Bump opentelemetry to latest version (#66)

Author: sfc-gh-bnandibhatla

NOTICE.txt

@@ -6,7 +6,7 @@ The following notices are required by licensors of software used in the Snowflak
 --------------------------------------------------------------------------------
 
 This library includes software which is copied from or derived from the OpenTelemetry Python API and SDK.
-OpenTelemetry Python v1.26.0
+OpenTelemetry Python v1.35.0
 https://github.com/open-telemetry/opentelemetry-python
 
                                  Apache License

anaconda/meta.yaml

@@ -14,8 +14,8 @@ requirements:
         - setuptools >=40.0.0
     run:
         - python
-        - opentelemetry-api ==1.26.0
-        - opentelemetry-sdk ==1.26.0
+        - opentelemetry-api ==1.35.0
+        - opentelemetry-sdk ==1.35.0
 
 about:
     home: https://www.snowflake.com/

scripts/vendor_otlp_proto_common.sh

@@ -9,7 +9,7 @@
 # fixes needed in the OTLP exporter.
 
 # Pinned commit/branch/tag for the current version used in opentelemetry-proto python package.
-REPO_BRANCH_OR_COMMIT="v1.26.0"
+REPO_BRANCH_OR_COMMIT="v1.35.0"
 
 set -e
 

setup.py

@@ -30,8 +30,8 @@
     long_description=LONG_DESCRIPTION,
     python_requires=REQUIRED_PYTHON_VERSION,
     install_requires=[
-        "opentelemetry-api == 1.26.0",
-        "opentelemetry-sdk == 1.26.0",
+        "opentelemetry-api == 1.35.0",
+        "opentelemetry-sdk == 1.35.0",
     ],
     packages=find_namespace_packages(
         where='src'

src/snowflake/telemetry/_internal/exporter/otlp/proto/logs/__init__.py

@@ -15,11 +15,7 @@
 
 import abc
 import logging
-import logging.config
-import threading
 import typing
-import opentelemetry.sdk.util.instrumentation as otel_instrumentation
-import opentelemetry.sdk._logs._internal as _logs_internal
 
 from snowflake.telemetry._internal.opentelemetry.exporter.otlp.proto.common._log_encoder import (
     encode_logs,
@@ -84,29 +80,40 @@ class SnowflakeLoggingHandler(_logs.LoggingHandler):
     discarded by the original implementation.
     """
 
-    LOGGER_NAME_TEMP_ATTRIBUTE = "__snow.logging.temp.logger_name"
+    CODE_FILEPATH: typing.Final = "code.filepath"
+    CODE_FILE_PATH: typing.Final = "code.file.path"
+    CODE_FUNCTION: typing.Final = "code.function"
+    CODE_FUNCTION_NAME: typing.Final = "code.function.name"
+    CODE_LINENO: typing.Final = "code.lineno"
+    CODE_LINE_NUMBER: typing.Final = "code.line.number"
 
     def __init__(
             self,
             log_writer: LogWriter,
         ):
         exporter = _ProtoLogExporter(log_writer)
-        provider = _SnowflakeTelemetryLoggerProvider()
-        provider.add_log_record_processor(
-            export.SimpleLogRecordProcessor(exporter)
+        processor = export.SimpleLogRecordProcessor(exporter)
+        provider = _logs.LoggerProvider(
+            resource=Resource.get_empty(),
+            multi_log_record_processor=processor
         )
         super().__init__(logger_provider=provider)
 
     @staticmethod
     def _get_attributes(record: logging.LogRecord) -> types.Attributes:
         attributes = _logs.LoggingHandler._get_attributes(record) # pylint: disable=protected-access
 
