snowpark-checkpoints-validators 0.2.0rc1__py3-none-any.whl → 0.2.1__py3-none-any.whl

snowflake/snowpark_checkpoints/utils/telemetry.py

@@ -0,0 +1,900 @@
+# Copyright 2025 Snowflake Inc.
+# SPDX-License-Identifier: Apache-2.0
+
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+
+# http://www.apache.org/licenses/LICENSE-2.0
+
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+import atexit
+import datetime
+import hashlib
+import inspect
+import json
+import os
+import re
+
+from contextlib import suppress
+from enum import IntEnum
+from functools import wraps
+from pathlib import Path
+from platform import python_version
+from sys import platform
+from typing import Any, Callable, Optional, TypeVar
+from uuid import getnode
+
+from snowflake.connector import (
+    SNOWFLAKE_CONNECTOR_VERSION,
+    time_util,
+)
+from snowflake.connector.constants import DIRS as SNOWFLAKE_DIRS
+from snowflake.connector.network import SnowflakeRestful
+from snowflake.connector.telemetry import TelemetryClient
+from snowflake.snowpark import VERSION as SNOWPARK_VERSION
+from snowflake.snowpark import dataframe as snowpark_dataframe
+from snowflake.snowpark.session import Session
+
+
+try:
+    from pyspark.sql import dataframe as spark_dataframe
+
+    def _is_spark_dataframe(df: Any) -> bool:
+        return isinstance(df, spark_dataframe.DataFrame)
+
+    def _get_spark_schema_types(df: spark_dataframe.DataFrame) -> list[str]:
+        return [str(schema_type.dataType) for schema_type in df.schema.fields]
+
+except Exception:
+
+    def _is_spark_dataframe(df: Any):
+        pass
+
+    def _get_spark_schema_types(df: Any):
+        pass
+
+
+VERSION_VARIABLE_PATTERN = r"^__version__ = ['\"]([^'\"]*)['\"]"
+VERSION_FILE_NAME = "__version__.py"
+
+
+class TelemetryManager(TelemetryClient):
+    def __init__(
+        self, rest: Optional[SnowflakeRestful] = None, is_telemetry_enabled: bool = True
+    ):
+        """TelemetryManager class to log telemetry events."""
+        super().__init__(rest)
+        self.sc_folder_path = (
+            Path(SNOWFLAKE_DIRS.user_config_path) / "snowpark-checkpoints-telemetry"
+        )
+        self.sc_sf_path_telemetry = "/telemetry/send"
+        self.sc_flush_size = 25
+        self.sc_is_enabled = is_telemetry_enabled
+        self.sc_is_testing = self._sc_is_telemetry_testing()
+        self.sc_memory_limit = 5 * 1024 * 1024
+        self._sc_upload_local_telemetry()
+        self.sc_log_batch = []
+        self.sc_hypothesis_input_events = []
+        self.sc_version = _get_version()
+        if rest:
+            atexit.register(self._sc_close_at_exit)
+
+    def set_sc_output_path(self, path: Path) -> None:
+        """Set the output path for testing.
+
+        Args:
+            path: path to write telemetry.
+
+        """
+        os.makedirs(path, exist_ok=True)
+        self.sc_folder_path = path
+
+    def sc_log_error(
+        self, event_name: str, parameters_info: Optional[dict] = None
+    ) -> None:
+        """Log an error telemetry event.
+
+        Args:
+            event_name (str): The name of the event.
+            parameters_info (dict, optional): Additional parameters for the event. Defaults to None.
+
+        """
+        if event_name is not None:
+            self._sc_log_telemetry(event_name, "error", parameters_info)
+
+    def sc_log_info(
+        self, event_name: str, parameters_info: Optional[dict] = None
+    ) -> None:
+        """Log an information telemetry event.
+
+        Args:
+            event_name (str): The name of the event.
+            parameters_info (dict, optional): Additional parameters for the event. Defaults to None.
+
+        """
+        if event_name is not None:
+            self._sc_log_telemetry(event_name, "info", parameters_info)
+
+    def _sc_log_telemetry(
+        self, event_name: str, event_type: str, parameters_info: Optional[dict] = None
+    ) -> dict:
+        """Log a telemetry event if is enabled.
