fix(ingestion/lookml): emit dummy sql condition for lookml custom con…

…dition tag (#11008)

Co-authored-by: Harshal Sheth <[email protected]>

metadata-ingestion/src/datahub/ingestion/source/looker/looker_liquid_tag.py

@@ -1,5 +1,5 @@
 from functools import lru_cache
-from typing import ClassVar, Optional, TextIO, cast
+from typing import ClassVar, Optional, TextIO
 
 from liquid import Environment
 from liquid.ast import Node
@@ -25,18 +25,9 @@ def __init__(self, tok: Token, sql_or_lookml_reference: str, filter_name: str):
         self.filter_name = filter_name
 
     def render_to_output(self, context: Context, buffer: TextIO) -> Optional[bool]:
-        filter_value: Optional[str] = cast(
-            str, context.globals.get(self.filter_name)
-        )  # to silent lint
-
-        if filter_value is None:
-            raise CustomTagException(
-                f'filter {self.filter_name} value is not provided for "condition" tag'
-            )
-
-        filter_value = filter_value.strip()
-
-        buffer.write(f"{self.sql_or_lookml_reference}='{filter_value}'")
+        # This implementation will make sure that sql parse work correctly if looker condition tag
+        # is used in lookml sql field
+        buffer.write(f"{self.sql_or_lookml_reference}='dummy_value'")
 
         return True
 

metadata-ingestion/src/datahub/ingestion/source/looker/looker_template_language.py

@@ -9,6 +9,7 @@
     CustomTagException,
     create_template,
 )
+from datahub.ingestion.source.looker.lookml_config import DERIVED_VIEW_PATTERN
 from datahub.ingestion.source.looker.str_functions import (
     remove_extra_spaces_and_newlines,
 )
@@ -94,6 +95,24 @@ def resolve_liquid_variable(text: str, liquid_variable: Dict[Any, Any]) -> str:
     return text
 
 
+def _complete_incomplete_sql(raw_view: dict, sql: str) -> str:
+
+    # Looker supports sql fragments that omit the SELECT and FROM parts of the query
+    # Add those in if we detect that it is missing
+    sql_query: str = sql
+
+    if not re.search(r"SELECT\s", sql_query, flags=re.I):
+        # add a SELECT clause at the beginning
+        sql_query = f"SELECT {sql}"
+
+    if not re.search(r"FROM\s", sql_query, flags=re.I):
+        # add a FROM clause at the end
+        sql_query = f"{sql_query} FROM {raw_view['name']}"
+
+    # Drop ${ and }
+    return re.sub(DERIVED_VIEW_PATTERN, r"\1", sql_query)
+
+
 def resolve_liquid_variable_in_view_dict(
     raw_view: dict, liquid_variable: Dict[Any, Any]
 ) -> None:
@@ -102,14 +121,18 @@ def resolve_liquid_variable_in_view_dict(
 
     for view in raw_view["views"]:
         if "sql_table_name" in view:
-            view["sql_table_name"] = resolve_liquid_variable(
+            view["datahub_transformed_sql_table_name"] = resolve_liquid_variable(
                 text=remove_extra_spaces_and_newlines(view["sql_table_name"]),
                 liquid_variable=liquid_variable,
-            )
+            )  # keeping original sql_table_name as is to avoid any visualization issue later
 
         if "derived_table" in view and "sql" in view["derived_table"]:
             # In sql we don't need to remove the extra spaces as sql parser takes care of extra spaces and \n
             # while generating URN from sql
-            view["derived_table"]["sql"] = resolve_liquid_variable(
+            view["derived_table"]["datahub_transformed_sql"] = resolve_liquid_variable(
                 text=view["derived_table"]["sql"], liquid_variable=liquid_variable
+            )  # keeping original sql as is, so that on UI sql will be shown same is it is visible on looker portal
+
+            view["derived_table"]["datahub_transformed_sql"] = _complete_incomplete_sql(
+                raw_view=view, sql=view["derived_table"]["datahub_transformed_sql"]
             )

metadata-ingestion/src/datahub/ingestion/source/looker/lookml_concept_context.py

@@ -266,15 +266,25 @@ def sql_table_name(self) -> str:
         sql_table_name: Optional[str] = self._get_sql_table_name_field()
         # if sql_table_name field is not set then the table name is equal to view-name
         if sql_table_name is None:
-            return self.raw_view[NAME].lower()
+            sql_table_name = self.raw_view[NAME].lower()
+
+        return sql_table_name
+
+    def datahub_transformed_sql_table_name(self) -> str:
+        table_name: Optional[str] = self.raw_view.get(
+            "datahub_transformed_sql_table_name"
+        )
+
+        if not table_name:
+            table_name = self.sql_table_name()
 
