@weaxs/openclaw-stock-analysis 0.2.11 → 0.2.13
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.
- package/README.md +11 -52
- package/dist/index.js +3423 -1546
- package/index.ts +74 -1648
- package/openclaw.plugin.json +3 -0
- package/package.json +1 -1
- package/scripts/cli-json.mjs +23 -0
- package/scripts/exec-process.mjs +89 -0
- package/scripts/setup-python.mjs +37 -4
- package/scripts/tool-calls.mjs +493 -0
- package/skills/bull-trend/SKILL.md +2 -2
- package/skills/bull-trend/references/fat-tails.md +76 -3
- package/skills/bull-trend/scripts/tail_analysis.py +92 -3
- package/skills/event-driven/SKILL.md +2 -2
- package/skills/growth-quality/SKILL.md +1 -1
- package/skills/market-review/SKILL.md +1 -0
- package/skills/stock-analysis/SKILL.md +12 -7
- package/skills/stock-analysis/references/stateless-analysis.md +9 -1
- package/skills/stock-screener/SKILL.md +1 -1
- package/skills/strategy-backtest/SKILL.md +1 -1
- package/tools/_cache.py +13 -0
- package/tools/_subproc.py +46 -0
- package/tools/alert_rules.py +2 -2
- package/tools/backtest.py +3 -3
- package/tools/capabilities.py +33 -9
- package/tools/cli_budgets.json +1 -0
- package/tools/credit_data.py +293 -0
- package/tools/diagnostics.py +3 -4
- package/tools/gather.py +2 -6
- package/tools/import_parser.py +3 -4
- package/tools/lixinger.py +119 -27
- package/tools/market_regime.py +5 -2
- package/tools/market_review.py +5 -6
- package/tools/name_resolver.py +45 -33
- package/tools/report_renderer.py +3 -3
- package/tools/requirements.txt +5 -4
- package/tools/risk_screening.py +3 -2
- package/tools/screener.py +41 -30
- package/tools/search_intel.py +79 -10
- package/tools/sec_edgar.py +268 -1
- package/tools/sec_holders.py +6 -0
- package/tools/signal_tracker.py +5 -5
- package/tools/stateless_analysis.py +132 -19
- package/tools/stock_data.py +1020 -817
- package/tools/tiaa_disclosures.py +4 -0
- package/tools/tool_schemas.json +2734 -0
- package/tools/trading_calendar.py +65 -30
- package/tools/value_data.py +423 -175
- package/tools/watchlist.py +43 -5
package/tools/value_data.py
CHANGED
|
@@ -6,7 +6,7 @@ import json
|
|
|
6
6
|
import os
|
|
7
7
|
import re
|
|
8
8
|
import sys
|
|
9
|
-
from datetime import date, timedelta
|
|
9
|
+
from datetime import date, datetime, timedelta
|
|
10
10
|
from html import unescape
|
|
11
11
|
from io import StringIO
|
|
12
12
|
from pathlib import Path
|
|
@@ -18,8 +18,19 @@ import pandas as pd
|
|
|
18
18
|
|
|
19
19
|
sys.path.insert(0, str(Path(__file__).resolve().parent))
|
|
20
20
|
from _cache import DAY, cached
|
|
21
|
-
from _subproc import scrub_secrets, socket_timeout, utf8_stdio
|
|
22
|
-
from stock_data import
|
|
21
|
+
from _subproc import error_message, scrub_secrets, socket_timeout, utf8_stdio
|
|
22
|
+
from stock_data import (
|
|
23
|
+
_A_INDEX_RE,
|
|
24
|
+
_cn_code,
|
|
25
|
+
_cn_now,
|
|
26
|
+
_em_datacenter,
|
|
27
|
+
_yf_hk_symbol,
|
|
28
|
+
detect_market,
|
|
29
|
+
normalize_stock_code,
|
|
30
|
+
)
|
|
31
|
+
from trading_calendar import _MARKET_TZ, market_now
|
|
32
|
+
|
|
33
|
+
YF_MARKETS = ("US", "JP", "KR", "TW")
|
|
23
34
|
|
|
24
35
|
|
|
25
36
|
def records(frame):
|
|
@@ -51,6 +62,24 @@ def capture(result, key, fetch):
|
|
|
51
62
|
result.setdefault("warnings", []).append(f"{key}: unavailable ({type(exc).__name__})")
|
|
52
63
|
|
|
53
64
|
|
|
65
|
+
def yf_symbol(symbol):
|
|
66
|
+
market = detect_market(symbol)
|
|
67
|
+
if market == "HK":
|
|
68
|
+
return _yf_hk_symbol(symbol)
|
|
69
|
+
if market not in YF_MARKETS or symbol.startswith("^"):
|
|
70
|
+
raise ValueError("a listed stock symbol is required")
|
|
71
|
+
if market == "US" and not re.fullmatch(r"[A-Za-z][A-Za-z0-9.-]{0,14}", symbol):
|
|
72
|
+
raise ValueError("a listed stock symbol is required (HK codes need .HK)") # "" / "00700" fall back to US
|
|
73
|
+
return symbol.upper().replace(".", "-") if market == "US" else symbol.upper()
|
|
74
|
+
|
|
75
|
+
|
|
76
|
+
def market_date(value, market):
|
|
77
|
+
parsed = pd.to_datetime(value, errors="coerce")
|
|
78
|
+
if pd.notna(parsed) and parsed.tzinfo is not None:
|
|
79
|
+
parsed = parsed.tz_convert(_MARKET_TZ["CN" if market == "A" else market])
|
|
80
|
+
return parsed.date().isoformat() if pd.notna(parsed) else None
|
|
81
|
+
|
|
82
|
+
|
|
54
83
|
def stock_identity(symbol):
|
|
55
84
|
market = detect_market(symbol)
|
|
56
85
|
if _A_INDEX_RE.fullmatch(symbol):
|
|
@@ -81,42 +110,91 @@ def valuation_summary(history, field):
|
|
|
81
110
|
}
|
|
82
111
|
|
|
83
112
|
|
|
113
|
+
def as_of(events, day):
|
|
114
|
+
"""Latest event filed on or before day (events sorted by filed)."""
|
|
115
|
+
found = None
|
|
116
|
+
for event in events:
|
|
117
|
+
if event["filed"] > day:
|
|
118
|
+
break
|
|
119
|
+
found = event
|
|
120
|
+
return found
|
|
121
|
+
|
|
122
|
+
|
|
123
|
+
def us_valuation_history(args):
|
|
124
|
+
import sec_edgar
|
|
125
|
+
import yfinance as yf
|
|
126
|
+
|
|
127
|
+
symbol = yf_symbol(args.symbol)
|
|
128
|
+
facts = sec_edgar.valuation_facts(symbol)
|
|
129
|
+
ticker = yf.Ticker(symbol)
|
|
130
|
+
start = (market_now("US").date() - pd.DateOffset(years=args.years)).date()
