warpzone-sdk 15.1.1.dev1__tar.gz → 15.1.2.dev1__tar.gz

{warpzone_sdk-15.1.1.dev1 → warpzone_sdk-15.1.2.dev1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: warpzone-sdk
-Version: 15.1.1.dev1
+Version: 15.1.2.dev1
 Summary: The main objective of this package is to centralize logic used to interact with Azure Functions, Azure Service Bus and Azure Table Storage
 Author: Team Enigma
 Author-email: enigma@energinet.dk
@@ -21,7 +21,7 @@ Requires-Dist: azure-servicebus (>=7.8.0)
 Requires-Dist: azure-storage-blob (>=12.14.1)
 Requires-Dist: cryptography (==43.0.3)
 Requires-Dist: datamazing (>=5.1.6)
-Requires-Dist: deltalake (==1.2.1)
+Requires-Dist: deltalake (>=1.4.0)
 Requires-Dist: numpy (>=1.26.4)
 Requires-Dist: obstore (>=0.8.2)
 Requires-Dist: opentelemetry-sdk (>=1.32.0)

{warpzone_sdk-15.1.1.dev1 → warpzone_sdk-15.1.2.dev1}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "warpzone-sdk"
-version = "15.1.1.dev1"
+version = "15.1.2.dev1"
 description = "The main objective of this package is to centralize logic used to interact with Azure Functions, Azure Service Bus and Azure Table Storage"
 authors = [{ name = "Team Enigma", email = "enigma@energinet.dk" }]
 requires-python = ">=3.10"
@@ -14,7 +14,7 @@ requires-poetry = ">=2.2"
 pandas = ">=2.0.3"
 polars = ">=1.33.1"
 obstore = ">=0.8.2"
-deltalake = "==1.2.1" # pin to avoid breaking changes in 1.3.0. Follow Github issue here https://github.com/delta-io/delta-rs/issues/3939
+deltalake = ">=1.4.0"
 pyarrow = ">=19.0.0"
 datamazing = ">=5.1.6"
 azure-core = ">=1.26.3"

{warpzone_sdk-15.1.1.dev1 → warpzone_sdk-15.1.2.dev1}/warpzone/db/client.py

@@ -195,12 +195,10 @@ class WarpzoneDatabaseClient:
             include_validity_period_columns (bool, optional): Whether to include
             validity period columns in the result;
             (`valid_from_time_utc`, `valid_to_time_utc`).
-            Defaults to False. If set to True while using `columns`-argument, make sure
-            to include these columns in the `columns`-list.
+            Defaults to False.
             include_generated_columns (bool, optional): Whether to include generated
             columns in the result; (e.g. `valid_from_time_utc`, `valid_to_time_utc`).
-            Defaults to False. If set to True while using `columns`-argument, make sure
-            to include these columns in the `columns`-list.
+            Defaults to False.
 
         Returns:
             pd.DataFrame: The result of the query.
@@ -252,10 +250,12 @@ class WarpzoneDatabaseClient:
             for field in table.schema().fields:
                 if field.generated_as is not None:
                     generated_cols.append(field.column_name)
-            pd_df = pd_df.drop(columns=generated_cols)
+            pd_df = pd_df.drop(columns=generated_cols, errors="ignore")
 
         # Drop valid-from/to columns
         if not include_validity_period_columns:
-            pd_df = pd_df.drop(columns=["valid_from_time_utc", "valid_to_time_utc"])
+            pd_df = pd_df.drop(
+                columns=["valid_from_time_utc", "valid_to_time_utc"], errors="ignore"
+            )
 
