From 4c88b38a7ea8c8f67f3ad70c4d437682f7af3197 Mon Sep 17 00:00:00 2001 From: xucheng Date: Thu, 6 Nov 2025 19:58:36 +0800 Subject: [PATCH] chore: sync backend/frontend/docs changes before push --- backend/app/data_manager.py | 47 +-- backend/app/data_providers/finnhub.py | 283 ++++++++++++++---- backend/app/data_providers/yfinance.py | 4 +- backend/app/routers/financial.py | 126 +++++++- docs/{ => 已完成任务}/tasks.md | 10 +- .../未完成任务/us_market_integration_tasks.md | 67 +++++ frontend/src/app/fonts/README.md | 2 + frontend/src/app/report/[symbol]/page.tsx | 68 ++--- frontend/src/hooks/useApi.ts | 46 ++- frontend/src/lib/prisma.ts | 2 + 10 files changed, 519 insertions(+), 136 deletions(-) rename docs/{ => 已完成任务}/tasks.md (88%) create mode 100644 docs/未完成任务/us_market_integration_tasks.md diff --git a/backend/app/data_manager.py b/backend/app/data_manager.py index 6664daf..785ae8e 100644 --- a/backend/app/data_manager.py +++ b/backend/app/data_manager.py @@ -2,6 +2,7 @@ import yaml import os import json from typing import Any, Dict, List, Optional +from numbers import Number from app.data_providers.base import BaseDataProvider from app.data_providers.tushare import TushareProvider # from app.data_providers.ifind import TonghsProvider @@ -44,20 +45,9 @@ class DataManager: self.providers = {} - # Build provider base config from environment variables and config/config.json, then initialize providers + # Build provider base config ONLY from config/config.json (do not read env vars) base_cfg: Dict[str, Any] = {"data_sources": {}} - # 1) Prefer env vars when present - for name, source_config in (self.config.get('data_sources') or {}).items(): - env_var = source_config.get('api_key_env') - if env_var: - api_key = os.getenv(env_var) - if api_key: - base_cfg["data_sources"][name] = {"api_key": api_key} - else: - logger.warning(f"Env var '{env_var}' for provider '{name}' not set; will try config.json.") - - # 2) Fallback to config/config.json if tokens are provided there try: # Use the same REPO_ROOT calculation as data_sources.yaml current_dir = os.path.dirname(__file__) @@ -76,7 +66,7 @@ class DataManager: cfg_json = json.load(jf) ds_from_json = (cfg_json.get("data_sources") or {}) for name, node in ds_from_json.items(): - if name not in base_cfg["data_sources"] and node.get("api_key"): + if node.get("api_key"): base_cfg["data_sources"][name] = {"api_key": node.get("api_key")} logger.info(f"Loaded API key for provider '{name}' from config.json") else: @@ -106,11 +96,9 @@ class DataManager: } for name, provider_class in provider_map.items(): - token = None + token = base_cfg.get("data_sources", {}).get(name, {}).get("api_key") source_config = self.config['data_sources'].get(name, {}) - if source_config and source_config.get('api_key_env'): - token = base_cfg.get("data_sources", {}).get(name, {}).get("api_key") - + # Initialize the provider if a token is found or not required if token or not source_config.get('api_key_env'): try: @@ -118,7 +106,7 @@ class DataManager: except Exception as e: logger.error(f"Failed to initialize provider '{name}': {e}") else: - logger.warning(f"Provider '{name}' requires token env '{source_config.get('api_key_env')}', but none provided. Skipping.") + logger.warning(f"Provider '{name}' requires API key but none provided in config.json. Skipping.") def _detect_market(self, stock_code: str) -> str: if stock_code.endswith(('.SH', '.SZ')): @@ -143,7 +131,17 @@ class DataManager: try: method = getattr(provider, method_name) data = await method(stock_code=stock_code, **kwargs) - if data is not None and (not isinstance(data, list) or data): + is_success = False + if data is None: + is_success = False + elif isinstance(data, list): + is_success = len(data) > 0 + elif isinstance(data, dict): + is_success = len(data) > 0 + else: + is_success = True + + if is_success: logger.info(f"Data successfully fetched from '{provider_name}' for '{stock_code}'.") return data except Exception as e: @@ -171,11 +169,18 @@ class DataManager: for key, value in report.items(): if key in ['ts_code', 'stock_code', 'year', 'end_date', 'period', 'ann_date', 'f_ann_date', 'report_type']: continue - if isinstance(value, (int, float)) and value is not None: + # Accept numpy/pandas numeric types as well as builtin numbers + if value is not None and isinstance(value, Number): if key not in series: series[key] = [] if not any(d['year'] == year for d in series[key]): - series[key].append({"year": year, "value": value}) + # Store as builtin float to avoid JSON serialization issues + try: + numeric_value = float(value) + except Exception: + # Fallback: skip if cannot coerce to float + continue + series[key].append({"year": year, "value": numeric_value}) return series else: return {} diff --git a/backend/app/data_providers/finnhub.py b/backend/app/data_providers/finnhub.py index f2e7c5e..f13481d 