@weaxs/openclaw-stock-analysis 0.2.8 → 0.2.10

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (46) hide show
  1. package/README.md +3 -1
  2. package/dist/index.js +154 -20
  3. package/index.ts +126 -6
  4. package/openclaw.plugin.json +4 -1
  5. package/package.json +4 -3
  6. package/scripts/profile-offline.py +127 -0
  7. package/scripts/profile-workflows.py +87 -0
  8. package/scripts/setup-python.mjs +11 -4
  9. package/scripts/venv-python.mjs +7 -0
  10. package/skills/event-driven/SKILL.md +2 -0
  11. package/skills/growth-quality/SKILL.md +2 -0
  12. package/skills/market-review/SKILL.md +1 -1
  13. package/skills/sector-rotation/SKILL.md +57 -0
  14. package/skills/sector-rotation/references/computation.md +83 -0
  15. package/skills/sector-rotation/references/strategy_rules.md +62 -0
  16. package/skills/sector-rotation/scripts/rotation_metrics.py +11 -0
  17. package/skills/stock-analysis/SKILL.md +15 -4
  18. package/skills/stock-analysis/references/host-workflows.md +45 -0
  19. package/skills/stock-analysis/references/stateless-analysis.md +16 -2
  20. package/skills/stock-screener/SKILL.md +6 -0
  21. package/skills/strategy-backtest/SKILL.md +10 -3
  22. package/tools/_cache.py +93 -21
  23. package/tools/_subproc.py +43 -8
  24. package/tools/anomaly_detect.py +11 -6
  25. package/tools/backtest.py +225 -70
  26. package/tools/capabilities.py +3 -0
  27. package/tools/cli_budgets.json +2 -1
  28. package/tools/gather.py +63 -8
  29. package/tools/market_regime.py +26 -15
  30. package/tools/market_review.py +2 -2
  31. package/tools/position_context.py +5 -2
  32. package/tools/report_renderer.py +4 -0
  33. package/tools/requirements.txt +2 -2
  34. package/tools/risk_screening.py +40 -15
  35. package/tools/screener.py +44 -18
  36. package/tools/search_intel.py +94 -32
  37. package/tools/sec_holders.py +27 -25
  38. package/tools/sector_rotation.py +466 -0
  39. package/tools/signal_tracker.py +100 -49
  40. package/tools/stateless_analysis.py +600 -33
  41. package/tools/stock_data.py +111 -26
  42. package/tools/technical.py +42 -40
  43. package/tools/trading_calendar.py +26 -21
  44. package/tools/value_data.py +6 -2
  45. package/tools/watchlist.py +24 -15
  46. package/tools/watchlist_context.py +63 -49
@@ -16,7 +16,7 @@ from pathlib import Path
16
16
  from zoneinfo import ZoneInfo
17
17
 
18
18
  sys.path.insert(0, str(Path(__file__).resolve().parent))
19
- from _cache import DAY, WEEK, cached # noqa: E402
19
+ from _cache import DAY, WEEK, atomic_write, cached # noqa: E402
20
20
  from _subproc import run_tool, scrub_secrets, socket_timeout, utf8_stdio # noqa: E402
21
21
 
22
22
  # Explicitly exchange-prefixed A-share codes (sh600519 / sz000858 / bj920001).
@@ -71,11 +71,12 @@ def normalize_stock_code(symbol: str) -> dict:
71
71
 
72
72
 
73
73
  def calc_limit_price(pre_close: float, ratio: float, direction: str = "up") -> float:
74
- """Calculate limit-up or limit-down price with exchange rounding (round-half-up: floor(x*100+0.5)/100)."""
75
- import numpy as np
74
+ """Calculate limit-up or limit-down price with decimal round-half-up."""
75
+ from decimal import ROUND_HALF_UP, Decimal
76
76
 
77
77
  sign = 1 if direction == "up" else -1
78
- return np.floor(pre_close * (1 + sign * ratio) * 100 + 0.5) / 100.0
78
+ price = Decimal(str(pre_close)) * (1 + sign * Decimal(str(ratio)))
79
+ return float(price.quantize(Decimal("0.01"), rounding=ROUND_HALF_UP))
79
80
 