+
+        Args:
+            event_name (str): The name of the event.
+            event_type (str): The type of the event (e.g., "error", "info").
+            parameters_info (dict, optional): Additional parameters for the event. Defaults to None.
+
+        Returns:
+            dict: The logged event.
+
+        """
+        if not self.sc_is_enabled:
+            return {}
+        event = _generate_event(
+            event_name, event_type, parameters_info, self.sc_version
+        )
+        self._sc_add_log_to_batch(event)
+        return event
+
+    def _sc_add_log_to_batch(self, event: dict) -> None:
+        """Add a log event to the batch. If the batch is full, send the events to the API.
+
+        Args:
+            event (dict): The event to add.
+
+        """
+        self.sc_log_batch.append(event)
+        if self.sc_is_testing:
+            self._sc_write_telemetry(self.sc_log_batch)
+            self.sc_log_batch = []
+            return
+
+        if len(self.sc_log_batch) >= self.sc_flush_size:
+            self.sc_send_batch(self.sc_log_batch)
+            self.sc_log_batch = []
+
+    def sc_send_batch(self, to_sent: list) -> bool:
+        """Send a request to the API to upload the events. If not have connection, write the events to local folder.
+
+        Args:
+            to_sent (list): The batch of events to send.
+
+        Returns:
+            bool: True if the batch was sent successfully, False otherwise.
+
+        """
+        if not self.sc_is_enabled:
+            return False
+        if self._rest is None:
+            self._sc_write_telemetry(to_sent)
+            self.sc_log_batch = []
+            return False
+        if to_sent == []:
+            return False
+        body = {"logs": to_sent}
+        ret = self._rest.request(
+            self.sc_sf_path_telemetry,
+            body=body,
+            method="post",
+            client=None,
+            timeout=5,
+        )
+        if not ret.get("success"):
+            self._sc_write_telemetry(to_sent)
+            self.sc_log_batch = []
+            return False
+        return True
+
+    def _sc_write_telemetry(self, batch: list) -> None:
+        """Write telemetry events to local folder. If the folder is full, free up space for the new events.
+
+        Args:
+            batch (list): The batch of events to write.
+
+        """
+        try:
+            os.makedirs(self.sc_folder_path, exist_ok=True)
+            for event in batch:
+                message = event.get("message")
+                if message is not None:
+                    file_path = (
+                        self.sc_folder_path
+                        / f'{datetime.datetime.now().strftime("%Y-%m-%d_%H-%M-%S-%f")}'
+                        f'_telemetry_{message.get("type")}.json'
+                    )
+                    json_content = self._sc_validate_folder_space(event)
+                    with open(file_path, "w") as json_file:
+                        json_file.write(json_content)
+        except Exception:
+            pass
+
+    def _sc_validate_folder_space(self, event: dict) -> str:
+        """Validate and manage folder space for the new telemetry events.
+
+        Args:
+            event (dict): The event to validate.
+
+        Returns:
+            str: The JSON content of the event.
+
+        """
+        json_content = json.dumps(event, indent=4, sort_keys=True)
+        new_json_file_size = len(json_content.encode("utf-8"))
+        telemetry_folder = self.sc_folder_path
+        folder_size = _get_folder_size(telemetry_folder)
+        if folder_size + new_json_file_size > self.sc_memory_limit:
+            _free_up_space(telemetry_folder, self.sc_memory_limit - new_json_file_size)
+        return json_content
+
+    def _sc_upload_local_telemetry(self) -> None:
+        """Send a request to the API to upload the local telemetry events."""