|
|
131
|
+
# Close is split-adjusted to today's share basis; scale filed share counts by splits after their as-of date.
|
|
132
|
+
closes = ticker.history(start=start.isoformat(), auto_adjust=False)["Close"].dropna()
|
|
133
|
+
if closes.empty:
|
|
134
|
+
raise ValueError("US price history unavailable")
|
|
135
|
+
closes = closes.resample("W-FRI").last().dropna() # weekly samples keep 10-year output bounded
|
|
136
|
+
splits = {market_date(stamp, "US"): float(ratio) for stamp, ratio in ticker.splits.items() if ratio > 0}
|
|
137
|
+
events = {key: sorted(facts[key], key=lambda e: e["filed"]) for key in ("ttm_net_income", "equity", "shares")}
|
|
138
|
+
history = []
|
|
139
|
+
for stamp, close in closes.items():
|
|
140
|
+
day = market_date(stamp, "US")
|
|
141
|
+
shares, income, equity = (as_of(events[key], day) for key in ("shares", "ttm_net_income", "equity"))
|
|
142
|
+
if not shares:
|
|
143
|
+
continue
|
|
144
|
+
factor = float(np.prod([ratio for split_day, ratio in splits.items() if split_day > shares["end"]]))
|
|
145
|
+
cap = float(close) * shares["val"] * factor
|
|
146
|
+
history.append(
|
|
147
|
+
{
|
|
148
|
+
"date": day,
|
|
149
|
+
"pe_ttm": round(cap / income["val"], 4) if income and income["val"] else None,
|
|
150
|
+
"pb": round(cap / equity["val"], 4) if equity and equity["val"] else None,
|
|
151
|
+
}
|
|
152
|
+
)
|
|
153
|
+
if not history:
|
|
154
|
+
raise ValueError("no US valuation history available")
|
|
155
|
+
return {
|
|
156
|
+
"symbol": args.symbol,
|
|
157
|
+
"market": "US",
|
|
158
|
+
"source": "sec_companyfacts+yfinance",
|
|
159
|
+
"requested_years": args.years,
|
|
160
|
+
"sampling": "weekly",
|
|
161
|
+
"share_basis": facts["share_basis"],
|
|
162
|
+
"history": history,
|
|
163
|
+
"metrics": {field: valuation_summary(history, field) for field in ("pe_ttm", "pb")},
|
|
164
|
+
"warnings": [
|
|
165
|
+
"Derived: market cap = weekly close x latest SEC-filed share count; PE uses TTM NetIncomeLoss attributable to parent, "
|
|
166
|
+
"PB uses StockholdersEquity, each effective from its filing date. "
|
|
167
|
+
"Multi-class issuers sum all classes at one class price."
|
|
168
|
+
],
|
|
169
|
+
}
|
|
170
|
+
|
|
171
|
+
|
|
84
172
|
def valuation_history(args):
|
|
85
173
|
import akshare as ak
|
|
86
174
|
|
|
87
|
-
market, symbol = stock_identity(args.symbol)
|
|
88
175
|
if not 1 <= args.years <= 10:
|
|
89
|
-
raise ValueError("
|
|
176
|
+
raise ValueError("years must be in 1..10")
|
|
177
|
+
if detect_market(args.symbol) == "US":
|
|
178
|
+
return us_valuation_history(args)
|
|
179
|
+
market, symbol = stock_identity(args.symbol)
|
|
90
180
|
warning = None
|
|
91
|
-
if
|
|
92
|
-
|
|
93
|
-
|
|
94
|
-
|
|
95
|
-
|
|
96
|
-
|
|
97
|
-
|
|
98
|
-
|
|
99
|
-
|
|
100
|
-
|
|
101
|
-
|
|
102
|
-
|
|
103
|
-
|
|
104
|
-
|
|
105
|
-
|
|
106
|
-
|
|
107
|
-
|
|
108
|
-
return {
|
|
109
|
-
"symbol": args.symbol,
|
|
110
|
-
"market": market,
|
|
111
|
-
"source": "lixinger",
|
|
112
|
-
"requested_years": args.years,
|
|
113
|
-
"history": history,
|
|
114
|
-
"metrics": {field: valuation_summary(history, field) for field in ("pe_ttm", "pb")},
|
|
115
|
-
}
|
|
116
|
-
except Exception:
|
|
117
|
-
warning = "Lixinger HK unavailable; Baidu history retained without mixing sources"
|
|
118
|
-
else:
|
|
119
|
-
warning = "LIXINGER_TOKEN not configured; HK history limited to available Baidu data"
|
|
181
|
+
if os.environ.get("LIXINGER_TOKEN"):
|
|
182
|
+
try:
|
|
183
|
+
import lixinger
|
|
184
|
+
|
|
185
|
+
history = lixinger.valuation_history("cn" if market == "A" else "hk", symbol, args.years)
|
|
186
|
+
return {
|
|
187
|
+
"symbol": args.symbol,
|
|
188
|
+
"market": market,
|
|
189
|
+
"source": "lixinger",
|
|
190
|
+
"requested_years": args.years,
|
|
191
|
+
"history": history,
|
|
192
|
+
"metrics": {field: valuation_summary(history, field) for field in ("pe_ttm", "pb")},
|
|
193
|
+
}
|
|
194
|
+
except Exception as e:
|
|
195
|
+
warning = f"Lixinger {market} unavailable: {error_message(e)}; Baidu history used without mixing sources"
|
|
196
|
+
elif market == "HK":
|
|
197
|
+
warning = "LIXINGER_TOKEN not configured; HK history limited to available Baidu data"
|
|
120
198
|
fetch = ak.stock_zh_valuation_baidu if market == "A" else ak.stock_hk_valuation_baidu
|
|
121
199
|
period = "近一年" if args.years == 1 else "近三年" if args.years <= 3 else "近五年" if args.years <= 5 else "近十年"
|
|
122
200
|
if market == "HK" and args.years > 3:
|
|
@@ -133,7 +211,7 @@ def valuation_history(args):
|
|
|
133
211
|
),
|
|
134
212
|
)
|
|
135
213
|
merged = {}
|
|
136
|
-
cutoff = (date
|
|
214
|
+
cutoff = (_cn_now().date() - timedelta(days=365.25 * args.years)).isoformat()
|
|
137
215
|
for field in ("pe_ttm", "pb"):
|
|
138
216
|
for row in series.get(field, []):
|
|
139
217
|
stamp = iso(row.get("date"))
|
|
@@ -160,28 +238,22 @@ INDEX_NAMES = {
|
|
|
160
238
|
}
|
|
161
239
|
|
|
162
240
|
|
|
241
|
+
def _csi_hist(ak):
