Crisismap docs

Collection of generic docs and relevant resources: standards, examples, data sources etc.

Generic scheme

Glossary

(since nearly everything in here deals with data, the word data is intentionally omitted in definitions)

Source

Any sort of raw data suppliers: platforms, services, publicly available APIs/DBs, aggregators over IoT sensors etc.

Provider

Essentially, a worker (COULD be daemonized script, crawler, tiny ETL pipeline etc.) aimed to access a single source, or even multiple sources (merging related data segments), retrieve (load, parse) raw data, process (filter, clean, normalize) and store it accordingly.

For now, storing SHOULD happen explicitly via INPUT queries, accessing back-end potential POST API surface is left for future considerations.

Providers COULD be derived from a generic provider (implementation details), exposing a set of templates to descendants.

Also, providers are about to be monitored and scheduled by an orchestrator - primarily a set of config declarations.

Back-end

Ideally, "almost" a logic-less proxying service acting in the middle of DB and clients.

If some comprehensive business logic appears, it COULD potentially be associated with this service OR derived into a dedicated module.

Main responsibilities:

• internally: storage accessor and data mapper, leveraged by models (representing multiple categories)
• externally: RESTful API wrapper (serving controllers and middlewares), together with a live update service (sockets) for clients

API itself acts "as a client", meaning serving JSON instead of explicit UI capabilities.

Categories

Existing and upcoming models/collections:

• news
  • news types
• incidents
  • incident types
• locations
• toponyms
• population
• reports

Clients

Front-end

Aimed for data representation and UX/UI (map/clustering + filtering/pagination).

External subscribers

Any app/service/crawler authorized to access aggregated data streams.

REST and RSS services are considered as first options.