aria-code 4.1.3__py3-none-any.whl

strategy_vault.py

@@ -0,0 +1,420 @@
+"""
+strategy_vault.py — Arthera Strategy Version Control System
+
+类似 Git，但专为量化策略设计：
+
+  /strategy save   "v1.2 加入波动率过滤"   → 保存当前策略快照
+  /strategy list                           → 列出所有版本
+  /strategy diff   v1 v2                   → 显示两版本差异
+  /strategy load   v3                      → 加载指定版本
+  /strategy review                         → AI 审查当前策略（过拟合/前视偏差）
+  /strategy bench  AAPL 2020-01-01         → 对标基准表现
+
+数据存储: ~/.arthera/strategy_vault.db (SQLite)
+"""
+
+from __future__ import annotations
+
+import difflib
+import hashlib
+import json
+import os
+import re
+import sqlite3
+import textwrap
+import time
+from dataclasses import dataclass
+from datetime import datetime
+from pathlib import Path
+from typing import Any, Dict, List, Optional, Tuple
+
+_VAULT_DIR = Path.home() / ".arthera" / "strategies"
+_VAULT_DB  = _VAULT_DIR / "vault.db"
+
+# ── Database setup ────────────────────────────────────────────────────────────
+
+def _get_db() -> sqlite3.Connection:
+    _VAULT_DIR.mkdir(parents=True, exist_ok=True)
+    conn = sqlite3.connect(str(_VAULT_DB))
+    conn.row_factory = sqlite3.Row
+    conn.executescript("""
+        CREATE TABLE IF NOT EXISTS strategies (
+            id          INTEGER PRIMARY KEY AUTOINCREMENT,
+            name        TEXT NOT NULL,
+            version_tag TEXT NOT NULL,
+            message     TEXT DEFAULT '',
+            code        TEXT NOT NULL,
+            metadata    TEXT DEFAULT '{}',
+            backtest_result TEXT DEFAULT NULL,
+            review_result   TEXT DEFAULT NULL,
+            code_hash   TEXT NOT NULL,
+            created_at  TEXT NOT NULL
+        );
+        CREATE INDEX IF NOT EXISTS idx_name ON strategies(name);
+        CREATE INDEX IF NOT EXISTS idx_hash ON strategies(code_hash);
+    """)
+    conn.commit()
+    return conn
+
+
+# ── Dataclass ─────────────────────────────────────────────────────────────────
+
+@dataclass
+class StrategyVersion:
+    id: int
+    name: str
+    version_tag: str
+    message: str
+    code: str
+    metadata: Dict[str, Any]
+    backtest_result: Optional[Dict]
+    review_result: Optional[Dict]
+    code_hash: str
+    created_at: str
+
+    @classmethod
+    def from_row(cls, row) -> "StrategyVersion":
+        meta = {}
+        try:
+            meta = json.loads(row["metadata"] or "{}")
+        except Exception:
+            pass
+        bt = None
+        try:
+            bt = json.loads(row["backtest_result"]) if row["backtest_result"] else None
+        except Exception:
+            pass
+        rv = None
+        try:
+            rv = json.loads(row["review_result"]) if row["review_result"] else None
+        except Exception:
+            pass
+        return cls(
+            id=row["id"], name=row["name"],
+            version_tag=row["version_tag"], message=row["message"],
+            code=row["code"], metadata=meta,
+            backtest_result=bt, review_result=rv,
+            code_hash=row["code_hash"], created_at=row["created_at"],
+        )
+
+    def summary_line(self) -> str:
+        tag  = f"[{self.version_tag}]"
+        ts   = self.created_at[:16]
+        bt   = ""
+        if self.backtest_result:
+            br = self.backtest_result
+            sharpe = br.get("sharpe_ratio")
+            ret    = br.get("total_return_pct")
+            if sharpe is not None:
+                bt = f"  sharpe={sharpe:.2f}  ret={ret:.1f}%"
+        reviewed = " ✓reviewed" if self.review_result else ""
+        msg = f"  {self.message[:50]}" if self.message else ""
+        return f"{self.id:4d}  {tag:12s}  {ts}{msg}{bt}{reviewed}"
+
+
+# ═══════════════════════════════════════════════════════════════════════════
+# StrategyVault
+# ═══════════════════════════════════════════════════════════════════════════
+
+class StrategyVault:
+
+    # ── Save ──────────────────────────────────────────────────────────────────
+
+    def save(
+        self,
+        code: str,
+        name: str = "strategy",
+        message: str = "",
+        metadata: Optional[Dict] = None,
+    ) -> StrategyVersion:
+        """Save a new version. Returns the saved StrategyVersion."""
