FastTrack connector¶

Status: implemented as a canonical REST provider.

FastTrack support uses the standard source.type: api runtime with provider-specific resources, authentication, and landing schema handling.

Runtime modules¶

Connector: dpone.runtime.connectors.api.fasttrack
Source: dpone.runtime.sources.api.fasttrack
Strategies: dpone.runtime.strategies.api.fasttrack
Registration: dpone.runtime.api_registry

Authentication modes¶

FastTrack supports two credential styles:

Resource-specific tokens such as dashboard_bot_key and flex_x_token.
Generic fallback token through token or backward-compatible api_key.

Recommended Vault secret:

{
  "endpoint": "https://api.example.com",
  "dashboard_bot_key": "...",
  "flex_x_token": "..."
}

Landing philosophy¶

FastTrack landing follows an ELT-safe pattern:

Preserve provider fields.
Do not perform semantic rename/drop in the landing layer.
Normalize column names mechanically to warehouse-safe snake_case.
Keep business normalization in downstream core models.

Temporal typing is enabled by default and can be controlled with provider options when a consumer needs raw string timestamps.

Manifest example¶

source:
  type: api
  api_type: fasttrack
  connection_id: fasttrack
  connection_type: vault
  resource: chat_sessions
  options:
    date_from: "2026-01-01"
    date_to: "2026-01-02"

sink:
  type: bigquery
  connection_id: bigquery_landing
  table:
    dataset: landing
    name: landing__fasttrack__chat_sessions
  strategy:
    mode: incremental_merge
    unique_key: [uuid_polzovatelya, data_nachala]

Testing¶

FastTrack has two useful test layers:

Mock integration tests with a local HTTP server.
Manual live tests against the vendor API through Vault credentials.

uv run pytest tests/integration -k fasttrack