         return pd_df

warpzone_sdk-15.1.2.dev1/warpzone/monitor/traces.py

@@ -0,0 +1,128 @@
+import logging
+import os
+import threading
+from contextlib import contextmanager
+from logging import StreamHandler
+
+from azure.monitor.opentelemetry.exporter import AzureMonitorTraceExporter
+from opentelemetry import context, trace
+from opentelemetry.sdk.resources import SERVICE_NAME, Resource
+from opentelemetry.sdk.trace import TracerProvider
+from opentelemetry.sdk.trace.export import BatchSpanProcessor
+from opentelemetry.sdk.trace.sampling import ALWAYS_ON
+from opentelemetry.trace.propagation.tracecontext import TraceContextTextMapPropagator
+
+logger = logging.getLogger(__name__)
+logger.addHandler(StreamHandler())
+
+tracer = trace.get_tracer(__name__)
+
+_TRACING_LOCK = threading.Lock()
+TRACING_IS_CONFIGURED = False
+
+
+def configure_tracing():
+    global TRACING_IS_CONFIGURED
+    # Add thread locking to avoid race conditions during setup
+    with _TRACING_LOCK:
+        if TRACING_IS_CONFIGURED:
+            # tracing should only be set up once
+            # to avoid duplicated trace handling.
+            # Global variables is the pattern used
+            # by opentelemetry, so we use the same
+            return
+
+        # set up tracer provider based on the Azure Function resource
+        # (this is make sure App Insights can track the trace source correctly)
+        # (https://learn.microsoft.com/en-us/azure/azure-monitor/app/opentelemetry-enable?tabs=net#set-the-cloud-role-name-and-the-cloud-role-instance).
+        # We use the ALWAYS ON sampler since otherwise spans will not be
+        # recording upon creation
+        # (https://anecdotes.dev/opentelemetry-on-google-cloud-unraveling-the-mystery-f61f044c18be)
+        service_name = os.getenv("WEBSITE_SITE_NAME") or "unknown-service"
+        resource = Resource.create({SERVICE_NAME: service_name})
+        trace.set_tracer_provider(
+            TracerProvider(
+                sampler=ALWAYS_ON,
+                resource=resource,
+            )
+        )
+
+        # setup azure monitor trace exporter to send telemetry to App Insights
+        try:
+            trace_exporter = AzureMonitorTraceExporter()
+        except ValueError:
+            logger.warning(
+                "Cant set up tracing to App Insights,"
+                " as no connection string is set."
+            )
+        else:
+            span_processor = BatchSpanProcessor(trace_exporter)
+            trace.get_tracer_provider().add_span_processor(span_processor)
+
+        TRACING_IS_CONFIGURED = True
+
+
+@contextmanager
+def set_trace_context(trace_parent: str, trace_state: str = ""):
+    """Context manager for setting the trace context
+
+    Args:
+        trace_parent (str): Trace parent ID
+        trace_state (str, optional): Trace state. Defaults to "".
+    """
+    carrier = {"traceparent": trace_parent, "tracestate": trace_state}
+    ctx = TraceContextTextMapPropagator().extract(carrier=carrier)
+
+    token = context.attach(ctx)  # attach context before run
+    try:
+        yield
+    finally:
+        context.detach(token)  # detach context after run
+
+
+def get_tracer(name: str):
+    tracer = trace.get_tracer(name)
+    return tracer
+
+
+def get_current_diagnostic_id() -> str:
+    """Gets diagnostic id from current span
+
+    The diagnostic id is a concatenation of operation-id and parent-id
+
+    Returns:
+        str: diagnostic id
+    """
+    span = trace.get_current_span()
+
+    if not span.is_recording():
+        return ""
+
+    operation_id = "{:016x}".format(span.context.trace_id)
+    parent_id = "{:016x}".format(span.context.span_id)
+
+    diagnostic_id = f"00-{operation_id}-{parent_id}-01"
+
+    return diagnostic_id
+
+
+# Service Bus trace constants (these were removed from azure-servicebus SDK)
+_SB_TRACE_NAMESPACE = "Microsoft.ServiceBus"
+
+
+@contextmanager
+def servicebus_send_span(subject: str) -> trace.Span:
+    """Start span for Service Bus message tracing.
+
+    Args:
+        subject: The message subject (used as span name for easy identification)
+
+    Yields:
+        trace.Span: the span
+    """
+    with tracer.start_as_current_span(
+        subject, kind=trace.SpanKind.PRODUCER
+    ) as msg_span:
+        msg_span.set_attributes({"az.namespace": _SB_TRACE_NAMESPACE})
+
+        yield msg_span

