interloper-api 0.6.0__tar.gz → 0.7.0__tar.gz

{interloper_api-0.6.0 → interloper_api-0.7.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: interloper-api
-Version: 0.6.0
+Version: 0.7.0
 Summary: Interloper FastAPI routes
 Author: Guillaume Onfroy
 Author-email: Guillaume Onfroy <guillaume@digitlcloud.com>

{interloper_api-0.6.0 → interloper_api-0.7.0}/pyproject.toml

@@ -3,7 +3,7 @@
 # ###############
 [project]
 name = "interloper-api"
-version = "0.6.0"
+version = "0.7.0"
 description = "Interloper FastAPI routes"
 readme = "README.md"
 authors = [{ name = "Guillaume Onfroy", email = "guillaume@digitlcloud.com" }]

{interloper_api-0.6.0 → interloper_api-0.7.0}/src/interloper_api/routes/runs.py

@@ -3,6 +3,7 @@
 from __future__ import annotations
 
 import datetime as dt
+from typing import Literal
 from uuid import UUID
 
 from fastapi import APIRouter, Depends, HTTPException, Response
@@ -11,10 +12,15 @@ from interloper_db import Profile, Store
 from interloper_db.models import Event, Run
 from pydantic import BaseModel
 
-from interloper_api.dependencies import get_org_id, get_store, require_viewer
+from interloper_api.dependencies import get_org_id, get_store, require_editor, require_viewer
 
 router = APIRouter()
 
+#: Hard cap on the number of events returned in a single page, regardless of
+#: the requested ``limit``. Keeps a pathological ``?limit=1000000`` from loading
+#: an entire run's history into memory at once.
+MAX_EVENTS_PAGE_SIZE = 1000
+
 
 class RunResponse(BaseModel):
     """Response body for a run."""
@@ -25,11 +31,20 @@ class RunResponse(BaseModel):
     backfill_id: UUID | None
     partition_date: dt.date | None
     status: str
+    retry_of: UUID | None = None
+    attempt: int = 1
+    retry_scope: str | None = None
     started_at: str | None = None
     completed_at: str | None = None
     created_at: str | None = None
 
 
+class RetryRequest(BaseModel):
+    """Request body for retrying a failed run."""
+
+    scope: Literal["all", "failed"] = "all"
+
+
 class AssetExecutionResponse(BaseModel):
     """Response body for an asset execution (from the asset_executions view)."""
 
@@ -75,6 +90,9 @@ def _run_to_response(run: Run) -> RunResponse:
         backfill_id=run.backfill_id,
         partition_date=run.partition_date,
         status=run.status,
+        retry_of=run.retry_of,
+        attempt=run.attempt,
+        retry_scope=run.retry_scope,
         started_at=str(run.started_at) if run.started_at else None,
         completed_at=str(run.completed_at) if run.completed_at else None,
         created_at=str(run.created_at) if run.created_at else None,
@@ -172,13 +190,49 @@ def list_asset_executions(
     ]
 
 
+@router.post("/{run_id}/retry")
+def retry_run(
+    run_id: UUID,
+    body: RetryRequest | None = None,
+    user: Profile = Depends(require_editor),
+    store: Store = Depends(get_store),
+) -> dict[str, str]:
+    """Queue a retry of a failed run.
+
+    Creates a new run linked to the original via ``retry_of``. With
+    ``scope="all"`` the whole DAG re-runs; with ``scope="failed"`` only the
+    previously failed/cancelled assets re-run.
+    """
+    scope = body.scope if body else "all"
+    try:
+        run = store.retry_run(run_id, scope=scope)
+    except NotFoundError:
+        raise HTTPException(status_code=404, detail=f"Run {run_id} not found")
+    except ValueError as e:
+        raise HTTPException(status_code=409, detail=str(e))
+    return {"status": "queued", "run_id": str(run.id)}
+
+
 @router.get("/{run_id}/events")
 def list_run_events(
     run_id: UUID,
+    response: Response,
     limit: int = 100,
+    offset: int = 0,
     user: Profile = Depends(require_viewer),
     store: Store = Depends(get_store),
 ) -> list[EventResponse]:
-    """List events for a run."""
-    events = store.list_events(run_id=run_id, limit=limit)
+    """List events for a run, oldest first.
+
+    Events are ordered ``timestamp ASC`` and paged with ``limit``/``offset``.
+    The total number of events for the run (ignoring ``limit``/``offset``) is
+    returned in the ``X-Total-Count`` response header so clients can page
+    through every event — including the terminal/outcome events
+    (``asset_completed``, ``asset_failed``, ``run_failed``, …) that sort last.
+    """
+    limit = max(1, min(limit, MAX_EVENTS_PAGE_SIZE))
+    offset = max(0, offset)
+    total = store.count_events(run_id=run_id)
+    response.headers["X-Total-Count"] = str(total)
+    events = store.list_events(run_id=run_id, limit=limit, offset=offset)
     return [_event_to_response(e) for e in events]