breaking: Remove the content_types module

doc/v2.rst

@@ -70,6 +70,28 @@ Please use :func:`sagemaker.predictor.Predictor.delete_endpoint` instead.
 The ``update_endpoint`` argument in ``deploy()`` methods for estimators and models has been deprecated.
 Please use :func:`sagemaker.predictor.Predictor.update_endpoint` instead.
 
+``sagemaker.content_types``
+---------------------------
+
+The ``sagemaker.content_types`` module is removed in v2.0 and later of the
+SageMaker Python SDK.
+
+Instead of importing constants from ``sagemaker.content_types``, explicitly
+write MIME types as a string,
+
++-------------------------------+--------------------------------+
+| v1.x                          | v2.0 and later                 |
++===============================+================================+
+| ``CONTENT_TYPE_JSON``         | ``"application/json"``         |
++-------------------------------+--------------------------------+
+| ``CONTENT_TYPE_CSV``          | ``"text/csv"``                 |
++-------------------------------+--------------------------------+
+| ``CONTENT_TYPE_OCTET_STREAM`` | ``"application/octet-stream"`` |
++-------------------------------+--------------------------------+
+| ``CONTENT_TYPE_NPY``          | ``"application/x-npy"``        |
++-------------------------------+--------------------------------+
+
+
 Require ``framework_version`` and ``py_version`` for Frameworks
 ===============================================================
 

src/sagemaker/sparkml/model.py

@@ -14,7 +14,6 @@
 from __future__ import absolute_import
 
 from sagemaker import Model, Predictor, Session, image_uris
-from sagemaker.content_types import CONTENT_TYPE_CSV
 from sagemaker.serializers import CSVSerializer
 
 framework_name = "sparkml-serving"
@@ -50,7 +49,7 @@ def __init__(self, endpoint_name, sagemaker_session=None):
             endpoint_name=endpoint_name,
             sagemaker_session=sagemaker_session,
             serializer=CSVSerializer(),
-            content_type=CONTENT_TYPE_CSV,
+            content_type="text/csv",
         )
 
 

src/sagemaker/tensorflow/model.py

@@ -17,7 +17,6 @@
 
 import sagemaker
 from sagemaker import image_uris
-from sagemaker.content_types import CONTENT_TYPE_JSON
 from sagemaker.deserializers import JSONDeserializer
 from sagemaker.predictor import Predictor
 from sagemaker.serializers import JSONSerializer
@@ -96,7 +95,7 @@ def _classify_or_regress(self, data, method):
         if method not in ["classify", "regress"]:
             raise ValueError("invalid TensorFlow Serving method: {}".format(method))
 
-        if self.content_type != CONTENT_TYPE_JSON:
+        if self.content_type != "application/json":
             raise ValueError("The {} api requires json requests.".format(method))
 
         args = {"CustomAttributes": "tfs-method={}".format(method)}

tests/integ/test_inference_pipeline.py

@@ -23,7 +23,6 @@
 )
 
 from sagemaker import image_uris
-from sagemaker.content_types import CONTENT_TYPE_CSV
 from sagemaker.model import Model
 from sagemaker.pipeline import PipelineModel
 from sagemaker.predictor import Predictor
@@ -87,9 +86,7 @@ def test_inference_pipeline_batch_transform(sagemaker_session, cpu_instance_type
     with timeout_and_delete_model_with_transformer(
         transformer, sagemaker_session, minutes=TRANSFORM_DEFAULT_TIMEOUT_MINUTES
     ):
-        transformer.transform(
-            transform_input, content_type=CONTENT_TYPE_CSV, job_name=batch_job_name
-        )
+        transformer.transform(transform_input, content_type="text/csv", job_name=batch_job_name)
         transformer.wait()
 
 
@@ -134,8 +131,8 @@ def test_inference_pipeline_model_deploy(sagemaker_session, cpu_instance_type):
             endpoint_name=endpoint_name,
             sagemaker_session=sagemaker_session,
             serializer=JSONSerializer,
-            content_type=CONTENT_TYPE_CSV,
-            accept=CONTENT_TYPE_CSV,
+            content_type="text/csv",
+            accept="text/csv",
         )
 
         with open(VALID_DATA_PATH, "r") as f:

tests/integ/test_multi_variant_endpoint.py

@@ -23,7 +23,6 @@
 from sagemaker.s3 import S3Uploader
 from sagemaker.session import production_variant
 from sagemaker.sparkml import SparkMLModel
-from sagemaker.content_types import CONTENT_TYPE_CSV
 from sagemaker.utils import unique_name_from_base
 from sagemaker.predictor import Predictor
 from sagemaker.serializers import CSVSerializer
@@ -153,17 +152,17 @@ def test_target_variant_invocation(sagemaker_session, multi_variant_endpoint):
     response = sagemaker_session.sagemaker_runtime_client.invoke_endpoint(
         EndpointName=multi_variant_endpoint.endpoint_name,
         Body=TEST_CSV_DATA,
-        ContentType=CONTENT_TYPE_CSV,
-        Accept=CONTENT_TYPE_CSV,
+        ContentType="text/csv",
+        Accept="text/csv",
         TargetVariant=TEST_VARIANT_1,
     )
     assert response["InvokedProductionVariant"] == TEST_VARIANT_1
 
     response = sagemaker_session.sagemaker_runtime_client.invoke_endpoint(
         EndpointName=multi_variant_endpoint.endpoint_name,
         Body=TEST_CSV_DATA,
-        ContentType=CONTENT_TYPE_CSV,
-        Accept=CONTENT_TYPE_CSV,
+        ContentType="text/csv",
+        Accept="text/csv",
         TargetVariant=TEST_VARIANT_2,
     )
     assert response["InvokedProductionVariant"] == TEST_VARIANT_2
@@ -174,8 +173,8 @@ def test_predict_invocation_with_target_variant(sagemaker_session, multi_variant
         endpoint_name=multi_variant_endpoint.endpoint_name,
         sagemaker_session=sagemaker_session,
         serializer=CSVSerializer(),
-        content_type=CONTENT_TYPE_CSV,
-        accept=CONTENT_TYPE_CSV,
+        content_type="text/csv",
+        accept="text/csv",
     )
 
     # Validate that no exception is raised when the target_variant is specified.
@@ -191,8 +190,8 @@ def test_variant_traffic_distribution(sagemaker_session, multi_variant_endpoint)
         response = sagemaker_session.sagemaker_runtime_client.invoke_endpoint(
             EndpointName=multi_variant_endpoint.endpoint_name,
             Body=TEST_CSV_DATA,
-            ContentType=CONTENT_TYPE_CSV,
-            Accept=CONTENT_TYPE_CSV,
+            ContentType="text/csv",
+            Accept="text/csv",
         )
         if response["InvokedProductionVariant"] == TEST_VARIANT_1:
             variant_1_invocation_count += 1
@@ -274,17 +273,17 @@ def test_target_variant_invocation_local_mode(sagemaker_session, multi_variant_e
     response = sagemaker_session.sagemaker_runtime_client.invoke_endpoint(
         EndpointName=multi_variant_endpoint.endpoint_name,
         Body=TEST_CSV_DATA,
-        ContentType=CONTENT_TYPE_CSV,
-        Accept=CONTENT_TYPE_CSV,
+        ContentType="text/csv",
+        Accept="text/csv",
         TargetVariant=TEST_VARIANT_1,
     )
     assert response["InvokedProductionVariant"] == TEST_VARIANT_1
 
     response = sagemaker_session.sagemaker_runtime_client.invoke_endpoint(
         EndpointName=multi_variant_endpoint.endpoint_name,
         Body=TEST_CSV_DATA,
-        ContentType=CONTENT_TYPE_CSV,
-        Accept=CONTENT_TYPE_CSV,
+        ContentType="text/csv",
+        Accept="text/csv",
         TargetVariant=TEST_VARIANT_2,
     )
     assert response["InvokedProductionVariant"] == TEST_VARIANT_2
@@ -302,8 +301,8 @@ def test_predict_invocation_with_target_variant_local_mode(
         endpoint_name=multi_variant_endpoint.endpoint_name,
         sagemaker_session=sagemaker_session,
         serializer=CSVSerializer(),
-        content_type=CONTENT_TYPE_CSV,
-        accept=CONTENT_TYPE_CSV,
+        content_type="text/csv",
+        accept="text/csv",
     )
 
     # Validate that no exception is raised when the target_variant is specified.

tests/integ/test_tuner_multi_algo.py

@@ -19,7 +19,6 @@
 
 from sagemaker import image_uris, utils
 from sagemaker.analytics import HyperparameterTuningJobAnalytics
-from sagemaker.content_types import CONTENT_TYPE_JSON
 from sagemaker.deserializers import JSONDeserializer
 from sagemaker.estimator import Estimator
 from sagemaker.tuner import ContinuousParameter, IntegerParameter, HyperparameterTuner
@@ -215,7 +214,7 @@ def _create_training_inputs(sagemaker_session):
 
 def _make_prediction(predictor, data):
     predictor.serializer = _prediction_data_serializer
-    predictor.content_type = CONTENT_TYPE_JSON
+    predictor.content_type = "application/json"
     predictor.deserializer = JSONDeserializer()
     return predictor.predict(data)