warpzone-sdk 15.0.0.dev13__py3-none-any.whl → 15.0.1.dev1__py3-none-any.whl

warpzone/db/client.py

@@ -32,29 +32,51 @@ class WarpzoneDatabaseClient:
         self,
         path: str,
         storage_options: dict[str, str] | None = None,
+        table_prefix: str = "",
     ):
         self.store = Store(
             path=path,
             storage_options=storage_options,
         )
+        self.table_prefix = table_prefix
 
     @classmethod
-    def from_storage_account(
+    def from_resource_name(
         cls,
         storage_account: str,
         container_name: str = "datasets",
+        sub_path: str = "",
+        table_prefix: str = "",
         credential: (
             AzureNamedKeyCredential | AzureSasCredential | TokenCredential
         ) = DefaultAzureCredential(),
     ):
+        """Create a WarpzoneDatabaseClient from resource name (storage account).
+        This assumes the path of the delta lake is of the form:
+        abfss://{container_name}@{storage_account}.dfs.core.windows.net/{sub_path}
+
+        Args:
+            storage_account (str): Storage account name.
+            container_name (str, optional): Container name. Defaults to "datasets".
+            sub_path (str, optional): Sub-path within the container. Defaults to "".
+            table_prefix (str, optional): Table prefix to use (e.g. `mz_` for archive).
+            Defaults to "".
+            credential (optional): Azure credential to use.
+            Defaults to DefaultAzureCredential().
+        """
         path = f"abfss://{container_name}@{storage_account}.dfs.core.windows.net"
+        if sub_path:
+            path += f"/{sub_path}"
+
         token = credential.get_token("https://storage.azure.com/.default")
         storage_options = {
             "account_name": storage_account,
             "token": token.token,
         }
 
-        return cls(path=path, storage_options=storage_options)
+        return cls(
+            path=path, storage_options=storage_options, table_prefix=table_prefix
+        )
 
     def get_unit_and_multiple(self, timedelta: pd.Timedelta) -> tuple[str | None, int]:
         """
@@ -148,7 +170,35 @@ class WarpzoneDatabaseClient:
         time_interval: Optional[pdz.TimeInterval] = None,
         time_travel: Optional[pdz.TimeTravel] = None,
         filters: Optional[dict[str, object]] = None,
+        include_validity_period_columns: bool = False,
+        include_generated_columns: bool = False,
     ) -> pd.DataFrame:
+        """Query table.
+        Query defaults are set to match old Table Storage client behavior.
+        Time travel defaults to "as of now"
+        Validity period columns are dropped by default.
+        Generated columns are dropped by default.
+
+        Args:
+            table_name (str): Name of the table
+            time_interval (Optional[pdz.TimeInterval], optional): Time interval for the
+            query. Defaults to None.
+            time_travel (Optional[pdz.TimeTravel], optional): Time travel information.
+            Defaults to None.
+            filters (Optional[dict[str, object]], optional): Filters to apply to the
+            query.
+            Defaults to None.
+            include_validity_period_columns (bool, optional): Whether to include
+            validity period columns in the result;
+            (e.g. `valid_from_time_utc`, `valid_to_time_utc`).
+            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.
+
+        Returns:
+            pd.DataFrame: The result of the query.
+        """
         # We do 'camelCaseToSnake_case' conversion here because the old
         # naming convention used in WarpZone was CamelCase, while the new
         # naming convention is snake_case. The goal is to remove this
@@ -191,16 +241,16 @@ class WarpzoneDatabaseClient:
         for col in pd_df.select_dtypes(include=["datetime", "datetimetz"]).columns:
             pd_df[col] = pd_df[col].dt.floor("s").dt.as_unit("ns")
 