80
81
 
81
82
  # Chains that get sticky ordering: the CN free-source chains, where failures are
@@ -145,8 +146,37 @@ def _akshare_retry(fn, *args, retries=2, delay=1, **kwargs):
145
146
  for attempt in range(retries + 1):
146
147
  try:
147
148
  return fn(*args, **kwargs)
148
- except Exception:
149
- if attempt == retries:
149
+ except Exception as exc:
150
+ from requests.exceptions import ChunkedEncodingError, HTTPError, Timeout
151
+ from requests.exceptions import ConnectionError as RequestsConnectionError
152
+ from requests.exceptions import JSONDecodeError as RequestsJSONDecodeError
153
+
154
+ transport_errors = (
155
+ ConnectionError,
156
+ TimeoutError,
157
+ RequestsConnectionError,
158
+ Timeout,
159
+ ChunkedEncodingError,
160
+ RequestsJSONDecodeError,
161
+ json.JSONDecodeError,
162
+ )
163
+ http_errors = (HTTPError,)
164
+ if type(exc).__module__.startswith("curl_cffi."):
165
+ from curl_cffi.requests.exceptions import ConnectionError as CurlConnectionError
166
+ from curl_cffi.requests.exceptions import HTTPError as CurlHTTPError
167
+ from curl_cffi.requests.exceptions import Timeout as CurlTimeout
168
+
169
+ transport_errors += (CurlConnectionError, CurlTimeout)
170
+ http_errors += (CurlHTTPError,)
171
+ transient = isinstance(exc, transport_errors)
172
+ if isinstance(exc, http_errors):
173
+ status = getattr(exc.response, "status_code", None)
174
+ transient = status in (408, 429) or isinstance(status, int) and 500 <= status < 600
175
+ if type(exc).__module__.startswith("yfinance."):
176
+ from yfinance import exceptions as yf_errors
177
+
178
+ transient = isinstance(exc, getattr(yf_errors, "YFRateLimitError", ()))
179
+ if not transient or attempt == retries:
150
180
  raise
151
181
  time.sleep(delay)
152
182
 
@@ -260,6 +290,7 @@ def _quote_efinance(symbol: str) -> dict:
260
290
  with contextlib.suppress(ValueError, TypeError):
261
291
  val = float(val) if en_key not in ("symbol", "name") else val
262
292
  result[en_key] = val
293
+ result["pe_basis"] = "dynamic"
263
294
  return _clean_row(result)
264
295
 