warpzone_sdk-15.1.1.dev1/warpzone/monitor/traces.py

@@ -1,228 +0,0 @@
-import inspect
-import logging
-import os
-import threading
-from contextlib import contextmanager
-from functools import wraps
-from logging import StreamHandler
-from typing import Callable
-
-from azure.monitor.opentelemetry.exporter import AzureMonitorTraceExporter
-from opentelemetry import context, trace
-from opentelemetry.sdk.resources import SERVICE_NAME, Resource
-from opentelemetry.sdk.trace import Tracer, TracerProvider
-from opentelemetry.sdk.trace.export import BatchSpanProcessor
-from opentelemetry.sdk.trace.sampling import ALWAYS_ON
-from opentelemetry.trace.propagation.tracecontext import TraceContextTextMapPropagator
-
-logger = logging.getLogger(__name__)
-logger.addHandler(StreamHandler())
-
-
-class WarpzoneTracer:
-    """Wrapper around OpenTelemetry tracer with additional trace decorator method"""
-
-    def __init__(self, otel_tracer: Tracer):
-        # Store the original tracer instead of calling super().__init__
-        # because we want to wrap an existing tracer, not create a new one
-        self._tracer = otel_tracer
-
-    def __getattr__(self, name):
-        """Delegate all attributes to the underlying tracer"""
-        return getattr(self._tracer, name)
-
-    def trace_function(
-        self,
-        name: str = None,
-        set_args_as_attributes: bool = False,
-        on_input: Callable = None,
-        on_output: Callable = None,
-    ):
-        """
-        Decorator to trace a function using this tracer object.
-
-        This decorator wraps functions with OpenTelemetry tracing, allowing for:
-        - Automatically create spans for function execution
-        - Customize the span name
-        - Add custom attributes to the span
-        - Add custom logic for inputs and outputs
-
-        Args:
-            name: Optional name for the span. If not provided, uses the function name.
-            set_args_as_attributes: If True, sets function arguments as attributes.
-            on_input: Optional callback called with (span, *args, **kwargs) before.
-            on_output: Optional callback called with (span, result) after.
-
-        Example:
-            # Simple tracing with function name
-            tracer = get_tracer(__name__)
-
-            @tracer.trace_function()
-            def my_function():
-                pass
-
-            # Custom tracing with input/output callbacks
-            @tracer.trace_function(
-                on_input=lambda span, new_data, existing_data, now: (
-                    span.set_attribute("new_records", len(new_data)),
-                    span.set_attribute("existing_records", len(existing_data)),
-                ),
-                on_output=lambda span, result: (
-                    span.set_attribute("merged_records", len(result))
-                ),
-            )
-            def merge_new_and_existing(new_data, existing_data, now):
-                pass
-        """
-
-        def decorator(func: Callable) -> Callable:
-            @wraps(func)
-            def wrapper(*args, **kwargs):
-                span_name = name or func.__name__
-                with self._tracer.start_as_current_span(span_name) as span:
-                    if set_args_as_attributes:
-                        # Get parameter names from function signature
-                        sig = inspect.signature(func)
-                        param_names = list(sig.parameters.keys())
-
-                        # Set positional arguments with their parameter names
-                        for i, arg in enumerate(args):
-                            if i < len(param_names):
-                                span.set_attribute(param_names[i], str(arg))
-                            else:
-                                # Fallback for *args if there are more args than params
-                                span.set_attribute(f"arg_{i}", str(arg))
-
-                        # Set keyword arguments
-                        for key, value in kwargs.items():
-                            span.set_attribute(str(key), str(value))