|
|
242
|
+
# akshare 1.20.44 renamed it; py3.10 stays below 1.20.40 with the old name.
|
|
243
|
+
return getattr(ak, "index_hist_csindex", None) or ak.stock_zh_index_hist_csindex
|
|
244
|
+
|
|
245
|
+
|
|
163
246
|
def lixinger_index_history(code, start, end):
|
|
164
|
-
import
|
|
247
|
+
import lixinger
|
|
165
248
|
|
|
166
249
|
def fetch():
|
|
167
|
-
|
|
168
|
-
"
|
|
169
|
-
|
|
170
|
-
"token": os.environ["LIXINGER_TOKEN"],
|
|
171
|
-
"stockCodes": [code],
|
|
172
|
-
"startDate": start,
|
|
173
|
-
"endDate": end,
|
|
174
|
-
"metricsList": ["pe_ttm.mcw", "pb.mcw"],
|
|
175
|
-
},
|
|
176
|
-
timeout=20,
|
|
250
|
+
data = lixinger.post(
|
|
251
|
+
"cn/index/fundamental",
|
|
252
|
+
{"stockCodes": [code], "startDate": start, "endDate": end, "metricsList": ["pe_ttm.mcw", "pb.mcw"]},
|
|
177
253
|
)
|
|
178
|
-
response.raise_for_status()
|
|
179
|
-
payload = response.json()
|
|
180
|
-
if payload.get("code") != 1:
|
|
181
|
-
raise ValueError("Lixinger index data unavailable; check token permissions")
|
|
182
254
|
rows = [
|
|
183
255
|
{"date": iso(r.get("date")), "pe_ttm": number(r.get("pe_ttm.mcw")), "pb": number(r.get("pb.mcw"))}
|
|
184
|
-
for r in
|
|
256
|
+
for r in data
|
|
185
257
|
if r.get("stockCode") == code
|
|
186
258
|
]
|
|
187
259
|
if not rows or not any(r["date"] and (r["pe_ttm"] is not None or r["pb"] is not None) for r in rows):
|
|
@@ -201,7 +273,7 @@ def index_valuation(args):
|
|
|
201
273
|
raise ValueError("years must be in 1..10")
|
|
202
274
|
result = {"symbol": code, "name": name, "source": "legulegu", "requested_years": args.years}
|
|
203
275
|
merged = {}
|
|
204
|
-
cutoff = (pd.Timestamp(date
|
|
276
|
+
cutoff = (pd.Timestamp(_cn_now().date()) - pd.DateOffset(years=args.years)).date().isoformat()
|
|
205
277
|
for field, fn, column in [("pe_ttm", ak.stock_index_pe_lg, "滚动市盈率"), ("pb", ak.stock_index_pb_lg, "市净率")]:
|
|
206
278
|
section = {}
|
|
207
279
|
capture(
|
|
@@ -211,7 +283,7 @@ def index_valuation(args):
|
|
|
211
283
|
)
|
|
212
284
|
for row in section.get("rows", []):
|
|
213
285
|
stamp = iso(row.get("日期"))
|
|
214
|
-
if stamp and cutoff <= stamp <= date
|
|
286
|
+
if stamp and cutoff <= stamp <= _cn_now().date().isoformat():
|
|
215
287
|
merged.setdefault(stamp, {"date": stamp})[field] = number(row.get(column))
|
|
216
288
|
result.setdefault("warnings", []).extend(section.get("warnings", []))
|
|
217
289
|
if not any(number(r.get("pe_ttm")) is not None for r in merged.values()) and re.fullmatch(
|
|
@@ -223,24 +295,27 @@ def index_valuation(args):
|
|
|
223
295
|
"history",
|
|
224
296
|
lambda: dataset(
|
|
225
297
|
f"csi-history:{code}:{args.years}",
|
|
226
|
-
lambda: ak
|
|
227
|
-
symbol=code, start_date=cutoff.replace("-", ""), end_date=date
|
|
298
|
+
lambda: _csi_hist(ak)(
|
|
299
|
+
symbol=code, start_date=cutoff.replace("-", ""), end_date=_cn_now().date().strftime("%Y%m%d")
|
|
228
300
|
),
|
|
229
301
|
),
|
|
230
302
|
)
|
|
231
303
|
for row in section.get("history", []):
|
|
232
304
|
stamp = iso(row.get("日期"))
|
|
233
305
|
pe = number(row.get("滚动市盈率"))
|
|
234
|
-
if stamp and cutoff <= stamp <= date
|
|
306
|
+
if stamp and cutoff <= stamp <= _cn_now().date().isoformat() and pe is not None:
|
|
235
307
|
merged.setdefault(stamp, {"date": stamp})["pe_ttm"] = pe
|
|
236
308
|
capture(
|
|
237
309
|
section,
|
|
238
310
|
"recent",
|
|
239
|
-
lambda: dataset(
|
|
311
|
+
lambda: dataset(
|
|
312
|
+
f"csi-valuation:{code}",
|
|
313
|
+
lambda: (getattr(ak, "index_value_csindex", None) or ak.stock_zh_index_value_csindex)(symbol=code),
|
|
314
|
+
),
|
|
240
315
|
)
|
|
241
316
|
for row in section.get("recent", []):
|
|
242
317
|
stamp = iso(row.get("日期"))
|
|
243
|
-
if stamp and cutoff <= stamp <= date
|
|
318
|
+
if stamp and cutoff <= stamp <= _cn_now().date().isoformat():
|
|
244
319
|
merged.setdefault(stamp, {"date": stamp}).update(
|
|
245
320
|
pe=number(row.get("市盈率1")), dividend_yield=number(row.get("股息率1"))
|
|
246
321
|
)
|
|
@@ -258,9 +333,9 @@ def index_valuation(args):
|
|
|
258
333
|
missing = [field for field in ("pe_ttm", "pb") if field not in result["metric_sources"]]
|
|
259
334
|
if missing and os.environ.get("LIXINGER_TOKEN") and re.fullmatch(r"\d{6}|H\d{5}", code, re.I):
|
|
260
335
|
section = {}
|
|
261
|
-
capture(section, "rows", lambda: lixinger_index_history(code, cutoff, date
|
|
336
|
+
capture(section, "rows", lambda: lixinger_index_history(code, cutoff, _cn_now().date().isoformat()))
|
|
262
337
|
for row in section.get("rows", []):
|
|
263
|
-
if row["date"] and cutoff <= row["date"] <= date
|
|
338
|
+
if row["date"] and cutoff <= row["date"] <= _cn_now().date().isoformat():
|
|
264
339
|
for field in missing:
|
|
265
340
|
if row[field] is not None:
|
|
266
341
|
merged.setdefault(row["date"], {"date": row["date"]})[field] = row[field]
|
|
@@ -328,38 +403,41 @@ def hk_board_meetings(board):
|
|
|
328
403
|
|
|
329
404
|
|
|
330
405
|
def earnings_calendar(args):
|
|
331
|
-
import akshare as ak
|
|
332
|
-
|
|
333
406
|
stamp = iso(args.report_date)
|
|
334
|
-
suffix = {"03-31": "一季", "06-30": "半年报", "09-30": "三季", "12-31": "年报"}
|
|
335
407
|
if not 1 <= args.top <= 500 or (args.report_date and not stamp):
|
|
336
408
|
raise ValueError("report_date must be a valid date and top must be in 1..500")
|
|
337
|
-
|
|
409
|
+
market = detect_market(args.symbol) if args.symbol else "A"
|
|
410
|
+
if market in ("KR", "TW"):
|
|
411
|
+
raise ValueError("earnings calendar covers A/HK/US/JP only")
|
|