         # sql_table_name is in the format "${view-name}.SQL_TABLE_NAME"
         # remove extra characters
         if self._is_dot_sql_table_name_present():
-            sql_table_name = re.sub(DERIVED_VIEW_PATTERN, r"\1", sql_table_name)
+            table_name = re.sub(DERIVED_VIEW_PATTERN, r"\1", table_name)
 
         # Some sql_table_name fields contain quotes like: optimizely."group", just remove the quotes
-        return sql_table_name.replace('"', "").replace("`", "").lower()
+        return table_name.replace('"', "").replace("`", "").lower()
 
     def derived_table(self) -> Dict[Any, Any]:
         """
@@ -296,30 +306,21 @@ def explore_source(self) -> Dict[Any, Any]:
 
         return derived_table["explore_source"]
 
-    def sql(self, transformed: bool = True) -> str:
+    def sql(self) -> str:
         """
         This function should only be called if is_sql_based_derived_case return true
         """
         derived_table = self.derived_table()
 
-        # Looker supports sql fragments that omit the SELECT and FROM parts of the query
-        # Add those in if we detect that it is missing
-        sql_query: str = derived_table["sql"]
-
-        if transformed:  # update the original sql attribute only if transformed is true
-            if not re.search(r"SELECT\s", sql_query, flags=re.I):
-                # add a SELECT clause at the beginning
-                sql_query = f"SELECT {sql_query}"
+        return derived_table["sql"]
 
-            if not re.search(r"FROM\s", sql_query, flags=re.I):
-                # add a FROM clause at the end
-                sql_query = f"{sql_query} FROM {self.name()}"
-                # Get the list of tables in the query
-
-            # Drop ${ and }
-            sql_query = re.sub(DERIVED_VIEW_PATTERN, r"\1", sql_query)
+    def datahub_transformed_sql(self) -> str:
+        """
+        This function should only be called if is_sql_based_derived_case return true
+        """
+        derived_table = self.derived_table()
 
-        return sql_query
+        return derived_table["datahub_transformed_sql"]
 
     def name(self) -> str:
         return self.raw_view[NAME]

metadata-ingestion/src/datahub/ingestion/source/looker/lookml_config.py

@@ -11,8 +11,10 @@
 from datahub.configuration.git import GitInfo
 from datahub.configuration.source_common import EnvConfigMixin
 from datahub.configuration.validate_field_rename import pydantic_renamed_field
-from datahub.ingestion.source.looker.looker_config import LookerCommonConfig
-from datahub.ingestion.source.looker.looker_connection import LookerConnectionDefinition
+from datahub.ingestion.source.looker.looker_config import (
+    LookerCommonConfig,
+    LookerConnectionDefinition,
+)
 from datahub.ingestion.source.looker.looker_lib_wrapper import (
     LookerAPI,
     LookerAPIConfig,

metadata-ingestion/src/datahub/ingestion/source/looker/lookml_source.py

@@ -201,8 +201,7 @@ def from_looker_dict(
         view_logic = view_context.view_file.raw_file_content[:max_file_snippet_length]
 
         if view_context.is_sql_based_derived_case():
-            view_logic = view_context.sql(transformed=False)
-            # Parse SQL to extract dependencies.
+            view_logic = view_context.sql()
             view_details = ViewProperties(
                 materialized=False,
                 viewLogic=view_logic,

metadata-ingestion/src/datahub/ingestion/source/looker/view_upstream.py

@@ -206,18 +206,21 @@ class AbstractViewUpstream(ABC):
     view_context: LookerViewContext
     looker_view_id_cache: LookerViewIdCache
     config: LookMLSourceConfig
+    reporter: LookMLSourceReport
     ctx: PipelineContext
 
     def __init__(
         self,
         view_context: LookerViewContext,
         looker_view_id_cache: LookerViewIdCache,
         config: LookMLSourceConfig,
+        reporter: LookMLSourceReport,
         ctx: PipelineContext,
     ):
         self.view_context = view_context
         self.looker_view_id_cache = looker_view_id_cache
         self.config = config
+        self.reporter = reporter
         self.ctx = ctx
 
