Add date/time range support to get_historical_candles
- _fetch_candles now accepts explicit from_ts/to_ts timestamps; falls back to count-based window when omitted - get_historical_candles accepts from_date/to_date in YYYY-MM-DD, YYYY-MM-DD HH:MM, or YYYY-MM-DD HH:MM:SS formats (all UTC) - Yahoo Finance fallback switched from range= to period1/period2 so it respects the exact window in both modes - Added 11 tests covering date range, datetime, end-of-day bump, invalid format, count cap, and pattern detection toggle - gitignore uv.lock (Dockerfile uses pip, not uv sync)
This commit is contained in:
Achmad
+2
-1
@@ -4,4 +4,5 @@ __pycache__/
|
||||
.pytest_cache/
|
||||
*.pyc
|
||||
.vscode/
|
||||
opencode.json
|
||||
opencode.json
|
||||
uv.lock
|
||||
@@ -311,13 +311,17 @@ def _fetch_candles(
|
||||
ticker: str,
|
||||
resolution: str = "D",
|
||||
count: int = 30,
|
||||
from_ts: int | None = None,
|
||||
to_ts: int | None = None,
|
||||
) -> list[dict[str, Any]]:
|
||||
# Try TradingView chart API first
|
||||
sym = _resolve_symbol(ticker, "stocks")
|
||||
to_ts = int(time.time())
|
||||
resolution_seconds = {"1": 60, "5": 300, "15": 900, "30": 1800, "60": 3600,
|
||||
"240": 14400, "D": 86400, "W": 604800, "M": 2592000}
|
||||
from_ts = to_ts - (count * resolution_seconds.get(resolution, 86400))
|
||||
if to_ts is None:
|
||||
to_ts = int(time.time())
|
||||
if from_ts is None:
|
||||
resolution_seconds = {"1": 60, "5": 300, "15": 900, "30": 1800, "60": 3600,
|
||||
"240": 14400, "D": 86400, "W": 604800, "M": 2592000}
|
||||
from_ts = to_ts - (count * resolution_seconds.get(resolution, 86400))
|
||||
|
||||
tv_urls = [
|
||||
"https://chart-data.tradingview.com/history",
|
||||
@@ -349,24 +353,14 @@ def _fetch_candles(
|
||||
interval_map = {"1": "1m", "5": "5m", "15": "15m", "30": "30m",
|
||||
"60": "60m", "240": "4h", "D": "1d", "W": "1wk", "M": "1mo"}
|
||||
interval = interval_map.get(resolution, "1d")
|
||||
range_map = {30: "1mo", 60: "3mo", 120: "6mo", 250: "1y", 500: "2y"}
|
||||
y_range = "1mo"
|
||||
for cnt, r in sorted(range_map.items()):
|
||||
if count <= cnt:
|
||||
y_range = r
|
||||
break
|
||||
else:
|
||||
y_range = "5y"
|
||||
|
||||
url = f"https://query1.finance.yahoo.com/v8/finance/chart/{yahoo_sym}"
|
||||
params = {"interval": interval, "range": y_range}
|
||||
params: dict[str, Any] = {"interval": interval, "period1": from_ts, "period2": to_ts}
|
||||
headers = {"User-Agent": TV_CHART_HEADERS["User-Agent"]}
|
||||
resp = requests.get(url, headers=headers, params=params, timeout=15)
|
||||
data = resp.json()
|
||||
|
||||
result = data.get("chart", {}).get("result", [None])[0]
|
||||
if not result:
|
||||
result = data.get("chart", {}).get("result", [None])[0]
|
||||
if result:
|
||||
timestamps = result.get("timestamp", [])
|
||||
quotes = result.get("indicators", {}).get("quote", [None])[0]
|
||||
@@ -385,7 +379,7 @@ def _fetch_candles(
|
||||
"time": timestamps[i],
|
||||
"open": o, "high": h, "low": l, "close": c, "volume": v,
|
||||
})
|
||||
return candles[-count:]
|
||||
return candles
|
||||
except requests.RequestException:
|
||||
pass
|
||||
|
||||
@@ -494,6 +488,8 @@ def get_historical_candles(
|
||||
count: int = 30,
|
||||
market_type: str = "stocks",
|
||||
detect_patterns: bool = True,
|
||||
from_date: str | None = None,
|
||||
to_date: str | None = None,
|
||||
) -> str:
|
||||
"""Fetch historical OHLCV candlestick data for candle pattern and trend analysis.
|
||||
|
||||
@@ -501,12 +497,40 @@ def get_historical_candles(
|
||||
ticker: Full ticker (e.g. "IDX:BBRI", "NASDAQ:AAPL", "IDX:BUMI").
|
||||
resolution: Candle resolution — 1, 3, 5, 15, 30, 60, 240 (minutes),
|
||||
D (daily, default), W (weekly), M (monthly).
|
||||
count: Number of candles to fetch (max 500, default 30).