412
|
+
if market in ("HK", "US", "JP"):
|
|
338
413
|
import yfinance as yf
|
|
339
414
|
|
|
340
|
-
|
|
341
|
-
|
|
342
|
-
|
|
343
|
-
|
|
344
|
-
|
|
345
|
-
|
|
346
|
-
|
|
347
|
-
|
|
348
|
-
|
|
349
|
-
|
|
350
|
-
|
|
351
|
-
|
|
352
|
-
|
|
353
|
-
|
|
415
|
+
code = yf_symbol(args.symbol)
|
|
416
|
+
today = market_now(market).date().isoformat()
|
|
417
|
+
sections, events = {}, []
|
|
418
|
+
if market == "HK":
|
|
419
|
+
hk_code = stock_identity(args.symbol)[1]
|
|
420
|
+
for board in ("main", "gem"):
|
|
421
|
+
capture(sections, board, lambda board=board: hk_board_meetings(board))
|
|
422
|
+
events = [
|
|
423
|
+
r
|
|
424
|
+
for board in ("main", "gem")
|
|
425
|
+
for r in sections.get(board, [])
|
|
426
|
+
if r["symbol"] == hk_code + ".HK"
|
|
427
|
+
and r["scheduled_date"] >= today
|
|
428
|
+
and (not stamp or r["report_date"] == stamp)
|
|
429
|
+
]
|
|
430
|
+
events = list({(r["scheduled_date"], r["purpose"], r["report_date"]): r for r in events}.values())
|
|
431
|
+
events.sort(key=lambda r: r["scheduled_date"])
|
|
354
432
|
|
|
355
433
|
def fetch():
|
|
356
|
-
frame = yf.Ticker(
|
|
434
|
+
frame = yf.Ticker(code).get_earnings_dates(limit=12)
|
|
357
435
|
if frame is None:
|
|
358
|
-
raise ValueError("
|
|
359
|
-
return sorted({
|
|
436
|
+
raise ValueError(f"{market} earnings schedule unavailable")
|
|
437
|
+
return sorted({market_date(d, market) for d in frame.index if market_date(d, market)})
|
|
360
438
|
|
|
361
439
|
if not events:
|
|
362
|
-
capture(sections, "dates", lambda: cached(f"
|
|
440
|
+
capture(sections, "dates", lambda: cached(f"{market.lower()}-earnings-schedule:{code}", DAY, fetch))
|
|
363
441
|
provider_events = [
|
|
364
442
|
{
|
|
365
443
|
"symbol": args.symbol,
|
|
@@ -370,13 +448,13 @@ def earnings_calendar(args):
|
|
|
370
448
|
"date_kind": "provider_earnings_schedule",
|
|
371
449
|
}
|
|
372
450
|
for d in sections.get("dates", [])
|
|
373
|
-
if d >=
|
|
451
|
+
if d >= today
|
|
374
452
|
]
|
|
375
453
|
if not stamp and not events:
|
|
376
454
|
events = provider_events
|
|
377
455
|
return {
|
|
378
456
|
"symbol": args.symbol,
|
|
379
|
-
"market":
|
|
457
|
+
"market": market,
|
|
380
458
|
"source": "HKEX board-meeting notices"
|
|
381
459
|
if any(r["confirmation"] == "official_board_meeting_notice" for r in events)
|
|
382
460
|
else "yfinance earnings schedule",
|
|
@@ -387,43 +465,48 @@ def earnings_calendar(args):
|
|
|
387
465
|
"forecasts_total": 0,
|
|
388
466
|
"unverified_schedules": provider_events[: args.top] if stamp else [],
|
|
389
467
|
"warnings": sections.get("warnings", [])
|
|
390
|
-
+
|
|
391
|
-
|
|
392
|
-
|
|
393
|
-
|
|
394
|
-
|
|
468
|
+
+ (
|
|
469
|
+
[
|
|
470
|
+
"official dates are scheduled board meetings approving results, not actual disclosure dates; notice list may not be exhaustive"
|
|
471
|
+
]
|
|
472
|
+
if market == "HK"
|
|
473
|
+
else []
|
|
474
|
+
)
|
|
475
|
+
+ ["provider-only schedules have unverified fiscal periods and are not assigned to requested report_date"],
|
|
476
|
+
"as_of": today,
|
|
395
477
|
}
|
|
396
|
-
if not stamp or stamp[5:] not in
|
|
478
|
+
if not stamp or stamp[5:] not in ("03-31", "06-30", "09-30", "12-31"):
|
|
397
479
|
raise ValueError("report_date must be a quarter end for A-share earnings calendars")
|
|
398
480
|
symbol = ""
|
|
399
481
|
if args.symbol:
|
|
400
482
|
market, symbol = stock_identity(args.symbol)
|
|
401
483
|
if market != "A":
|
|
402
484
|
raise ValueError("earnings calendar covers A shares only")
|
|
403
|
-
result = {"report_date": stamp, "source": "
|
|
404
|
-
|
|
405
|
-
|
|
406
|
-
|
|
407
|
-
|
|
408
|
-
|
|
409
|
-
|
|
410
|
-
|
|
411
|
-
)
|
|
412
|
-
|
|
413
|
-
|
|
414
|
-
|
|
485
|
+
result = {"report_date": stamp, "source": "eastmoney"}
|
|
486
|
+
# Server-side filter + limit: one page holds the requested rows, `count` the full total.
|
|
487
|
+
condition = f"(REPORT_DATE='{stamp}')" + (f'(SECURITY_CODE="{symbol}")' if symbol else "")
|
|
488
|
+
totals = {}
|
|
489
|
+
queries = {
|
|
490
|
+
"disclosures": ("RPT_PUBLIC_BS_APPOIN", "APPOINT_PUBLISH_DATE,SECURITY_CODE", "1,1"),
|
|
491
|
+
"forecasts": ("RPT_PUBLIC_OP_NEWPREDICT", "NOTICE_DATE,SECURITY_CODE", "-1,-1"),
|
|
492
|
+
}
|
|
493
|
+
for key, (report, sort, order) in queries.items():
|
|
494
|
+
|
|
495
|
+
def fetch(report=report, sort=sort, order=order, key=key):
|
|
496
|
+
totals[key], rows = _em_datacenter(report, condition, sort, args.top, order)
|
|
497
|
+
return rows
|
|
498
|
+
|
|
499
|
+
capture(result, key, fetch)
|
|
415
500
|
for key in ("disclosures", "forecasts"):
|
|
416
|
-
rows = result.get(key, [])
|
|
417
|
-
if symbol:
|
|
418
|
-
rows = [r for r in rows if str(r.get("股票代码", "")).zfill(6) == symbol]
|
|
419
501
|
output = []
|
|
420
|
-
for row in
|
|
421
|
-
item = {"symbol": str(row.get("
|
|
502
|
+
for row in result.get(key, []):
|
|
503
|
+
item = {"symbol": str(row.get("SECURITY_CODE", "")).zfill(6), "name": row.get("SECURITY_NAME_ABBR")}
|
|
422
504
|
if key == "disclosures":
|
|
423
|
-
|
|
505
|
+
columns = ("FIRST_APPOINT_DATE", "FIRST_CHANGE_DATE", "SECOND_CHANGE_DATE", "THIRD_CHANGE_DATE")
|
|
506
|
+
revised = [iso(row.get(c)) for c in columns]
|
|
424
507
|
item.update(
|
|
425
508
|
scheduled_date=next((d for d in reversed(revised) if d), None),
|
|
426
|
-
actual_date=iso(row.get("
|
|
509
|
+
actual_date=iso(row.get("ACTUAL_PUBLISH_DATE")),
|
|
427
510
|
original_scheduled_date=revised[0],
|
|
428
511
|
scheduled_date_revisions=[
|