+        code_hash = hashlib.sha256(code.encode()).hexdigest()[:12]
+        # Auto version tag: v{n+1}
+        with _get_db() as conn:
+            row = conn.execute(
+                "SELECT COUNT(*) as c FROM strategies WHERE name=?", (name,)
+            ).fetchone()
+            n   = row["c"] if row else 0
+            tag = f"v{n + 1}"
+            conn.execute(
+                """INSERT INTO strategies
+                   (name, version_tag, message, code, metadata, code_hash, created_at)
+                   VALUES (?, ?, ?, ?, ?, ?, ?)""",
+                (name, tag, message, code,
+                 json.dumps(metadata or {}),
+                 code_hash, datetime.now().isoformat()),
+            )
+            conn.commit()
+            last_id = conn.execute("SELECT last_insert_rowid() as lid").fetchone()["lid"]
+            row = conn.execute("SELECT * FROM strategies WHERE id=?", (last_id,)).fetchone()
+            return StrategyVersion.from_row(row)
+
+    # ── List ──────────────────────────────────────────────────────────────────
+
+    def list(self, name: str = "strategy", limit: int = 20) -> List[StrategyVersion]:
+        with _get_db() as conn:
+            rows = conn.execute(
+                "SELECT * FROM strategies WHERE name=? ORDER BY id DESC LIMIT ?",
+                (name, limit),
+            ).fetchall()
+        return [StrategyVersion.from_row(r) for r in rows]
+
+    def list_all_names(self) -> List[str]:
+        with _get_db() as conn:
+            rows = conn.execute(
+                "SELECT DISTINCT name FROM strategies ORDER BY name"
+            ).fetchall()
+        return [r["name"] for r in rows]
+
+    # ── Load ──────────────────────────────────────────────────────────────────
+
+    def load(self, name: str = "strategy",
+             version_tag: Optional[str] = None,
+             version_id: Optional[int] = None) -> Optional[StrategyVersion]:
+        with _get_db() as conn:
+            if version_id:
+                row = conn.execute(
+                    "SELECT * FROM strategies WHERE id=?", (version_id,)
+                ).fetchone()
+            elif version_tag:
+                row = conn.execute(
+                    "SELECT * FROM strategies WHERE name=? AND version_tag=? ORDER BY id DESC LIMIT 1",
+                    (name, version_tag),
+                ).fetchone()
+            else:
+                # Latest
+                row = conn.execute(
+                    "SELECT * FROM strategies WHERE name=? ORDER BY id DESC LIMIT 1",
+                    (name,),
+                ).fetchone()
+        return StrategyVersion.from_row(row) if row else None
+
+    # ── Diff ──────────────────────────────────────────────────────────────────
+
+    def diff(
+        self,
+        name: str = "strategy",
+        tag_a: Optional[str] = None,
+        tag_b: Optional[str] = None,
+    ) -> str:
+        """Unified diff between two versions (default: last two)."""