|
||||
count: Number of candles to fetch when no date range given (max 500, default 30).
|
||||
market_type: Market type (stocks, crypto, forex, etc.).
|
||||
detect_patterns: Whether to detect candlestick patterns (default True).
|
||||
from_date: Start date/datetime (e.g. "2026-05-08" or "2026-05-17 14:00"). When
|
||||
provided, fetches candles from this point instead of counting back from now.
|
||||
to_date: End date/datetime (e.g. "2026-05-13" or "2026-05-17 14:05"). Defaults to
|
||||
now when from_date is given but to_date is omitted.
|
||||
"""
|
||||
import calendar
|
||||
|
||||
def _parse_dt(s: str, end_of_period: bool = False) -> int:
|
||||
for fmt in ("%Y-%m-%d %H:%M:%S", "%Y-%m-%d %H:%M", "%Y-%m-%d"):
|
||||
try:
|
||||
t = time.strptime(s, fmt)
|
||||
ts = int(calendar.timegm(t))
|
||||
# For date-only strings, end_of_period bumps to 23:59:59
|
||||
if end_of_period and fmt == "%Y-%m-%d":
|
||||
ts += 86399
|
||||
return ts
|
||||
except ValueError:
|
||||
continue
|
||||
raise ValueError(f"Unrecognised date/datetime format: {s!r}. Use YYYY-MM-DD or YYYY-MM-DD HH:MM")
|
||||
|
||||
from_ts: int | None = None
|
||||
to_ts: int | None = None
|
||||
if from_date:
|
||||
from_ts = _parse_dt(from_date)
|
||||
if to_date:
|
||||
to_ts = _parse_dt(to_date, end_of_period=True)
|
||||
elif from_date:
|
||||
to_ts = int(time.time())
|
||||
|
||||
count = min(count, 500)
|
||||
candles = _fetch_candles(ticker, resolution, count)
|
||||
candles = _fetch_candles(ticker, resolution, count, from_ts=from_ts, to_ts=to_ts)
|
||||
if not candles:
|
||||
return f"No historical data returned for {ticker}"
|
||||
|
||||
@@ -518,7 +542,8 @@ def get_historical_candles(
|
||||
avg_vol = sum(c["volume"] for c in candles) / len(candles)
|
||||
|
||||
lines: list[str] = []
|
||||
lines.append(f"📊 {ticker} — {count}candle chart ({resolution} resolution)")
|
||||
candle_label = f"{from_date} → {to_date or 'now'}" if from_date else f"{len(candles)} candles"
|
||||
lines.append(f"📊 {ticker} — {candle_label} ({resolution} resolution)")
|
||||
lines.append(f" Period: {time.strftime('%Y-%m-%d', time.gmtime(first['time']))}"
|
||||
f" → {time.strftime('%Y-%m-%d', time.gmtime(last['time']))}")
|
||||
lines.append(f" Latest: O={last['open']:.2f} H={last['high']:.2f} L={last['low']:.2f} C={last['close']:.2f}")
|
||||
|
||||
@@ -1,6 +1,8 @@
|
||||
from __future__ import annotations
|
||||
|
||||
import calendar
|
||||
import os
|
||||
import time
|
||||
from unittest.mock import patch
|
||||
|
||||
import pytest
|
||||
@@ -15,6 +17,7 @@ from server import (
|
||||
find_top_gainers,
|
||||
find_top_losers,
|
||||
fundamental_scan,
|
||||
get_historical_candles,
|
||||
get_stock_quotes,
|
||||
list_fields,
|
||||
list_markets,
|
||||
@@ -468,3 +471,91 @@ class TestSessionTools:
|
||||
server._session_cookies = None
|
||||
result = clear_session()
|
||||
assert "cleared" in result
|
||||
|
||||
|
||||
# ---------------------------------------------------------------------------
|
||||
# get_historical_candles
|
||||
# ---------------------------------------------------------------------------
|
||||
|
||||
_MOCK_CANDLES = [
|
||||
{"time": 1746662400, "open": 100.0, "high": 110.0, "low": 95.0, "close": 105.0, "volume": 1_000_000},
|
||||
{"time": 1746748800, "open": 105.0, "high": 115.0, "low": 100.0, "close": 108.0, "volume": 1_200_000},
|
||||
{"time": 1746835200, "open": 108.0, "high": 112.0, "low": 104.0, "close": 106.0, "volume": 900_000},
|
||||
]
|
||||
|
||||
|
||||
class TestGetHistoricalCandles:
|
||||
@patch("server._fetch_candles", return_value=_MOCK_CANDLES)
|
||||
def test_basic_count_based(self, mock_fetch):
|
||||
result = get_historical_candles("NASDAQ:AAPL", resolution="D", count=3)
|
||||
mock_fetch.assert_called_once_with("NASDAQ:AAPL", "D", 3, from_ts=None, to_ts=None)
