chore(source-s3): Update CDK to v5 (old)

airbyte-integrations/connectors/source-s3/integration_tests/spec.json

@@ -1,5 +1,7 @@
 {
   "documentationUrl": "https://docs.airbyte.com/integrations/sources/s3",
+  "supportsNormalization": false,
+  "supportsDBT": false,
   "connectionSpecification": {
     "title": "Config",
     "description": "NOTE: When this Spec is changed, legacy_config_transformer.py must also be modified to uptake the changes\nbecause it is responsible for converting legacy S3 v3 configs into v4 configs using the File-Based CDK.",
@@ -338,6 +340,21 @@
                   },
                   "description": "Extract text from document formats (.pdf, .docx, .md, .pptx) and emit as one record per file.",
                   "required": ["filetype"]
+                },
+                {
+                  "title": "Excel Format",
+                  "type": "object",
+                  "properties": {
+                    "filetype": {
+                      "title": "Filetype",
+                      "default": "excel",
+                      "const": "excel",
+                      "type": "string"
+                    }
+                  },
+                  "required": [
+                    "filetype"
+                  ]
                 }
               ]
             },

airbyte-integrations/connectors/source-s3/metadata.yaml

@@ -10,7 +10,7 @@ data:
   connectorSubtype: file
   connectorType: source
   definitionId: 69589781-7828-43c5-9f63-8925b1c1ccc2
-  dockerImageTag: 4.8.0
+  dockerImageTag: 4.9.0
   dockerRepository: airbyte/source-s3
   documentationUrl: https://docs.airbyte.com/integrations/sources/s3
   githubIssueLabel: source-s3

airbyte-integrations/connectors/source-s3/pyproject.toml

@@ -3,7 +3,7 @@ requires = [ "poetry-core>=1.0.0",]
 build-backend = "poetry.core.masonry.api"
 
 [tool.poetry]
-version = "4.8.0"
+version = "4.9.0"
 name = "source-s3"
 description = "Source implementation for S3."
 authors = [ "Airbyte <[email protected]>",]
@@ -23,7 +23,7 @@ python-snappy = "==0.6.1"
 dill = "==0.3.4"
 transformers = "4.38.2"
 urllib3 = "<2"
-airbyte-cdk = {extras = ["file-based"], version = "^3"}
+airbyte-cdk = {extras = ["file-based"], version = "^5"}
 
 [tool.poetry.scripts]
 source-s3 = "source_s3.run:run"

airbyte-integrations/connectors/source-s3/source_s3/run.py

@@ -3,12 +3,13 @@
 #
 
 import sys
+import time
 import traceback
-from datetime import datetime
 from typing import List
 
 from airbyte_cdk import AirbyteEntrypoint, AirbyteMessage, Type, launch
-from airbyte_cdk.models import AirbyteErrorTraceMessage, AirbyteTraceMessage, TraceType
+from airbyte_cdk.models import AirbyteErrorTraceMessage, AirbyteMessageSerializer, AirbyteTraceMessage, TraceType
+from orjson import orjson
 from source_s3.v4 import Config, Cursor, SourceS3, SourceS3StreamReader
 
 
@@ -27,17 +28,21 @@ def get_source(args: List[str]):
         )
     except Exception:
         print(
-            AirbyteMessage(
-                type=Type.TRACE,
-                trace=AirbyteTraceMessage(
-                    type=TraceType.ERROR,
-                    emitted_at=int(datetime.now().timestamp() * 1000),
-                    error=AirbyteErrorTraceMessage(
-                        message="Error starting the sync. This could be due to an invalid configuration or catalog. Please contact Support for assistance.",
-                        stack_trace=traceback.format_exc(),
-                    ),
-                ),
-            ).json()
+            orjson.dumps(
+                AirbyteMessageSerializer.dump(
+                    AirbyteMessage(
+                        type=Type.TRACE,
+                        trace=AirbyteTraceMessage(
+                            type=TraceType.ERROR,
+                            emitted_at=time.time_ns() // 1_000_000,
+                            error=AirbyteErrorTraceMessage(
+                                message="Error starting the sync. This could be due to an invalid configuration or catalog. Please contact Support for assistance.",
+                                stack_trace=traceback.format_exc(),
+                            ),
+                        ),
+                    )
+                )
+            ).decode()
         )
         return None
 

airbyte-integrations/connectors/source-s3/v4_main.py

@@ -3,12 +3,13 @@
 #
 
 import sys
+import time
 import traceback
-from datetime import datetime
 from typing import List
 
 from airbyte_cdk import AirbyteEntrypoint, AirbyteMessage, Type, launch
-from airbyte_cdk.models import AirbyteErrorTraceMessage, AirbyteTraceMessage, TraceType
+from airbyte_cdk.models import AirbyteErrorTraceMessage, AirbyteMessageSerializer, AirbyteTraceMessage, TraceType
+from orjson import orjson
 from source_s3.v4 import Config, Cursor, SourceS3, SourceS3StreamReader
 
 
@@ -27,17 +28,21 @@ def get_source(args: List[str]):
         )
     except Exception:
         print(
-            AirbyteMessage(
-                type=Type.TRACE,
-                trace=AirbyteTraceMessage(
-                    type=TraceType.ERROR,
-                    emitted_at=int(datetime.now().timestamp() * 1000),
-                    error=AirbyteErrorTraceMessage(
-                        message="Error starting the sync. This could be due to an invalid configuration or catalog. Please contact Support for assistance.",
-                        stack_trace=traceback.format_exc(),
-                    ),
-                ),
-            ).json()
+            orjson.dumps(
+                AirbyteMessageSerializer.dump(
+                    AirbyteMessage(
+                        type=Type.TRACE,
+                        trace=AirbyteTraceMessage(
+                            type=TraceType.ERROR,
+                            emitted_at=time.time_ns() // 1_000_000,
+                            error=AirbyteErrorTraceMessage(
+                                message="Error starting the sync. This could be due to an invalid configuration or catalog. Please contact Support for assistance.",
+                                stack_trace=traceback.format_exc(),
+                            ),
+                        ),
+                    )
+                )
+            ).decode()
         )
         return None