+        versions = self.list(name, limit=10)
+        if len(versions) < 2:
+            return "需要至少2个版本才能比较差异。"
+
+        if tag_a and tag_b:
+            va = next((v for v in versions if v.version_tag == tag_a), None)
+            vb = next((v for v in versions if v.version_tag == tag_b), None)
+        elif tag_a:
+            va = next((v for v in versions if v.version_tag == tag_a), None)
+            vb = versions[0]
+        else:
+            vb = versions[0]
+            va = versions[1]
+
+        if not va or not vb:
+            return f"找不到指定版本: {tag_a} / {tag_b}"
+
+        diff_lines = list(difflib.unified_diff(
+            va.code.splitlines(keepends=True),
+            vb.code.splitlines(keepends=True),
+            fromfile=f"{name} {va.version_tag} ({va.created_at[:10]})",
+            tofile=f"{name} {vb.version_tag} ({vb.created_at[:10]})",
+            n=3,
+        ))
+        if not diff_lines:
+            return f"{va.version_tag} 与 {vb.version_tag} 代码完全相同。"
+        return "".join(diff_lines)
+
+    # ── Store backtest / review results ───────────────────────────────────────
+
+    def save_backtest(self, version_id: int, result: Dict):
+        with _get_db() as conn:
+            conn.execute(
+                "UPDATE strategies SET backtest_result=? WHERE id=?",
+                (json.dumps(result), version_id),
+            )
+            conn.commit()
+
+    def save_review(self, version_id: int, review: Dict):
+        with _get_db() as conn:
+            conn.execute(
+                "UPDATE strategies SET review_result=? WHERE id=?",
+                (json.dumps(review), version_id),
+            )
+            conn.commit()
+
+
+# ═══════════════════════════════════════════════════════════════════════════
+# AI Strategy Reviewer (过拟合 / 前视偏差 / 代码质量)
+# ═══════════════════════════════════════════════════════════════════════════
+
+# 静态规则检测（不需要 AI，速度快）
+
+_LOOKAHEAD_PATTERNS = [
+    (r"\.shift\(-\d+\)",           "负向shift：使用了未来数据"),
+    (r"df\[.+\]\s*=.*\.shift\(0\)", "shift(0)可能引入当日收盘前视偏差"),
+    (r"fillna\(method=['\"]bfill",  "后向填充(bfill)引入前视偏差"),
+    (r"\.rolling\([^)]+\)\.apply\(", "rolling.apply 注意是否含未来数据"),
+    (r"target\s*=\s*.+\[.+[+>]\d", "目标变量使用未来标签"),
+]
+
+_OVERFIT_PATTERNS = [
+    (r"\.optimize\(",               "参数优化可能导致过拟合"),
+    (r"for.*param.*range\(",        "网格搜索参数可能过拟合"),
+    (r"if.*sharpe.*>[5-9]",         "Sharpe>5 在样本外几乎不可能，可能过拟合"),
+    (r"n_estimators\s*=\s*[5-9]\d{2,}", "超大集成树数量，可能过拟合"),
+    (r"in_sample.*backtest",        "样本内回测不代表真实表现"),
+]
+
+def static_review(code: str) -> Dict[str, Any]:
+    """快速静态规则审查，不调用 AI。"""
+    warnings = []
+    errors   = []
+
+    for pattern, msg in _LOOKAHEAD_PATTERNS:
+        if re.search(pattern, code, re.IGNORECASE):
+            errors.append({"type": "look_ahead_bias", "detail": msg, "pattern": pattern})
+
+    for pattern, msg in _OVERFIT_PATTERNS:
+        if re.search(pattern, code, re.IGNORECASE):
+            warnings.append({"type": "overfit_risk", "detail": msg, "pattern": pattern})
+
+    # 回测质量检查
+    quality_checks = []
+    if "train_test_split" not in code and "out_of_sample" not in code:
+        quality_checks.append("未见样本外验证（train/test split 或 walk-forward）")
+    if "commission" not in code and "slippage" not in code:
+        quality_checks.append("未设置手续费/滑点，回测结果偏乐观")
+    if "random_state" in code and "shuffle" in code:
+        quality_checks.append("时序数据注意不要随机shuffle，会引入未来信息")
+
+    risk_score = len(errors) * 3 + len(warnings) + len(quality_checks)
+    grade = "A" if risk_score == 0 else "B" if risk_score <= 2 else "C" if risk_score <= 5 else "D"
+
+    return {
+        "grade":          grade,
+        "risk_score":     risk_score,
+        "errors":         errors,      # 严重（前视偏差）
+        "warnings":       warnings,    # 中等（过拟合风险）
+        "quality_checks": quality_checks,  # 建议改进
+        "summary": (f"发现 {len(errors)} 个严重问题, "
+                    f"{len(warnings)} 个警告, "
+                    f"{len(quality_checks)} 条优化建议"),
+    }
+
+
+async def ai_review_strategy(
+    code: str,
+    backtest_result: Optional[Dict],
+    ollama_url: str,
+    model: str,
+    on_token: Optional[callable] = None,
+) -> Dict[str, Any]:
+    """AI 深度审查策略代码（结合回测结果）."""