+        if not self.sc_is_enabled or self.sc_is_testing or not self._rest:
+            return
+        batch = []
+        for file in self.sc_folder_path.glob("*.json"):
+            with open(file) as json_file:
+                data_dict = json.load(json_file)
+                batch.append(data_dict)
+        if batch == []:
+            return
+        body = {"logs": batch}
+        ret = self._rest.request(
+            self.sc_sf_path_telemetry,
+            body=body,
+            method="post",
+            client=None,
+            timeout=5,
+        )
+        if ret.get("success"):
+            for file in self.sc_folder_path.glob("*.json"):
+                file.unlink()
+
+    def _sc_is_telemetry_testing(self) -> bool:
+        is_testing = os.getenv("SNOWPARK_CHECKPOINTS_TELEMETRY_TESTING") == "true"
+        if is_testing:
+            local_telemetry_path = (
+                Path(os.getcwd()) / "snowpark-checkpoints-output" / "telemetry"
+            )
+            self.set_sc_output_path(local_telemetry_path)
+            self.sc_is_enabled = True
+        return is_testing
+
+    def _sc_is_telemetry_manager(self) -> bool:
+        """Check if the class is telemetry manager.
+
+        Returns:
+            bool: True if the class is telemetry manager, False otherwise.
+
+        """
+        return True
+
+    def sc_is_hypothesis_event_logged(self, event_name: tuple[str, int]) -> bool:
+        """Check if a Hypothesis event is logged.
+
+        Args:
+            event_name (tuple[str, int]): A tuple containing the name of the event and an integer identifier
+            (0 for info, 1 for error).
+
+        Returns:
+            bool: True if the event is logged, False otherwise.
+
+        """
+        return event_name in self.sc_hypothesis_input_events
+
+    def _sc_close(self) -> None:
+        """Close the telemetry manager and upload collected events.
+
+        This function closes the telemetry manager, uploads any collected events,
+        and performs any necessary cleanup to ensure no data is lost.
+        """
+        atexit.unregister(self._sc_close_at_exit)
+        if self.sc_log_batch and self.sc_is_enabled and not self.sc_is_testing:
+            self.sc_send_batch(self.sc_log_batch)
+
+    def _sc_close_at_exit(self) -> None:
+        """Close the telemetry manager at exit and upload collected events.
+
+        This function ensures that the telemetry manager is closed and all collected events
+        are uploaded when the program exits, preventing data loss.
+        """
+        with suppress(Exception):
+            self._sc_close()
+
+
+def _generate_event(
+    event_name: str,
+    event_type: str,
+    parameters_info: Optional[dict] = None,
+    sc_version: Optional[str] = None,
+) -> dict:
+    """Generate a telemetry event.
+
+    Args:
+        event_name (str): The name of the event.
+        event_type (str): The type of the event (e.g., "error", "info").
+        parameters_info (dict, optional): Additional parameters for the event. Defaults to None.
+        sc_version (str, optional): The version of the package. Defaults to None.
+
+    Returns:
+        dict: The generated event.
+
+    """
+    metadata = _get_metadata()
+    if sc_version is not None:
+        metadata["snowpark_checkpoints_version"] = sc_version
+    message = {
+        "event_type": event_type,
+        "type": "snowpark-checkpoints",
+        "event_name": event_name,
+        "driver_type": "PythonConnector",
+        "driver_version": SNOWFLAKE_CONNECTOR_VERSION,
+        "metadata": metadata,
+        "data": json.dumps(parameters_info or {}),
+    }
+    timestamp = time_util.get_time_millis()
+    event_base = {"message": message, "timestamp": str(timestamp)}
+
+    return event_base
+
+
+def _get_metadata() -> dict:
+    """Get metadata for telemetry events.
+
+    Returns:
+        dict: The metadata including OS version, Python version, and device ID.
+
+    """
+    return {
+        "os_version": platform,
+        "python_version": python_version(),
+        "snowpark_version": ".".join(str(x) for x in SNOWPARK_VERSION if x is not None),
+        "device_id": _get_unique_id(),
+    }
+
+
+def _get_version() -> str:
+    """Get the version of the package.
+
+    Returns:
+        str: The version of the package.
+
+    """
+    try:
+        directory_levels_up = 1
+        project_root = Path(__file__).resolve().parents[directory_levels_up]
+        version_file_path = project_root / VERSION_FILE_NAME
+        with open(version_file_path) as file:
+            content = file.read()
+            version_match = re.search(VERSION_VARIABLE_PATTERN, content, re.MULTILINE)
+            if version_match:
+                return version_match.group(1)
+        return None
+    except Exception:
+        return None
+
+
+def _get_folder_size(folder_path: Path) -> int:
+    """Get the size of a folder. Only considers JSON files.