-        # Temporarily storing logger's name in record's attributes.
-        # This attribute will be removed by the logger.
-        #
-        # TODO (SNOW-1235374): opentelemetry-python issue #2485: Record logger
-        # name as the instrumentation scope name
-        attributes[SnowflakeLoggingHandler.LOGGER_NAME_TEMP_ATTRIBUTE] = record.name
+        # Preserving old naming conventions for code attributes that were changed as part of
+        # https://github.com/open-telemetry/opentelemetry-python/commit/1b1e8d80c764ad3aa76abfb56a7002ddea11fdb5 in
+        # order to avoid a behavior change for Snowflake customers.
+        if SnowflakeLoggingHandler.CODE_FILE_PATH in attributes:
+            attributes[SnowflakeLoggingHandler.CODE_FILEPATH] = attributes.pop(SnowflakeLoggingHandler.CODE_FILE_PATH)
+        if SnowflakeLoggingHandler.CODE_FUNCTION_NAME in attributes:
+            attributes[SnowflakeLoggingHandler.CODE_FUNCTION] = attributes.pop(
+                SnowflakeLoggingHandler.CODE_FUNCTION_NAME)
+        if SnowflakeLoggingHandler.CODE_LINE_NUMBER in attributes:
+            attributes[SnowflakeLoggingHandler.CODE_LINENO] = attributes.pop(SnowflakeLoggingHandler.CODE_LINE_NUMBER)
+
         return attributes
 
     def _translate(self, record: logging.LogRecord) -> _logs.LogRecord:
@@ -115,75 +122,6 @@ def _translate(self, record: logging.LogRecord) -> _logs.LogRecord:
         return otel_record
 
 
-class _SnowflakeTelemetryLogger(_logs.Logger):
-    """
-    An Open Telemetry Logger which creates an InstrumentationScope for each
-    logger name it encounters.
-    """
-
-    def __init__(
-        self,
-        resource: Resource,
-        multi_log_record_processor: typing.Union[
-            _logs_internal.SynchronousMultiLogRecordProcessor,
-            _logs_internal.ConcurrentMultiLogRecordProcessor,
-        ],
-        instrumentation_scope: otel_instrumentation.InstrumentationScope,
-    ):
-        super().__init__(resource, multi_log_record_processor, instrumentation_scope)
-        self._lock = threading.Lock()
-        self.cached_scopes = {}
-
-    def emit(self, record: _logs.LogRecord):
-        if SnowflakeLoggingHandler.LOGGER_NAME_TEMP_ATTRIBUTE not in record.attributes:
-            # The record doesn't contain our custom attribute with a logger name,
-            # so we can call the superclass's `emit` method. It will emit a log
-            # record with the default instrumentation scope.
-            super().emit(record)
-            return
-
-        # Creating an InstrumentationScope for each logger name,
-        # and caching those scopes.
-        logger_name = record.attributes[SnowflakeLoggingHandler.LOGGER_NAME_TEMP_ATTRIBUTE]
-        del record.attributes[SnowflakeLoggingHandler.LOGGER_NAME_TEMP_ATTRIBUTE]
-        with self._lock:
-            if logger_name in self.cached_scopes:
-                current_scope = self.cached_scopes[logger_name]
-            else:
-                current_scope = otel_instrumentation.InstrumentationScope(logger_name)
-                self.cached_scopes[logger_name] = current_scope
-
-        # Emitting a record with a scope that corresponds to the logger
-        # that logged it. NOT calling the superclass here for two reasons:
-        #  1. Logger.emit takes a LogRecord, not LogData.
-        #  2. It would emit a log record with the default instrumentation scope,
-        #     not with the scope we want.
-        log_data = _logs.LogData(record, current_scope)
-        self._multi_log_record_processor.emit(log_data)
-
-
-class _SnowflakeTelemetryLoggerProvider(_logs.LoggerProvider):
-    """
-    A LoggerProvider that creates SnowflakeTelemetryLoggers
-    """
-
-    def get_logger(
-            self, name: str,
-            version: types.Optional[str] = None,
-            schema_url: types.Optional[str] = None,
-            attributes: types.Optional[types.Attributes] = None,
-    ) -> _logs.Logger:
-        return _SnowflakeTelemetryLogger(
-            Resource.get_empty(),
-            self._multi_log_record_processor,
-            otel_instrumentation.InstrumentationScope(
-                name,
-                version,
-                schema_url,
-            ),
-        )
-
-
 __all__ = [
     "LogWriter",
     "SnowflakeLoggingHandler",

src/snowflake/telemetry/_internal/opentelemetry/exporter/otlp/proto/common/__init__.py