     @abstractmethod
@@ -244,9 +247,10 @@ def __init__(
         view_context: LookerViewContext,
         looker_view_id_cache: LookerViewIdCache,
         config: LookMLSourceConfig,
+        reporter: LookMLSourceReport,
         ctx: PipelineContext,
     ):
-        super().__init__(view_context, looker_view_id_cache, config, ctx)
+        super().__init__(view_context, looker_view_id_cache, config, reporter, ctx)
         # These are the function where we need to catch the response once calculated
         self._get_spr = lru_cache(maxsize=1)(self.__get_spr)
         self._get_upstream_dataset_urn = lru_cache(maxsize=1)(
@@ -259,25 +263,14 @@ def __get_spr(self) -> Optional[SqlParsingResult]:
             return None
 
         spr = create_lineage_sql_parsed_result(
-            query=self.view_context.sql(),
+            query=self.view_context.datahub_transformed_sql(),
             default_schema=self.view_context.view_connection.default_schema,
             default_db=self.view_context.view_connection.default_db,
             platform=self.view_context.view_connection.platform,
             platform_instance=self.view_context.view_connection.platform_instance,
             env=self.view_context.view_connection.platform_env or self.config.env,
             graph=self.ctx.graph,
         )
-
-        if (
-            spr.debug_info.table_error is not None
-            or spr.debug_info.column_error is not None
-        ):
-            logging.debug(
-                f"Failed to parsed the sql query. table_error={spr.debug_info.table_error} and "
-                f"column_error={spr.debug_info.column_error}"
-            )
-            return None
-
         return spr
 
     def __get_upstream_dataset_urn(self) -> List[Urn]:
@@ -286,6 +279,15 @@ def __get_upstream_dataset_urn(self) -> List[Urn]:
         if sql_parsing_result is None:
             return []
 
+        if sql_parsing_result.debug_info.table_error is not None:
+            self.reporter.report_warning(
+                title="Table Level Lineage Missing",
+                message="Error in parsing derived sql",
+                context=f"View-name: {self.view_context.name()}",
+                exc=sql_parsing_result.debug_info.table_error,
+            )
+            return []
+
         upstream_dataset_urns: List[str] = [
             _drop_hive_dot(urn) for urn in sql_parsing_result.in_tables
         ]
@@ -306,6 +308,15 @@ def create_fields(self) -> List[ViewField]:
         if spr is None:
             return []
 
+        if spr.debug_info.column_error is not None:
+            self.reporter.report_warning(
+                title="Column Level Lineage Missing",
+                message="Error in parsing derived sql for CLL",
+                context=f"View-name: {self.view_context.name()}",
+                exc=spr.debug_info.column_error,
+            )
+            return []
+
         fields: List[ViewField] = []
 
         column_lineages: List[ColumnLineageInfo] = (
@@ -336,6 +347,15 @@ def get_upstream_column_ref(
         if sql_parsing_result is None:
             return []
 
+        if sql_parsing_result.debug_info.column_error is not None:
+            self.reporter.report_warning(
+                title="Column Level Lineage Missing",
+                message="Error in parsing derived sql for CLL",
+                context=f"View-name: {self.view_context.name()}. "
+                f"Error: {sql_parsing_result.debug_info.column_error}",
+            )
+            return []
+
         upstreams_column_refs: List[ColumnRef] = []
         if sql_parsing_result.column_lineage:
             for cll in sql_parsing_result.column_lineage:
@@ -384,9 +404,11 @@ def __init__(
         view_context: LookerViewContext,
         looker_view_id_cache: LookerViewIdCache,
         config: LookMLSourceConfig,
+        reporter: LookMLSourceReport,
         ctx: PipelineContext,
     ):
-        super().__init__(view_context, looker_view_id_cache, config, ctx)
+        super().__init__(view_context, looker_view_id_cache, config, reporter, ctx)
+
         self._get_upstream_dataset_urn = lru_cache(maxsize=1)(
             self.__get_upstream_dataset_urn
         )
@@ -402,7 +424,7 @@ def __get_upstream_dataset_urn(self) -> List[str]:
             base_folder_path=self.view_context.base_folder_path,
         )
 
