phase 1

core/api/detect/sources.py

@@ -0,0 +1,108 @@
+"""
+Source browser for detection pipeline.
+
+Lists available media sources from blob storage (MinIO).
+
+GET  /detect/sources                              — list chunk jobs
+GET  /detect/sources/{job_id}/chunks              — list chunks for a job
+GET  /detect/sources/{job_id}/chunks/{name}/url   — presigned preview URL
+"""
+
+from __future__ import annotations
+
+import logging
+
+from fastapi import APIRouter, HTTPException
+from pydantic import BaseModel
+
+logger = logging.getLogger(__name__)
+
+router = APIRouter(prefix="/detect", tags=["detect"])
+
+
+class ChunkInfoResponse(BaseModel):
+    filename: str
+    key: str
+    size_bytes: int
+
+
+class SourceInfoResponse(BaseModel):
+    job_id: str
+    source_type: str = "chunk_job"
+    chunk_count: int
+    total_bytes: int = 0
+
+
+def _list_sources() -> list[SourceInfoResponse]:
+    """List chunk jobs from blob storage."""
+    from core.storage.blob import get_store
+
+    store = get_store("out")
+    try:
+        objects = store.list(prefix="chunks/")
+    except Exception as e:
+        logger.warning("Failed to list blob sources: %s", e)
+        return []
+
+    jobs: dict[str, int] = {}
+    job_bytes: dict[str, int] = {}
+    for obj in objects:
+        rel_key = obj.key.removeprefix(store.prefix)
+        parts = rel_key.split("/")
+        if len(parts) >= 3 and parts[0] == "chunks":
+            job_id = parts[1]
+            jobs[job_id] = jobs.get(job_id, 0) + 1
+            job_bytes[job_id] = job_bytes.get(job_id, 0) + obj.size_bytes
+
+    sources = []
+    for job_id, count in sorted(jobs.items()):
+        source = SourceInfoResponse(
+            job_id=job_id,
+            source_type="chunk_job",
+            chunk_count=count,
+            total_bytes=job_bytes.get(job_id, 0),
+        )
+        sources.append(source)
+    return sources
+
+
+@router.get("/sources", response_model=list[SourceInfoResponse])
+def list_sources():
+    """List available chunk jobs from blob storage."""
+    return _list_sources()
+
+
+@router.get("/sources/{source_job_id}/chunks", response_model=list[ChunkInfoResponse])
+def list_chunks(source_job_id: str):
+    """List chunks for a specific source job."""
+    from core.storage.blob import get_store
+
+    store = get_store("out")
+    try:
+        objects = store.list(prefix=f"chunks/{source_job_id}/", extensions={".mp4"})
+    except Exception as e:
+        logger.warning("Failed to list chunks for %s: %s", source_job_id, e)
+        raise HTTPException(status_code=503, detail=f"Blob storage unavailable: {e}")
+
+    if not objects:
+        raise HTTPException(status_code=404, detail=f"Source not found: {source_job_id}")
+
+    chunks = []
+    for obj in objects:
+        info = ChunkInfoResponse(filename=obj.filename, key=obj.key, size_bytes=obj.size_bytes)
+        chunks.append(info)
+    return sorted(chunks, key=lambda c: c.filename)
+
+
+@router.get("/sources/{source_job_id}/chunks/{filename}/url")
+def get_chunk_url(source_job_id: str, filename: str):
+    """Return a presigned URL for previewing a chunk in the browser."""
+    from core.storage.blob import get_store
+
+    store = get_store("out")
+    key = f"chunks/{source_job_id}/{filename}"
+    try:
+        url = store.get_url(key, expires=3600)
+    except Exception as e:
+        raise HTTPException(status_code=503, detail=f"Could not generate URL: {e}")
+    return {"url": url}