real handler, sign_pdfs CodeUri, env vars, S3 IAM, fail-loud config

functions/sign_pdfs/handler.py

@@ -22,6 +22,7 @@ Production refinements vs sign_pdfs_v1 (see docs/lambdas-md/lambda-20-sign-pdfs.
   8. N concurrent consumers via asyncio.gather — presign throughput scales with
      the concurrency knob in the event.
 """
+
 import asyncio
 import json
 import os
@@ -35,12 +36,17 @@ _DONE = object()
 
 
 def _cfg(event: dict) -> dict:
+    # bucket and prefix are required — fail loud if neither event nor env supplies
+    # them. The function deliberately has no fallback default (no "2026/04/" baked
+    # in) because that's a deployment-time concern, not handler code.
     return {
-        "bucket": event.get("bucket") or os.environ.get("BUCKET_NAME", "my-company-reports-bucket"),
-        "prefix": event.get("prefix") or os.environ.get("PREFIX", "2026/04/"),
-        "expiry": int(event.get("expiry_seconds") or os.environ.get("URL_EXPIRY_SECONDS", "900")),
+        "bucket": event.get("bucket") or os.environ["BUCKET_NAME"],
+        "prefix": event.get("prefix") or os.environ["PREFIX"],  # e.g. "2026/04/"
+        "expiry": int(
+            event.get("expiry_seconds") or os.environ.get("URL_EXPIRY_SECONDS", "900")
+        ),
         "endpoint": os.environ.get("S3_ENDPOINT_URL") or None,
-        "page_size": int(event.get("page_size") or 1000),
+        "page_size": int(event.get("page_size") or 100),
         "concurrency": int(event.get("concurrency") or 4),
         "queue_max": int(os.environ.get("QUEUE_MAX", "2000")),
     }
@@ -51,26 +57,42 @@ def _log(event_type: str, **fields):
 
 
 def _emit_emf(metrics: dict, **dims):
-    print(json.dumps({
-        "_aws": {
-            "Timestamp": int(time.time() * 1000),
-            "CloudWatchMetrics": [{
-                "Namespace": "eth/sign_pdfs",
-                "Dimensions": [list(dims.keys())],
-                "Metrics": [
-                    {"Name": k, "Unit": "Bytes" if k.endswith("Bytes") else "Count"}
-                    for k in metrics
-                ],
-            }],
-        },
-        **dims, **metrics,
-    }))
+    print(
+        json.dumps(
+            {
+                "_aws": {
+                    "Timestamp": int(time.time() * 1000),
+                    "CloudWatchMetrics": [
+                        {
+                            "Namespace": "eth/sign_pdfs",
+                            "Dimensions": [list(dims.keys())],
+                            "Metrics": [
+                                {
+                                    "Name": k,
+                                    "Unit": "Bytes" if k.endswith("Bytes") else "Count",
+                                }
+                                for k in metrics
+                            ],
+                        }
+                    ],
+                },
+                **dims,
+                **metrics,
+            }
+        )
+    )
 
 
 async def _run(event: dict, request_id: str):
     cfg = _cfg(event)
-    _log("start", request_id=request_id, bucket=cfg["bucket"], prefix=cfg["prefix"],
-         page_size=cfg["page_size"], concurrency=cfg["concurrency"])
+    _log(
+        "start",
+        request_id=request_id,
+        bucket=cfg["bucket"],
+        prefix=cfg["prefix"],
+        page_size=cfg["page_size"],
+        concurrency=cfg["concurrency"],
+    )
     t0 = time.monotonic()
 
     pages = 0
@@ -85,7 +107,8 @@ async def _run(event: dict, request_id: str):
             try:
                 paginator = s3.get_paginator("list_objects_v2")
                 async for page in paginator.paginate(
-                    Bucket=cfg["bucket"], Prefix=cfg["prefix"],
+                    Bucket=cfg["bucket"],
+                    Prefix=cfg["prefix"],
                     PaginationConfig={"PageSize": cfg["page_size"]},
                 ):
                     pages += 1
@@ -120,10 +143,17 @@ async def _run(event: dict, request_id: str):
                         local += 1
                     except Exception as exc:
                         errors += 1
-                        _log("presign_error", request_id=request_id, key=item, error=str(exc))
+                        _log(
+                            "presign_error",
+                            request_id=request_id,
+                            key=item,
+                            error=str(exc),
+                        )
 
             prod_task = asyncio.create_task(producer())
-            counts = await asyncio.gather(*(consumer() for _ in range(cfg["concurrency"])))
+            counts = await asyncio.gather(
+                *(consumer() for _ in range(cfg["concurrency"]))
+            )
             await prod_task
 
         count = sum(counts)
@@ -134,7 +164,9 @@ async def _run(event: dict, request_id: str):
         # the whole manifest in memory like `body = await f.read()` would.
         with open(manifest_path, "rb") as f:
             await s3.put_object(
-                Bucket=cfg["bucket"], Key=manifest_key, Body=f,
+                Bucket=cfg["bucket"],
+                Key=manifest_key,
+                Body=f,
                 ContentType="application/x-ndjson",
             )
 
@@ -154,13 +186,24 @@ async def _run(event: dict, request_id: str):
         response_bytes = len(json.dumps(result))
         duration_ms = (time.monotonic() - t0) * 1000
 
-        _log("complete", request_id=request_id, count=count, errors=errors,
-             pages=pages, duration_ms=round(duration_ms, 2))
-        _emit_emf({
-            "PDFsProcessed": count, "S3ListPages": pages,
-            "PresignCount": count, "ManifestBytes": manifest_bytes,
-            "ResponseBytes": response_bytes,
-        }, Function="sign_pdfs")
+        _log(
+            "complete",
+            request_id=request_id,
+            count=count,
+            errors=errors,
+            pages=pages,
+            duration_ms=round(duration_ms, 2),
+        )
+        _emit_emf(
+            {
+                "PDFsProcessed": count,
+                "S3ListPages": pages,
+                "PresignCount": count,
+                "ManifestBytes": manifest_bytes,
+                "ResponseBytes": response_bytes,
+            },
+            Function="sign_pdfs",
+        )
 
         return result