|
429
512
|
{"revision": i, "scheduled_date": d} for i, d in enumerate(revised[1:], 1) if d
|
|
@@ -431,16 +514,16 @@ def earnings_calendar(args):
|
|
|
431
514
|
)
|
|
432
515
|
else:
|
|
433
516
|
item.update(
|
|
434
|
-
announcement_date=iso(row.get("
|
|
435
|
-
forecast_type=row.get("
|
|
436
|
-
indicator=row.get("
|
|
437
|
-
forecast_value=number(row.get("
|
|
438
|
-
change_pct=number(row.get("
|
|
439
|
-
reason=row.get("
|
|
517
|
+
announcement_date=iso(row.get("NOTICE_DATE")),
|
|
518
|
+
forecast_type=row.get("PREDICT_TYPE"),
|
|
519
|
+
indicator=row.get("PREDICT_FINANCE"),
|
|
520
|
+
forecast_value=number(row.get("FORECAST_JZ")),
|
|
521
|
+
change_pct=number(row.get("INCREASE_JZ")),
|
|
522
|
+
reason=row.get("CHANGE_REASON_EXPLAIN"),
|
|
440
523
|
)
|
|
441
524
|
output.append(item)
|
|
442
|
-
result[key + "_total"] =
|
|
443
|
-
result[key] = output
|
|
525
|
+
result[key + "_total"] = totals.get(key, 0)
|
|
526
|
+
result[key] = output
|
|
444
527
|
if result.get("warnings") and not result["disclosures"] and not result["forecasts"]:
|
|
445
528
|
raise ValueError("earnings calendar sources unavailable")
|
|
446
529
|
return result
|
|
@@ -495,8 +578,10 @@ def northbound_holdings(symbol, report_date=""):
|
|
|
495
578
|
exchange = _cn_code(symbol)[:2]
|
|
496
579
|
if exchange not in ("sh", "sz"):
|
|
497
580
|
raise ValueError("Stock Connect holdings cover Shanghai/Shenzhen securities only")
|
|
498
|
-
if report_date
|
|
499
|
-
|
|
581
|
+
if report_date:
|
|
582
|
+
report_date = iso(report_date)
|
|
583
|
+
if not report_date or report_date[5:] not in ("03-31", "06-30", "09-30", "12-31"):
|
|
584
|
+
raise ValueError("report_date must be a quarter end")
|
|
500
585
|
|
|
501
586
|
def fetch():
|
|
502
587
|
if exchange == "sh":
|
|
@@ -547,7 +632,7 @@ def northbound_holdings(symbol, report_date=""):
|
|
|
547
632
|
|
|
548
633
|
stamp = report_date
|
|
549
634
|
if not stamp:
|
|
550
|
-
end = date(date
|
|
635
|
+
end = date(_cn_now().date().year, ((_cn_now().date().month - 1) // 3) * 3 + 1, 1) - timedelta(days=1)
|
|
551
636
|
for _ in range(2):
|
|
552
637
|
candidate = end.isoformat()
|
|
553
638
|
if query(candidate, "").get("data"):
|
|
@@ -566,7 +651,7 @@ def northbound_holdings(symbol, report_date=""):
|
|
|
566
651
|
source = "SZSE quarterly Stock Connect holdings"
|
|
567
652
|
if (
|
|
568
653
|
not stamp
|
|
569
|
-
or stamp > date
|
|
654
|
+
or stamp > _cn_now().date().isoformat()
|
|
570
655
|
or (report_date and stamp != report_date)
|
|
571
656
|
or shares is None
|
|
572
657
|
or shares < 0
|
|
@@ -581,7 +666,7 @@ def northbound_holdings(symbol, report_date=""):
|
|
|
581
666
|
"note": "exchange-disclosed aggregate Stock Connect quantity for this security; not a top-ten shareholder inference",
|
|
582
667
|
}
|
|
583
668
|
|
|
584
|
-
return cached(f"northbound-official:{symbol}:{report_date or
|
|
669
|
+
return cached(f"northbound-official:{symbol}:{report_date or ''}", DAY, fetch)
|
|
585
670
|
|
|
586
671
|
|
|
587
672
|
def shareholder_structure(args):
|
|
@@ -671,7 +756,7 @@ def shareholder_structure(args):
|
|
|
671
756
|
def hk_dividend_ranking(report_date=""):
|
|
672
757
|
import requests
|
|
673
758
|
|
|
674
|
-
stamp = iso(report_date) if report_date else f"{date
|
|
759
|
+
stamp = iso(report_date) if report_date else f"{_cn_now().date().year - 1}-12-31"
|
|
675
760
|
if not stamp or stamp[5:] != "12-31":
|
|
676
761
|
raise ValueError("HK dividend ranking requires an annual year-end report_date")
|
|
677
762
|
|
|
@@ -761,32 +846,43 @@ def hk_dividend_ranking(report_date=""):
|
|
|
761
846
|
return cached(f"hk-dividend-ranking:{stamp}", DAY, fetch)
|
|
762
847
|
|
|
763
848
|
|
|
764
|
-
def dividend_ranking(report_date="", market="A"):
|
|
765
|
-
|
|
849
|
+
def dividend_ranking(report_date="", market="A", top=500, codes=None):
|
|
850
|
+
"""Implemented cash-dividend plans ranked by yield, sorted and limited server-side.
|
|
766
851
|
|
|
852
|
+
`codes` (A only) restricts the query to those stocks — the screener asks only for its candidates."""
|
|
767
853
|
if market == "HK":
|
|
768
854
|
return hk_dividend_ranking(report_date)
|
|
769
855
|
if market != "A":
|
|
770
856
|
raise ValueError("dividend rankings cover A/HK only")
|
|
771
|
-
stamp = iso(report_date) if report_date else f"{date
|
|
857
|
+
stamp = iso(report_date) if report_date else f"{_cn_now().date().year - 1}-12-31"
|
|
772
858
|
if not stamp or stamp[5:] not in ("03-31", "06-30", "09-30", "12-31"):
|
|
773
859
|
raise ValueError("dividend report_date must be a quarter end")
|
|
774
|
-
|
|
860
|
+
base = f"(REPORT_DATE='{stamp}')(ASSIGN_PROGRESS=\"实施分配\")"
|
|
861
|
+
if codes is None:
|
|
862
|
+
queries = [(base, top)]
|
|
863
|
+
else:
|
|
864
|
+
codes = [c for c in codes if re.fullmatch(r"\d{6}", str(c))]
|
|
865
|
+
chunks = [codes[i : i + 200] for i in range(0, len(codes), 200)]
|
|
866
|
+
queries = [(base + "(SECURITY_CODE in (" + ",".join(f'"{c}"' for c in chunk) + "))", 500) for chunk in chunks]
|
|
867
|
+
rows = []
|
|
868
|
+
for condition, size in queries:
|
|
869
|
+
rows += _em_datacenter("RPT_SHAREBONUS_DET", condition, "DIVIDENT_RATIO", size)[1]
|
|
775
870
|
return {
|
|
776
871
|
"report_date": stamp,
|
|
777
872
|
"source": "eastmoney",
|
|
778
873
|
"yield_basis": "provider report-period cash dividend yield, percent",
|
|
779
874
|
"items": [
|
|
780
875
|
{
|
|
781
|
-
"symbol": str(r.get("
|
|
782
|
-
"name": r.get("
|
|
783
|
-
|
|
784
|
-
"
|
|
785
|
-
"
|
|
786
|
-
"