265
296
 
@@ -328,6 +359,7 @@ def _tencent_payload_row(f: list, symbol: str) -> dict:
328
359
  "prev_close": _parse_float(f[4]),
329
360
  "market_cap": market_cap * 1e8 if market_cap is not None else None,
330
361
  "pe": _parse_float(f[39]),
362
+ "pe_basis": "unknown",
331
363
  # Tencent's HK map labels f[47] DY and f[49] WK52_Low; neither is an A-share ratio.
332
364
  "pb": None if hk else _parse_float(f[46]),
333
365
  "turnover_rate": None if hk else _parse_float(f[38]),
@@ -626,6 +658,9 @@ def _kline_tushare(symbol: str, period: str, count: int) -> list:
626
658
  df = df.rename(columns={f"{c}_qfq": c for c in prices})
627
659
  if not df[prices].abs().lt(float("inf")).all().all():
628
660
  raise ValueError("tushare returned invalid adjusted prices")
661
+ df["trade_date"] = pd.to_datetime(df["trade_date"], format="%Y%m%d")
662
+ if "amount" in df.columns:
663
+ df["amount"] = pd.to_numeric(df["amount"], errors="coerce") * 1000
629
664
  col_map = {"trade_date": "date", "vol": "volume", "amount": "turnover", "pct_chg": "change_pct"}
630
665
  df = df.rename(columns=col_map)
631
666
  df = df.sort_values("date")
@@ -798,7 +833,7 @@ def _kline_finnhub(symbol: str, period: str, count: int) -> list:
798
833
  rows.append(
799
834
  _clean_row(
800
835
  {
801
- "date": datetime.fromtimestamp(candles["t"][i]).strftime("%Y-%m-%d"),
836
+ "date": datetime.fromtimestamp(candles["t"][i], timezone.utc).strftime("%Y-%m-%d"),
802
837
  "open": candles["o"][i],
803
838
  "high": candles["h"][i],
804
839
  "low": candles["l"][i],
@@ -1130,6 +1165,7 @@ def _quote_akshare(symbol: str) -> dict:
1130
1165
  "prev_close": r.get("昨收"),
1131
1166
  "market_cap": r.get("总市值"),
1132
1167
  "pe": r.get("市盈率-动态"),
1168
+ "pe_basis": "dynamic",
1133
1169
  "pb": r.get("市净率"),
1134
1170
  "turnover_rate": r.get("换手率"),
1135
1171
  "volume_ratio": r.get("量比"),
@@ -1161,6 +1197,7 @@ def _quote_yfinance(symbol: str) -> dict:
1161
1197
  "prev_close": info.get("regularMarketPreviousClose"),
1162
1198
  "market_cap": info.get("marketCap"),
1163
1199
  "pe": info.get("trailingPE"),
1200
+ "pe_basis": "trailing",
1164
1201
  "pb": info.get("priceToBook"),
1165
1202
  }
1166
1203
  )
@@ -1370,8 +1407,10 @@ def cmd_capital_flow(args):
1370
1407
  summary = {"symbol": args.symbol}
1371
1408
  for days, label in [(5, "5d"), (10, "10d"), (20, "20d")]:
1372
1409
  tail = df_all.tail(days)
1373
- summary[f"main_net_{label}"] = _sanitize(tail.get("main_net_inflow", pd.Series()).sum())
1374
- summary[f"super_large_net_{label}"] = _sanitize(tail.get("super_large_net", pd.Series()).sum())
1410
+ summary[f"main_net_{label}"] = _sanitize(tail.get("main_net_inflow", pd.Series()).sum(min_count=days))
1411
+ summary[f"super_large_net_{label}"] = _sanitize(
1412
+ tail.get("super_large_net", pd.Series()).sum(min_count=days)
1413
+ )
1375
1414
  recent = df_all["main_net_inflow"].tail(5) if "main_net_inflow" in df_all.columns else pd.Series()
1376
1415
  if len(recent) >= 3:
1377
1416
  pos_count = (recent > 0).sum()
@@ -1566,6 +1605,13 @@ def financials_yf(symbol: str, periods: int = 1, annual: bool = False) -> dict:
1566
1605
  "roe": info.get("returnOnEquity"),
1567
1606
  "debt_to_equity": info.get("debtToEquity"),
1568
1607
  "dividend_yield": info.get("dividendYield"),
1608
+ "snapshot_units": {
1609
+ "pe": "trailing",
1610
+ "roe": "ratio",
1611
+ "profit_margin": "ratio",
1612
+ "debt_to_equity": "percent",
1613
+ "dividend_yield": "ratio",
1614
+ },
1569
1615
  }
1570
1616
  )
1571
1617
  if warnings:
@@ -1703,7 +1749,8 @@ def _snapshot_efinance() -> list:
1703
1749
  "昨收": "prev_close",
1704
1750
  }
1705
1751
  df = df.rename(columns=col_map)
1706
- keep = [c for c in col_map.values() if c in df.columns]
1752
+ df["pe_basis"] = "dynamic"
1753
+ keep = [c for c in col_map.values() if c in df.columns] + ["pe_basis"]
1707
1754
  return [_clean_row(r) for r in df[keep].to_dict("records")]
1708
1755
 
1709
1756
 
@@ -1815,7 +1862,8 @@ def _snapshot_akshare() -> list:
1815
1862
  "60日涨跌幅": "change_pct_60d",
1816
1863
  }
1817
1864
  df = df.rename(columns=col_map)
1818
- keep = [c for c in col_map.values() if c in df.columns]
1865
+ df["pe_basis"] = "dynamic"
1866
+ keep = [c for c in col_map.values() if c in df.columns] + ["pe_basis"]
1819
1867
  df = df[keep]
1820
1868
  return [_clean_row(r) for r in df.to_dict("records")]
1821
1869
 
@@ -1837,7 +1885,11 @@ def snapshot_hk() -> list:
1837
1885
  "总市值": "market_cap",
1838
1886
  }
1839
1887
  df = df.rename(columns=col_map)
1840
- keep = [c for c in col_map.values() if c in df.columns]
1888
+ if "symbol" in df.columns:
1889
+ symbols = df["symbol"].astype(str).str.upper()
1890
+ df["symbol"] = symbols.map(lambda s: s if s.endswith(".HK") else f"{s}.HK")
1891
+ df["pe_basis"] = "unknown"
1892
+ keep = [c for c in col_map.values() if c in df.columns] + ["pe_basis"]
1841
1893
  return [_clean_row(r) for r in df[keep].to_dict("records")]
1842
1894
  except Exception as e:
1843
1895
  # summary only (type + short message) — never a stack trace or response body
@@ -1860,7 +1912,11 @@ def snapshot_us() -> list:
1860
1912
  "总市值": "market_cap",
1861
1913
  }
1862
1914
  df = df.rename(columns=col_map)
1863
- keep = [c for c in col_map.values() if c in df.columns]
1915
+ if "symbol" in df.columns:
1916
+ symbols = df["symbol"].astype(str).str.upper().str.replace(r"\.(?:OQ|N|A)$", "", regex=True)
1917
+ df["symbol"] = symbols.str.replace(".", "-", regex=False)
1918
+ df["pe_basis"] = "unknown"
1919
+ keep = [c for c in col_map.values() if c in df.columns] + ["pe_basis"]
1864
1920
  return [_clean_row(r) for r in df[keep].to_dict("records")]
1865
1921
  except Exception as e:
1866
1922
  return [{"error": f"US snapshot unavailable: {type(e).__name__}: {scrub_secrets(str(e))}"[:200]}]
@@ -2154,8 +2210,8 @@ def _disk_cache_get(key: str):
2154
2210
 