-        # We remove the valid-from and valid-to columns, as well
-        # as any generated columns, as this was not present
-        # in the old solution (Azure Table Stroage)
-        generated_cols = []
-        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=["valid_from_time_utc", "valid_to_time_utc"] + generated_cols
-        )
+        # Drop generated columns
+        if not include_generated_columns:
+            generated_cols = []
+            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)
+
+        # 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"])
 
         return pd_df

warpzone/function/integrations.py

@@ -62,7 +62,7 @@ def get_db_client() -> WarpzoneDatabaseClient:
 
 
 def get_delta_db_client() -> WarpzoneDeltaDatabaseClient:
-    db_client = WarpzoneDeltaDatabaseClient.from_storage_account(
+    db_client = WarpzoneDeltaDatabaseClient.from_resource_name(
         os.environ["OPERATIONAL_DATA_STORAGE_ACCOUNT"],
         credential=_credential,
     )

warpzone/function/monitor.py

@@ -1,5 +1,4 @@
 import asyncio
-import logging
 from contextlib import contextmanager
 from typing import Callable
 
@@ -18,16 +17,7 @@ def configure_monitoring():
     """
     Configure logging and tracing on Azure Function to
     - export telemetry to App Insights
-    - suppress spamming logs
     """
-    # disable logging for HTTP calls to avoid log spamming
-    logging.getLogger("azure.core.pipeline.policies.http_logging_policy").setLevel(
-        logging.WARNING
-    )
-
-    # disable logging for Service Bus underlying uAMQP library to avoid log spamming
-    logging.getLogger("uamqp").setLevel(logging.WARNING)
-
     # configure tracer provider
     traces.configure_tracing()
 

warpzone/monitor/logs.py

@@ -1,6 +1,7 @@
 # NOTE: OpenTelemetry logging to Azure is still in EXPERIMENTAL mode!
 import logging
 import os
+import threading
 from logging import StreamHandler
 
 from azure.monitor.opentelemetry.exporter import AzureMonitorLogExporter
@@ -27,42 +28,45 @@ _NOISY_LOGGERS = [
 for _logger_name in _NOISY_LOGGERS:
     logging.getLogger(_logger_name).setLevel(logging.WARNING)
 
+_LOGGING_LOCK = threading.Lock()
 LOGGING_IS_CONFIGURED = False
 
 
 def configure_logging():
     global LOGGING_IS_CONFIGURED
-    if LOGGING_IS_CONFIGURED:
-        # logging should only be set up once
-        # to avoid duplicated log handling.
-        # Global variables is the pattern used
-        # by opentelemetry, so we use the same
-        return
+    # Add thread locking to avoid race conditions during setup
+    with _LOGGING_LOCK:
+        if LOGGING_IS_CONFIGURED:
+            # logging should only be set up once
+            # to avoid duplicated log handling.
+            # Global variables is the pattern used
+            # by opentelemetry, so we use the same
+            return
 
-    # set up logger provider based on the Azure Function resource
-    # (this is make sure App Insights can track the log 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)
-    resource = Resource.create({SERVICE_NAME: os.getenv("WEBSITE_SITE_NAME")})
-    logs.set_logger_provider(
-        LoggerProvider(
-            resource=resource,
+        # set up logger provider based on the Azure Function resource
+        # (this is make sure App Insights can track the log 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)
+        service_name = os.getenv("WEBSITE_SITE_NAME") or "unknown-service"
+        resource = Resource.create({SERVICE_NAME: service_name})
+        logs.set_logger_provider(
+            LoggerProvider(
+                resource=resource,
+            )
         )
-    )
 
-    # setup azure monitor log exporter to send telemetry to App Insights
-    try:
-        log_exporter = AzureMonitorLogExporter()
-    except ValueError:
-        # if no App Insights instrumentation key is set (e.g. when running unit tests),
-        # the exporter creation will fail. In this case we skip it
-        logger.warning(
-            "Cant set up logging to App Insights, as no instrumentation key is set."
-        )
-    else:
-        log_record_processor = BatchLogRecordProcessor(log_exporter)
-        logs.get_logger_provider().add_log_record_processor(log_record_processor)
+        # setup azure monitor log exporter to send telemetry to App Insights
+        try:
+            log_exporter = AzureMonitorLogExporter()
+        except ValueError:
+            logger.warning(
+                "Cant set up logging to App Insights,"
+                " as no connection string is set."
+            )
+        else:
+            log_record_processor = BatchLogRecordProcessor(log_exporter)
+            logs.get_logger_provider().add_log_record_processor(log_record_processor)
 