|
|
876
|
+
"symbol": str(r.get("SECURITY_CODE", "")).zfill(6),
|
|
877
|
+
"name": r.get("SECURITY_NAME_ABBR"),
|
|
878
|
+
# Eastmoney reports this ratio as a fraction (0.0101 = 1.01%).
|
|
879
|
+
"dividend_yield": round(number(r.get("DIVIDENT_RATIO")) * 100, 6),
|
|
880
|
+
"ex_date": iso(r.get("EX_DIVIDEND_DATE")),
|
|
881
|
+
"announcement_date": iso(r.get("NOTICE_DATE")),
|
|
882
|
+
"status": r.get("ASSIGN_PROGRESS"),
|
|
787
883
|
}
|
|
788
884
|
for r in rows
|
|
789
|
-
if
|
|
885
|
+
if number(r.get("DIVIDENT_RATIO")) is not None
|
|
790
886
|
],
|
|
791
887
|
}
|
|
792
888
|
|
|
@@ -797,14 +893,15 @@ def dividends(args):
|
|
|
797
893
|
if args.mode == "ranking":
|
|
798
894
|
if args.symbol:
|
|
799
895
|
raise ValueError("ranking covers all A shares; omit symbol")
|
|
800
|
-
result = dividend_ranking(args.report_date, getattr(args, "market", "A"))
|
|
896
|
+
result = dividend_ranking(args.report_date, getattr(args, "market", "A"), top=args.top)
|
|
801
897
|
result["items"] = sorted(result["items"], key=lambda r: r["dividend_yield"], reverse=True)[: args.top]
|
|
802
898
|
return result
|
|
803
|
-
market
|
|
899
|
+
market = detect_market(args.symbol)
|
|
900
|
+
market, symbol = (market, yf_symbol(args.symbol)) if market in YF_MARKETS else stock_identity(args.symbol)
|
|
804
901
|
import akshare as ak
|
|
805
902
|
from _subproc import run_tool
|
|
806
903
|
|
|
807
|
-
today = date
|
|
904
|
+
today = market_now(market).date()
|
|
808
905
|
result = {
|
|
809
906
|
"symbol": args.symbol,
|
|
810
907
|
"market": market,
|
|
@@ -841,32 +938,47 @@ def dividends(args):
|
|
|
841
938
|
import yfinance as yf
|
|
842
939
|
|
|
843
940
|
sources = {}
|
|
844
|
-
|
|
845
|
-
|
|
846
|
-
|
|
847
|
-
|
|
848
|
-
|
|
849
|
-
|
|
850
|
-
|
|
851
|
-
|
|
852
|
-
|
|
853
|
-
|
|
854
|
-
|
|
855
|
-
|
|
856
|
-
|
|
857
|
-
|
|
858
|
-
|
|
941
|
+
if market == "HK":
|
|
942
|
+
capture(
|
|
943
|
+
sources,
|
|
944
|
+
"calendar",
|
|
945
|
+
lambda: dataset(
|
|
946
|
+
f"hk-dividend-calendar:{symbol}", lambda: ak.stock_hk_dividend_payout_em(symbol=symbol)
|
|
947
|
+
),
|
|
948
|
+
)
|
|
949
|
+
calendar = [
|
|
950
|
+
{
|
|
951
|
+
"ex_date": iso(r.get("除净日")),
|
|
952
|
+
"payment_date": iso(r.get("发放日")),
|
|
953
|
+
"announcement_date": iso(r.get("最新公告日期")),
|
|
954
|
+
"plan": r.get("分红方案"),
|
|
955
|
+
"fiscal_year": r.get("财政年度"),
|
|
956
|
+
}
|
|
957
|
+
for r in sources.get("calendar", [])
|
|
958
|
+
]
|
|
959
|
+
ticker = yf.Ticker(_yf_hk_symbol(args.symbol))
|
|
960
|
+
else:
|
|
961
|
+
ticker = yf.Ticker(symbol)
|
|
962
|
+
capture(sources, "calendar", lambda: ticker.calendar)
|
|
963
|
+
upcoming = sources.get("calendar") or {}
|
|
964
|
+
calendar = [
|
|
965
|
+
{
|
|
966
|
+
"ex_date": iso(upcoming.get("Ex-Dividend Date")),
|
|
967
|
+
"payment_date": iso(upcoming.get("Dividend Date")),
|
|
968
|
+
}
|
|
969
|
+
]
|
|
970
|
+
if not calendar[0]["ex_date"]:
|
|
971
|
+
calendar = []
|
|
859
972
|
if args.mode == "calendar":
|
|
860
973
|
if sources.get("warnings"):
|
|
861
|
-
raise ValueError("
|
|
974
|
+
raise ValueError(f"{market} dividend calendar unavailable")
|
|
862
975
|
return {
|
|
863
976
|
"symbol": args.symbol,
|
|
864
977
|
"market": market,
|
|
865
|
-
"source": "eastmoney",
|
|
978
|
+
"source": "eastmoney" if market == "HK" else "yfinance",
|
|
866
979
|
"as_of": today.isoformat(),
|
|
867
980
|
"calendar": [r for r in calendar if r.get("ex_date") and r["ex_date"] > today.isoformat()],
|
|
868
981
|
}
|
|
869
|
-
ticker = yf.Ticker(_yf_hk_symbol(args.symbol))
|
|
870
982
|
|
|
871
983
|
def cash_history():
|
|
872
984
|
raw = ticker.dividends
|
|
@@ -877,17 +989,17 @@ def dividends(args):
|
|
|
877
989
|
)
|
|
878
990
|
return frame.reset_index()
|
|
879
991
|
|
|
880
|
-
capture(sources, "cash", lambda: dataset(f"
|
|
992
|
+
capture(sources, "cash", lambda: dataset(f"{market.lower()}-dividend-cash:{symbol}", cash_history))
|
|
881
993
|
capture(sources, "quote", lambda: ticker.info)
|
|
882
994
|
paid = sources.get("cash", [])
|
|
883
995
|
info = sources.get("quote") or {}
|
|
884
996
|
if sources.get("warnings"):
|
|
885
997
|
result["warnings"] = sources["warnings"]
|
|
886
998
|
if not paid and not calendar:
|
|
887
|
-
raise ValueError("
|
|
999
|
+
raise ValueError(f"{market} dividend sources unavailable")
|
|
888
1000
|
events = [
|
|
889
1001
|
{
|
|
890
|
-
"ex_date":
|
|
1002
|
+
"ex_date": market_date(r.get("Date"), market),
|
|
891
1003
|
"cash_per_share": number(r.get("cash_per_share")),
|
|
892
1004
|
"currency": r.get("currency") or info.get("currency"),
|
|
893
1005
|
"status": "paid",
|
|
@@ -895,7 +1007,11 @@ def dividends(args):
|
|
|
895
1007
|
for r in paid
|
|
896
1008
|
]
|
|
897
1009
|
price = number(info.get("currentPrice") or info.get("regularMarketPrice"))
|
|
898
|
-
result.update(
|
|
1010
|
+
result.update(
|
|
1011
|
+
source="yfinance cash / eastmoney calendar" if market == "HK" else "yfinance",
|
|
1012
|
+
currency=info.get("currency"),
|
|
1013
|
+
price=price,
|
|
1014
|
+
)
|
|
899
1015
|
result["calendar"] = [r for r in calendar if r["ex_date"] and r["ex_date"] > today.isoformat()]
|
|
900
1016
|
cutoff = (today - timedelta(days=365.25 * args.years)).isoformat()
|
|
901
1017
|
events = sorted(
|
|
@@ -1052,15 +1168,15 @@ def etf_basket(symbol):
|
|
|
1052
1168
|
"SHOWTYPE": "JSON",
|
|
1053
1169
|
"CATALOGID": "sgshqd",
|
|
1054
1170
|
"txtJCorDH": symbol,
|
|
1055
|
-
"txtStart": (date
|
|
1056
|
-
"txtEnd": date
|
|
1171
|
+
"txtStart": (_cn_now().date() - timedelta(days=30)).isoformat(),
|
|
1172
|
+
"txtEnd": _cn_now().date().isoformat(),
|
|
1057
1173
|
},
|
|
1058
1174
|
headers={"Referer": "https://www.szse.cn/"},
|
|
1059
1175
|
timeout=20,
|
|
1060
1176
|
)
|
|
1061
1177
|
response.raise_for_status()
|
|
1062
1178
|
