hatchet-sdk 1.14.2__py3-none-any.whl → 1.14.3__py3-none-any.whl

hatchet_sdk/__init__.py

@@ -11,9 +11,9 @@ from hatchet_sdk.clients.listeners.run_event_listener import (
     StepRunEventType,
     WorkflowRunEventType,
 )
-from hatchet_sdk.clients.rest.models.accept_invite_request import AcceptInviteRequest
 
 # import models into sdk package
+from hatchet_sdk.clients.rest.models.accept_invite_request import AcceptInviteRequest
 from hatchet_sdk.clients.rest.models.api_error import APIError
 from hatchet_sdk.clients.rest.models.api_errors import APIErrors
 from hatchet_sdk.clients.rest.models.api_meta import APIMeta
@@ -166,6 +166,7 @@ from hatchet_sdk.runnables.types import (
 )
 from hatchet_sdk.runnables.workflow import TaskRunRef
 from hatchet_sdk.utils.opentelemetry import OTelAttribute
+from hatchet_sdk.utils.serde import remove_null_unicode_character
 from hatchet_sdk.worker.worker import Worker, WorkerStartOptions, WorkerStatus
 from hatchet_sdk.workflow_run import WorkflowRunRef
 
@@ -290,5 +291,6 @@ __all__ = [
     "WorkflowVersionDefinition",
     "WorkflowVersionMeta",
     "or_",
+    "remove_null_unicode_character",
     "workflow",
 ]

hatchet_sdk/utils/serde.py

@@ -0,0 +1,52 @@
+from typing import Any, TypeVar, cast, overload
+
+T = TypeVar("T")
+K = TypeVar("K")
+
+
+@overload
+def remove_null_unicode_character(data: str) -> str: ...
+
+
+@overload
+def remove_null_unicode_character(data: dict[K, T]) -> dict[K, T]: ...
+
+
+@overload
+def remove_null_unicode_character(data: list[T]) -> list[T]: ...
+
+
+@overload
+def remove_null_unicode_character(data: tuple[T, ...]) -> tuple[T, ...]: ...
+
+
+def remove_null_unicode_character(
+    data: str | dict[K, T] | list[T] | tuple[T, ...],
+) -> str | dict[K, T] | list[T] | tuple[T, ...]:
+    """
+    Recursively traverse a dictionary (a task's output) and remove the unicode escape sequence \\u0000 which will cause unexpected behavior in Hatchet.
+
+    Needed as Hatchet does not support \\u0000 in task outputs
+
+    :param data: The task output (a JSON-serializable dictionary or mapping)
+    :return: The same dictionary with all \\u0000 characters removed from strings, and nested dictionaries/lists processed recursively.
+    :raises TypeError: If the input is not a string, dictionary, list, or tuple.
+    """
+    if isinstance(data, str):
+        return data.replace("\u0000", "")
+
+    if isinstance(data, dict):
+        return {
+            key: remove_null_unicode_character(cast(Any, value))
+            for key, value in data.items()
+        }
+
+    if isinstance(data, list):
+        return [remove_null_unicode_character(cast(Any, item)) for item in data]
+
+    if isinstance(data, tuple):
+        return tuple(remove_null_unicode_character(cast(Any, item)) for item in data)
+
+    raise TypeError(
+        f"Unsupported type {type(data)}. Expected str, dict, list, or tuple."
+    )

hatchet_sdk/worker/runner/runner.py

@@ -4,9 +4,9 @@ import functools
 import json
 from collections.abc import Callable
 from concurrent.futures import ThreadPoolExecutor
-from contextlib import suppress
 from enum import Enum
 from multiprocessing import Queue
+from textwrap import dedent
 from threading import Thread, current_thread
 from typing import Any, Literal, cast, overload
 
@@ -49,6 +49,7 @@ from hatchet_sdk.runnables.contextvars import (
 )
 from hatchet_sdk.runnables.task import Task
 from hatchet_sdk.runnables.types import R, TWorkflowInput
+from hatchet_sdk.utils.serde import remove_null_unicode_character
 from hatchet_sdk.worker.action_listener_process import ActionEvent
 from hatchet_sdk.worker.runner.utils.capture_logs import (
     AsyncLogSender,
@@ -410,7 +411,7 @@ class Runner:
         )
 