@@ -14,7 +14,7 @@
 #
 # This file has been modified from the original source code at
 #
-#     https://github.com/open-telemetry/opentelemetry-python/tree/v1.26.0
+#     https://github.com/open-telemetry/opentelemetry-python/tree/v1.35.0
 #
 # by Snowflake Inc.
 

src/snowflake/telemetry/_internal/opentelemetry/exporter/otlp/proto/common/_internal/__init__.py

@@ -14,42 +14,42 @@
 #
 # This file has been modified from the original source code at
 #
-#     https://github.com/open-telemetry/opentelemetry-python/tree/v1.26.0
+#     https://github.com/open-telemetry/opentelemetry-python/tree/v1.35.0
 #
 # by Snowflake Inc.
 
 
+from __future__ import annotations
+
 import logging
 from collections.abc import Sequence
-from itertools import count
 from typing import (
     Any,
+    Callable,
+    Dict,
+    List,
     Mapping,
     Optional,
-    List,
-    Callable,
     TypeVar,
-    Dict,
-    Iterator,
 )
 
-from opentelemetry.sdk.util.instrumentation import InstrumentationScope
+from snowflake.telemetry._internal.opentelemetry.proto.common.v1.common_marshaler import AnyValue as PB2AnyValue
 from snowflake.telemetry._internal.opentelemetry.proto.common.v1.common_marshaler import (
-    InstrumentationScope as PB2InstrumentationScope,
+    ArrayValue as PB2ArrayValue,
 )
-from snowflake.telemetry._internal.opentelemetry.proto.resource.v1.resource_marshaler import (
-    Resource as PB2Resource,
+from snowflake.telemetry._internal.opentelemetry.proto.common.v1.common_marshaler import (
+    InstrumentationScope as PB2InstrumentationScope,
 )
-from snowflake.telemetry._internal.opentelemetry.proto.common.v1.common_marshaler import AnyValue as PB2AnyValue
 from snowflake.telemetry._internal.opentelemetry.proto.common.v1.common_marshaler import KeyValue as PB2KeyValue
 from snowflake.telemetry._internal.opentelemetry.proto.common.v1.common_marshaler import (
     KeyValueList as PB2KeyValueList,
 )
-from snowflake.telemetry._internal.opentelemetry.proto.common.v1.common_marshaler import (
-    ArrayValue as PB2ArrayValue,
+from snowflake.telemetry._internal.opentelemetry.proto.resource.v1.resource_marshaler import (
+    Resource as PB2Resource,
 )
 from opentelemetry.sdk.trace import Resource
-from opentelemetry.util.types import Attributes
+from opentelemetry.sdk.util.instrumentation import InstrumentationScope
+from opentelemetry.util.types import _ExtendedAttributes
 
 _logger = logging.getLogger(__name__)
 
@@ -65,14 +65,19 @@ def _encode_instrumentation_scope(
     return PB2InstrumentationScope(
         name=instrumentation_scope.name,
         version=instrumentation_scope.version,
+        attributes=_encode_attributes(instrumentation_scope.attributes),
     )
 
 
 def _encode_resource(resource: Resource) -> PB2Resource:
     return PB2Resource(attributes=_encode_attributes(resource.attributes))
 