2155
2211
  def _disk_cache_set(key: str, data):
2156
2212
  try:
2157
- _DATA_CACHE_DIR.mkdir(parents=True, exist_ok=True)
2158
- (_DATA_CACHE_DIR / f"{key}.json").write_text(json.dumps(data, ensure_ascii=False), encoding="utf-8")
2213
+ with atomic_write(_DATA_CACHE_DIR / f"{key}.json") as stream:
2214
+ json.dump(data, stream, ensure_ascii=False, allow_nan=False)
2159
2215
  except Exception:
2160
2216
  pass
2161
2217
 
@@ -2711,6 +2767,9 @@ def compute_market_stats(data: list) -> dict:
2711
2767
  code_info = normalize_stock_code(sym)
2712
2768
  lp = code_info.get("limit_pct")
2713
2769
  if lp:
2770
+ name = s.get("name", "")
2771
+ if name and "ST" in name.upper() and code_info["board"] == "main":
2772
+ lp = 0.05
2714
2773
  limit_up = calc_limit_price(float(prev), lp, "up")
2715
2774
  limit_down = calc_limit_price(float(prev), lp, "down")
2716
2775
  if abs(float(price) - limit_up) < 0.01:
@@ -2775,6 +2834,7 @@ def cmd_fundamental_context(args):
2775
2834
  quote = quote_a(args.symbol)
2776
2835
  result["valuation"] = {
2777
2836
  "pe": quote.get("pe"),
2837
+ "pe_basis": quote.get("pe_basis"),
2778
2838
  "pb": quote.get("pb"),
2779
2839
  "market_cap": quote.get("market_cap"),
2780
2840
  }
@@ -3149,15 +3209,12 @@ def cmd_margin_trading(args):
3149
3209
  try:
3150
3210
  import concurrent.futures
3151
3211
 
3152
- import akshare as ak
3153
3212
  from trading_calendar import prev_trading_days
3154
3213
 
3155
3214
  if exchange == "sh":
3156
3215
  col_map = _MARGIN_DETAIL_SSE_COLS
3157
- detail_fn = ak.stock_margin_detail_sse
3158
3216
  elif exchange == "sz":
3159
3217
  col_map = _MARGIN_DETAIL_SZSE_COLS
3160
- detail_fn = ak.stock_margin_detail_szse
3161
3218
  else:
3162
3219
  return {"error": f"margin_trading: {args.symbol} is a BSE code; only SSE/SZSE margin detail is available"}
3163
3220
  trade_days = prev_trading_days("CN", days)
@@ -3165,17 +3222,45 @@ def cmd_margin_trading(args):
3165
3222
  return {"error": "margin_trading: trading calendar unavailable"}
3166
3223
 