+
+    Args:
+        folder_path (Path): The path to the folder.
+
+    Returns:
+        int: The size of the folder in bytes.
+
+    """
+    return sum(f.stat().st_size for f in folder_path.glob("*.json") if f.is_file())
+
+
+def _free_up_space(folder_path: Path, max_size: int) -> None:
+    """Free up space in a folder by deleting the oldest files. Only considers JSON files.
+
+    Args:
+        folder_path (Path): The path to the folder.
+        max_size (int): The maximum allowed size of the folder in bytes.
+
+    """
+    files = sorted(folder_path.glob("*.json"), key=lambda f: f.stat().st_mtime)
+    current_size = _get_folder_size(folder_path)
+    for file in files:
+        if current_size <= max_size:
+            break
+        current_size -= file.stat().st_size
+        file.unlink()
+
+
+def _get_unique_id() -> str:
+    """Get a unique device ID. The ID is generated based on the hashed MAC address.
+
+    Returns:
+        str: The hashed device ID.
+
+    """
+    node_id_str = str(getnode())
+    hashed_id = hashlib.sha256(node_id_str.encode()).hexdigest()
+    return hashed_id
+
+
+def get_telemetry_manager() -> TelemetryManager:
+    """Get the telemetry manager.
+
+    Returns:
+        TelemetryManager: The telemetry manager.
+
+    """
+    try:
+        connection = Session.builder.getOrCreate().connection
+        if not hasattr(connection._telemetry, "_sc_is_telemetry_manager"):
+            connection._telemetry = TelemetryManager(
+                connection._rest, connection.telemetry_enabled
+            )
+        return connection._telemetry
+    except Exception:
+        telemetry_manager = TelemetryManager(None, is_telemetry_enabled=True)
+        telemetry_manager.sc_flush_size = 1
+        return telemetry_manager
+
+
+def get_snowflake_schema_types(df: snowpark_dataframe.DataFrame) -> list[str]:
+    """Extract the data types of the schema fields from a Snowflake DataFrame.
+
+    Args:
+        df (snowpark_dataframe.DataFrame): The Snowflake DataFrame.
+
+    Returns:
+        list[str]: A list of data type names of the schema fields.
+
+    """
+    return [str(schema_type.datatype) for schema_type in df.schema.fields]
+
+
+def _is_snowpark_dataframe(df: Any) -> bool:
+    """Check if the given dataframe is a Snowpark dataframe.
+
+    Args:
+        df: The dataframe to check.
+
+    Returns:
+        bool: True if the dataframe is a Snowpark dataframe, False otherwise.
+
+    """
+    return isinstance(df, snowpark_dataframe.DataFrame)
+
+
+def get_load_json(json_schema: str) -> dict:
+    """Load and parse a JSON schema file.
+
+    Args:
+        json_schema (str): The path to the JSON schema file.
+
+    Returns:
+        dict: The parsed JSON content.
+
+    Raises:
+        ValueError: If there is an error reading or parsing the JSON file.
+
+    """
+    try:
+        with open(json_schema, encoding="utf-8") as file:
+            return json.load(file)
+    except (OSError, json.JSONDecodeError) as e:
+        raise ValueError(f"Error reading JSON schema file: {e}") from None
+
+
+def extract_parameters(
+    func: Callable, args: tuple, kwargs: dict, params_list: Optional[list[str]]
+) -> dict:
+    """Extract parameters from the function arguments.
+
+    Args:
+        func (Callable): The function being decorated.
+        args (tuple): The positional arguments passed to the function.
+        kwargs (dict): The keyword arguments passed to the function.
+        params_list (list[str]): The list of parameters to extract.
+
+    Returns:
+        dict: A dictionary of extracted parameters.
+
+    """
+    parameters = inspect.signature(func).parameters
+    param_data = {}
+    if params_list:
+        for _, param in enumerate(params_list):
+            if len(args) > 0:
+                index = list(parameters.keys()).index(param)
+                param_data[param] = args[index]
+            else:
+                if kwargs[param]:
+                    param_data[param] = kwargs[param]
+    return param_data
+
+
+def check_dataframe_schema_event(
+    telemetry_data: dict, param_data: dict
+) -> tuple[str, dict]:
+    """Handle telemetry event for checking dataframe schema.