100644 --- a/backend/app/data_providers/finnhub.py +++ b/backend/app/data_providers/finnhub.py @@ -5,6 +5,7 @@ import pandas as pd from datetime import datetime, timedelta import asyncio import logging +import httpx logger = logging.getLogger(__name__) @@ -14,11 +15,37 @@ class FinnhubProvider(BaseDataProvider): if not self.token: raise ValueError("Finnhub API key not provided.") self.client = finnhub.Client(api_key=self.token) + try: + masked = f"***{self.token[-4:]}" if isinstance(self.token, str) and len(self.token) >= 4 else "***" + logger.info(f"[Finnhub] client initialized (token={masked})") + except Exception: + # 避免日志失败影响初始化 + pass async def get_stock_basic(self, stock_code: str) -> Optional[Dict[str, Any]]: - async def _fetch(): + def _fetch(): try: - profile = self.client.company_profile2(symbol=stock_code) + profile = None + try: + profile = self.client.company_profile2(symbol=stock_code) + logger.debug(f"[Finnhub] SDK company_profile2 ok symbol={stock_code} name={profile.get('name') if isinstance(profile, dict) else None}") + except Exception as e: + logger.warning(f"[Finnhub] SDK company_profile2 failed for {stock_code}: {e}") + # Fallback to direct HTTP if SDK call fails + try: + resp = httpx.get( + 'https://finnhub.io/api/v1/stock/profile2', + params={'symbol': stock_code}, + headers={'X-Finnhub-Token': self.token}, + timeout=20.0, + ) + logger.debug(f"[Finnhub] HTTP profile2 status={resp.status_code} len={len(resp.text)}") + if resp.status_code == 200: + profile = resp.json() + else: + logger.error(f"[Finnhub] HTTP profile2 failed status={resp.status_code} body={resp.text[:200]}") + except Exception: + profile = None if not profile: return None @@ -39,13 +66,18 @@ class FinnhubProvider(BaseDataProvider): return await loop.run_in_executor(None, _fetch) async def get_daily_price(self, stock_code: str, start_date: str, end_date: str) -> List[Dict[str, Any]]: - async def _fetch(): + def _fetch(): try: start_ts = int(datetime.strptime(start_date, '%Y%m%d').timestamp()) end_ts = int(datetime.strptime(end_date, '%Y%m%d').timestamp()) + logger.debug(f"[Finnhub] stock_candles symbol={stock_code} D {start_date}->{end_date}") res = self.client.stock_candles(stock_code, 'D', start_ts, end_ts) if res.get('s') != 'ok': + try: + logger.warning(f"[Finnhub] stock_candles not ok symbol={stock_code} status={res.get('s')}") + except Exception: + pass return [] df = pd.DataFrame(res) @@ -68,76 +100,211 @@ class FinnhubProvider(BaseDataProvider): return await loop.run_in_executor(None, _fetch) async def get_financial_statements(self, stock_code: str, report_dates: List[str]) -> List[Dict[str, Any]]: - async def _fetch(): + def _fetch(): try: - # Finnhub provides financials as a whole, not by specific date ranges in one call - # We fetch all available and then filter. - # Note: 'freq' can be 'annual' or 'quarterly'. We'll use annual. - res = self.client.financials_reported(symbol=stock_code, freq='annual') + # 1) 拉取年度报表（financials_reported, annual） + res = None + try: + res = self.client.financials_reported(symbol=stock_code, freq='annual') + except Exception as e: + logger.warning(f"[Finnhub] SDK financials_reported failed for {stock_code}: {e}") + # Fallback: direct HTTP + try: + r = httpx.get( + 'https://finnhub.io/api/v1/stock/financials-reported', + params={'symbol': stock_code, 'freq': 'annual'}, + headers={'X-Finnhub-Token': self.token}, + timeout=30.0, + ) + logger.debug(f"[Finnhub] HTTP financials-reported status={r.status_code} len={len(r.text)}") + if r.status_code == 200: + res = r.json() + else: + logger.error(f"[Finnhub] HTTP financials-reported failed status={r.status_code} body={r.text[:300]}") + except Exception: + res = None if not res or not res.get('data'): + logger.warning(f"[Finnhub] financials-reported empty for {stock_code}") return [] df = pd.DataFrame(res['data']) + if df.empty: + logger.warning(f"[Finnhub] financials-reported dataframe empty for {stock_code}") + return [] - # Filter by requested dates - years_to_fetch = {date[:4] for date in report_dates} - df = df[df['year'].astype(str).isin(years_to_fetch)] + # 2) 仅保留请求的年份 + years_to_fetch = {str(date)[:4] for date in report_dates} + logger.debug(f"[Finnhub] filter years {sorted(list(years_to_fetch))} before={len(df)}") + if 'year' in df.columns: + df = df[df['year'].astype(str).isin(years_to_fetch)] + # 兜底：如果缺失 year 列，则用 endDate 推断 + if 'year' not in df.columns and 'endDate' in df.columns: + df = df[df['endDate'].astype(str).str[:4].isin(years_to_fetch)] - # The data is deeply nested in 'report'. We need to extract and pivot it. - all_reports = [] - for index, row in df.iterrows(): - report_data = {'ts_code': stock_code, 'end_date': row['endDate']} - - # Extract concepts from balance sheet, income statement, and cash flow - for item in row['report'].get('bs', []): - report_data[item['concept']] = item['value'] - for item in row['report'].get('ic', []): - report_data[item['concept']] = item['value'] - for item in row['report'].get('cf', []): - report_data[item['concept']] = item['value'] - - all_reports.append(report_data) + if df.empty: + logger.warning(f"[Finnhub] financials-reported no rows after filter for {stock_code}") + return [] - # Further normalization of keys would be needed here to match a common format - # e.g. 