3167
3224
  def fetch_day(date_iso):
3168
- """Fetch + filter one day's all-stock table, fault-isolated per day."""
3169
- try:
3225
+ """Cache the published exchange-day table, then select this stock."""
3226
+
3227
+ def fetch_table():
3228
+ import akshare as ak
3229
+
3230
+ detail_fn = ak.stock_margin_detail_sse if exchange == "sh" else ak.stock_margin_detail_szse
3170
3231
  df = _akshare_retry(detail_fn, date=date_iso.replace("-", ""))
3232
+ if df is None or df.empty:
3233
+ return []
3234
+ df = df.rename(columns=col_map)
3235
+ required = {"symbol", "margin_buy", "margin_balance"}
3236
+ if not required.issubset(df.columns):
3237
+ raise ValueError("margin detail table missing financial columns")
3238
+ df = df[df["symbol"].astype(str).str.fullmatch(r"\d{6}")]
3239
+ if "信用交易日期" in df.columns:
3240
+ source_days = df["信用交易日期"].map(lambda v: re.sub(r"\D", "", str(v))[:8])
3241
+ if not (source_days == date_iso.replace("-", "")).all():
3242
+ raise ValueError("margin detail table date does not match requested day")
3243
+ keep = [c for c in col_map.values() if c in df.columns]
3244
+ table = [{"date": date_iso, **_clean_row(r)} for r in df[keep].to_dict("records")]
3245
+ for row in table:
3246
+ values = [_parse_float(row.get(field)) for field in ("margin_buy", "margin_balance")]
3247
+ if any(value is None or not math.isfinite(value) for value in values):
3248
+ # Keep usable stocks without persisting an incomplete exchange snapshot.
3249
+ return {"rows": table, "warnings": ["incomplete financial values"]}
3250
+ return table
3251
+
3252
+ try:
3253
+ today = datetime.now(ZoneInfo("Asia/Shanghai")).date().isoformat()
3254
+ table = (
3255
+ cached(f"margin-table:{exchange}:{date_iso}", DAY, fetch_table)
3256
+ if date_iso < today
3257
+ else fetch_table()
3258
+ )
3259
+ if isinstance(table, dict):
3260
+ table = table["rows"]
3261
+ return date_iso, [r for r in table if str(r.get("symbol")) == bare], None
3171
3262
  except Exception as e:
3172
3263
  return date_iso, [], f"{date_iso}: {type(e).__name__}: {scrub_secrets(str(e))}"
3173
- if df is None or df.empty:
3174
- return date_iso, [], None
3175
- df = df.rename(columns=col_map)
3176
- keep = [c for c in col_map.values() if c in df.columns]
3177
- hit = df[df["symbol"].astype(str) == bare] if "symbol" in df.columns else df.iloc[0:0]
3178
- return date_iso, [{"date": date_iso, **_clean_row(r)} for r in hit[keep].to_dict("records")], None
3179
3264
 
3180
3265
  rows = []
3181
3266
  failures = []
@@ -231,12 +231,13 @@ def macd_series(close: pd.Series) -> tuple:
231
231
 
232
232
 
233
233
  def rsi_series(close: pd.Series, period: int) -> pd.Series:
234
- """Full RSI series (SMA of gains/losses over `period`). NaN where there were no losses."""
234
+ """Full RSI series (SMA of gains/losses over `period`)."""
235
235
  delta = close.diff()
236
236
  gain = delta.where(delta > 0, 0.0).rolling(period).mean()
237
237
  loss = (-delta.where(delta < 0, 0.0)).rolling(period).mean()
238
238
  rs = gain / loss.replace(0, np.nan)
239
- return 100 - 100 / (1 + rs)
239
+ rsi = (100 - 100 / (1 + rs)).mask((loss == 0) & (gain > 0), 100)
240
+ return rsi.mask((loss == 0) & (gain == 0), 50)
240
241
 
241
242
 
242
243
  def bollinger_series(close: pd.Series, period: int = 20) -> tuple:
@@ -257,22 +258,23 @@ def kdj_series(high: pd.Series, low: pd.Series, close: pd.Series) -> tuple:
257
258
  return k, d, 3 * k - 2 * d
258
259
 
259
260
 
261
+ def _ma_arrangement(values):
262
+ valid = [value for value in values if value is not None]
263
+ if len(valid) < 3:
264
+ return "insufficient_data"
265
+ if all(a > b for a, b in zip(valid, valid[1:])):
266
+ return "bullish"
267
+ if all(a < b for a, b in zip(valid, valid[1:])):
268
+ return "bearish"
269
+ return "mixed"
270
+
271
+
260
272
  def calc_ma(close: pd.Series) -> dict:
261
273
  mas = {}
262
274
  for n in [5, 10, 20, 60]:
263
275
  ma = ma_series(close, n)
264
276
  mas[f"ma{n}"] = round(float(ma.iloc[-1]), 2) if len(ma.dropna()) > 0 else None
265
- vals = [mas.get(f"ma{n}") for n in [5, 10, 20, 60]]
266
- valid = [v for v in vals if v is not None]
267
- if len(valid) >= 3:
268
- if valid == sorted(valid, reverse=True):
269
- mas["ma_arrangement"] = "bullish"
270
- elif valid == sorted(valid):
271
- mas["ma_arrangement"] = "bearish"
272
- else:
273
- mas["ma_arrangement"] = "mixed"
274
- else:
275
- mas["ma_arrangement"] = "insufficient_data"
277
+ mas["ma_arrangement"] = _ma_arrangement([mas.get(f"ma{n}") for n in [5, 10, 20, 60]])
276
278
  return mas
277
279
 
278
280
 
@@ -353,11 +355,13 @@ def calc_kdj(high: pd.Series, low: pd.Series, close: pd.Series) -> dict:
353
355
 
