Add configurable 30-day NewBook rate rescrape on 2/4/6/12h intervals

Intraday rescrape jobs are distributed evenly between the main nightly run
(05:20) and cover only the next 30 days — lightweight complement to the
full 720-day nightly sweep. Interval is configurable from the Newbook tab
in Settings and takes effect immediately without a container restart.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
This commit is contained in:
jtricerolph 2026-07-15 09:44:18 +00:00
parent 7d0b7d2d3b
commit 078cb47b16
3 changed files with 238 additions and 1 deletions

View file

@ -781,6 +781,90 @@ async def get_occupancy_history(
} }
# ============================================
# RESCRAPE INTERVAL CONFIG
# ============================================
def _rescrape_times_from_config(interval_hours: int, time_str: str) -> list:
"""Compute extra rescrape fire times (HH:MM strings) from interval and base time."""
try:
parts = time_str.split(':')
base_hour, base_minute = int(parts[0]), int(parts[1])
except (ValueError, IndexError):
base_hour, base_minute = 5, 20
if interval_hours not in (2, 4, 6, 12):
return []
return [
f"{(base_hour + offset) % 24:02d}:{base_minute:02d}"
for offset in range(interval_hours, 24, interval_hours)
]
@router.get("/config/rescrape-schedule")
async def get_rescrape_schedule(
db: AsyncSession = Depends(get_db),
current_user: dict = Depends(get_current_user)
):
"""Return current 30-day rescrape interval config and computed fire times."""
result = await db.execute(
text("""
SELECT config_key, config_value FROM system_config
WHERE config_key IN ('newbook_rescrape_interval_hours', 'sync_newbook_current_rates_time')
""")
)
cfg = {row.config_key: row.config_value for row in result.fetchall()}
try:
interval_hours = int(cfg.get('newbook_rescrape_interval_hours') or '0')
except ValueError:
interval_hours = 0
time_str = cfg.get('sync_newbook_current_rates_time') or '05:20'
return {
"interval_hours": interval_hours,
"base_time": time_str,
"rescrape_times": _rescrape_times_from_config(interval_hours, time_str),
}
class RescrapeIntervalBody(BaseModel):
interval_hours: int
@router.post("/config/rescrape-interval")
async def set_rescrape_interval(
body: RescrapeIntervalBody,
db: AsyncSession = Depends(get_db),
current_user: dict = Depends(get_current_user)
):
"""Save 30-day rescrape interval and live-reload the scheduler jobs."""
if body.interval_hours not in (0, 2, 4, 6, 12):
raise HTTPException(status_code=400, detail="interval_hours must be 0, 2, 4, 6, or 12")
await db.execute(
text("""
INSERT INTO system_config (config_key, config_value)
VALUES ('newbook_rescrape_interval_hours', :val)
ON CONFLICT (config_key) DO UPDATE SET config_value = EXCLUDED.config_value
"""),
{"val": str(body.interval_hours)}
)
await db.commit()
from scheduler import apply_newbook_rescrape_schedule
apply_newbook_rescrape_schedule()
time_result = await db.execute(
text("SELECT config_value FROM system_config WHERE config_key = 'sync_newbook_current_rates_time'")
)
time_row = time_result.fetchone()
time_str = (time_row.config_value if time_row and time_row.config_value else None) or '05:20'
return {
"status": "saved",
"interval_hours": body.interval_hours,
"base_time": time_str,
"rescrape_times": _rescrape_times_from_config(body.interval_hours, time_str),
}
@router.post("/refresh-date/{rate_date}") @router.post("/refresh-date/{rate_date}")
async def refresh_single_date( async def refresh_single_date(
rate_date: str, rate_date: str,

View file

@ -61,6 +61,53 @@ async def run_scheduled_direct_scrape():
await loop.run_in_executor(None, run_scrape_all_direct) await loop.run_in_executor(None, run_scrape_all_direct)
async def run_short_newbook_rescrape():
"""Fetch next-30-day NewBook rates — lightweight complement to the full nightly run."""
from jobs.fetch_current_rates import run_fetch_current_rates
await run_fetch_current_rates(horizon_days=30)
def _compute_rescrape_times(base_hour: int, base_minute: int, interval_hours: int) -> list:
"""Return (hour, minute) tuples evenly spaced around the clock, excluding the base (main) run."""
return [
((base_hour + offset) % 24, base_minute)
for offset in range(interval_hours, 24, interval_hours)
]
def apply_newbook_rescrape_schedule():
"""Read config and (re)register 30-day NewBook rescrape jobs without a scheduler restart."""
for job in list(scheduler.get_jobs()):
if job.id.startswith('nb_rescrape_'):
scheduler.remove_job(job.id)
interval_str = get_config_value('newbook_rescrape_interval_hours', '0')
try:
interval_hours = int(interval_str or '0')
except ValueError:
interval_hours = 0
if interval_hours not in (2, 4, 6, 12):
logger.info("NewBook 30-day rescrape disabled")
return
rates_hour, rates_minute = get_sync_time('newbook_current_rates', 5, 20)
times = _compute_rescrape_times(rates_hour, rates_minute, interval_hours)
for i, (h, m) in enumerate(times):
scheduler.add_job(
run_short_newbook_rescrape,
CronTrigger(hour=h, minute=m),
id=f'nb_rescrape_{i}',
replace_existing=True,
)
logger.info(
f"NewBook rescrape: {len(times)} extra run(s) at {interval_hours}h intervals "
f"(base {rates_hour:02d}:{rates_minute:02d}): "
f"{[f'{h:02d}:{m:02d}' for h, m in times]}"
)
async def run_scheduled_parity_check(): async def run_scheduled_parity_check():
from jobs.check_rate_parity import run_parity_check from jobs.check_rate_parity import run_parity_check
import asyncio import asyncio
@ -159,6 +206,9 @@ def start_scheduler():
scheduler.start() scheduler.start()
logger.info(f"Scheduler started: booking scrape at {scrape_hour:02d}:{scrape_minute:02d}, rates fetch at {rates_hour:02d}:{rates_minute:02d}, direct scrape at 06:00, parity check at 06:45, watchdog every 30m") logger.info(f"Scheduler started: booking scrape at {scrape_hour:02d}:{scrape_minute:02d}, rates fetch at {rates_hour:02d}:{rates_minute:02d}, direct scrape at 06:00, parity check at 06:45, watchdog every 30m")
# 30-day NewBook rescrape — optional intraday refresh, interval from config
apply_newbook_rescrape_schedule()
def shutdown_scheduler(): def shutdown_scheduler():
if scheduler.running: if scheduler.running:

View file

@ -1,7 +1,7 @@
import { useState, useEffect } from 'react' import { useState, useEffect } from 'react'
import { useParams, useNavigate } from 'react-router-dom' import { useParams, useNavigate } from 'react-router-dom'
import { useQuery, useMutation, useQueryClient } from '@tanstack/react-query' import { useQuery, useMutation, useQueryClient } from '@tanstack/react-query'
import { Save, RefreshCw, Database, Clock, BedDouble, ChevronUp, ChevronDown, Network, ShieldCheck, AlertTriangle } from 'lucide-react' import { Save, RefreshCw, Database, Clock, BedDouble, ChevronUp, ChevronDown, Network, ShieldCheck, AlertTriangle, Timer } from 'lucide-react'
import api from '../api' import api from '../api'
const TABS = [ const TABS = [
@ -664,6 +664,109 @@ function NewbookTab({ config, isLoading, onSave, onSyncNow, saving, syncing }: N
</div> </div>
<RoomCategoriesCard /> <RoomCategoriesCard />
<RescrapeCard />
</div>
)
}
// ─── 30-day Rescrape Schedule ─────────────────────────────────────────────────
interface RescrapeSchedule {
interval_hours: number
base_time: string
rescrape_times: string[]
}
function RescrapeCard() {
const qc = useQueryClient()
const [interval, setInterval] = useState<number>(0)
const [loaded, setLoaded] = useState(false)
const { data, isLoading } = useQuery<RescrapeSchedule>({
queryKey: ['rescrape-schedule'],
queryFn: () => api.get('/bookability/config/rescrape-schedule').then(r => r.data),
})
useEffect(() => {
if (data && !loaded) {
setInterval(data.interval_hours)
setLoaded(true)
}
}, [data, loaded])
const save = useMutation({
mutationFn: () =>
api.post('/bookability/config/rescrape-interval', { interval_hours: interval }).then(r => r.data as RescrapeSchedule),
onSuccess: () => qc.invalidateQueries({ queryKey: ['rescrape-schedule'] }),
})
const previewTimes = (data && loaded)
? (() => {
const baseTime = data.base_time
if (interval === 0) return []
const [bh, bm] = baseTime.split(':').map(Number)
const times: string[] = []
for (let offset = interval; offset < 24; offset += interval) {
const h = (bh + offset) % 24
times.push(`${String(h).padStart(2, '0')}:${String(bm).padStart(2, '0')}`)
}
return times
})()
: data?.rescrape_times ?? []
if (isLoading) return null
return (
<div className="card">
<div className="card-header">
<span style={{ display: 'flex', alignItems: 'center', gap: 8 }}>
<Timer size={16} strokeWidth={1.75} />
30-day Rate Rescrape
</span>
<span className={`badge ${interval > 0 ? 'badge-success' : 'badge-neutral'}`}>
{interval > 0 ? `Every ${interval}h` : 'Disabled'}
</span>
</div>
<div className="card-body" style={{ display: 'flex', flexDirection: 'column', gap: 14 }}>
<p style={{ fontSize: 13, color: 'var(--text-mid)', margin: 0 }}>
Re-fetches the next 30 days of NewBook rates throughout the day, between the main
nightly run at <strong>{data?.base_time ?? '05:20'}</strong>. Useful for keeping
Bookability fresh on days when tariff availability changes.
</p>
<div style={{ display: 'flex', gap: 8, alignItems: 'center' }}>
<select
style={{ width: 200 }}
value={interval}
onChange={e => setInterval(Number(e.target.value))}
>
<option value={0}>Disabled</option>
<option value={2}>Every 2 hours</option>
<option value={4}>Every 4 hours</option>
<option value={6}>Every 6 hours</option>
<option value={12}>Every 12 hours</option>
</select>
<button
className="btn btn-primary btn-sm"
disabled={save.isPending}
onClick={() => save.mutate()}
>
<Save size={13} strokeWidth={1.75} />
{save.isPending ? 'Saving…' : 'Save'}
</button>
{save.isSuccess && (
<span style={{ fontSize: 12, color: 'var(--success)' }}> Saved</span>
)}
</div>
{previewTimes.length > 0 && (
<div style={{ fontSize: 12, color: 'var(--text-mid)' }}>
<span style={{ fontWeight: 600, marginRight: 6 }}>Rescrape runs at:</span>
{previewTimes.join(' · ')}
<span style={{ marginLeft: 6 }}>(in addition to the {data?.base_time} main run)</span>
</div>
)}
</div>
</div> </div>
) )
} }