'AssetsTotal' -> 'total_assets' - # This is a complex task and depends on the desired final schema. - - # We will now normalize and calculate derived metrics - normalized_reports = [] - for report in all_reports: - normalized_report = { - "ts_code": report.get("ts_code"), - "end_date": report.get("end_date"), - # Balance Sheet - "total_assets": report.get("AssetsTotal"), - "total_liabilities": report.get("LiabilitiesTotal"), - "equity": report.get("StockholdersEquityTotal"), - # Income Statement - "revenue": report.get("RevenuesTotal"), - "net_income": report.get("NetIncomeLoss"), - "gross_profit": report.get("GrossProfit"), - # Cash Flow - "net_cash_flow_operating": report.get("NetCashFlowOperating"), + def _normalize_key(s: Optional[str]) -> str: + if not isinstance(s, str): + return "" + return ''.join(ch.lower() for ch in s if ch.isalnum()) + + def pick(report_block: List[Dict[str, Any]], concept_candidates: List[str], label_candidates: List[str] = []) -> Optional[float]: + if not report_block: + return None + try: + by_concept = { _normalize_key(item.get('concept')): item.get('value') for item in report_block if isinstance(item, dict) } + by_label = { _normalize_key(item.get('label')): item.get('value') for item in report_block if isinstance(item, dict) } + except Exception: + return None + for key in concept_candidates: + v = by_concept.get(_normalize_key(key)) + if v is not None: + try: + return float(v) + except Exception: + continue + for key in label_candidates: + v = by_label.get(_normalize_key(key)) + if v is not None: + try: + return float(v) + except Exception: + continue + return None + + # 3) 遍历年度记录，展开并标准化字段名 + flat_reports: List[Dict[str, Any]] = [] + for _, row in df.iterrows(): + bs = (row.get('report') or {}).get('bs', []) + ic = (row.get('report') or {}).get('ic', []) + cf = (row.get('report') or {}).get('cf', []) + + end_date = str(row.get('endDate') or '') + + revenue = pick( + ic, + concept_candidates=['Revenues', 'RevenueFromContractWithCustomerExcludingAssessedTax', 'SalesRevenueNet', 'Revenue', 'RevenuesNet', 'SalesRevenueGoodsNet'], + label_candidates=['Total revenue', 'Revenue', 'Sales revenue'] + ) + net_income = pick( + ic, + concept_candidates=['NetIncomeLoss', 'ProfitLoss', 'NetIncomeLossAvailableToCommonStockholdersBasic', 'NetIncomeLossAvailableToCommonStockholdersDiluted'], + label_candidates=['Net income', 'Net income (loss)'] + ) + gross_profit = pick( + ic, + concept_candidates=['GrossProfit'], + label_candidates=['Gross profit'] + ) + + total_assets = pick( + bs, + concept_candidates=['Assets', 'AssetsTotal', 'AssetsCurrentAndNoncurrent', 'AssetsIncludingAssetsMeasuredAtFairValue'], + label_candidates=['Total assets'] + ) + total_equity = pick( + bs, + concept_candidates=['StockholdersEquityIncludingPortionAttributableToNoncontrollingInterest', 'StockholdersEquity', 'StockholdersEquityTotal', 'Equity'], + label_candidates=['Total equity', "Stockholders' equity"] + ) + goodwill = pick( + bs, + concept_candidates=['Goodwill', 'GoodwillAndIntangibleAssets'], + label_candidates=['Goodwill', 'Goodwill and intangible assets'] + ) + + n_cashflow_act = pick( + cf, + concept_candidates=['NetCashProvidedByUsedInOperatingActivities', 'NetCashProvidedByUsedInOperatingActivitiesContinuingOperations', 'NetCashFlowOperating'], + label_candidates=['Net cash provided by operating activities'] + ) + capex = pick( + cf, + concept_candidates=['CapitalExpenditures', 'PaymentsToAcquirePropertyPlantAndEquipment', 'PaymentsToAcquireProductiveAssets'], + label_candidates=['Capital expenditures'] + ) + + # 计算衍生指标 + free_cash_flow = None + if isinstance(n_cashflow_act, (int, float)) and isinstance(capex, (int, float)): + free_cash_flow = n_cashflow_act - capex + + normalized = { + # 基本元字段 + 'ts_code': stock_code, + 'end_date': end_date, # DataManager 会从这里抽取 year + + # 标准命名（见 financial_data_dictionary） + 'revenue': revenue, + 'n_income': net_income, + 'gross_profit': gross_profit, + + 'total_assets': total_assets, + 'total_hldr_eqy_exc_min_int': total_equity, + 'goodwill': goodwill, + + 'n_cashflow_act': n_cashflow_act, + 'c_pay_acq_const_fiolta': capex, + '__free_cash_flow': free_cash_flow, } - # Calculate derived metrics - if normalized_report["revenue"] and normalized_report["revenue"] > 0: - normalized_report["gross_margin"] = (normalized_report["gross_profit"] / normalized_report["revenue"]) if normalized_report["gross_profit"] else