-    LOGGING_IS_CONFIGURED = True
+        LOGGING_IS_CONFIGURED = True
 
 
 def get_logger(name: str):

warpzone/monitor/traces.py

@@ -1,5 +1,6 @@
 import logging
 import os
+import threading
 from contextlib import contextmanager
 from logging import StreamHandler
 
@@ -16,6 +17,7 @@ logger.addHandler(StreamHandler())
 
 tracer = trace.get_tracer(__name__)
 
+_TRACING_LOCK = threading.Lock()
 TRACING_IS_CONFIGURED = False
 
 
@@ -28,34 +30,43 @@ def configure_tracing():
         # 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)
-    resource = Resource.create({SERVICE_NAME: os.getenv("WEBSITE_SITE_NAME")})
-    trace.set_tracer_provider(
-        TracerProvider(
-            sampler=ALWAYS_ON,
-            resource=resource,
+    # 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:
-        # if no App Insights instrumentation key is set (e.g. when running unit tests),
-        # the exporter creation will fail. In this case we skip it
-        logger.warning(
-            "Cant set up tracing to App Insights, as no instrumentation key is set."
-        )
-    else:
-        span_processor = BatchSpanProcessor(trace_exporter)
-        trace.get_tracer_provider().add_span_processor(span_processor)
-
-    TRACING_IS_CONFIGURED = True
+        # 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

warpzone/tablestorage/db/client.py

@@ -12,13 +12,10 @@ from azure.identity import DefaultAzureCredential
 
 from warpzone.blobstorage.client import WarpzoneBlobClient
 from warpzone.healthchecks import HealthCheckResult, check_health_of
-from warpzone.monitor import traces
 from warpzone.tablestorage.db import base_client
 from warpzone.tablestorage.db.table_config import DataType, TableMetadata
 from warpzone.tablestorage.tables.client import WarpzoneTableClient
 
-tracer = traces.get_tracer(__name__)
-
 
 class WarpzoneDatabaseClient:
     """Class to interact with Azure Table Storage for database queries
@@ -160,31 +157,22 @@ class WarpzoneDatabaseClient:
         filters: Optional[dict[str, object]] = None,
         use_cache: Optional[bool] = True,
     ) -> pd.DataFrame:
-        with tracer.start_as_current_span(
-            "WarpzoneDatabaseClient.query",
-            attributes={
-                "table_name": table_name,
-                "use_cache": use_cache,
-                "has_filters": filters is not None,
-                "has_time_interval": time_interval is not None,
-            },
-        ):
-            table_metadata = self.get_table_metadata(table_name)
-
-            match table_metadata.data_type:
-                case DataType.TIME_SERIES:
-                    df = self._query_time_series(
-                        table_metadata, time_interval, filters, use_cache
+        table_metadata = self.get_table_metadata(table_name)
+
+        match table_metadata.data_type:
+            case DataType.TIME_SERIES:
+                df = self._query_time_series(
+                    table_metadata, time_interval, filters, use_cache
+                )
+            case _:
+                if time_interval:
+                    raise ValueError(
+                        f"Table {table_name} is not a time series table,"
+                        " and cannot be queried with a time interval."
                     )
-                case _:
-                    if time_interval:
-                        raise ValueError(
-                            f"Table {table_name} is not a time series table,"
-                            " and cannot be queried with a time interval."
-                        )
-                    df = self._query_generic(table_metadata, None, filters, use_cache)
-
-            return df
+                df = self._query_generic(table_metadata, None, filters, use_cache)
+
+        return df
 
     def list_tables(self):
         return self._table_client.list_tables()

{warpzone_sdk-15.0.0.dev13.dist-info → warpzone_sdk-15.0.1.dev1.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: warpzone-sdk
-Version: 15.0.0.dev13
+Version: 15.0.1.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

{warpzone_sdk-15.0.0.dev13.dist-info → warpzone_sdk-15.0.1.dev1.dist-info}/RECORD

