feat(telemetry): Implement telemetry message notification (kedro-org#760

) * Implement telemetry message notification Signed-off-by: Dmitry Sorokin <dmd40in@gmail.com> * Fix docs build Signed-off-by: Dmitry Sorokin <dmd40in@gmail.com> --------- Signed-off-by: Dmitry Sorokin <dmd40in@gmail.com> Signed-off-by: Merel Theisen <merel.theisen@quantumblack.com>
galenseilis · Aug 27, 2024 · 021cb73 · 021cb73
1 parent 7303e1f
commit 021cb73
Show file tree

Hide file tree

Showing 5 changed files with 112 additions and 75 deletions.
diff --git a/kedro-datasets/docs/source/api/kedro_datasets.rst b/kedro-datasets/docs/source/api/kedro_datasets.rst
@@ -11,56 +11,56 @@ kedro_datasets
    :toctree:
    :template: autosummary/class.rst
 
-   kedro_datasets.api.APIDataset
-   kedro_datasets.biosequence.BioSequenceDataset
-   kedro_datasets.dask.CSVDataset
-   kedro_datasets.dask.ParquetDataset
-   kedro_datasets.databricks.ManagedTableDataset
-   kedro_datasets.email.EmailMessageDataset
-   kedro_datasets.geopandas.GeoJSONDataset
-   kedro_datasets.holoviews.HoloviewsWriter
-   kedro_datasets.huggingface.HFDataset
-   kedro_datasets.huggingface.HFTransformerPipelineDataset
-   kedro_datasets.ibis.TableDataset
-   kedro_datasets.json.JSONDataset
-   kedro_datasets.matlab.MatlabDataset
-   kedro_datasets.matplotlib.MatplotlibWriter
-   kedro_datasets.networkx.GMLDataset
-   kedro_datasets.networkx.GraphMLDataset
-   kedro_datasets.networkx.JSONDataset
-   kedro_datasets.pandas.CSVDataset
-   kedro_datasets.pandas.DeltaTableDataset
-   kedro_datasets.pandas.ExcelDataset
-   kedro_datasets.pandas.FeatherDataset
-   kedro_datasets.pandas.GBQQueryDataset
-   kedro_datasets.pandas.GBQTableDataset
-   kedro_datasets.pandas.GenericDataset
-   kedro_datasets.pandas.HDFDataset
-   kedro_datasets.pandas.JSONDataset
-   kedro_datasets.pandas.ParquetDataset
-   kedro_datasets.pandas.SQLQueryDataset
-   kedro_datasets.pandas.SQLTableDataset
-   kedro_datasets.pandas.XMLDataset
-   kedro_datasets.partitions.IncrementalDataset
-   kedro_datasets.partitions.PartitionedDataset
-   kedro_datasets.pickle.PickleDataset
-   kedro_datasets.pillow.ImageDataset
-   kedro_datasets.plotly.JSONDataset
-   kedro_datasets.plotly.PlotlyDataset
-   kedro_datasets.polars.CSVDataset
-   kedro_datasets.polars.EagerPolarsDataset
-   kedro_datasets.polars.LazyPolarsDataset
-   kedro_datasets.redis.PickleDataset
-   kedro_datasets.snowflake.SnowparkTableDataset
-   kedro_datasets.spark.DeltaTableDataset
-   kedro_datasets.spark.SparkDataset
-   kedro_datasets.spark.SparkHiveDataset
-   kedro_datasets.spark.SparkJDBCDataset
-   kedro_datasets.spark.SparkStreamingDataset
-   kedro_datasets.svmlight.SVMLightDataset
-   kedro_datasets.tensorflow.TensorFlowModelDataset
-   kedro_datasets.text.TextDataset
-   kedro_datasets.tracking.JSONDataset
-   kedro_datasets.tracking.MetricsDataset
-   kedro_datasets.video.VideoDataset
-   kedro_datasets.yaml.YAMLDataset
+   api.APIDataset
+   biosequence.BioSequenceDataset
+   dask.CSVDataset
+   dask.ParquetDataset
+   databricks.ManagedTableDataset
+   email.EmailMessageDataset
+   geopandas.GeoJSONDataset
+   holoviews.HoloviewsWriter
+   huggingface.HFDataset
+   huggingface.HFTransformerPipelineDataset
+   ibis.TableDataset
+   json.JSONDataset
+   matlab.MatlabDataset
+   matplotlib.MatplotlibWriter
+   networkx.GMLDataset
+   networkx.GraphMLDataset
+   networkx.JSONDataset
+   pandas.CSVDataset
+   pandas.DeltaTableDataset
+   pandas.ExcelDataset
+   pandas.FeatherDataset
+   pandas.GBQQueryDataset
+   pandas.GBQTableDataset
+   pandas.GenericDataset
+   pandas.HDFDataset
+   pandas.JSONDataset
+   pandas.ParquetDataset
+   pandas.SQLQueryDataset
+   pandas.SQLTableDataset
+   pandas.XMLDataset
+   partitions.IncrementalDataset
+   partitions.PartitionedDataset
+   pickle.PickleDataset
+   pillow.ImageDataset
+   plotly.JSONDataset
+   plotly.PlotlyDataset
+   polars.CSVDataset
+   polars.EagerPolarsDataset
+   polars.LazyPolarsDataset
+   redis.PickleDataset
+   snowflake.SnowparkTableDataset
+   spark.DeltaTableDataset
+   spark.SparkDataset
+   spark.SparkHiveDataset
+   spark.SparkJDBCDataset
+   spark.SparkStreamingDataset
+   svmlight.SVMLightDataset
+   tensorflow.TensorFlowModelDataset
+   text.TextDataset
+   tracking.JSONDataset
+   tracking.MetricsDataset
+   video.VideoDataset
+   yaml.YAMLDataset
diff --git a/kedro-datasets/docs/source/api/kedro_datasets_experimental.rst b/kedro-datasets/docs/source/api/kedro_datasets_experimental.rst
@@ -11,9 +11,9 @@ kedro_datasets_experimental
    :toctree:
    :template: autosummary/class.rst
 