None - normalized_report["net_margin"] = (normalized_report["net_income"] / normalized_report["revenue"]) if normalized_report["net_income"] else None - - if normalized_report["total_assets"] and normalized_report["total_assets"] > 0: - normalized_report["roa"] = (normalized_report["net_income"] / normalized_report["total_assets"]) if normalized_report["net_income"] else None + # 一些常用比率（若有足够数据则计算），命名对齐文档 + if isinstance(revenue, (int, float)) and revenue > 0 and isinstance(gross_profit, (int, float)): + normalized['grossprofit_margin'] = gross_profit / revenue + if isinstance(revenue, (int, float)) and revenue > 0 and isinstance(net_income, (int, float)): + normalized['netprofit_margin'] = net_income / revenue + if isinstance(total_assets, (int, float)) and total_assets > 0 and isinstance(net_income, (int, float)): + normalized['roa'] = net_income / total_assets + if isinstance(total_equity, (int, float)) and total_equity > 0 and isinstance(net_income, (int, float)): + normalized['roe'] = net_income / total_equity - if normalized_report["equity"] and normalized_report["equity"] > 0: - normalized_report["roe"] = (normalized_report["net_income"] / normalized_report["equity"]) if normalized_report["net_income"] else None - - normalized_reports.append(normalized_report) + flat_reports.append(normalized) + try: + logger.debug( + f"[Finnhub] row endDate={end_date} revenue={revenue} net_income={net_income} gross_profit={gross_profit} " + f"assets={total_assets} equity={total_equity} goodwill={goodwill} n_cfo={n_cashflow_act} capex={capex}" + ) + except Exception: + pass - return normalized_reports + # Convert flat reports to series dict directly to match DataManager expected format + series: Dict[str, List[Dict[str, Any]]] = {} + for report in flat_reports: + end_date = str(report.get('end_date') or '') + year = end_date[:4] if len(end_date) >= 4 else None + if not year: + continue + period = f"{year}1231" + + for key, value in report.items(): + if key in ['ts_code', 'end_date']: + continue + # Only collect numeric values + try: + if value is None: + continue + num = float(value) + except Exception: + continue + if key not in series: + series[key] = [] + # Avoid duplicate period entries + exists = any(dp.get('period') == period for dp in series[key]) + if not exists: + series[key].append({'period': period, 'value': num}) + + try: + total_points = sum(len(v) for v in series.values()) + logger.info(f"[Finnhub] built series for {stock_code} keys={len(series)} points={total_points}") + except Exception: + pass + return series except Exception as e: logger.error(f"Finnhub get_financial_statements failed for {stock_code}: {e}") return [] - + loop = asyncio.get_event_loop() return await loop.run_in_executor(None, _fetch) diff --git a/backend/app/data_providers/yfinance.py b/backend/app/data_providers/yfinance.py index 1c92de9..29962f6 100644 --- a/backend/app/data_providers/yfinance.py +++ b/backend/app/data_providers/yfinance.py @@ -21,7 +21,7 @@ class YfinanceProvider(BaseDataProvider): return stock_code async def get_stock_basic(self, stock_code: str) -> Optional[Dict[str, Any]]: - async def _fetch(): + def _fetch(): try: ticker = yf.Ticker(self._map_stock_code(stock_code)) info = ticker.info @@ -44,7 +44,7 @@ class YfinanceProvider(BaseDataProvider): return await loop.run_in_executor(None, _fetch) async def get_daily_price(self, stock_code: str, start_date: str, end_date: str) -> List[Dict[str, Any]]: - async def _fetch(): + def _fetch(): try: # yfinance date format is YYYY-MM-DD start_fmt = datetime.strptime(start_date, '%Y%m%d').strftime('%Y-%m-%d') diff --git a/backend/app/routers/financial.py b/backend/app/routers/financial.py index d1529e2..544c2b7 100644 --- a/backend/app/routers/financial.py +++ b/backend/app/routers/financial.py @@ -5,6 +5,7 @@ import json import os import time from datetime import datetime, timezone, timedelta +from enum import Enum from typing import Dict, List from fastapi import APIRouter, HTTPException, Query @@ -46,6 +47,13 @@ def get_dm(): router = APIRouter() + +class MarketEnum(str, Enum): + cn = "cn" + us = "us" + hk = "hk" + jp = "jp" + # Load metric config from file (project root is repo root, not backend/) # routers/ -> app/ -> backend/ -> repo root REPO_ROOT = os.path.abspath(os.path.join(os.path.dirname(__file__), "..", "..", "..")) @@ -257,9 +265,10 @@ async def get_financial_config(): return FinancialConfigResponse(api_groups=api_groups) -@router.get("/china/{ts_code}", response_model=BatchFinancialDataResponse) -async def get_china_financials( - ts_code: str, +@router.get("/{market}/{stock_code}", response_model=BatchFinancialDataResponse) +async def get_financials( + market: MarketEnum, + stock_code: str, years: int = Query(10, ge=1, le=10), ): # Load metric config @@ -274,11 +283,11 @@ async def get_china_financials( steps: List[StepRecord] = [] # Get company name - company_name = ts_code + company_name = stock_code try: - basic_data = await get_dm().get_stock_basic(stock_code=ts_code) + basic_data = await get_dm().get_stock_basic(stock_code=stock_code) if basic_data: - company_name = basic_data.get("name", ts_code) + company_name = basic_data.get("name", stock_code) except Exception: pass # Continue without it @@ -295,7 +304,7 @@ async def get_china_financials( steps.append(step_financials) # Fetch all financial statements at once (already in series format from provider) - series = await get_dm().get_financial_statements(stock_code=ts_code, report_dates=report_dates) + series = await get_dm().get_financial_statements(stock_code=stock_code, report_dates=report_dates) # Get the latest current year report period for market data latest_current_year_report = None @@ -322,7 +331,8 @@ async def get_china_financials( has_daily_basic = bool(api_groups.get("daily_basic")) has_daily = bool(api_groups.get("daily")) - if has_daily_basic or has_daily: + # 目前仅对中国市场启用 daily_basic/daily 数据拉取，其他市场由对应 provider 后续实现 + if market == MarketEnum.cn and (has_daily_basic or has_daily): step_market = StepRecord(name="拉取市值与股价", start_ts=datetime.now(timezone.utc).isoformat(), status="running") steps.append(step_market) @@ -339,7 +349,7 @@ async def get_china_financials( try: if has_daily_basic: - db_rows = await get_dm().get_data('get_daily_basic_points', stock_code=ts_code, trade_dates=market_dates) + db_rows = await get_dm().get_data('get_daily_basic_points', stock_code=stock_code, trade_dates=market_dates) if isinstance(db_rows, list): for row in db_rows: trade_date = row.get('trade_date') or row.get('trade_dt') or row.get('date') @@ -369,7 +379,7 @@ async def get_china_financials( else: series[key].append({"period": period, "value": value}) if has_daily: - d_rows = await get_dm().get_data('get_daily_points', stock_code=ts_code, trade_dates=market_dates) + d_rows = await get_dm().get_data('get_daily_points', stock_code=stock_code, trade_dates=market_dates) if isinstance(d_rows, list): for row in d_rows: trade_date = row.get('trade_date') or row.get('trade_dt') or row.get('date') @@ -413,10 +423,23 @@ async def get_china_financials( if not series: raise HTTPException(status_code=502, detail={"message": "No data returned from any data source", "errors": errors}) - # Truncate periods and sort (the data should already be mostly correct, but we ensure) - for key, arr in series.items(): - # Deduplicate and sort desc by period, then cut to requested periods, and return asc - uniq = {item["period"]: item for item in arr} + # 统一 period 字段；若仅有 year 则映射为 YYYY1231；然后去重与排序 + for key, arr in list(series.items()): + normalized: List[Dict] = [] + for item in arr: + period = item.get("period") + if not period: + year = item.get("year") + if year: + period = f"{str(year)}1231" + if not period: + # 跳过无法确定 period 的项 + continue + value = item.get("value") + normalized.append({"period": str(period), "value": value}) + + # Deduplicate by period + uniq = {it["period"]: it for it in normalized} arr_sorted_desc = sorted(uniq.values(), key=lambda x: x["period"], reverse=True) arr_limited = arr_sorted_desc[:years] arr_sorted = sorted(arr_limited, key=lambda x: x["period"]) @@ -438,7 +461,7 @@ async def get_china_financials( steps=steps, ) - return BatchFinancialDataResponse(ts_code=ts_code, name=company_name, series=series, meta=meta) + return BatchFinancialDataResponse(ts_code=stock_code, name=company_name, series=series, meta=meta) @router.get("/china/{ts_code}/company-profile", response_model=CompanyProfileResponse) @@ -800,6 +823,66 @@ async def get_today_snapshot(ts_code: str): raise HTTPException(status_code=500, detail=f"Failed to fetch snapshot: {e}") +@router.get("/{market}/{stock_code}/snapshot", response_model=TodaySnapshotResponse) +async def get_market_snapshot(market: MarketEnum, stock_code: str): + """ + 市场无关的“昨日快照”接口。 + - CN: 复用中国市场的快照逻辑（daily_basic/daily）。 + - 其他市场: 兜底使用日行情获取最近交易日收盘价；其余字段暂返回空值。 + """ + if market == MarketEnum.cn: + return await get_today_snapshot(stock_code) + + try: + # 公司名称（可选） + company_name = None + try: + basic = await get_dm().get_stock_basic(stock_code=stock_code) + if basic: + company_name = basic.get("name") + except Exception: + company_name = None + + base_dt = (datetime.now() - timedelta(days=1)).date() + base_str = base_dt.strftime("%Y%m%d") + + # 为了稳妥拿到最近交易日，回看近 10 天 + start_dt = base_dt - timedelta(days=10) + start_str = start_dt.strftime("%Y%m%d") + end_dt = base_dt + timedelta(days=1) + end_str = end_dt.strftime("%Y%m%d") + + rows = await get_dm().get_daily_price(stock_code=stock_code, start_date=start_str, end_date=end_str) + trade_date = None + close = None + if isinstance(rows, list) and rows: + # 选择 <= base_str 的最后一条记录 + try: + candidates = [r for r in rows if str(r.get("trade_date") or