@@ -2,7 +2,7 @@ warpzone/__init__.py,sha256=Ay7znIadokCdHHqsPlWAqhdg-zdWhXHNtxVinViYm7o,1533
 warpzone/blobstorage/__init__.py,sha256=lnc0uiaGLF0qMi_rWhCpRSFvaj0CJEiMCAl6Yqn1ZiA,21
 warpzone/blobstorage/client.py,sha256=YwDV83acoCeHS_D_ydsTYwnf56rSTy9CKpsxqeoXmBs,4638
 warpzone/db/__init__.py,sha256=lnc0uiaGLF0qMi_rWhCpRSFvaj0CJEiMCAl6Yqn1ZiA,21
-warpzone/db/client.py,sha256=_LYIZIYwaii9dyTwpqoOCYbqzzcbJNgflzSvUqeTto8,6919
+warpzone/db/client.py,sha256=DitLBzMEtcZAfBMRc7iTyFLVB8al7lD2D39CgRPu9ms,9183
 warpzone/deltastorage/__init__.py,sha256=cV8sGT2N_N5Z-E179NMW5O7q3FUDrb3j5f-yVNlNPv0,152
 warpzone/deltastorage/data_types.py,sha256=tWjLO_0ig7-tYxSHShvrd0znA7FFDnS-wBuFClUQG2U,2059
 warpzone/deltastorage/generated_columns.py,sha256=Dr_bihM7v9JKCgBXxc3JQC2P2mUGobXsReRKOl6jDO4,4765
@@ -16,8 +16,8 @@ warpzone/enums/topicenum.py,sha256=cj9F7kYzwvDeHN5jJXwgnHXTh6E64VwRNpFepNthIyk,2
 warpzone/function/__init__.py,sha256=rJOZBpWsUgjMc7YtXMJ1rLGm45KB1AhDJ_Y2ISiSISc,35
 warpzone/function/checks.py,sha256=B9YqThymf16ac_fVAYKilv20ru5v9nwXgHlbxYIaG98,1018
 warpzone/function/functionize.py,sha256=bSV0QvwKbD9Vo3a_8cc1rgV2rzTdMMvidinyXItBfvs,2128
-warpzone/function/integrations.py,sha256=sDt2BTx6a4mVc-33wTITP9XQVPustwj7rkX4urTyOqo,4018
-warpzone/function/monitor.py,sha256=szPn6jYfWcdypZVjqy3Md7SKpKoZ979PZ8px4PFIm5U,2223
+warpzone/function/integrations.py,sha256=Law-0TI_tbm8rq5XXuilDH69_0LhoqaZhDbTL498Qik,4016
+warpzone/function/monitor.py,sha256=xD13d4795a9qgGphOywFl4sOXRQjKypiL2ozQFOseqQ,1862
 warpzone/function/process.py,sha256=nbUVywM8ChfUwuaqFisgaD98aNRgeZkK4g5sbtuBdRs,2339
 warpzone/function/processors/__init__.py,sha256=DhIdSWLBcIeSO8IJdxPqGIhgwwnkDN6_Xqwy93BCLeA,46
 warpzone/function/processors/dependencies.py,sha256=m17BwdKyQEvzCPxpQZpAW5l1uYRIHWmweDz3XJskmpA,1259
@@ -28,8 +28,8 @@ warpzone/function/types.py,sha256=5m2hRrnLC3eqIlAH5-MM9_wKjMZ6lYawZtCOVStyFuY,72
 warpzone/healthchecks/__init__.py,sha256=9gc_Mt2szs8sDSwy0V4l3JZ6d9hX41xTpZCkDP2qsY4,2108
 warpzone/healthchecks/model.py,sha256=mM7DnrirLbUpBPPfi82MUPP654D0eOR2_F65TmzsPD0,1187
 warpzone/monitor/__init__.py,sha256=gXT2cxR4tlZER54zd7D49ZQBVyitLaqj13_cUoILuyM,109
-warpzone/monitor/logs.py,sha256=HVlLWv_NbJAmm4g43SlM4zYz2F_pRJrI11pjOEAnlzc,2987