354
356
 
355
357
  def calc_volume(volume: pd.Series) -> dict:
356
- vol_ma5 = float(volume.rolling(5).mean().iloc[-1])
357
- vol_ma10 = float(volume.rolling(10).mean().iloc[-1])
358
+ vol_ma5 = float(volume.rolling(5).mean().shift(1).iloc[-1])
359
+ vol_ma10 = float(volume.rolling(10).mean().shift(1).iloc[-1])
358
360
  cur = float(volume.iloc[-1])
359
- ratio = cur / vol_ma5 if vol_ma5 > 0 else 0
360
- if ratio > 3:
361
+ ratio = cur / vol_ma5 if np.isfinite(vol_ma5) and vol_ma5 > 0 else None
362
+ if ratio is None:
363
+ signal = "insufficient_data"
364
+ elif ratio > 3:
361
365
  signal = "extremely_heavy"
362
366
  elif ratio > 2:
363
367
  signal = "heavy"
@@ -367,9 +371,9 @@ def calc_volume(volume: pd.Series) -> dict:
367
371
  signal = "normal"
368
372
  return {
369
373
  "current": int(cur),
370
- "vol_ma5": round(vol_ma5, 0),
371
- "vol_ma10": round(vol_ma10, 0),
372
- "volume_ratio": round(ratio, 2),
374
+ "vol_ma5": round(vol_ma5, 0) if np.isfinite(vol_ma5) else None,
375
+ "vol_ma10": round(vol_ma10, 0) if np.isfinite(vol_ma10) else None,
376
+ "volume_ratio": round(ratio, 2) if ratio is not None else None,
373
377
  "signal": signal,
374
378
  }
375
379
 
@@ -411,9 +415,11 @@ def calc_bias(close: pd.Series, ma_data: dict) -> dict:
411
415
 
412
416
  def calc_volume_direction(close: pd.Series, volume: pd.Series) -> str:
413
417
  if len(close) < 2:
414
- return "normal"
418
+ return "insufficient_data"
415
419
  price_up = float(close.iloc[-1]) > float(close.iloc[-2])
416
- vol_ma5 = float(volume.rolling(5).mean().iloc[-1])
420
+ vol_ma5 = float(volume.rolling(5).mean().shift(1).iloc[-1])
421
+ if not np.isfinite(vol_ma5) or vol_ma5 <= 0:
422
+ return "insufficient_data"
417
423
  cur_vol = float(volume.iloc[-1])
418
424
  heavy = cur_vol > vol_ma5 * 1.5
419
425
  shrink = cur_vol < vol_ma5 * 0.7
@@ -461,6 +467,8 @@ def generate_signal_score(
461
467
  # --- Trend score (30 points) ---
462
468
  arrangement = ma_data.get("ma_arrangement", "mixed")
463
469
  overall = trend_data.get("overall", "neutral")
470
+ ma_values = [value for value in (ma_data.get(f"ma{n}") for n in (5, 10, 20, 60)) if value is not None]
471
+ flat_ma = len(ma_values) >= 3 and len(set(ma_values)) == 1
464
472
  short_t = trend_data.get("short_term", "neutral")
465
473
  medium_t = trend_data.get("medium_term", "neutral")
466
474
  long_t = trend_data.get("long_term", "neutral")
@@ -479,15 +487,15 @@ def generate_signal_score(
479
487
  reasons.append("短中期趋势偏多")
480
488
  elif arrangement == "mixed":
481
489
  trend_score = 12
482
- elif bear_count >= 2:
483
- trend_score = 8
484
- risks.append("中长期趋势偏空")
485
490
  elif arrangement == "bearish" and bear_count == 3:
486
491
  trend_score = 0
487
492
  risks.append("均线空头排列,趋势极弱")
488
493
  elif arrangement == "bearish":
489
494
  trend_score = 4
490
495
  risks.append("均线空头排列,不宜做多")
496
+ elif bear_count >= 2:
497
+ trend_score = 8
498
+ risks.append("中长期趋势偏空")
491
499
  else:
492
500
  trend_score = 12
493
501
  score += trend_score
@@ -604,9 +612,9 @@ def generate_signal_score(
604
612
  score += rsi_score
605
613
 
606
614
  # --- Buy signal classification ---
607
- if score >= 75 and overall == "bullish":
615
+ if score >= 75 and overall == "bullish" and not flat_ma:
608
616
  buy_signal = "STRONG_BUY"
609
- elif score >= 60 and overall in ("bullish", "neutral"):
617
+ elif score >= 60 and overall in ("bullish", "neutral") and not flat_ma:
610
618
  buy_signal = "BUY"
611
619
  elif score >= 45:
612
620
  buy_signal = "HOLD"
@@ -640,7 +648,9 @@ def generate_signals(
640
648
  signals.append("MACD空头排列(DIF < DEA)")
641
649
 