-    kedro_datasets_experimental.langchain.ChatAnthropicDataset
-    kedro_datasets_experimental.langchain.ChatCohereDataset
-    kedro_datasets_experimental.langchain.ChatOpenAIDataset
-    kedro_datasets_experimental.langchain.OpenAIEmbeddingsDataset
-    kedro_datasets_experimental.netcdf.NetCDFDataset
-    kedro_datasets_experimental.rioxarray.GeoTIFFDataset
+   langchain.ChatAnthropicDataset
+   langchain.ChatCohereDataset
+   langchain.ChatOpenAIDataset
+   langchain.OpenAIEmbeddingsDataset
+   netcdf.NetCDFDataset
+   rioxarray.GeoTIFFDataset
diff --git a/kedro-telemetry/kedro_telemetry/__init__.py b/kedro-telemetry/kedro_telemetry/__init__.py
@@ -1,3 +1,7 @@
 """Kedro plugin for collecting Kedro usage data."""
 
 __version__ = "0.5.0"
+
+import logging
+
+logging.getLogger(__name__).setLevel(logging.INFO)
diff --git a/kedro-telemetry/kedro_telemetry/plugin.py b/kedro-telemetry/kedro_telemetry/plugin.py
@@ -163,12 +163,20 @@ def before_command_run(
 
             consent = _check_for_telemetry_consent(project_metadata.project_path)
             if not consent:
-                logger.debug(
+                logger.info(
                     "Kedro-Telemetry is installed, but you have opted out of "
                     "sharing usage analytics so none will be collected.",
                 )
                 return
 
+            logger.info(
+                "Kedro is sending anonymous usage data with the sole purpose of improving the product. "
+                "No personal data or IP addresses are stored on our side. "
+                "If you want to opt out, set the `KEDRO_DISABLE_TELEMETRY` or `DO_NOT_TRACK` environment variables, "
+                "or create a `.telemetry` file in the current working directory with the contents `consent: false`. "
+                "Read more at https://docs.kedro.org/en/stable/configuration/telemetry.html"
+            )
+
             # get KedroCLI and its structure from actual project root
             cli = KedroCLI(project_path=project_metadata.project_path)
             cli_struct = _get_cli_structure(cli_obj=cli, get_help=False)
@@ -177,7 +185,6 @@ def before_command_run(
             )
             main_command = masked_command_args[0] if masked_command_args else "kedro"
 
-            logger.debug("You have opted into product usage analytics.")
             user_uuid = _get_or_create_uuid()
             project_properties = _get_project_properties(
                 user_uuid, project_metadata.project_path / PYPROJECT_CONFIG_NAME
@@ -219,15 +226,10 @@ def after_context_created(self, context):
 