-warpzone/monitor/traces.py,sha256=ARd8IW1BqCRgOXu-A_guOAOGpGvpKhJRTzjJgbAgx5Y,3994
+warpzone/monitor/logs.py,sha256=q3SUQCtG1ii0B9GkVs2l8kgZ5b5bI2qI6L0SoOW2QTY,3147
+warpzone/monitor/traces.py,sha256=aJQIFk0VWN-plchsEZcrUjSJgiq2v2aG53ZBjoqsg2M,4401
 warpzone/servicebus/data/__init__.py,sha256=lnc0uiaGLF0qMi_rWhCpRSFvaj0CJEiMCAl6Yqn1ZiA,21
 warpzone/servicebus/data/client.py,sha256=zECS3JwedhYnDk8PntYgIYpBF_uu9YN38KzpPFK7CKs,6511
 warpzone/servicebus/events/__init__.py,sha256=lnc0uiaGLF0qMi_rWhCpRSFvaj0CJEiMCAl6Yqn1ZiA,21
@@ -37,7 +37,7 @@ warpzone/servicebus/events/client.py,sha256=8v8XsF-2RwzKIi_93IzR_eR-BZTGXXHSuV4P
 warpzone/servicebus/events/triggers.py,sha256=_QuPTBbje7LrBoz0qhhgrtDZOcE6x1S9GNu-WJUQ8bY,2626
 warpzone/tablestorage/db/__init__.py,sha256=lnc0uiaGLF0qMi_rWhCpRSFvaj0CJEiMCAl6Yqn1ZiA,21
 warpzone/tablestorage/db/base_client.py,sha256=ropKO6z0UXqBl38NuGYV4VZ_ZFm4w1d84ReOLYoBKLY,2376
-warpzone/tablestorage/db/client.py,sha256=w8p3ghtpFnbDQtqqAQTFnC346lLDjBKmi3rRjNgSrz0,7053
+warpzone/tablestorage/db/client.py,sha256=-OchZI7x--Z9msNp912ggZ_sNGwNDXO-i7VGV6INSrg,6587
 warpzone/tablestorage/db/table_config.py,sha256=PC45dnr3vVMVr8ktu3GWoH8u3JKJaCRAvMCroUaC1NE,1336
 warpzone/tablestorage/tables/__init__.py,sha256=l_8wElG1oam39fmXqIXDAnCYqbKXEyX1G0cD6JCCv4s,78
 warpzone/tablestorage/tables/client.py,sha256=chvmTz5S8M1-IWYgfJHyTCwLZVsVzlchMuCYwHzQhCE,3636
@@ -52,6 +52,6 @@ warpzone/tools/copy.py,sha256=5fddotMZkXZO8avzUbGOhvs0cp8mce95pNpy0oPVjnQ,2596
 warpzone/transform/__init__.py,sha256=ruGa7tl-v4ndlWpULE1jSGU_a4_iRc3V6eyNr5xKP9E,27
 warpzone/transform/data.py,sha256=Abb8PcrgMbbNCJkkIUdtrTHdlY0OfXid387qw1nDpFY,2362
 warpzone/transform/schema.py,sha256=nbSQtDMvXkyqGKuwhuFCF0WsEDsaNyoPYpMKvbsKlv8,2423
-warpzone_sdk-15.0.0.dev13.dist-info/METADATA,sha256=diG21hT6W77E7aoY22H7XGDO4U4cyAnPp7_8GRRzZLY,7285
-warpzone_sdk-15.0.0.dev13.dist-info/WHEEL,sha256=3ny-bZhpXrU6vSQ1UPG34FoxZBp3lVcvK0LkgUz6VLk,88
-warpzone_sdk-15.0.0.dev13.dist-info/RECORD,,
+warpzone_sdk-15.0.1.dev1.dist-info/METADATA,sha256=tBjfrjx9z8oq9nfdZWwoam-XvrVe8_T8_DqvW6PMJTQ,7284
+warpzone_sdk-15.0.1.dev1.dist-info/WHEEL,sha256=3ny-bZhpXrU6vSQ1UPG34FoxZBp3lVcvK0LkgUz6VLk,88
+warpzone_sdk-15.0.1.dev1.dist-info/RECORD,,