+    import aiohttp
+
+    # 先跑静态检测
+    static = static_review(code)
+
+    bt_summary = ""
+    if backtest_result:
+        br = backtest_result
+        bt_summary = f"""
+【回测结果】
+  年化收益: {br.get('annual_return_pct', 'N/A')}%
+  夏普比率: {br.get('sharpe_ratio', 'N/A')}
+  最大回撤: {br.get('max_drawdown_pct', 'N/A')}%
+  总收益: {br.get('total_return_pct', 'N/A')}%
+  胜率: {br.get('win_rate_pct', 'N/A')}%
+  交易次数: {br.get('total_trades', 'N/A')}"""
+
+    static_issues = ""
+    if static["errors"]:
+        static_issues += "\n静态检测发现严重问题:\n"
+        for e in static["errors"]:
+            static_issues += f"  ❌ {e['detail']}\n"
+    if static["warnings"]:
+        static_issues += "静态检测警告:\n"
+        for w in static["warnings"]:
+            static_issues += f"  ⚠️  {w['detail']}\n"
+
+    code_preview = code[:2000] + ("...(截断)" if len(code) > 2000 else "")
+
+    prompt = f"""你是量化策略审查专家，请对以下策略进行深度代码审查。
+
+{bt_summary}
+{static_issues}
+
+【策略代码】
+```python
+{code_preview}
+```
+
+请从以下5个维度审查（每项给出1-2个具体问题或确认通过）：
+
+**1. 前视偏差 (Look-Ahead Bias)**
+检查是否使用了未来数据（如未来收益、未来价格、负向lag等）
+
+**2. 过拟合风险 (Overfitting)**
+参数数量是否合理？是否有样本外验证？夏普比率是否过高？
+
+**3. 执行可行性 (Execution Reality)**
+交易成本、流动性、订单执行假设是否合理？
+
+**4. 策略逻辑 (Logic Soundness)**
+策略的核心假设是否有理论支撑？边界条件是否处理？
+
+**5. 改进建议 (Improvements)**
+最重要的2-3个改进点
+
+最后给出:**综合评级** [A/B/C/D] 和 **是否推荐实盘测试** [是/否/待改进后]
+
+中文回答，专业且直接。"""
+
+    messages = [
+        {"role": "system", "content": "你是严格的量化策略审查专家，专注于发现策略缺陷，防止真金白银的亏损。"},
+        {"role": "user",   "content": prompt},
+    ]
+
+    full_text = ""
+    try:
+        async with aiohttp.ClientSession() as sess:
+            async with sess.post(
+                f"{ollama_url}/api/chat",
+                json={"model": model, "messages": messages, "stream": True,
+                      "options": {"temperature": 0.2, "num_predict": 800}},
+                timeout=aiohttp.ClientTimeout(total=120),
+            ) as resp:
+                async for line in resp.content:
+                    text = line.decode("utf-8", errors="ignore").strip()
+                    if not text:
+                        continue
+                    try:
+                        data = json.loads(text)
+                        tok = data.get("message", {}).get("content", "")
+                        if tok:
+                            full_text += tok
+                            if on_token:
+                                on_token(tok)
+                    except Exception:
+                        pass
+    except Exception as e:
+        full_text = f"[AI审查失败: {e}]"
+
+    return {
+        "static":     static,
+        "ai_review":  full_text,
+        "reviewed_at": datetime.now().isoformat(),
+    }
+
+
+# ── Singleton ─────────────────────────────────────────────────────────────────
+
+_vault: Optional[StrategyVault] = None
+
+def get_vault() -> StrategyVault:
+    global _vault
+    if _vault is None:
+        _vault = StrategyVault()
+    return _vault

ui/__init__.py

@@ -0,0 +1,100 @@
+"""Aria Code — UI package.