+
+    Args:
+        telemetry_data (dict): The telemetry data dictionary.
+        param_data (dict): The parameter data dictionary.
+
+    Returns:
+        tuple: A tuple containing the event name and telemetry data.
+
+    """
+    telemetry_data[MODE_KEY] = CheckpointMode.SCHEMA.value
+    try:
+        telemetry_data[STATUS_KEY] = param_data.get(STATUS_KEY)
+        pandera_schema = param_data.get(PANDERA_SCHEMA_PARAM_NAME)
+        schema_types = []
+        for schema_type in pandera_schema.columns.values():
+            if schema_type.dtype is not None:
+                schema_types.append(str(schema_type.dtype))
+        if schema_types:
+            telemetry_data[SCHEMA_TYPES_KEY] = schema_types
+        return DATAFRAME_VALIDATOR_SCHEMA, telemetry_data
+    except Exception:
+        if param_data.get(STATUS_KEY):
+            telemetry_data[STATUS_KEY] = param_data.get(STATUS_KEY)
+        pandera_schema = param_data.get(PANDERA_SCHEMA_PARAM_NAME)
+        if pandera_schema:
+            schema_types = []
+            for schema_type in pandera_schema.columns.values():
+                if schema_type.dtype is not None:
+                    schema_types.append(str(schema_type.dtype))
+            if schema_types:
+                telemetry_data[SCHEMA_TYPES_KEY] = schema_types
+        return DATAFRAME_VALIDATOR_ERROR, telemetry_data
+
+
+def check_output_or_input_schema_event(
+    telemetry_data: dict, param_data: dict
+) -> tuple[str, dict]:
+    """Handle telemetry event for checking output or input schema.
+
+    Args:
+        telemetry_data (dict): The telemetry data dictionary.
+        param_data (dict): The parameter data dictionary.
+
+    Returns:
+        tuple: A tuple containing the event name and telemetry data.
+
+    """
+    try:
+        pandera_schema = param_data.get(PANDERA_SCHEMA_PARAM_NAME)
+        schema_types = []
+        for schema_type in pandera_schema.columns.values():
+            if schema_type.dtype is not None:
+                schema_types.append(str(schema_type.dtype))
+        if schema_types:
+            telemetry_data[SCHEMA_TYPES_KEY] = schema_types
+        return DATAFRAME_VALIDATOR_SCHEMA, telemetry_data
+    except Exception:
+        return DATAFRAME_VALIDATOR_ERROR, telemetry_data
+
+
+def collect_dataframe_checkpoint_mode_schema_event(
+    telemetry_data: dict, param_data: dict
+) -> tuple[str, dict]:
+    """Handle telemetry event for collecting dataframe checkpoint mode schema.
+
+    Args:
+        telemetry_data (dict): The telemetry data dictionary.
+        param_data (dict): The parameter data dictionary.
+
+    Returns:
+        tuple: A tuple containing the event name and telemetry data.
+
+    """
+    telemetry_data[MODE_KEY] = CheckpointMode.SCHEMA.value
+    try:
+        schema_types = param_data.get("column_type_dict")
+        telemetry_data[SCHEMA_TYPES_KEY] = [
+            schema_types[schema_type].dataType.typeName()
+            for schema_type in schema_types
+        ]
+        return DATAFRAME_COLLECTION_SCHEMA, telemetry_data
+    except Exception:
+        return DATAFRAME_COLLECTION_ERROR, telemetry_data
+
+
+def collect_dataframe_checkpoint_mode_dataframe_event(
+    telemetry_data: dict, param_data: dict
+) -> tuple[str, dict]:
+    """Handle telemetry event for collecting dataframe checkpoint mode dataframe.
+
+    This function processes telemetry data for a dataframe checkpoint mode event. It updates the telemetry data
+    with the mode and schema types of the Spark DataFrame being collected.
+
+    Args:
+        telemetry_data (dict): The telemetry data dictionary to be updated.