r.get("date") or "") <= base_str] + if candidates: + last = sorted(candidates, key=lambda r: str(r.get("trade_date") or r.get("date") or ""))[-1] + trade_date = str(last.get("trade_date") or last.get("date") or base_str) + close = last.get("close") + except Exception: + pass + + if trade_date is None: + trade_date = base_str + + return TodaySnapshotResponse( + ts_code=stock_code, + trade_date=trade_date, + name=company_name, + close=close, + pe=None, + pb=None, + dv_ratio=None, + total_mv=None, + ) + except Exception as e: + raise HTTPException(status_code=500, detail=f"Failed to fetch snapshot: {e}") + + @router.get("/china/{ts_code}/analysis/{analysis_type}/stream") async def stream_analysis( ts_code: str, @@ -925,3 +1008,16 @@ async def stream_analysis( "X-Accel-Buffering": "no", } return StreamingResponse(streamer(), media_type="text/plain; charset=utf-8", headers=headers) + + +@router.get("/{market}/{stock_code}/analysis/{analysis_type}/stream") +async def stream_analysis_market( + market: MarketEnum, + stock_code: str, + analysis_type: str, + company_name: str = Query(None, description="Company name for better context"), +): + """ + 市场无关的分析流接口。逻辑与中国市场一致，仅路径不同。 + """ + return await stream_analysis(stock_code, analysis_type, company_name) diff --git a/docs/tasks.md b/docs/已完成任务/tasks.md similarity index 88% rename from docs/tasks.md rename to docs/已完成任务/tasks.md index 250a722..b8bedb1 100644 --- a/docs/tasks.md +++ b/docs/已完成任务/tasks.md @@ -19,9 +19,9 @@ - [x] **T2.1 [Backend/DB]**: 根据设计文档，使用SQLAlchemy ORM定义`Report`, `AnalysisModule`, `ProgressTracking`, `SystemConfig`四个核心数据模型。 **[完成 - 2025-10-21]** - [x] **T2.2 [Backend/DB]**: 创建第一个Alembic迁移脚本，在数据库中生成上述四张表。 **[完成 - 2025-10-21]** - [x] **T2.3 [Backend]**: 实现`ConfigManager`服务，完成从`config.json`加载配置并与数据库配置合并的逻辑。 **[完成 - 2025-10-21]** -- **T2.4 [Backend/API]**: 创建Pydantic Schema，用于配置接口的请求和响应 (`ConfigResponse`, `ConfigUpdateRequest`, `ConfigTestRequest`, `ConfigTestResponse`)。 -- **T2.5 [Backend/API]**: 实现`/api/config`的`GET`和`PUT`端点，用于读取和更新系统配置。 -- **T2.6 [Backend/API]**: 实现`/api/config/test`的`POST`端点，用于验证数据库连接等配置的有效性。 +- [x] **T2.4 [Backend/API]**: 创建Pydantic Schema，用于配置接口的请求和响应 (`ConfigResponse`, `ConfigUpdateRequest`, `ConfigTestRequest`, `ConfigTestResponse`)。 +- [x] **T2.5 [Backend/API]**: 实现`/api/config`的`GET`和`PUT`端点，用于读取和更新系统配置。 +- [x] **T2.6 [Backend/API]**: 实现`/api/config/test`的`POST`端点，用于验证数据库连接等配置的有效性。 ## Phase 3: 前端基础与配置页面 (P1) @@ -39,8 +39,8 @@ 此阶段是项目的核心，重点开发后端的报告生成流程和前端的实时进度展示。 -- **T4.1 [Backend/Service]**: 实现`DataSourceManager`，封装对Tushare和Yahoo Finance的数据获取逻辑。 -- **T4.2 [Backend/Service]**: 实现`AIService`，封装对Google Gemini API的调用逻辑，包括Token使用统计。 +- [x] **T4.1 [Backend/Service]**: 实现`DataSourceManager`，封装对Tushare和Yahoo Finance的数据获取逻辑。 +- [x] **T4.2 [Backend/Service]**: 实现`AIService`，封装对Google Gemini API的调用逻辑，包括Token使用统计。 - **T4.3 [Backend/Service]**: 实现`ProgressTracker`服务，提供`initialize`, `start_step`, `complete_step`, `get_progress`等方法，并与数据库交互。 - **T4.4 [Backend/Service]**: 定义`AnalysisModule`的基类/接口，并初步实现一到两个模块（如`FinancialDataModule`）作为示例。 - **T4.5 [Backend/Service]**: 实现核心的`ReportGenerator`服务，编排数据获取、各分析模块调用、进度更新的完整流程。 diff --git a/docs/未完成任务/us_market_integration_tasks.md b/docs/未完成任务/us_market_integration_tasks.md new file mode 100644 index 0000000..763ff4d --- /dev/null +++ b/docs/未完成任务/us_market_integration_tasks.md @@ -0,0 +1,67 @@ +# 美国市场数据集成任务清单 + +本文档用于跟踪和管理为项目集成美国市场数据（使用 Finnhub 作为数据源）所需的各项开发任务。 + +## 任务列表 + +- [x] **后端：实现 FinnhubProvider 数据映射** + - **目标**：根据 `docs/financial_data_dictionary.md` 中的定义，在 `backend/app/data_providers/finnhub.py` 文件中，完成从 Finnhub API 原始数据到系统标准字段的完整映射。 + - **关键点**： + - [x] 处理直接映射的字段。 + - [x] 实现所有需要通过计算得出的衍生指标。 + - [x] 确保处理 `null` 或空值，避免计算错误。 + - [x] 验证返回的数据结构符合 `DataManager` 的预期。 + +- [x] **后端：按市场分段的 API 路由** + - **目标**：在 `backend/app/routers/financial.py` 中，将现有的 `/api/v1/financials/china/{ts_code}` 改为按市场分段：`/api/v1/financials/{market}/{stock_code}`（示例：`/api/v1/financials/us/AAPL`，`/api/v1/financials/cn/600519.SH`）。 + - **关键点**： + - [x] 去除硬编码的 `china`，新增路径参数 `market`，并对取值做校验（`cn/us/hk/jp`）。 + - [x] 使用单一处理函数，根据 `market` 分派到相应的数据提供方与代码格式规范。 + +- [x] **前端：更新 API 调用** + - **目标**：修改前端调用，基于用户选择的市场与股票代码，请求新的按市场分段路由。 + - **关键点**： + - [x] 替换 `useChinaFinancials`，新增通用 `useFinancials(market, stockCode, years)`。 + - [x] 将请求路径改为 `/api/financials/{market}/{stock_code}?years=...`（代理到后端对应的 `/api/v1/financials/{market}/{stock_code}`）。 + - [ ] 确保展示与错误处理兼容美国、香港、日本等市场。 + +- [ ] **测试与验证** + - **目标**：对整个流程进行端到端测试，确保两个市场的功能都稳定可靠。 + - **关键点**： + - [ ] **中国市场回归测试**：使用多个中国 A 股代码测试，确保原有功能不受影响。 + - [ ] **美国市场功能测试**：使用多个美国股票代码（如 `AAPL`, `MSFT`）测试，验证报告能否成功生成。 + - [ ] **数据一致性验证**：抽样对比 Finnhub 返回的数据和前端展示的数据，确保映射和计算的准确性。 + - [ ] **错误处理测试**：测试无效的股票代码，检查系统是否能给出清晰的错误提示。 + + - **前置条件**： + - [ ] 在 `config/config.json` 或环境变量中配置 `FINNHUB_API_KEY`。 + - [ ] 后端已启动（默认 `http://127.0.0.1:8000/api`），前端已启动（默认 `http://127.0.0.1:3000`）。 + + - **接口用例（后端）**： + - [ ] GET `/api/v1/financials/cn/600519.SH?years=10` + - 期望：`200`；返回 `ts_code`、`name`、`series`（含 `revenue`、`n_income` 等关键指标，period/年序列齐全）。 + - [ ] GET `/api/v1/financials/cn/000001.SZ?years=5` + - 期望：`200`；返回与上同，近 5 年序列。 + - [ ] GET `/api/v1/financials/us/AAPL?years=10` + - 期望：`200`；`series` 至少包含：`revenue`、`n_income`、`total_assets`、`total_hldr_eqy_exc_min_int`、`__free_cash_flow`、`grossprofit_margin`、`netprofit_margin`、`roe`、`roa`。 + - [ ] GET `/api/v1/financials/us/MSFT?years=10` + - 期望：`200`；字段与口径同 AAPL。 + - [ ] GET `/api/v1/financials/us/INVALID?years=10` + - 期望：`4xx/5xx`；`detail.message` 含可读错误。 + + - **页面用例（前端）**： + - [ ] 打开 `/report/600519.SH?market=cn` + - 期望：基本信息与“昨日快照”显示；“财务数据（来自 Tushare）”表格展示 10 期内主要指标。 + - [ ] 打开 `/report/000001.SZ?market=cn` + - 期望：与上同；代码规范化逻辑（无后缀时自动补 `.SZ/.SH`）正常。 + - [ ] 打开 `/report/AAPL?market=us` + - 期望：“股价图表”正常；“财务数据”表格展示主要指标（含自由现金流、毛利率、净利率、ROA、ROE）。 + - [ ] 打开 `/report/MSFT?market=us` + - 期望：与上同。 + - [ ] 打开 `/report/INVALID?market=us` + - 期望：顶部状态为“读取失败”并有错误提示文案。 + + - **验收标准**： + - [ ] 中国市场功能无回归；美国市场关键指标齐全、值域合理（百分比类 ∈ [-1000%, 1000%]，金额类为有限数）。 + - [ ] 报错信息清晰可读；网络/密钥缺失时提示明确。 + - [ ] 页内主要表格不出现 `NaN/Infinity`；空值以 `-` 展示。 diff --git a/frontend/src/app/fonts/README.md b/frontend/src/app/fonts/README.md index 5c00bab..f0dbba5 100644 --- a/frontend/src/app/fonts/README.md +++ b/frontend/src/app/fonts/README.md @@ -18,3 +18,5 @@ + + diff --git a/frontend/src/app/report/[symbol]/page.tsx b/frontend/src/app/report/[symbol]/page.tsx index 7d28f39..079042e 100644 --- a/frontend/src/app/report/[symbol]/page.tsx +++ b/frontend/src/app/report/[symbol]/page.tsx @@ -1,7 +1,7 @@ 'use client'; import { useParams, useSearchParams } from 'next/navigation'; -import { useChinaFinancials, useFinancialConfig, useAnalysisConfig, generateFullAnalysis, useChinaSnapshot } from '@/hooks/useApi'; +import { useChinaFinancials, useFinancials, useFinancialConfig, useAnalysisConfig, generateFullAnalysis, useSnapshot } from '@/hooks/useApi'; import { Spinner } from '@/components/ui/spinner'; import { Button } from '@/components/ui/button'; import { CheckCircle, XCircle, RotateCw } from 'lucide-react'; @@ -20,11 +20,18 @@ export default function ReportPage() { const params = useParams(); const searchParams = useSearchParams(); const symbol = params.symbol as string; - const market = (searchParams.get('market') || '').toLowerCase(); + const marketParam = (searchParams.get('market') || '').toLowerCase(); + const normalizedMarket = (() => { + if (marketParam === 'usa') return 'us'; + if (marketParam === 'china') return 'cn'; + if (marketParam === 'hkex') return 'hk'; + if (marketParam === 'jpn') return 'jp'; + return marketParam; + })(); - const displayMarket = market === 'china' ? '中国' : market; + const displayMarket = marketParam === 'china' ? '中国' : marketParam; - const isChina = market === 'china' || market === 'cn'; + const isChina = normalizedMarket === 'cn'; // 规范化中国市场 ts_code：若为6位数字或无后缀，自动推断交易所 const normalizedTsCode = (() => { @@ -40,8 +47,13 @@ export default function ReportPage() { return symbol.toUpperCase(); })(); - const { data: financials, error, isLoading } = useChinaFinancials(isChina ? normalizedTsCode : undefined, 10); - const { data: snapshot, error: snapshotError, isLoading: snapshotLoading } = useChinaSnapshot(isChina ? normalizedTsCode : undefined); + const chinaFin = useChinaFinancials(isChina ? normalizedTsCode : undefined, 10); + const otherFin = useFinancials(!isChina ? normalizedMarket : undefined, !isChina ? symbol : undefined, 10); + const financials = (chinaFin.data ?? otherFin.data) as any; + const error = chinaFin.error ?? otherFin.error; + const isLoading = chinaFin.isLoading || otherFin.isLoading; + const unifiedSymbol = isChina ? normalizedTsCode : symbol; + const { data: snapshot, error: snapshotError, isLoading: snapshotLoading } = useSnapshot(normalizedMarket, unifiedSymbol); const { data: financialConfig } = useFinancialConfig(); const { data: analysisConfig } = useAnalysisConfig(); @@ -139,7 +151,7 @@ export default function ReportPage() { } const runFullAnalysis = async () => { - if (!isChina || !financials || !analysisConfig?.analysis_modules || isAnalysisRunningRef.current) { + if (!financials || !analysisConfig?.analysis_modules || isAnalysisRunningRef.current) { return; } @@ -321,7 +333,7 @@ export default function ReportPage() { }, [startTime]); const retryAnalysis = async (analysisType: string) => { - if (!isChina || !financials || !analysisConfig?.analysis_modules) { + if (!financials || !analysisConfig?.analysis_modules) { return; } analysisFetchedRefs.current[analysisType] = false; @@ -344,7 +356,7 @@ export default function ReportPage() { try { const startedMsLocal = Date.now(); const response = await fetch( - `/api/financials/china/${normalizedTsCode}/analysis/${analysisType}/stream?company_name=${encodeURIComponent(financials?.name || normalizedTsCode)}` + `/api/financials/${normalizedMarket}/${unifiedSymbol}/analysis/${analysisType}/stream?company_name=${encodeURIComponent(financials?.name || unifiedSymbol)}` ); if (!response.ok) { throw new Error(`HTTP error! status: ${response.status}`); @@ -420,7 +432,7 @@ export default function ReportPage() { }; useEffect(() => { - if (!isChina || isLoading || error || !financials || !analysisConfig?.analysis_modules || analysisTypes.length === 0) { + if (isLoading || error || !financials || !analysisConfig?.analysis_modules || analysisTypes.length === 0) { return; } if (isAnalysisRunningRef.current) { @@ -477,7 +489,7 @@ export default function ReportPage() { abortControllerRef.current = new AbortController(); const startedMsLocal = Date.now(); const response = await fetch( - `/api/financials/china/${normalizedTsCode}/analysis/${analysisType}/stream?company_name=${encodeURIComponent(financials?.name || normalizedTsCode)}`, + `/api/financials/${normalizedMarket}/${unifiedSymbol}/analysis/${analysisType}/stream?company_name=${encodeURIComponent(financials?.name || unifiedSymbol)}`, { signal: abortControllerRef.current.signal } ); if (!response.ok) { @@ -572,7 +584,7 @@ export default function ReportPage() { } }; runAnalysesSequentially(); - }, [isChina, isLoading, error, financials, analysisConfig, analysisTypes, normalizedTsCode, manualRunKey]); + }, [isLoading, error, financials, analysisConfig, analysisTypes, normalizedTsCode, manualRunKey]); const stopAll = () => { stopRequestedRef.current = true; @@ -603,7 +615,7 @@ export default function ReportPage() {
股票代码: - {normalizedTsCode} + {unifiedSymbol}
交易市场: @@ -627,8 +639,7 @@ export default function ReportPage() { - {isChina && ( - + 昨日快照 @@ -714,9 +725,7 @@ export default function ReportPage() {
- )} - {isChina && ( - + - )} - {isChina && ( - +
任务状态 @@ -774,11 +781,9 @@ export default function ReportPage() { )} - )}
- {isChina && ( - + 股价图表 财务数据 @@ -795,19 +800,19 @@ export default function ReportPage() {
- 实时股价图表 - {normalizedTsCode} + 实时股价图表 - {unifiedSymbol}
-

财务数据（来自 Tushare）

+

财务数据

{isLoading ? ( @@ -817,11 +822,7 @@ export default function ReportPage() { )}
- {isLoading - ? '正在读取 Tushare 数据…' - : error - ? '读取失败' - : '读取完成'} + {isLoading ? '正在读取数据…' : error ? '读取失败' : '读取完成'}
{error &&

加载失败

} @@ -1527,7 +1528,6 @@ export default function ReportPage() { )} - )} ); } diff --git a/frontend/src/hooks/useApi.ts b/frontend/src/hooks/useApi.ts index b9bd0e3..cb89f86 100644 --- a/frontend/src/hooks/useApi.ts +++ b/frontend/src/hooks/useApi.ts @@ -63,7 +63,7 @@ export function useFinancialConfig() { export function useChinaFinancials(ts_code?: string, years: number = 10) { return useSWR( - ts_code ? `/api/financials/china/${encodeURIComponent(ts_code)}?years=${encodeURIComponent(String(years))}` : null, + ts_code ? `/api/financials/cn/${encodeURIComponent(ts_code)}?years=${encodeURIComponent(String(years))}` : null, fetcher, { revalidateOnFocus: false, // 不在窗口聚焦时重新验证 @@ -74,6 +74,28 @@ export function useChinaFinancials(ts_code?: string, years: number = 10) { ); } +export function useFinancials(market?: string, stockCode?: string, years: number = 10) { + const normalizeMarket = (m?: string) => { + const t = (m || '').toLowerCase(); + if (t === 'usa') return 'us'; + if (t === 'china') return 'cn'; + if (t === 'hkex') return 'hk'; + if (t === 'jpn') return 'jp'; + return t; + }; + const mkt = normalizeMarket(market); + return useSWR( + mkt && stockCode ? `/api/financials/${encodeURIComponent(mkt)}/${encodeURIComponent(stockCode)}?years=${encodeURIComponent(String(years))}` : null, + fetcher, + { + revalidateOnFocus: false, + revalidateOnReconnect: false, + dedupingInterval: 300000, + errorRetryCount: 1, + } + ); +} + export function useAnalysisConfig() { return useSWR('/api/financials/analysis-config', fetcher); } @@ -124,3 +146,25 @@ export function useChinaSnapshot(ts_code?: string) { } ); } + +export function useSnapshot(market?: string, stockCode?: string) { + const normalizeMarket = (m?: string) => { + const t = (m || '').toLowerCase(); + if (t === 'usa') return 'us'; + if (t === 'china') return 'cn'; + if (t === 'hkex') return 'hk'; + if (t === 'jpn') return 'jp'; + return t; + }; + const mkt = normalizeMarket(market); + return useSWR( + mkt && stockCode ? `/api/financials/${encodeURIComponent(mkt)}/${encodeURIComponent(stockCode)}/snapshot` : null, + fetcher, + { + revalidateOnFocus: false, + revalidateOnReconnect: false, + dedupingInterval: 120000, + errorRetryCount: 1, + } + ); +} diff --git a/frontend/src/lib/prisma.ts b/frontend/src/lib/prisma.ts index 149adb5..ee17f49 100644 --- a/frontend/src/lib/prisma.ts +++ b/frontend/src/lib/prisma.ts @@ -40,3 +40,5 @@ if (process.env.NODE_ENV !== 'production') globalForPrisma.prisma = prisma + +