerts:
+                self.sent.append(text)
+                for n in self.notifiers.values():
+                    n.send(text)
+        return alerts

tdxalert-0.1.0/src/tdxalert/notify.py

@@ -0,0 +1,78 @@
+"""tdxalert.notify — 推送渠道（可插拔，纯标准库实现）。"""
+
+from __future__ import annotations
+
+import json
+import urllib.parse
+import urllib.request
+from typing import Any, Dict, Protocol
+
+
+class Notifier(Protocol):
+    def send(self, text: str) -> None: ...
+
+
+class ConsoleNotifier:
+    """本地控制台输出（默认兜底渠道）。"""
+
+    def send(self, text: str) -> None:
+        print(f"[tdxalert] {text}")
+
+
+class ServerChanNotifier:
+    """Server酱·Turbo 版：https://sct.ftqq.com/ 微信服务号消息。"""
+
+    def __init__(self, key: str):
+        self.key = key
+
+    def send(self, text: str) -> None:
+        data = urllib.parse.urlencode({"title": text[:32], "desp": text}).encode()
+        urllib.request.urlopen(f"https://sctapi.ftqq.com/{self.key}.send",
+                               data=data, timeout=10)
+
+
+class PushPlusNotifier:
+    """PushPlus：https://www.pushplus.plus/ 微信公众号消息。"""
+
+    def __init__(self, token: str):
+        self.token = token
+
+    def send(self, text: str) -> None:
+        body = json.dumps({"token": self.token, "title": text[:64],
+                           "content": text, "template": "txt"}).encode()
+        req = urllib.request.Request("https://www.pushplus.plus/send", data=body,
+                                     headers={"Content-Type": "application/json"})
+        urllib.request.urlopen(req, timeout=10)
+
+
+class WebhookNotifier:
+    """通用 webhook（钉钉/企微机器人等 POST JSON 接口）。"""
+
+    def __init__(self, url: str):
+        self.url = url
+
+    def send(self, text: str) -> None:
+        body = json.dumps({"msgtype": "text", "text": {"content": text}}).encode()
+        req = urllib.request.Request(self.url, data=body,
+                                     headers={"Content-Type": "application/json"})
+        urllib.request.urlopen(req, timeout=10)
+
+
+_BUILDERS = {
+    "console": lambda cfg: ConsoleNotifier(),
+    "serverchan": lambda cfg: ServerChanNotifier(cfg["key"]),
+    "pushplus": lambda cfg: PushPlusNotifier(cfg["token"]),
+    "webhook": lambda cfg: WebhookNotifier(cfg["url"]),
+}
+
+
+def build_notifiers(cfg: Dict[str, Any]) -> Dict[str, Notifier]:
+    out: Dict[str, Notifier] = {}
+    if not cfg:
+        out["console"] = ConsoleNotifier()
+        return out
+    for name, c in cfg.items():
+        if name not in _BUILDERS:
+            raise ValueError(f"未知推送渠道: {name}（支持 {list(_BUILDERS)}）")
+        out[name] = _BUILDERS[name](c if isinstance(c, dict) else {"key": c})
+    return out