+        param_data (dict): The parameter data dictionary containing the DataFrame information.
+
+    Returns:
+        tuple: A tuple containing the event name and the updated telemetry data dictionary.
+
+    """
+    telemetry_data[MODE_KEY] = CheckpointMode.DATAFRAME.value
+    try:
+        if _is_spark_dataframe(param_data.get(DF_PARAM_NAME)):
+            telemetry_data[SPARK_SCHEMA_TYPES_KEY] = _get_spark_schema_types(
+                param_data.get(DF_PARAM_NAME)
+            )
+            return DATAFRAME_COLLECTION_DF, telemetry_data
+    except Exception:
+        return DATAFRAME_COLLECTION_ERROR, telemetry_data
+
+
+def assert_return_event(telemetry_data: dict, param_data: dict) -> tuple[str, dict]:
+    """Handle telemetry event for asserting return values.
+
+    Args:
+        telemetry_data (dict): The telemetry data dictionary.
+        param_data (dict): The parameter data dictionary.
+
+    Returns:
+        tuple: A tuple containing the event name and telemetry data.
+
+    """
+    if param_data.get(STATUS_KEY) is not None:
+        telemetry_data[STATUS_KEY] = param_data.get(STATUS_KEY, None)
+    try:
+        if _is_snowpark_dataframe(
+            param_data.get(SNOWPARK_RESULTS_PARAM_NAME)
+        ) and _is_spark_dataframe(param_data.get(SPARK_RESULTS_PARAM_NAME)):
+            telemetry_data[SNOWFLAKE_SCHEMA_TYPES_KEY] = get_snowflake_schema_types(
+                param_data.get(SNOWPARK_RESULTS_PARAM_NAME)
+            )
+            telemetry_data[SPARK_SCHEMA_TYPES_KEY] = _get_spark_schema_types(
+                param_data.get(SPARK_RESULTS_PARAM_NAME)
+            )
+            return DATAFRAME_VALIDATOR_MIRROR, telemetry_data
+        else:
+            return VALUE_VALIDATOR_MIRROR, telemetry_data
+    except Exception:
+        if _is_snowpark_dataframe(param_data.get(SNOWPARK_RESULTS_PARAM_NAME)):
+            telemetry_data[SNOWFLAKE_SCHEMA_TYPES_KEY] = get_snowflake_schema_types(
+                param_data.get(SNOWPARK_RESULTS_PARAM_NAME)
+            )
+        if _is_spark_dataframe(param_data.get(SPARK_RESULTS_PARAM_NAME)):
+            telemetry_data[SPARK_SCHEMA_TYPES_KEY] = _get_spark_schema_types(
+                param_data.get(SPARK_RESULTS_PARAM_NAME)
+            )
+        return DATAFRAME_VALIDATOR_ERROR, telemetry_data
+
+
+def dataframe_strategy_event(
+    telemetry_data: dict, param_data: dict, telemetry_m: TelemetryManager
+) -> tuple[Optional[str], Optional[dict]]:
+    """Handle telemetry event for dataframe strategy.
+
+    Args:
+        telemetry_data (dict): The telemetry data dictionary.
+        param_data (dict): The parameter data dictionary.
+        telemetry_m (TelemetryManager): The telemetry manager.
+
+    Returns:
+        tuple: A tuple containing the event name and telemetry data.