     ## IMPORTANT: Keep this method's signature in sync with the wrapper in the OTel instrumentor
-    async def handle_start_step_run(self, action: Action) -> None:
+    async def handle_start_step_run(self, action: Action) -> Exception | None:
         action_name = action.action_id
 
         # Find the corresponding action function from the registry
@@ -444,8 +445,11 @@ class Runner:
 
             ## FIXME: Handle cancelled exceptions and other special exceptions
             ## that we don't want to suppress here
-            with suppress(Exception):
+            try:
                 await task
+            except Exception as e:
+                ## Used for the OTel instrumentor to capture exceptions
+                return e
 
         ## Once the step run completes, we need to remove the workflow spawn index
         ## so we don't leak memory
@@ -453,6 +457,8 @@ class Runner:
             async with spawn_index_lock:
                 workflow_spawn_indices.pop(action.key)
 
+        return None
+
     ## IMPORTANT: Keep this method's signature in sync with the wrapper in the OTel instrumentor
     async def handle_start_group_key_run(self, action: Action) -> Exception | None:
         action_name = action.action_id
@@ -557,14 +563,30 @@ class Runner:
                 f"Tasks must return either a dictionary or a Pydantic BaseModel which can be serialized to a JSON object. Got object of type {type(output)} instead."
             )
 
-        if output is not None:
-            try:
-                return json.dumps(output, default=str)
-            except Exception as e:
-                logger.error(f"Could not serialize output: {e}")
-                return str(output)
+        if output is None:
+            return ""
+
+        try:
+            serialized_output = json.dumps(output, default=str)
+        except Exception as e:
+            logger.error(f"Could not serialize output: {e}")
+            serialized_output = str(output)
+
+        if "\\u0000" in serialized_output:
+            raise IllegalTaskOutputError(
+                dedent(
+                    f"""
+                Task outputs cannot contain the unicode null character \\u0000
+
+                Please see this Discord thread: https://discord.com/channels/1088927970518909068/1384324576166678710/1386714014565928992
+                Relevant Postgres documentation: https://www.postgresql.org/docs/current/datatype-json.html
+
+                Use `hatchet_sdk.{remove_null_unicode_character.__name__}` to sanitize your output if you'd like to remove the character.
+                """
+                )
+            )
 
-        return ""
+        return serialized_output
 
     async def wait_for_tasks(self) -> None:
         running = len(self.tasks.keys())

{hatchet_sdk-1.14.2.dist-info → hatchet_sdk-1.14.3.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: hatchet-sdk
-Version: 1.14.2
+Version: 1.14.3
 Summary: 
 License: MIT
 Author: Alexander Belanger