 
-def _encode_value(value: Any) -> PB2AnyValue:
+def _encode_value(
+    value: Any, allow_null: bool = False
+) -> Optional[PB2AnyValue]:
+    if allow_null is True and value is None:
+        return None
     if isinstance(value, bool):
         return PB2AnyValue(bool_value=value)
     if isinstance(value, str):
@@ -81,21 +86,49 @@ def _encode_value(value: Any) -> PB2AnyValue:
         return PB2AnyValue(int_value=value)
     if isinstance(value, float):
         return PB2AnyValue(double_value=value)
+    if isinstance(value, bytes):
+        return PB2AnyValue(bytes_value=value)
     if isinstance(value, Sequence):
         return PB2AnyValue(
-            array_value=PB2ArrayValue(values=[_encode_value(v) for v in value])
+            array_value=PB2ArrayValue(
+                values=_encode_array(value, allow_null=allow_null)
+            )
         )
     elif isinstance(value, Mapping):
         return PB2AnyValue(
             kvlist_value=PB2KeyValueList(
-                values=[_encode_key_value(str(k), v) for k, v in value.items()]
+                values=[
+                    _encode_key_value(str(k), v, allow_null=allow_null)
+                    for k, v in value.items()
+                ]
             )
         )
     raise Exception(f"Invalid type {type(value)} of value {value}")
 
 
-def _encode_key_value(key: str, value: Any) -> PB2KeyValue:
-    return PB2KeyValue(key=key, value=_encode_value(value))
+def _encode_key_value(
+    key: str, value: Any, allow_null: bool = False
+) -> PB2KeyValue:
+    return PB2KeyValue(
+        key=key, value=_encode_value(value, allow_null=allow_null)
+    )
+
+
+def _encode_array(
+    array: Sequence[Any], allow_null: bool = False
+) -> Sequence[PB2AnyValue]:
+    if not allow_null:
+        # Let the exception get raised by _encode_value()
+        return [_encode_value(v, allow_null=allow_null) for v in array]
+
+    return [
+        _encode_value(v, allow_null=allow_null)
+        if v is not None
+        # Use an empty AnyValue to represent None in an array. Behavior may change pending
+        # https://github.com/open-telemetry/opentelemetry-specification/issues/4392
+        else PB2AnyValue()
+        for v in array
+    ]
 
 
 def _encode_span_id(span_id: int) -> bytes:
@@ -107,14 +140,17 @@ def _encode_trace_id(trace_id: int) -> bytes:
 
 
 def _encode_attributes(
-    attributes: Attributes,
+    attributes: _ExtendedAttributes,
+    allow_null: bool = False,
 ) -> Optional[List[PB2KeyValue]]:
     if attributes:
         pb2_attributes = []
         for key, value in attributes.items():
             # pylint: disable=broad-exception-caught
             try:
-                pb2_attributes.append(_encode_key_value(key, value))
+                pb2_attributes.append(
+                    _encode_key_value(key, value, allow_null=allow_null)
+                )
             except Exception as error:
                 _logger.exception("Failed to encode key %s: %s", key, error)
     else:
@@ -145,38 +181,3 @@ def _get_resource_data(
             )
         )
     return resource_data
-
-
-def _create_exp_backoff_generator(max_value: int = 0) -> Iterator[int]:
-    """
-    Generates an infinite sequence of exponential backoff values. The sequence starts
-    from 1 (2^0) and doubles each time (2^1, 2^2, 2^3, ...). If a max_value is specified
-    and non-zero, the generated values will not exceed this maximum, capping at max_value
-    instead of growing indefinitely.
-
-    Parameters:
-    - max_value (int, optional): The maximum value to yield. If 0 or not provided, the
-      sequence grows without bound.
-
-    Returns:
-    Iterator[int]: An iterator that yields the exponential backoff values, either uncapped or
-    capped at max_value.
-
-    Example:
-    ```
-    gen = _create_exp_backoff_generator(max_value=10)
-    for _ in range(5):
-        print(next(gen))
-    ```
-    This will print:
-    1
-    2
-    4
-    8
-    10
-
-    Note: this functionality used to be handled by the 'backoff' package.
-    """
-    for i in count(0):
-        out = 2**i
-        yield min(out, max_value) if max_value else out