+
+    """
+    try:
+        test_function_name = inspect.stack()[2].function
+        is_logged = telemetry_m.sc_is_hypothesis_event_logged((test_function_name, 0))
+        if not is_logged:
+            schema_param = param_data.get(DATAFRAME_STRATEGY_SCHEMA_PARAM_NAME)
+            if isinstance(schema_param, str):
+                json_data = get_load_json(schema_param)["custom_data"]["columns"]
+                telemetry_data[SCHEMA_TYPES_KEY] = [
+                    column["type"] for column in json_data
+                ]
+            else:
+                schema_types = []
+                for schema_type in schema_param.columns.values():
+                    if schema_type.dtype is not None:
+                        schema_types.append(str(schema_type.dtype))
+                if schema_types:
+                    telemetry_data[SCHEMA_TYPES_KEY] = schema_types
+            telemetry_m.sc_hypothesis_input_events.append((test_function_name, 0))
+            if None in telemetry_data[SCHEMA_TYPES_KEY]:
+                telemetry_m.sc_log_error(HYPOTHESIS_INPUT_SCHEMA_ERROR, telemetry_data)
+            else:
+                telemetry_m.sc_log_info(HYPOTHESIS_INPUT_SCHEMA, telemetry_data)
+            telemetry_m.sc_send_batch(telemetry_m.sc_log_batch)
+        return None, None
+    except Exception:
+        test_function_name = inspect.stack()[2].function
+        is_logged = telemetry_m.sc_is_hypothesis_event_logged((test_function_name, 1))
+        if not is_logged:
+            telemetry_m.sc_hypothesis_input_events.append((test_function_name, 0))
+            telemetry_m.sc_log_error(HYPOTHESIS_INPUT_SCHEMA_ERROR, telemetry_data)
+            telemetry_m.sc_send_batch(telemetry_m.sc_log_batch)
+        return None, None
+
+
+def compare_data_event(telemetry_data: dict, param_data: dict) -> tuple[str, dict]:
+    """Handle telemetry event for comparing data.
+
+    This function processes telemetry data for a data comparison event. It updates the telemetry data
+    with the mode, status, and schema types of the Snowflake DataFrame being compared.
+
+    Args:
+        telemetry_data (dict): The telemetry data dictionary to be updated.
+        param_data (dict): The parameter data dictionary containing the DataFrame and status information.
+
+    Returns:
+        tuple: A tuple containing the event name and the updated telemetry data dictionary.
+
+    """
+    telemetry_data[MODE_KEY] = CheckpointMode.DATAFRAME.value
+    telemetry_data[STATUS_KEY] = param_data.get(STATUS_KEY, None)
+    try:
+        telemetry_data[SCHEMA_TYPES_KEY] = get_snowflake_schema_types(
+            param_data.get("df")
+        )
+        return DATAFRAME_VALIDATOR_DF, telemetry_data
+    except Exception:
+        return DATAFRAME_VALIDATOR_ERROR, telemetry_data
+
+
+def handle_result(
+    func_name: str,
+    result: Any,
+    param_data: dict,
+    multiple_return: bool,
+    telemetry_m: TelemetryManager,
+    return_indexes: Optional[list[tuple[str, int]]] = None,
+) -> tuple[Optional[str], Optional[dict]]:
+    """Handle the result of the function and collect telemetry data.
+
+    Args:
+        func_name (str): The name of the function.
+        result: The result of the function.
+        param_data (dict): The extracted parameters.
+        multiple_return (bool): Whether the function returns multiple values.
+        telemetry_m (TelemetryManager): The telemetry manager.
+        return_indexes (list[tuple[str, int]]): The list of return values to report. Defaults to None.
+
+    Returns:
+        tuple: A tuple containing the event name (str) and telemetry data (dict).
+
+    """
+    if result is not None and return_indexes is not None:
+        if multiple_return:
+            for name, index in return_indexes:
+                param_data[name] = result[index]
+        else:
+            param_data[return_indexes[0][0]] = result[return_indexes[0][1]]
+
+    telemetry_data = {
+        FUNCTION_KEY: func_name,
+    }
+
+    telemetry_event = None
+    data = None
+    if func_name == "_check_dataframe_schema":
+        telemetry_event, data = check_dataframe_schema_event(telemetry_data, param_data)
+    elif func_name in ["check_output_schema", "check_input_schema"]:
+        telemetry_event, data = check_output_or_input_schema_event(
+            telemetry_data, param_data
+        )
+    if func_name == "_compare_data":
+        telemetry_event, data = compare_data_event(telemetry_data, param_data)
+    elif func_name == "_collect_dataframe_checkpoint_mode_schema":
+        telemetry_event, data = collect_dataframe_checkpoint_mode_schema_event(
+            telemetry_data, param_data
+        )
+    elif func_name == "_collect_dataframe_checkpoint_mode_dataframe":
+        telemetry_event, data = collect_dataframe_checkpoint_mode_dataframe_event(
+            telemetry_data, param_data
+        )
+    elif func_name == "_assert_return":
+        telemetry_event, data = assert_return_event(telemetry_data, param_data)
+    elif func_name == "dataframe_strategy":
+        telemetry_event, data = dataframe_strategy_event(
+            telemetry_data, param_data, telemetry_m
+        )
+    return telemetry_event, data
+
+
+fn = TypeVar("fn", bound=Callable)
+
+
+def report_telemetry(
+    params_list: list[str] = None,
+    return_indexes: list[tuple[str, int]] = None,
+    multiple_return: bool = False,
+) -> Callable[[fn], fn]:
+    """Report telemetry events for a function.