+
+Public surface:
+
+    from ui import console, HAS_RICH, HAS_PT        # shared console + flags
+    from ui import _EscWatcher, _esc_watcher         # ESC cancellation
+    from ui import AriaPTCompleter, ARIA_PT_STYLE    # prompt_toolkit completer
+    from ui import arrow_select, run_picker_in_thread # arrow-key picker
+    from ui import PanelInputConfig, run_panel_input  # structured input panel
+    from ui.render import render_verdict_banner, ...  # Rich output renderers
+    from ui.banner import render_full_banner, ...     # startup banner
+    from ui.render.output import print_error, print_tool_result, ...
+"""
+
+from .console import (
+    console,
+    HAS_RICH,
+    HAS_PT,
+    _SYNTAX_THEME,
+    _EscWatcher,
+    _esc_watcher,
+    _HAS_TERMIOS,
+)
+
+from .completer import AriaPTCompleter, ARIA_PT_STYLE
+
+from .picker import (
+    arrow_select,
+    run_picker_in_thread,
+    _arrow_select,
+    _run_picker_in_thread,
+)
+
+from .input_box import (
+    PanelInputConfig,
+    run_panel_input,
+    detect_terminal_theme,
+    PromptAndPlaceholderProcessor,
+    PlaceholderProcessor,
+)
+
+from .robot import (
+    RobotState,
+    set_robot_state,
+    get_robot_state,
+    get_robot_row,
+    get_robot_frame,
+    get_status_dot,
+)
+
+from .banner import (
+    privacy_status_label,
+    control_status_label,
+    ollama_status_label,
+    bottom_toolbar_parts,
+    render_compact_banner,
+    render_full_banner,
+    render_try_hints,
+)
+
+from .render.output import (
+    FINANCE_TOOL_NAMES,
+    clean_tool_error_message,
+    error_hint,
+    print_error,
+    print_tool_result,
+    print_tool_activity_group,
+    print_fallback_toast,
+    print_context_warning,
+    print_tool_blocked,
+    print_thinking_header,
+    print_done_footer,
+)
+
+__all__ = [
+    # console
+    "console", "HAS_RICH", "HAS_PT", "_SYNTAX_THEME",
+    "_EscWatcher", "_esc_watcher", "_HAS_TERMIOS",
+    # completer
+    "AriaPTCompleter", "ARIA_PT_STYLE",
+    # picker
+    "arrow_select", "run_picker_in_thread",
+    "_arrow_select", "_run_picker_in_thread",
+    # input panel
+    "PanelInputConfig", "run_panel_input",
+    "detect_terminal_theme",
+    "PromptAndPlaceholderProcessor", "PlaceholderProcessor",
+    # robot mascot
+    "RobotState", "set_robot_state", "get_robot_state",
+    "get_robot_row", "get_robot_frame", "get_status_dot",
+    # banner
+    "privacy_status_label", "control_status_label", "ollama_status_label",
+    "bottom_toolbar_parts", "render_compact_banner", "render_full_banner", "render_try_hints",
+    # output rendering
+    "FINANCE_TOOL_NAMES", "clean_tool_error_message", "error_hint",
+    "print_error", "print_tool_result",
+    "print_tool_activity_group", "print_fallback_toast",
+    "print_context_warning", "print_tool_blocked",
+    "print_thinking_header", "print_done_footer",
+]