hotel-auto-x-meituan-collector

name	hotel-auto-x-meituan-collector
description	Use when working on hotel-auto-x Meituan collection, including Meituan review collection, bad reviews, review score, full daily collection, realtime traffic/order metrics, ad collection, Meituan session/profile problems, and Meituan data mapping into daily_report.

hotel-auto-x Meituan Collector

Use this skill for Meituan collection work in /home/qing/hotel-auto-x/hotel_auto_x.

Safety

• Do not run full all-store collection during busy scheduler windows unless explicitly requested.
• Do not touch other platform Profiles when fixing one Meituan store.
• Use stealth=False for Meituan browser contexts unless existing code says otherwise.
• Do not bypass process_result() for daily_report or bad-review writes.

Project Map

• Main collector: platforms/meituan.py::MeituanCollector
• Shared identity helpers: platforms/meituan_common.py
• Realtime collector: platforms/meituan_realtime.py::collect_meituan_realtime
• Ad collector: platforms/ad_meituan.py::collect_meituan_ad
• Review scheduler: scheduler/engine.py::task_collect_meituan_review
• Full scheduler: scheduler/engine.py::task_collect_meituan_full
• Realtime scheduler: scheduler/realtime_engine.py::task_realtime_meituan
• Ad scheduler: scheduler/ad_engine.py::task_collect_meituan_ad
• ETL: pipeline/etl.py

Main Data Types

Daily/full collection can fill:

• Meituan score and review counters.
• Bad review details through _bad_reviews, written into bad_reviews.
• Data-center business metrics when captured or fetched by fallback.

Realtime collection fills realtime_snapshot fields such as exposure, browsing/intention UV, pay orders, pay rooms, conversion, rankings, and related peer values.

Ad collection writes ad daily data through scheduler/ad pipeline, not the normal review ETL path.

Review-Only Flow

Production entry: MeituanCollector().run_review_only(store_id, stat_date).

Scheduler path:

1. Filter active stores with meituan_account and collect_meituan=True.
2. Skip circuit-open or locked login Profile.
3. Run review-only collector with timeout.
4. Call process_result(result).
5. Mark session active/expired based on result.
6. Retry retryable failures once.

Full Flow

Production entry: MeituanCollector().run(store_id, stat_date).

Full collection uses API capture plus fallback paths, then writes via ETL. Account groups are collected serially inside each account group.

Realtime Flow

Production entry: collect_meituan_realtime(store_id).

Validate that returned data has at least one core field:

• mt_intention_uv
• mt_pay_orders
• mt_pay_rooms
• mt_exposure

Repeated no_data, load_timeout, network_error, anti_bot, or session_expired triggers platform/store cooldowns in scheduler/realtime_engine.py.

Known Timing Rule

Some Meituan traffic fields for the current day can show -. For yesterday data, collect or audit on the next day after platform settlement. Do not classify all current-day - exposure as a permanent data gap.

Minimal Diagnostic Pattern

from platforms.meituan import MeituanCollector

result = MeituanCollector().run_review_only("MD00001", "2026-07-06")
print(result.success, result.error, result.method, result.data)

Only write with:

from pipeline.etl import process_result
process_result(result)

when the user explicitly wants data repaired or imported.

Failure Meaning

• session_expired: re-login the store Profile.
• anti_bot: stop aggressive retries and inspect platform page/state.
• api_empty or no_data: inspect identity extraction, request params, and whether the date is settled.
• resource_busy_login: user login window is open; skip instead of forcing close.

Reporting

Report store_id, target date, collection phase, method, key fields, _bad_reviews count if present, and ETL/write status.