{hatchet_sdk-1.14.2.dist-info → hatchet_sdk-1.14.3.dist-info}/RECORD

@@ -1,4 +1,4 @@
-hatchet_sdk/__init__.py,sha256=3iVKyZ4OnGxZWU0YtXQG9imlljPNBdPARqqCm6Ft03o,10640
+hatchet_sdk/__init__.py,sha256=LdT8xJ7-95gwksRZuCzuKDM5bfLx8uW8gXGPL7VMqwc,10743
 hatchet_sdk/client.py,sha256=OXb2hOJ5p7pY5QMlM4cydb4aGyf6bDdbyWQjPMVCe64,2413
 hatchet_sdk/clients/admin.py,sha256=t8GXhjMiFp9iL0ofISNxSOKlfwaOTX2iQbZfz1G0pHU,16936
 hatchet_sdk/clients/dispatcher/action_listener.py,sha256=tFknXH9iSP0OFYYVcKeDZVrcDNIz00ZQVTxSbHpbKhI,13863
@@ -283,6 +283,7 @@ hatchet_sdk/utils/backoff.py,sha256=6B5Rb5nLKw_TqqgpJMYjIBV1PTTtbOMRZCveisVhg_I,
 hatchet_sdk/utils/datetimes.py,sha256=vIZNEX8tt-bknaIuTmoLEmKVt18dBjClH3urYtCJAys,775
 hatchet_sdk/utils/opentelemetry.py,sha256=64TVwCLrUzEmcL2BUNPV_QubfiR5jajOZtVeGYLnEEA,1226
 hatchet_sdk/utils/proto_enums.py,sha256=v2gp_ZmIhPxURVXwz5lscllXwZXDl5XGXeL6gezw3o0,1241
+hatchet_sdk/utils/serde.py,sha256=d2iypdK2UQCPA19NgYa0Tr7CTbk25KPbaCMqcqN3CYk,1645
 hatchet_sdk/utils/timedelta_to_expression.py,sha256=YujnBnGn7lxtkUdKIeqmOiN_ZCGBpRPjCCSzcD3jxzA,644
 hatchet_sdk/utils/typing.py,sha256=FgYnZyJSoRjNVFodxlI9gn0X8ST1KFed7xfUynIxa2U,978
 hatchet_sdk/v0/__init__.py,sha256=r3Q7l2RsLgdIkK2jjiz7-JJpD1T_Zy--Oa9MN5n_yEs,9654
@@ -509,11 +510,11 @@ hatchet_sdk/v0/workflow_run.py,sha256=jsEZprXshrSV7i_TtL5uoCL03D18zQ3NeJCq7mp97D
 hatchet_sdk/worker/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 hatchet_sdk/worker/action_listener_process.py,sha256=Xzzn1dDFJrqnC9HBsh3fYI8lfpOD4Ecze47qmm_XUWE,12923
 hatchet_sdk/worker/runner/run_loop_manager.py,sha256=GeILClNXaDbsjXCQb0bBdgeyAwZGem2JdaH0t6wz__I,4053
-hatchet_sdk/worker/runner/runner.py,sha256=rTD0OrPMyFY0k4DDcKpuz67K1YpHdZUZSd1o2L_QlYs,20781
+hatchet_sdk/worker/runner/runner.py,sha256=Btz7wd0DQ-5aPEZLN3T5X5__8DjuxN6H4yQEVX0eZiI,21678
 hatchet_sdk/worker/runner/utils/capture_logs.py,sha256=DKw6boqVsSCM1XvBWYuc833MZxCdSpMxg3l4aAqKPyw,3465
 hatchet_sdk/worker/worker.py,sha256=nDuRo_LishRuOCTnDonc7G7qeOoW93nRHGd-fQOE_bs,16541
 hatchet_sdk/workflow_run.py,sha256=KcylcqRwKADtnzOTjoiVr1vdr7qTZFtDeD5aRS6A4Y8,2823
-hatchet_sdk-1.14.2.dist-info/METADATA,sha256=Z5hZyudaXspH-TNuGTWoEtVcGtrV5UGeImzNo3ePanI,3636
-hatchet_sdk-1.14.2.dist-info/WHEEL,sha256=FMvqSimYX_P7y0a7UY-_Mc83r5zkBZsCYPm7Lr0Bsq4,88
-hatchet_sdk-1.14.2.dist-info/entry_points.txt,sha256=Un_76pcLse-ZGBlwebhQpnTPyQrripeHW8J7qmEpGOk,1400
-hatchet_sdk-1.14.2.dist-info/RECORD,,
+hatchet_sdk-1.14.3.dist-info/METADATA,sha256=3D-GMkNjhRG37vaE8W4lXoG1MMu_2MpnKcAwK5elTkY,3636
+hatchet_sdk-1.14.3.dist-info/WHEEL,sha256=FMvqSimYX_P7y0a7UY-_Mc83r5zkBZsCYPm7Lr0Bsq4,88
+hatchet_sdk-1.14.3.dist-info/entry_points.txt,sha256=Un_76pcLse-ZGBlwebhQpnTPyQrripeHW8J7qmEpGOk,1400
+hatchet_sdk-1.14.3.dist-info/RECORD,,