-        # Current view will always be present in cache. The assert  will silence the lint
+        # Current view will always be present in cache. assert  will silence the lint
         assert current_view_id
 
         # We're creating a "LookerExplore" just to use the urn generator.
@@ -467,9 +489,10 @@ def __init__(
         view_context: LookerViewContext,
         looker_view_id_cache: LookerViewIdCache,
         config: LookMLSourceConfig,
+        reporter: LookMLSourceReport,
         ctx: PipelineContext,
     ):
-        super().__init__(view_context, looker_view_id_cache, config, ctx)
+        super().__init__(view_context, looker_view_id_cache, config, reporter, ctx)
         self.upstream_dataset_urn = None
 
         self._get_upstream_dataset_urn = lru_cache(maxsize=1)(
@@ -478,9 +501,9 @@ def __init__(
 
     def __get_upstream_dataset_urn(self) -> Urn:
         # In regular case view's upstream dataset is either same as view-name or mentioned in "sql_table_name" field
-        # view_context.sql_table_name() handle this condition to return dataset name
+        # view_context.datahub_transformed_sql_table_name() handle this condition to return dataset name
         qualified_table_name: str = _generate_fully_qualified_name(
-            sql_table_name=self.view_context.sql_table_name(),
+            sql_table_name=self.view_context.datahub_transformed_sql_table_name(),
             connection_def=self.view_context.view_connection,
             reporter=self.view_context.reporter,
         )
@@ -522,20 +545,21 @@ def __init__(
         view_context: LookerViewContext,
         looker_view_id_cache: LookerViewIdCache,
         config: LookMLSourceConfig,
+        reporter: LookMLSourceReport,
         ctx: PipelineContext,
     ):
-        super().__init__(view_context, looker_view_id_cache, config, ctx)
+        super().__init__(view_context, looker_view_id_cache, config, reporter, ctx)
         self.upstream_dataset_urn = []
 
         self._get_upstream_dataset_urn = lru_cache(maxsize=1)(
             self.__get_upstream_dataset_urn
         )
 
     def __get_upstream_dataset_urn(self) -> List[Urn]:
-        # In this case view_context.sql_table_name() refers to derived view name
+        # In this case view_context.datahub_transformed_sql_table_name() refers to derived view name
         looker_view_id = get_derived_looker_view_id(
             qualified_table_name=_generate_fully_qualified_name(
-                self.view_context.sql_table_name(),
+                self.view_context.datahub_transformed_sql_table_name(),
                 self.view_context.view_connection,
                 self.view_context.reporter,
             ),
@@ -591,6 +615,7 @@ def create_view_upstream(
         return RegularViewUpstream(
             view_context=view_context,
             config=config,
+            reporter=reporter,
             ctx=ctx,
             looker_view_id_cache=looker_view_id_cache,
         )
@@ -599,6 +624,7 @@ def create_view_upstream(
         return DotSqlTableNameViewUpstream(
             view_context=view_context,
             config=config,
+            reporter=reporter,
             ctx=ctx,
             looker_view_id_cache=looker_view_id_cache,
         )
@@ -610,6 +636,7 @@ def create_view_upstream(
         return SqlBasedDerivedViewUpstream(
             view_context=view_context,
             config=config,
+            reporter=reporter,
             ctx=ctx,
             looker_view_id_cache=looker_view_id_cache,
         )
@@ -618,6 +645,7 @@ def create_view_upstream(
         return NativeDerivedViewUpstream(
             view_context=view_context,
             config=config,
+            reporter=reporter,
             ctx=ctx,
             looker_view_id_cache=looker_view_id_cache,
         )
@@ -631,6 +659,7 @@ def create_view_upstream(
     return EmptyImplementation(
         view_context=view_context,
         config=config,
+        reporter=reporter,
         ctx=ctx,
         looker_view_id_cache=looker_view_id_cache,
     )