642
650
  rsi_sig = rsi_data.get("signal", "")
643
- if "overbought" in rsi_sig:
651
+ if rsi_sig == "insufficient_data":
652
+ signals.append("RSI数据不足,无法判断超买超卖")
653
+ elif "overbought" in rsi_sig:
644
654
  signals.append(f"RSI超买区间(RSI6={rsi_data.get('rsi6')}),注意回调风险")
645
655
  elif "oversold" in rsi_sig:
646
656
  signals.append(f"RSI超卖区间(RSI6={rsi_data.get('rsi6')}),关注反弹机会")
@@ -663,7 +673,9 @@ def generate_signals(
663
673
  signals.append(f"KDJ超卖(J={kdj_data['j']}),关注低位机会")
664
674
 
665
675
  vs = vol_data.get("signal", "")
666
- if "heavy" in vs:
676
+ if vs == "insufficient_data":
677
+ signals.append("成交量数据不足,无法判断放量或缩量")
678
+ elif "heavy" in vs:
667
679
  signals.append(f"成交量放大(量比={vol_data['volume_ratio']}),市场关注度高")
668
680
  elif vs == "light":
669
681
  signals.append(f"缩量交易(量比={vol_data['volume_ratio']}),市场观望情绪浓")
@@ -712,17 +724,7 @@ def calculate_ma_standalone(
712
724
  else:
713
725
  bias[f"bias_ma{n}"] = None
714
726
 
715
- valid_mas = [(n, mas[f"ma{n}"]) for n in periods if mas.get(f"ma{n}") is not None]
716
- if len(valid_mas) >= 3:
717
- vals = [v for _, v in valid_mas]
718
- if vals == sorted(vals, reverse=True):
719
- alignment = "bullish"
720
- elif vals == sorted(vals):
721
- alignment = "bearish"
722
- else:
723
- alignment = "mixed"
724
- else:
725
- alignment = "insufficient_data"
727
+ alignment = _ma_arrangement([mas[f"ma{n}"] for n in periods if mas.get(f"ma{n}") is not None])
726
728
 
727
729
  cross_signals = []
728
730
  for fast, slow in [(5, 10), (5, 20), (10, 20), (20, 60)]:
@@ -4,10 +4,12 @@
4
4
  import argparse
5
5
  import json
6
6
  import sys
7
- from datetime import datetime, timedelta
7
+ from datetime import date, datetime, timedelta
8
8
  from pathlib import Path
9
+ from zoneinfo import ZoneInfo
9
10
 
10
11
  sys.path.insert(0, str(Path(__file__).resolve().parent))
12
+ from _cache import DAY, cached
11
13
  from _subproc import socket_timeout, utf8_stdio
12
14
 
13
15
  # Supported markets. argparse uses type=str.upper so lowercase input is accepted too.
@@ -18,10 +20,8 @@ def _is_weekend(d: datetime) -> bool:
18
20
  return d.weekday() >= 5
19
21
 
20
22
 
21
- # Process-local memo: one CLI call = one process, and the CN path
22
- # (is_trading_day / the prev|next_trading_days walk-back) would otherwise
23
- # re-fetch the sina calendar per step — up to ~5 fetches for one stale check.
24
- _CN_TRADE_DATES: dict[int, set[str]] = {}
23
+ # Year views reuse a daily cached full calendar across CLI processes.
24
+ _CN_TRADE_DATES: dict[tuple[str, int], set[str]] = {}
25
25
 