tdxalert-0.1.0/src/tdxalert/rules.py

@@ -0,0 +1,104 @@
+"""tdxalert.rules — 基础预警规则集。
+
+规则在每帧快照上求值，返回告警消息（str）或 None。
+
+所有规则通用字段：
+    code      证券代码，"*" 表示全市场
+    cooldown  同一规则对同一代码的告警冷却秒数（默认 300）
+
+内置类型：
+    price_above / price_below    价格上破 / 下破阈值
+    pct_above / pct_below        涨跌幅超阈（%，正涨负跌）
+    vol_spike                    成交量较滚动均值放大 ratio 倍
+    change_dir                   方向反转（price 上穿/下穿 ref_price）
+"""
+
+from __future__ import annotations
+
+from dataclasses import dataclass, field
+from datetime import datetime
+from typing import Any, Dict, List, Optional
+
+
+@dataclass
+class Rule:
+    type: str
+    code: str = "*"
+    value: float = 0.0
+    window: int = 20          # vol_spike 滚动窗口
+    ratio: float = 3.0        # vol_spike 放大倍数
+    cooldown: int = 300
+    name: str = ""
+
+    # 运行时状态
+    last_fired: Dict[str, str] = field(default_factory=dict)   # code -> ISO 时间
+    vol_hist: Dict[str, List[float]] = field(default_factory=dict)
+
+    def matches(self, code: str) -> bool:
+        return self.code in ("*", code)
+
+    def _cooled(self, code: str) -> bool:
+        t = self.last_fired.get(code)
+        if not t:
+            return True
+        return (datetime.now() - datetime.fromisoformat(t)).total_seconds() >= self.cooldown
+
+    def _fire(self, code: str) -> None:
+        self.last_fired[code] = datetime.now().isoformat(timespec="seconds")
+
+
+def parse(cfg: Dict[str, Any]) -> Rule:
+    known = {"type", "code", "value", "window", "ratio", "cooldown", "name"}
+    extra = set(cfg) - known
+    if extra:
+        raise ValueError(f"规则含未知字段: {extra}（规则: {cfg}）")
+    r = Rule(type=cfg["type"], **{k: v for k, v in cfg.items() if k not in ("type", "name")})
+    r.name = cfg.get("name") or cfg["type"]
+    return r
+
+
+def evaluate(rule: Rule, code: str, q: Dict[str, Any]) -> Optional[str]:
+    """对单只股票快照求值。q 至少含 price；pct/vol/amount 可选。"""
+    if not rule.matches(code) or not rule._cooled(code):
+        return None
+    price = q.get("price")
+    if price is None:
+        return None
+    msg: Optional[str] = None
+
+    if rule.type == "price_above" and price > rule.value:
+        msg = f"价格 {price} 上破 {rule.value}"
+    elif rule.type == "price_below" and price < rule.value:
+        msg = f"价格 {price} 下破 {rule.value}"
+    elif rule.type == "pct_above" and (q.get("pct") or 0) > rule.value:
+        msg = f"涨幅 {q['pct']}% > {rule.value}%"
+    elif rule.type == "pct_below" and (q.get("pct") or 0) < rule.value:
+        msg = f"跌幅 {q['pct']}% < {rule.value}%"
+    elif rule.type == "change_dir":
+        ref = rule.value
+        hist = rule.vol_hist.setdefault(code, [])
+        prev = hist[-1] if hist else None
+        hist.append(price)
+        if prev is not None:
+            if prev <= ref < price:
+                msg = f"价格上穿 {ref}（{prev} -> {price}）"
+            elif prev >= ref > price:
+                msg = f"价格下穿 {ref}（{prev} -> {price}）"
+    elif rule.type == "vol_spike":
+        vol = q.get("vol")
+        if vol:
+            h = rule.vol_hist.setdefault(code, [])
+            h.append(vol)
+            if len(h) > rule.window:
+                h.pop(0)
+            if len(h) >= max(3, rule.window // 2):
+                avg = sum(h[:-1]) / len(h[:-1])
+                if avg > 0 and vol >= avg * rule.ratio:
+                    msg = f"成交量 {vol:.0f} 为前{len(h)-1}根均值 {avg:.0f} 的 {vol/avg:.1f} 倍"
+
+    if msg:
+        rule._fire(code)
+        if rule.type == "change_dir":
+            rule.vol_hist[code] = [price]      # 重置基准，避免同一方向重复触发
+        return f"[{rule.name}] {code} {msg}（现价 {price}）"
+    return None