dates = re.findall(rf"pcf_{re.escape(symbol)}_(\d{{8}})", json.dumps(response.json()))
|
|
1063
|
-
dates = [d for d in dates if iso(d) and iso(d) <= date
|
|
1179
|
+
dates = [d for d in dates if iso(d) and iso(d) <= _cn_now().date().isoformat()]
|
|
1064
1180
|
if not dates:
|
|
1065
1181
|
raise ValueError("no published SZSE PCF found")
|
|
1066
1182
|
response = requests.get(
|
|
@@ -1083,7 +1199,7 @@ def etf_basket(symbol):
|
|
|
1083
1199
|
if (
|
|
1084
1200
|
fund_code != symbol
|
|
1085
1201
|
or not stamp
|
|
1086
|
-
or stamp > date
|
|
1202
|
+
or stamp > _cn_now().date().isoformat()
|
|
1087
1203
|
or not unit
|
|
1088
1204
|
or unit <= 0
|
|
1089
1205
|
or count != len(components)
|
|
@@ -1141,7 +1257,7 @@ def etf_basket(symbol):
|
|
|
1141
1257
|
"basis": "complete published creation/redemption basket, not actual fund holdings; quantities per creation unit, rates are fractions",
|
|
1142
1258
|
}
|
|
1143
1259
|
|
|
1144
|
-
return cached(f"etf-basket:{symbol}
|
|
1260
|
+
return cached(f"etf-basket:{symbol}", DAY, fetch)
|
|
1145
1261
|
|
|
1146
1262
|
|
|
1147
1263
|
def fund_disclosed_equities(symbol):
|
|
@@ -1208,7 +1324,7 @@ def fund_disclosed_equities(symbol):
|
|
|
1208
1324
|
}
|
|
1209
1325
|
raise ValueError("requested semiannual/annual equity report not available")
|
|
1210
1326
|
|
|
1211
|
-
end = date(date
|
|
1327
|
+
end = date(_cn_now().date().year, ((_cn_now().date().month - 1) // 6) * 6 + 1, 1) - timedelta(days=1)
|
|
1212
1328
|
requested = end.isoformat()
|
|
1213
1329
|
for _ in range(2):
|
|
1214
1330
|
stamp = end.isoformat()
|
|
@@ -1227,20 +1343,28 @@ def etf_details(args):
|
|
|
1227
1343
|
symbol = _cn_code(args.symbol)[2:]
|
|
1228
1344
|
if detect_market(args.symbol) != "A" or not normalize_stock_code(symbol)["is_etf"] or not 10 <= args.days <= 252:
|
|
1229
1345
|
raise ValueError("ETF details requires an A-share ETF code and days in 10..252")
|
|
1346
|
+
benchmark = (args.benchmark or "").strip()
|
|
1347
|
+
if suffixed := re.fullmatch(r"(\d{6})\.(SH|SZ)", benchmark, re.I):
|
|
1348
|
+
benchmark = suffixed[2].lower() + suffixed[1]
|
|
1349
|
+
if benchmark and not re.fullmatch(
|
|
1350
|
+
r"(?i)(sh|sz)?\d{6}|H\d{5}|93\d{4}|\^?(?!(sh|sz)\d)[A-Z][A-Z0-9.=-]{0,14}", benchmark
|
|
1351
|
+
):
|
|
1352
|
+
raise ValueError("benchmark must be an index/ETF code such as sh000300, 399006, 000300.SH, H30269 or ^NDX")
|
|
1230
1353
|
result = {
|
|
1231
1354
|
"symbol": symbol,
|
|
1232
1355
|
"market": "A",
|
|
1233
1356
|
"source": "eastmoney",
|
|
1234
1357
|
"holdings_basis": "latest disclosed fund portfolio, not a complete live creation/redemption basket",
|
|
1235
1358
|
}
|
|
1236
|
-
end = date
|
|
1359
|
+
end = _cn_now().date()
|
|
1237
1360
|
start = end - timedelta(days=args.days * 2)
|
|
1361
|
+
# akshare 1.20.41 renamed fund_etf_fund_info_em; py3.10 is capped below 1.20.40 and keeps the old name
|
|
1238
1362
|
capture(
|
|
1239
1363
|
result,
|
|
1240
1364
|
"nav_history",
|
|
1241
1365
|
lambda: dataset(
|
|
1242
|
-
f"etf-nav:{symbol}:{args.days}",
|
|
1243
|
-
lambda: ak.fund_etf_fund_info_em(
|
|
1366
|
+
f"etf-nav:{symbol}:{args.days}:{end.isoformat()}",
|
|
1367
|
+
lambda: (getattr(ak, "fund_exchange_nav_hist_em", None) or ak.fund_etf_fund_info_em)(
|
|
1244
1368
|
fund=symbol, start_date=start.strftime("%Y%m%d"), end_date=end.strftime("%Y%m%d")
|
|
1245
1369
|
),
|
|
1246
1370
|
),
|
|
@@ -1310,13 +1434,30 @@ def etf_details(args):
|
|
|
1310
1434
|
capture(result, "basket", lambda: etf_basket(symbol))
|
|
1311
1435
|
if getattr(args, "include_full_holdings", False):
|
|
1312
1436
|
capture(result, "disclosed_equity_report", lambda: fund_disclosed_equities(symbol))
|
|
1313
|
-
if
|
|
1314
|
-
|
|
1315
|
-
|
|
1437
|
+
if benchmark:
|
|
1438
|
+
if re.fullmatch(r"H\d{5}|93\d{4}", benchmark.upper()):
|
|
1439
|
+
benchmark = code = benchmark.upper()
|
|
1440
|
+
capture(
|
|
1441
|
+
result,
|
|
1442
|
+
"benchmark_history",
|
|
1443
|
+
lambda: dataset(
|
|
1444
|
+
f"csi-close:{code}:{args.days}:{end.isoformat()}",
|
|
1445
|
+
lambda: _csi_hist(ak)(
|
|
1446
|
+
symbol=code, start_date=start.strftime("%Y%m%d"), end_date=end.strftime("%Y%m%d")
|
|
1447
|
+
),
|
|
1448
|
+
),
|
|
1449
|
+
)
|
|
1450
|
+
history = result.pop("benchmark_history", [])
|
|
1451
|
+
benchmark_rows = [{"date": iso(r.get("日期")), "close": number(r.get("收盘"))} for r in history]
|
|
1452
|
+
else:
|
|
1453
|
+
if re.fullmatch(r"\d{6}", benchmark):
|
|
1454
|
+
benchmark = ("sz" if benchmark.startswith(("399", "15", "16")) else "sh") + benchmark
|
|
1455
|
+
code = benchmark[2:] if re.fullmatch(r"(?i)(sh|sz)\d{6}", benchmark) else ""
|
|
1456
|
+
benchmark = benchmark.lower() if code else benchmark
|
|
1457
|
+
benchmark_rows = run_tool("stock_data.py", ["kline", benchmark, "--count", str(args.days + 5)])
|
|
1316
1458
|
result["tracking_error"] = tracking_error(nav, benchmark_rows if isinstance(benchmark_rows, list) else [])
|
|
1317
1459
|
result["tracking_error"]["benchmark"] = benchmark
|
|
1318
|
-
if
|
|
1319
|
-
code = benchmark[2:]
|
|
1460
|
+
if code:
|
|
1320
1461
|
capture(
|
|
1321
1462
|
result,
|
|
1322
1463
|
"benchmark_constituents",
|
|
@@ -1360,15 +1501,122 @@ ANNOUNCEMENT_CATEGORIES = {
|
|
|
1360
1501
|
"inquiry": ("", "问询"),
|
|
1361
1502
|
"dividend": ("category_qyfpxzcs_szsh", ""),
|
|
1362
1503
|
}
|
|
1504
|
+
HK_ANNOUNCEMENT_CATEGORIES = {
|
|
1505
|
+
"forecast": r"Profit Warning|Positive Profit Alert|Inside Information.*Profit",
|
|
1506
|
+
"annual": r"Final Results|Annual Report",
|
|
1507
|
+
"interim": r"Interim Results|Interim/Half-Year Report",
|
|
1508
|
+
"quarterly": r"Quarterly Results",
|
|
1509
|
+
"issuance": r"Issue of Securities|Placing|Rights Issue|Open Offer|Subscription",
|
|
1510
|
+
"dividend": r"Dividend",
|
|
1511
|
+
}
|
|
1512
|
+
|
|
1513
|
+
|
|
1514
|
+
def hk_announcements(args, symbol):
