跳到主要内容

提供商架构概述

Fonrex uses a distributed multi-provider scraping architecture to aggregate market indicators, fundamental statistics, and financial news without reliance on a single upstream vendor.

Execution Flow​

GET /fundamental?ticker=AIR.PA
       │
       ▼
FinancialProviderRunner.run()
       │
       ├─► Asynchronous Parallel Scraping (asyncio.gather)
       │    ├── ZoneBourse.fetch()       (Timeout: 8s)
       │    ├── Gurufocus.fetch()        (Timeout: 8s)
       │    ├── YahooFinance.fetch()     (Timeout: 5s)
       │    └── MorningStar.fetch()      (Timeout: 8s)
       │
       ▼
ValidationLayer.validate_results()
       │ Outlier Check & Range Limits (min/max boundaries)
       ▼
StandardFinancials Formatter
       │ Merges validated metrics across providers
       ▼
HTTP JSON Response

Resilience Strategy​

  1. Non-Blocking Gather: Providers execute in parallel via asyncio.gather(..., return_exceptions=True). A failure or timeout in one provider (e.g. ZoneBourse) does not affect the remaining providers.
  2. Provider Mapping Fallbacks: Resolves search parameters in this priority order: provider_url ➔ provider_ticker ➔ ISIN ➔ ticker.
  3. Consensus Outlier Filtering: The ValidationLayer filters out anomalous metrics returned by individual providers before merging responses.