     @hook_impl
     def after_catalog_created(self, catalog):
+        # The user notification message is sent only once per command during the before_command_run hook
         if not self.consent:
-            logger.debug(
-                "Kedro-Telemetry is installed, but you have opted out of "
-                "sharing usage analytics so none will be collected.",
-            )
             return
 
-        logger.debug("You have opted into product usage analytics.")
-
         default_pipeline = pipelines.get("__default__")  # __default__
         user_uuid = _get_or_create_uuid()
 

diff --git a/kedro-telemetry/tests/test_plugin.py b/kedro-telemetry/tests/test_plugin.py
@@ -1,3 +1,4 @@
+import logging
 import sys
 from pathlib import Path
 
@@ -121,7 +122,7 @@ def fake_sub_pipeline():
 
 
 class TestKedroTelemetryCLIHooks:
-    def test_before_command_run(self, mocker, fake_metadata):
+    def test_before_command_run(self, mocker, fake_metadata, caplog):
         mocker.patch(
             "kedro_telemetry.plugin._check_for_telemetry_consent", return_value=True
         )
@@ -139,9 +140,10 @@ def test_before_command_run(self, mocker, fake_metadata):
         )
 
         mocked_heap_call = mocker.patch("kedro_telemetry.plugin._send_heap_event")
-        telemetry_hook = KedroTelemetryCLIHooks()
-        command_args = ["--version"]
-        telemetry_hook.before_command_run(fake_metadata, command_args)
+        with caplog.at_level(logging.INFO):
+            telemetry_hook = KedroTelemetryCLIHooks()
+            command_args = ["--version"]
+            telemetry_hook.before_command_run(fake_metadata, command_args)
         expected_properties = {
             "username": "user_uuid",
             "project_id": "digested",
@@ -170,6 +172,20 @@ def test_before_command_run(self, mocker, fake_metadata):
             ),
         ]
         assert mocked_heap_call.call_args_list == expected_calls
+        assert any(
+            "Kedro is sending anonymous usage data with the sole purpose of improving the product. "
+            "No personal data or IP addresses are stored on our side. "
+            "If you want to opt out, set the `KEDRO_DISABLE_TELEMETRY` or `DO_NOT_TRACK` environment variables, "
+            "or create a `.telemetry` file in the current working directory with the contents `consent: false`. "
+            "Read more at https://docs.kedro.org/en/stable/configuration/telemetry.html"
+            in record.message
+            for record in caplog.records
+        )
+        assert not any(
+            "Kedro-Telemetry is installed, but you have opted out of "
+            "sharing usage analytics so none will be collected." in record.message
+            for record in caplog.records
+        )
 
     def test_before_command_run_with_tools(self, mocker, fake_metadata):
         mocker.patch(
@@ -276,17 +292,32 @@ def test_before_command_run_empty_args(self, mocker, fake_metadata):
 
         assert mocked_heap_call.call_args_list == expected_calls
 
-    def test_before_command_run_no_consent_given(self, mocker, fake_metadata):
+    def test_before_command_run_no_consent_given(self, mocker, fake_metadata, caplog):
         mocker.patch(
             "kedro_telemetry.plugin._check_for_telemetry_consent", return_value=False
         )
 
         mocked_heap_call = mocker.patch("kedro_telemetry.plugin._send_heap_event")
-        telemetry_hook = KedroTelemetryCLIHooks()
-        command_args = ["--version"]
-        telemetry_hook.before_command_run(fake_metadata, command_args)
+        with caplog.at_level(logging.INFO):
+            telemetry_hook = KedroTelemetryCLIHooks()
+            command_args = ["--version"]
+            telemetry_hook.before_command_run(fake_metadata, command_args)
 
         mocked_heap_call.assert_not_called()
+        assert not any(
+            "Kedro is sending anonymous usage data with the sole purpose of improving the product. "
+            "No personal data or IP addresses are stored on our side. "
+            "If you want to opt out, set the `KEDRO_DISABLE_TELEMETRY` or `DO_NOT_TRACK` environment variables, "
+            "or create a `.telemetry` file in the current working directory with the contents `consent: false`. "
+            "Read more at https://docs.kedro.org/en/latest/configuration/telemetry.html"
+            in record.message
+            for record in caplog.records
+        )
+        assert any(
+            "Kedro-Telemetry is installed, but you have opted out of "
+            "sharing usage analytics so none will be collected." in record.message
+            for record in caplog.records
+        )
 
     def test_before_command_run_connection_error(self, mocker, fake_metadata, caplog):
         mocker.patch(