+
+    Args:
+        params_list (list[str], optional): The list of parameters to report. Defaults to None.
+        return_indexes (list[tuple[str, int]], optional): The list of return values to report. Defaults to None.
+        multiple_return (bool, optional): Whether the function returns multiple values. Defaults to False.
+
+    Returns:
+        Callable[[fn], fn]: The decorator function.
+
+    """
+
+    def report_telemetry_decorator(func):
+        func_name = func.__name__
+
+        @wraps(func)
+        def wrapper(*args, **kwargs):
+            func_exception = None
+            result = None
+            try:
+                result = func(*args, **kwargs)
+            except Exception as err:
+                func_exception = err
+
+            if os.getenv("SNOWPARK_CHECKPOINTS_TELEMETRY_ENABLED") == "false":
+                return result
+            telemetry_event = None
+            data = None
+            telemetry_m = None
+            try:
+                param_data = extract_parameters(func, args, kwargs, params_list)
+                telemetry_m = get_telemetry_manager()
+                telemetry_event, data = handle_result(
+                    func_name,
+                    result,
+                    param_data,
+                    multiple_return,
+                    telemetry_m,
+                    return_indexes,
+                )
+            except Exception:
+                pass
+            finally:
+                if func_exception is not None:
+                    if telemetry_m is not None:
+                        telemetry_m.sc_log_error(telemetry_event, data)
+                    raise func_exception
+                if telemetry_m is not None:
+                    telemetry_m.sc_log_info(telemetry_event, data)
+
+            return result
+
+        return wrapper
+
+    return report_telemetry_decorator
+
+
+# Constants for telemetry
+DATAFRAME_COLLECTION_SCHEMA = "DataFrame_Collection_Schema"
+DATAFRAME_COLLECTION_DF = "DataFrame_Collection_DF"
+DATAFRAME_VALIDATOR_MIRROR = "DataFrame_Validator_Mirror"
+VALUE_VALIDATOR_MIRROR = "Value_Validator_Mirror"
+DATAFRAME_VALIDATOR_SCHEMA = "DataFrame_Validator_Schema"
+DATAFRAME_VALIDATOR_DF = "DataFrame_Validator_DF"
+HYPOTHESIS_INPUT_SCHEMA = "Hypothesis_Input_Schema"
+DATAFRAME_COLLECTION_ERROR = "DataFrame_Collection_Error"
+DATAFRAME_VALIDATOR_ERROR = "DataFrame_Validator_Error"
+HYPOTHESIS_INPUT_SCHEMA_ERROR = "Hypothesis_Input_Schema_Error"
+
+FUNCTION_KEY = "function"
+STATUS_KEY = "status"
+SCHEMA_TYPES_KEY = "schema_types"
+ERROR_KEY = "error"
+MODE_KEY = "mode"
+SNOWFLAKE_SCHEMA_TYPES_KEY = "snowflake_schema_types"
+SPARK_SCHEMA_TYPES_KEY = "spark_schema_types"
+
+DATAFRAME_STRATEGY_SCHEMA_PARAM_NAME = "schema"
+PANDERA_SCHEMA_PARAM_NAME = "pandera_schema"
+SNOWPARK_RESULTS_PARAM_NAME = "snowpark_results"
+SPARK_RESULTS_PARAM_NAME = "spark_results"
+DF_PARAM_NAME = "df"
+
+
+class CheckpointMode(IntEnum):
+    SCHEMA = 1
+    DATAFRAME = 2