tdxalert-0.1.0/src/tdxalert.egg-info/PKG-INFO

@@ -0,0 +1,121 @@
+Metadata-Version: 2.4
+Name: tdxalert
+Version: 0.1.0
+Summary: 盘中预警推送引擎：可插拔数据源 + 基础规则集 + 微信推送，零依赖
+Author: mengyiheng
+License: MIT
+Keywords: stock,alert,quant,wechat,monitor
+Classifier: Development Status :: 4 - Beta
+Classifier: Intended Audience :: Developers
+Classifier: Intended Audience :: Financial and Insurance Industry
+Classifier: License :: OSI Approved :: MIT License
+Classifier: Programming Language :: Python :: 3
+Classifier: Topic :: Office/Business :: Financial :: Investment
+Requires-Python: >=3.9
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Dynamic: license-file
+
+<p align="center"><img src="assets/cover.png" alt="tdxalert" width="720"/></p>
+
+# tdxalert — 盘中预警推送引擎
+
+把你的**全市场实时快照**接进来，按规则监控，触发即推送**微信**。
+
+- 🔌 **数据源可插拔**：引擎只吃标准化的快照帧——你自己的采集器、pytdx、
+  [tdxreplay](https://gitee.com/mengyiheng/tdxreplay) 回放流，都能接
+- 📏 **基础规则集**：价格上/下破、涨跌幅超阈、成交量异常放大、参考线上穿/下穿，
+  每条规则独立冷却时间
+- 💬 **微信推送**：Server酱 / PushPlus，另附钉钉/企微 webhook 与控制台渠道
+- 🪶 **零第三方依赖**，纯 Python 标准库
+- 🧪 规则引擎与推送解耦，方便先接回放源**离线验证预警逻辑**再上实盘
+
+## 安装
+
+```bash
+pip install tdxalert
+```
+
+## 快速开始
+
+### 1. 写配置 `alert.json`
+
+```json
+{
+  "notifiers": { "serverchan": { "key": "SCTxxxxxxxx 你的SendKey" } },
+  "rules": [
+    { "type": "price_above", "code": "600519", "value": 1500, "cooldown": 600 },
+    { "type": "pct_below",   "code": "*",      "value": -3 },
+    { "type": "vol_spike",   "code": "*",      "window": 20, "ratio": 3 },
+    { "type": "change_dir",  "code": "000001", "value": 10.5 }
+  ]
+}
+```
+
+### 2. 把你的快照流接进来（每行一帧 JSON）
+
+```
+{"600519": {"price": 1501.2, "pct": 1.2, "vol": 3200, "name": "贵州茅台"}, "000001": {...}}
+```
+
+```bash
+your_snapshot_source | tdxalert run --config alert.json
+```
+
+### 3. 或者只用 Python API
+
+```python
+from tdxalert import AlertEngine
+
+engine = AlertEngine.from_config(cfg)
+for frame in my_snapshot_loop():   # 你的 REA / 采集器
+    engine.feed(frame)             # 触发即推微信
+```
+
+## 规则类型
+
+| type | 含义 | 专属参数 |
+|---|---|---|
+| `price_above` / `price_below` | 价格上破 / 下破 | `value` 阈值 |
+| `pct_above` / `pct_below` | 涨跌幅超阈（%） | `value` |
+| `vol_spike` | 成交量较滚动均值放大 | `window` 窗口根数、`ratio` 倍数 |
+| `change_dir` | 上穿 / 下穿参考线 | `value` 参考价 |
+
+通用参数：`code`（`"*"`=全市场）、`cooldown` 冷却秒数（默认 300）。
+
+## 快照帧字段约定
+
+| 字段 | 必需 | 说明 |
+|---|---|---|
+| `price` | ✅ | 最新价 |
+| `pct` | 建议 | 涨跌幅 %（正涨负跌） |
+| `vol` | 建议 | 成交量（vol_spike 需要） |
+| `name` | 可选 | 证券名称（出现在告警文案里） |
+
+## 提示
+
+- **先离线验证再上实盘**：把 [tdxreplay](https://gitee.com/mengyiheng/tdxreplay)
+  的回放流转成快照帧喂进来，历史数据上确认规则触发符合预期后再接实时源；
+- Server酱 SendKey 获取：<https://sct.ftqq.com/>（微信扫码即用）；
+- 规则误报多就加大 `cooldown`；放量告警灵敏度高，`window` 建议不低于 10。
+
+## 免责声明
+
+本项目仅为个人研究工具，不构成任何投资建议；告警可能延迟、丢失或误报。
+使用前请[阅读完整免责声明](DISCLAIMER.md)。
+
+## License
+
+MIT
+
+## 量化相关定制需求
+
+预警引擎恰恰是最值得按个人习惯定制的模块：
+
+| 方向 | 示例 |
+|---|---|
+| 📏 规则扩展 | 接入你的策略条件、多规则组合、板块/概念联动预警 |
+| 💬 渠道扩展 | Telegram/邮件/短信、按告警级别分流推送 |
+| 🔌 源适配 | 把你现有的采集器/行情源接成标准快照帧 |
+
+**联系合作**：[Gitee Issues](https://gitee.com/mengyiheng/tdxalert/issues) 或 [Gitee 私信](https://gitee.com/mengyiheng)。规则类需求请附：监控对象、触发条件、期望推送方式。

