Skip to main content

Architecture Overview

LeBonTag v2 is built around a modular architecture combining:

  • OSM data ingestion
  • Validation workflows
  • Quality control processing
  • PostgreSQL/PostGIS data storage
  • Python-based background processing

High-level architecture​

LeBonTag Architecture

Main components​

1. OSM Data Sources​

  • OSM replication (FR mirror)
  • OSM API (changesets)
  • Periodic updates via cron jobs

These feed the ingestion layer.

2. Python CRON Processes​

Two main background pipelines:

  • OSM changes ingestion
  • Validation export (osmChange generation)

Responsibilities:

  • Synchronizing OSM data
  • Applying validation decisions
  • Feeding the reference schema (osm2pgsql)
  • Triggering quality rule re-evaluation

3. Reference Data (osm2pgsql)​

  • Stores up-to-date OSM data
  • Used as base dataset for quality control
  • Updated with validated changes from LeBonTag

4. LeBonTag Application Database​

Dedicated schema storing:

  • Validation workflow state
  • Quality rule definitions
  • Object groups and business groups
  • Audit and history

5. Web Interface​

Provides:

  • Object validation
  • Quality rule exploration
  • Administrative configuration
  • Role-based access control

Data flow summary​

  1. OSM changes are fetched.
  2. Data is stored in the reference schema.
  3. Quality rules analyze the dataset.
  4. Objects are flagged for validation.
  5. Validators review via UI.
  6. Approved changes generate osmChange exports.
  7. Reference dataset is updated accordingly.