-
-                    # Call on_input callback if provided
-                    if on_input:
-                        on_input(span, *args, **kwargs)
-
-                    result = func(*args, **kwargs)
-
-                    # Call on_output callback if provided
-                    if on_output:
-                        on_output(span, result)
-
-                    return result
-
-            return wrapper
-
-        return decorator
-
-
-tracer = WarpzoneTracer(trace.get_tracer(__name__))
-
-
-_TRACING_LOCK = threading.Lock()
-TRACING_IS_CONFIGURED = False
-
-
-def configure_tracing():
-    global TRACING_IS_CONFIGURED
-    # Add thread locking to avoid race conditions during setup
-    with _TRACING_LOCK:
-        if TRACING_IS_CONFIGURED:
-            # tracing should only be set up once
-            # to avoid duplicated trace handling.
-            # Global variables is the pattern used
-            # by opentelemetry, so we use the same
-            return
-
-        # set up tracer provider based on the Azure Function resource
-        # (this is make sure App Insights can track the trace source correctly)
-        # (https://learn.microsoft.com/en-us/azure/azure-monitor/app/opentelemetry-enable?tabs=net#set-the-cloud-role-name-and-the-cloud-role-instance).
-        # We use the ALWAYS ON sampler since otherwise spans will not be
-        # recording upon creation
-        # (https://anecdotes.dev/opentelemetry-on-google-cloud-unraveling-the-mystery-f61f044c18be)
-        service_name = os.getenv("WEBSITE_SITE_NAME") or "unknown-service"
-        resource = Resource.create({SERVICE_NAME: service_name})
-        trace.set_tracer_provider(
-            TracerProvider(
-                sampler=ALWAYS_ON,
-                resource=resource,
-            )
-        )
-
-        # setup azure monitor trace exporter to send telemetry to App Insights
-        try:
-            trace_exporter = AzureMonitorTraceExporter()
-        except ValueError:
-            logger.warning(
-                "Cant set up tracing to App Insights,"
-                " as no connection string is set."
-            )
-        else:
-            span_processor = BatchSpanProcessor(trace_exporter)
-            trace.get_tracer_provider().add_span_processor(span_processor)
-
-        TRACING_IS_CONFIGURED = True
-
-
-@contextmanager
-def set_trace_context(trace_parent: str, trace_state: str = ""):
-    """Context manager for setting the trace context
-
-    Args:
-        trace_parent (str): Trace parent ID
-        trace_state (str, optional): Trace state. Defaults to "".
-    """
-    carrier = {"traceparent": trace_parent, "tracestate": trace_state}
-    ctx = TraceContextTextMapPropagator().extract(carrier=carrier)
-
-    token = context.attach(ctx)  # attach context before run
-    try:
-        yield
-    finally:
-        context.detach(token)  # detach context after run
-
-
-def get_tracer(name: str):
-    otel_tracer = trace.get_tracer(name)
-    return WarpzoneTracer(otel_tracer)
-
-
-def get_current_diagnostic_id() -> str:
-    """Gets diagnostic id from current span
-
-    The diagnostic id is a concatenation of operation-id and parent-id
-
-    Returns:
-        str: diagnostic id
-    """
-    span = trace.get_current_span()
-
-    if not span.is_recording():
-        return ""
-
-    operation_id = "{:016x}".format(span.context.trace_id)
-    parent_id = "{:016x}".format(span.context.span_id)
-
-    diagnostic_id = f"00-{operation_id}-{parent_id}-01"
-
-    return diagnostic_id
-
-
-# Service Bus trace constants (these were removed from azure-servicebus SDK)
-_SB_TRACE_NAMESPACE = "Microsoft.ServiceBus"
-
-
-@contextmanager
-def servicebus_send_span(subject: str):
-    """Start span for Service Bus message tracing.
-
-    Args:
-        subject: The message subject (used as span name for easy identification)
-
-    Yields:
-        Span: the span
-    """
-    with tracer.start_as_current_span(
-        subject, kind=trace.SpanKind.PRODUCER
-    ) as msg_span:
-        msg_span.set_attributes({"az.namespace": _SB_TRACE_NAMESPACE})
-
-        yield msg_span