tdxalert-0.1.0/src/tdxalert.egg-info/SOURCES.txt

@@ -0,0 +1,15 @@
+LICENSE
+README.md
+pyproject.toml
+src/tdxalert/__init__.py
+src/tdxalert/__main__.py
+src/tdxalert/cli.py
+src/tdxalert/engine.py
+src/tdxalert/notify.py
+src/tdxalert/rules.py
+src/tdxalert.egg-info/PKG-INFO
+src/tdxalert.egg-info/SOURCES.txt
+src/tdxalert.egg-info/dependency_links.txt
+src/tdxalert.egg-info/entry_points.txt
+src/tdxalert.egg-info/top_level.txt
+tests/test_alert.py

tdxalert-0.1.0/src/tdxalert.egg-info/dependency_links.txt

@@ -0,0 +1 @@
+

tdxalert-0.1.0/src/tdxalert.egg-info/entry_points.txt

@@ -0,0 +1,2 @@
+[console_scripts]
+tdxalert = tdxalert.cli:main

tdxalert-0.1.0/src/tdxalert.egg-info/top_level.txt

@@ -0,0 +1 @@
+tdxalert

tdxalert-0.1.0/tests/test_alert.py

@@ -0,0 +1,63 @@
+from datetime import datetime
+
+import pytest
+
+from tdxalert import AlertEngine, parse
+from tdxalert.notify import ConsoleNotifier
+
+
+def _engine(rules, sent):
+    e = AlertEngine(rules, {})
+    e.notifiers = {"console": ConsoleNotifier()}
+    e.sent = sent
+    return e
+
+
+def test_price_threshold_and_cooldown():
+    cfg = [{"type": "price_above", "code": "600519", "value": 1500, "cooldown": 300}]
+    e = _engine([parse(c) for c in cfg], [])
+    q = {"600519": {"price": 1501, "name": "贵州茅台"}}
+    assert len(e.feed(q)) == 1
+    assert len(e.feed(q)) == 0          # 冷却期内不重复
+
+
+def test_pct_and_vol_spike():
+    cfg = [
+        {"type": "pct_below", "code": "*", "value": -2},
+        {"type": "vol_spike", "code": "000001", "window": 4, "ratio": 3},
+    ]
+    e = _engine([parse(c) for c in cfg], [])
+    base = {"000001": {"price": 10, "pct": -3, "vol": 100}}
+    assert len(e.feed(base)) == 1                       # 跌幅触发
+    for _ in range(3):                                  # 平量若干根
+        e.feed({"000001": {"price": 10, "pct": 0, "vol": 100}})
+    alerts = e.feed({"000001": {"price": 10, "pct": 0, "vol": 500}})
+    assert any("成交量" in a for a in alerts)           # 5 倍放量触发
+
+
+def test_change_dir_resets_baseline():
+    e = _engine([parse({"type": "change_dir", "code": "600519", "value": 10, "cooldown": 0})], [])
+    e.feed({"600519": {"price": 9}})
+    assert len(e.feed({"600519": {"price": 11}})) == 1   # 上穿
+    assert len(e.feed({"600519": {"price": 11}})) == 0   # 不在同一方向重复
+    assert len(e.feed({"600519": {"price": 9}})) == 1    # 回落再次下穿
+
+
+def test_wildcard_and_name_in_message():
+    e = _engine([parse({"type": "price_below", "code": "*", "value": 5})], [])
+    alerts = e.feed({"000001": {"price": 4.9, "name": "平安银行"}})
+    assert alerts and "平安银行" in alerts[0]
+
+
+def test_config_load():
+    cfg = {
+        "notifiers": {"console": {}},
+        "rules": [{"type": "price_above", "code": "600519", "value": 1}],
+    }
+    e = AlertEngine.from_config(cfg)
+    assert len(e.rules) == 1 and "console" in e.notifiers
+
+
+def test_unknown_rule_field_rejected():
+    with pytest.raises(ValueError):
+        parse({"type": "price_above", "code": "600519", "value": 1, "foo": 2})