|
|
1515
|
+
import requests
|
|
1516
|
+
|
|
1517
|
+
if args.category != "all" and args.category not in HK_ANNOUNCEMENT_CATEGORIES:
|
|
1518
|
+
raise ValueError("HK announcements support all/forecast/annual/interim/quarterly/issuance/dividend")
|
|
1519
|
+
end = date.fromisoformat(args.end_date) if args.end_date else market_now("HK").date()
|
|
1520
|
+
start = date.fromisoformat(args.start_date) if args.start_date else end - timedelta(days=90)
|
|
1521
|
+
if start > end or not 1 <= args.top <= 500:
|
|
1522
|
+
raise ValueError("start_date must not exceed end_date and top must be in 1..500")
|
|
1523
|
+
response = requests.get(
|
|
1524
|
+
"https://www1.hkexnews.hk/search/prefix.do",
|
|
1525
|
+
params={"callback": "callback", "lang": "EN", "type": "A", "name": symbol, "market": "SEHK"},
|
|
1526
|
+
timeout=20,
|
|
1527
|
+
)
|
|
1528
|
+
response.raise_for_status()
|
|
1529
|
+
stocks = json.loads(re.sub(r"^\s*callback\(|\);?\s*$", "", response.text)).get("stockInfo", [])
|
|
1530
|
+
stock = next((row for row in stocks if row.get("code") == symbol), None)
|
|
1531
|
+
if not stock:
|
|
1532
|
+
raise ValueError("stock not found in HKEXnews directory")
|
|
1533
|
+
response = requests.get(
|
|
1534
|
+
"https://www1.hkexnews.hk/search/titleSearchServlet.do",
|
|
1535
|
+
params={
|
|
1536
|
+
"sortDir": 0,
|
|
1537
|
+
"sortByOptions": "DateTime",
|
|
1538
|
+
"category": 0,
|
|
1539
|
+
"market": "SEHK",
|
|
1540
|
+
"stockId": stock["stockId"],
|
|
1541
|
+
"documentType": -1,
|
|
1542
|
+
"fromDate": start.strftime("%Y%m%d"),
|
|
1543
|
+
"toDate": end.strftime("%Y%m%d"),
|
|
1544
|
+
"title": "",
|
|
1545
|
+
"searchType": 0,
|
|
1546
|
+
"t1code": -2,
|
|
1547
|
+
"t2Gcode": -2,
|
|
1548
|
+
"t2code": -2,
|
|
1549
|
+
# shortcut: one 500-row page; narrow the date range when truncated.
|
|
1550
|
+
"rowRange": 500,
|
|
1551
|
+
"lang": "E",
|
|
1552
|
+
},
|
|
1553
|
+
timeout=20,
|
|
1554
|
+
)
|
|
1555
|
+
response.raise_for_status()
|
|
1556
|
+
data = response.json()
|
|
1557
|
+
items = json.loads(data.get("result") or "[]")
|
|
1558
|
+
rows = []
|
|
1559
|
+
for row in items:
|
|
1560
|
+
kind = unescape(re.sub(r"<[^>]+>", " ", row.get("LONG_TEXT") or row.get("SHORT_TEXT") or "")).strip()
|
|
1561
|
+
title = unescape(re.sub(r"<[^>]+>", " ", row.get("TITLE") or "")).strip()
|
|
1562
|
+
# One announcement can carry several headlines (e.g. Final Results + Dividend); match any of them.
|
|
1563
|
+
matches = [
|
|
1564
|
+
key for key, pattern in HK_ANNOUNCEMENT_CATEGORIES.items() if re.search(pattern, f"{kind} {title}", re.I)
|
|
1565
|
+
]
|
|
1566
|
+
inferred = args.category if args.category in matches else next(iter(matches), "other")
|
|
1567
|
+
path = row.get("FILE_LINK") or ""
|
|
1568
|
+
if args.category != "all" and inferred != args.category or not path.startswith("/listedco/"):
|
|
1569
|
+
continue
|
|
1570
|
+
rows.append(
|
|
1571
|
+
{
|
|
1572
|
+
"title": title,
|
|
1573
|
+
"date": datetime.strptime(row["DATE_TIME"], "%d/%m/%Y %H:%M").date().isoformat(),
|
|
1574
|
+
"category": inferred,
|
|
1575
|
+
"pdf_url": f"https://www1.hkexnews.hk{path}",
|
|
1576
|
+
"type": kind,
|
|
1577
|
+
}
|
|
1578
|
+
)
|
|
1579
|
+
total = int(data.get("recordCnt") or len(items))
|
|
1580
|
+
return {
|
|
1581
|
+
"symbol": args.symbol,
|
|
1582
|
+
"market": "HK",
|
|
1583
|
+
"source": "hkexnews",
|
|
1584
|
+
"start_date": start.isoformat(),
|
|
1585
|
+
"end_date": end.isoformat(),
|
|
1586
|
+
"total": total if args.category == "all" else len(rows),
|
|
1587
|
+
"total_basis": "provider_record_count" if args.category == "all" else "matched_in_loaded_rows",
|
|
1588
|
+
"announcements": rows[: args.top],
|
|
1589
|
+
"truncated": total > len(items) or len(rows) > args.top,
|
|
1590
|
+
"body_extracted": False,
|
|
1591
|
+
"requested_category": args.category,
|
|
1592
|
+
}
|
|
1363
1593
|
|
|
1364
1594
|
|
|
1365
1595
|
def announcements(args):
|
|
1366
1596
|
import requests
|
|
1367
1597
|
|
|
1598
|
+
if detect_market(args.symbol) == "US":
|
|
1599
|
+
import sec_edgar
|
|
1600
|
+
|
|
1601
|
+
if args.symbol.isdigit():
|
|
1602
|
+
# Bare digits are an unsuffixed HK code as often as a CIK; only CIK-prefixed numbers reach SEC.
|
|
1603
|
+
raise ValueError("numeric symbols need a CIK prefix (CIK320193) or HK .HK suffix (00700.HK)")
|
|
1604
|
+
end = date.fromisoformat(args.end_date) if args.end_date else market_now("US").date()
|
|
1605
|
+
start = date.fromisoformat(args.start_date) if args.start_date else end - timedelta(days=90)
|
|
1606
|
+
if start > end or not 1 <= args.top <= 500:
|
|
1607
|
+
raise ValueError("start_date must not exceed end_date and top must be in 1..500")
|
|
1608
|
+
symbol = (
|
|
1609
|
+
re.sub(r"^CIK", "", args.symbol, flags=re.I)
|
|
1610
|
+
if re.fullmatch(r"(?i)CIK[0-9]{1,10}", args.symbol)
|
|
1611
|
+
else args.symbol
|
|
1612
|
+
)
|
|
1613
|
+
return sec_edgar.filings(symbol, args.category, start, end, args.top)
|
|
1368
1614
|
market, symbol = stock_identity(args.symbol)
|
|
1615
|
+
if market == "HK":
|
|
1616
|
+
return hk_announcements(args, symbol)
|
|
1369
1617
|
if market != "A" or not 1 <= args.top <= 500 or args.category not in ANNOUNCEMENT_CATEGORIES:
|
|
1370
|
-
raise ValueError("announcements require an A-share stock, valid category and top in 1..500")
|
|
1371
|
-
end = date.fromisoformat(args.end_date) if args.end_date else date
|
|
1618
|
+
raise ValueError("announcements require an A-share or US stock, valid category and top in 1..500")
|
|
1619
|
+
end = date.fromisoformat(args.end_date) if args.end_date else _cn_now().date()
|
|
1372
1620
|
start = date.fromisoformat(args.start_date) if args.start_date else end - timedelta(days=90)
|
|
1373
1621
|
if start > end:
|
|
1374
1622
|
raise ValueError("start_date must not exceed end_date")
|