|
||||
assert "NASDAQ:AAPL" in result
|
||||
assert "3 candles" in result
|
||||
|
||||
@patch("server._fetch_candles", return_value=_MOCK_CANDLES)
|
||||
def test_from_date_only(self, mock_fetch):
|
||||
result = get_historical_candles("NASDAQ:AAPL", from_date="2026-05-08")
|
||||
kw = mock_fetch.call_args.kwargs
|
||||
expected_from = int(calendar.timegm(time.strptime("2026-05-08", "%Y-%m-%d")))
|
||||
assert kw["from_ts"] == expected_from
|
||||
assert kw["to_ts"] is not None # defaults to now
|
||||
assert "2026-05-08" in result
|
||||
|
||||
@patch("server._fetch_candles", return_value=_MOCK_CANDLES)
|
||||
def test_from_and_to_date(self, mock_fetch):
|
||||
result = get_historical_candles("NASDAQ:AAPL", from_date="2026-05-08", to_date="2026-05-13")
|
||||
kw = mock_fetch.call_args.kwargs
|
||||
expected_from = int(calendar.timegm(time.strptime("2026-05-08", "%Y-%m-%d")))
|
||||
expected_to = int(calendar.timegm(time.strptime("2026-05-13", "%Y-%m-%d"))) + 86399
|
||||
assert kw["from_ts"] == expected_from
|
||||
assert kw["to_ts"] == expected_to
|
||||
assert "2026-05-08 → 2026-05-13" in result
|
||||
|
||||
@patch("server._fetch_candles", return_value=_MOCK_CANDLES)
|
||||
def test_datetime_strings(self, mock_fetch):
|
||||
get_historical_candles("NASDAQ:AAPL", from_date="2026-05-17 14:00", to_date="2026-05-17 14:05")
|
||||
kw = mock_fetch.call_args.kwargs
|
||||
expected_from = int(calendar.timegm(time.strptime("2026-05-17 14:00", "%Y-%m-%d %H:%M")))
|
||||
expected_to = int(calendar.timegm(time.strptime("2026-05-17 14:05", "%Y-%m-%d %H:%M")))
|
||||
assert kw["from_ts"] == expected_from
|
||||
assert kw["to_ts"] == expected_to
|
||||
|
||||
@patch("server._fetch_candles", return_value=_MOCK_CANDLES)
|
||||
def test_datetime_with_seconds(self, mock_fetch):
|
||||
get_historical_candles("NASDAQ:AAPL", from_date="2026-05-17 14:00:30", to_date="2026-05-17 14:05:00")
|
||||
kw = mock_fetch.call_args.kwargs
|
||||
expected_from = int(calendar.timegm(time.strptime("2026-05-17 14:00:30", "%Y-%m-%d %H:%M:%S")))
|
||||
assert kw["from_ts"] == expected_from
|
||||
|
||||
@patch("server._fetch_candles", return_value=_MOCK_CANDLES)
|
||||
def test_date_only_to_date_end_of_day(self, mock_fetch):
|
||||
# date-only to_date should be bumped to end of that day (23:59:59)
|
||||
get_historical_candles("NASDAQ:AAPL", from_date="2026-05-08", to_date="2026-05-13")
|
||||
kw = mock_fetch.call_args.kwargs
|
||||
midnight = int(calendar.timegm(time.strptime("2026-05-13", "%Y-%m-%d")))
|
||||
assert kw["to_ts"] == midnight + 86399
|
||||
|
||||
@patch("server._fetch_candles", return_value=[])
|
||||
def test_no_data_returned(self, mock_fetch):
|
||||
result = get_historical_candles("NASDAQ:AAPL")
|
||||
assert "No historical data" in result
|
||||
|
||||
@patch("server._fetch_candles", return_value=_MOCK_CANDLES)
|
||||
def test_invalid_date_format_raises(self, mock_fetch):
|
||||
with pytest.raises(ValueError, match="Unrecognised date"):
|
||||
get_historical_candles("NASDAQ:AAPL", from_date="05/08/2026")
|
||||
|
||||
@patch("server._fetch_candles", return_value=_MOCK_CANDLES)
|
||||
def test_count_capped_at_500(self, mock_fetch):
|
||||
get_historical_candles("NASDAQ:AAPL", count=9999)
|
||||
assert mock_fetch.call_args.args[2] == 500
|
||||
|
||||
@patch("server._fetch_candles", return_value=_MOCK_CANDLES)
|
||||
def test_detect_patterns_included_by_default(self, mock_fetch):
|
||||
result = get_historical_candles("NASDAQ:AAPL")
|
||||
assert "Pattern Detection" in result
|
||||
|
||||
@patch("server._fetch_candles", return_value=_MOCK_CANDLES)
|
||||
def test_detect_patterns_disabled(self, mock_fetch):
|
||||
result = get_historical_candles("NASDAQ:AAPL", detect_patterns=False)
|
||||
assert "Pattern Detection" not in result
|
||||
|
||||
Reference in New Issue
Block a user