26
26
 
27
27
  def cn_trade_dates(year: int) -> set[str]:
@@ -30,25 +30,30 @@ def cn_trade_dates(year: int) -> set[str]:
30
30
  Empty set when the calendar can't be fetched (failures are not cached — the
31
31
  next call retries); callers treat empty as "weekends are the only non-trading
32
32
  days"."""
33
- cached = _CN_TRADE_DATES.get(year)
34
- if cached is not None:
35
- return cached
33
+ stamp = datetime.now(ZoneInfo("Asia/Shanghai")).date().isoformat()
34
+ memo = _CN_TRADE_DATES.get((stamp, year))
35
+ if memo is not None:
36
+ return memo
36
37
  try:
37
- import akshare as ak
38
-
39
- with socket_timeout(25):
40
- df = ak.tool_trade_date_hist_sina()
41
- dates = set()
42
- for _, row in df.iterrows():
43
- val = row.iloc[0]
44
- d = str(val)[:10]
45
- if d.startswith(str(year)):
46
- dates.add(d)
38
+
39
+ def fetch():
40
+ import akshare as ak
41
+
42
+ with socket_timeout(25):
43
+ df = ak.tool_trade_date_hist_sina()
44
+ if df is None or df.empty:
45
+ return []
46
+ return sorted({date.fromisoformat(str(value)[:10]).isoformat() for value in df.iloc[:, 0]})
47
+
48
+ dates = cached(f"cn-trade-calendar:v1:{stamp}", DAY, fetch)
49
+ grouped = {}
50
+ for day in dates:
51
+ parsed = date.fromisoformat(day).isoformat()
52
+ grouped.setdefault((stamp, int(parsed[:4])), set()).add(parsed)
53
+ _CN_TRADE_DATES.update(grouped)
47
54
  except Exception:
48
55
  return set()
49
- if dates:
50
- _CN_TRADE_DATES[year] = dates
51
- return dates
56
+ return _CN_TRADE_DATES.get((stamp, year), set())
52
57
 
53
58
 
54
59
  def _is_exchange_trading_day(exchange_code: str, date_str: str) -> bool:
@@ -1146,7 +1146,11 @@ def etf_basket(symbol):
1146
1146
 
1147
1147
  def fund_disclosed_equities(symbol):
1148
1148
  import requests
1149
- from akshare.utils import demjson
1149
+
1150
+ try:
1151
+ from akshare.utils.jsonish import loads as decode
1152
+ except ImportError:
1153
+ from akshare.utils.demjson import decode
1150
1154
  from bs4 import BeautifulSoup
1151
1155
 
1152
1156
  limit = 10000
@@ -1160,7 +1164,7 @@ def fund_disclosed_equities(symbol):
1160
1164
  )
1161
1165
  response.raise_for_status()
1162
1166
  text = response.text
1163
- payload = demjson.decode(text[text.index("{") : text.rindex("}") + 1])
1167
+ payload = decode(text[text.index("{") : text.rindex("}") + 1])
1164
1168
  soup = BeautifulSoup(payload["content"], "lxml")
1165
1169
  for block in soup.select("div.boxitem"):
1166
1170
  heading = block.find("h4")
@@ -6,11 +6,12 @@ import json
6
6
  import math
7
7
  import sys
8
8
  from concurrent.futures import ThreadPoolExecutor, as_completed
9
+ from contextlib import nullcontext
9
10
  from pathlib import Path
10
11
 
11
12
  sys.path.insert(0, str(Path(__file__).resolve().parent))
12
13
  from _subproc import utf8_stdio
13
- from gather import gather_analysis
14
+ from gather import Collection, gather_analysis
14
15
 
15
16
 
16
17
  def has_analysis_data(data) -> bool:
@@ -28,27 +29,35 @@ def has_analysis_data(data) -> bool:
28
29
  return False
29
30
 
30
31
 
31
- def analyze_watchlist(symbols: list[str], workers: int = 3) -> dict:
32
+ def analyze_watchlist(symbols: list[str], workers: int = 3, *, session=None, on_result=None) -> dict:
32
33
  if isinstance(workers, bool) or not isinstance(workers, int) or workers <= 0:
33
34
  return {"error": "workers must be a positive integer"}
34
35
  results = {}
35
36
  success = 0
36
37
  failed = 0
37
38
 
38
- with ThreadPoolExecutor(max_workers=workers) as pool:
39
- futures = {pool.submit(gather_analysis, sym): sym for sym in symbols}
40
- for future in as_completed(futures):
41
- sym = futures[future]
42
- try:
43
- data = future.result()
44
- results[sym] = data
45
- if has_analysis_data(data):
46
- success += 1
47
- else:
39
+ with Collection() if session is None else nullcontext(session) as session:
40
+
41
+ def analyze(sym):
42
+ data = gather_analysis(sym, session=session)
43
+ if on_result is not None:
44
+ on_result(sym, data)
45
+ return data
46
+
47
+ with ThreadPoolExecutor(max_workers=workers) as pool:
48
+ futures = {pool.submit(analyze, sym): sym for sym in symbols}
49
+ for future in as_completed(futures):
50
+ sym = futures[future]
51
+ try:
52
+ data = future.result()
53
+ results[sym] = data
54
+ if has_analysis_data(data):
55
+ success += 1
56
+ else:
57
+ failed += 1
58
+ except Exception:
59
+ results[sym] = None
48
60
  failed += 1
49
- except Exception:
50
- results[sym] = None
51
- failed += 1
52
61
 
53
62
  return {
54
63
  "results": results,