Minor: Add extended glossary (#14918)

* Add Nested glossary to extended sample data

* Fix lint

* Add Stress test config

* Fix Lint

ingestion/pipelines/extended_sample_data.yaml

@@ -9,6 +9,8 @@ source:
       connectionOptions:
         sampleDataFolder: "./examples/sample_data"
         extendedSampleDataFolder: "./examples/extended_sample_data"
+        includeGlossary: True
+        includeLineageStressTesting: True
   sourceConfig: {}
 sink:
   type: metadata-rest

ingestion/src/metadata/ingestion/source/database/extended_sample_data.py

@@ -24,6 +24,10 @@ from metadata.generated.schema.api.data.createDatabase import CreateDatabaseRequ
 from metadata.generated.schema.api.data.createDatabaseSchema import (
     CreateDatabaseSchemaRequest,
 )
+from metadata.generated.schema.api.data.createGlossary import CreateGlossaryRequest
+from metadata.generated.schema.api.data.createGlossaryTerm import (
+    CreateGlossaryTermRequest,
+)
 from metadata.generated.schema.api.data.createTable import CreateTableRequest
 from metadata.generated.schema.api.lineage.addLineage import AddLineageRequest
 from metadata.generated.schema.entity.data.dashboard import Dashboard
@@ -91,10 +95,20 @@ class ExtendedSampleDataSource(Source):  # pylint: disable=too-many-instance-att
         self.store_table_fqn = set()
         self.store_data_model_fqn = []
         self.store_dashboard_fqn = []
+        self.main_glossary = None
+        self.glossary_term_list = []
 
         sample_data_folder = self.service_connection.connectionOptions.__root__.get(
             "sampleDataFolder"
         )
+        self.include_glossary = self.service_connection.connectionOptions.__root__.get(
+            "includeGlossary"
+        )
+        self.include_lineage_stress_testing = (
+            self.service_connection.connectionOptions.__root__.get(
+                "includeLineageStressTesting"
+            )
+        )
         extneded_sample_data_folder = (
             self.service_connection.connectionOptions.__root__.get(
                 "extendedSampleDataFolder"
@@ -225,13 +239,20 @@ class ExtendedSampleDataSource(Source):  # pylint: disable=too-many-instance-att
             )
         )
 
-    def generate_sample_data(self):  # pylint: disable=too-many-locals
+    def generate_sample_data(
+        self,
+    ):  # pylint: disable=too-many-locals,too-many-statements
         """
         Generate sample data for dashboard and database service,
         with lineage between them, having long names, special characters and description
         """
-
+        if self.include_glossary:
-        db = self.create_database_request("extended_sample_data", self.generate_text())
+            yield from self.create_glossary()
+            yield from self.create_glossary_term()
+        if self.include_lineage_stress_testing:
+            db = self.create_database_request(
+                "extended_sample_data", self.generate_text()
+            )
             yield Either(right=db)
 
             schema = self.create_database_schema_request(
@@ -313,9 +334,13 @@ class ExtendedSampleDataSource(Source):  # pylint: disable=too-many-instance-att
                     schema_name=schema.name.__root__,
                     table_name=table_request.name.__root__,
                 )
-            from_table = self.metadata.get_by_name(entity=Table, fqn=table_entity_fqn)
+                from_table = self.metadata.get_by_name(
+                    entity=Table, fqn=table_entity_fqn
+                )
 
-            yield from self.create_depth_nodes(from_table=from_table, to_table=to_table)
+                yield from self.create_depth_nodes(
+                    from_table=from_table, to_table=to_table
+                )
 
             self.dashboard_service_json["name"] = name
             self.dashboard_service_json["description"] = text
@@ -404,6 +429,49 @@ class ExtendedSampleDataSource(Source):  # pylint: disable=too-many-instance-att
                             )
                         )
 
+    def create_glossary(self):
+        self.main_glossary = CreateGlossaryRequest(
+            name="NestedGlossaryTest",
+            displayName="NestedGlossaryTest",
+            description="Description of test glossary",
+        )
+        yield Either(right=self.main_glossary)
+
+    def create_glossary_term(self):
+        """
+        Create Glossary Terms
+        """
+        for _ in range(20):
+            random_name = self.fake.first_name()
+            yield Either(
+                right=CreateGlossaryTermRequest(
+                    glossary="NestedGlossaryTest",
+                    name=random_name,
+                    displayName=random_name,
+                    description="Test glossary term ",
+                    parent=self.glossary_term_list[-1]
+                    if len(self.glossary_term_list) > 3
+                    else None,
+                )
+            )
+            if len(self.glossary_term_list) > 3 and self.glossary_term_list[-1]:
+                self.glossary_term_list.append(
+                    f"{self.glossary_term_list[-1]}.{random_name}"
+                )
+            else:
+                self.glossary_term_list.append(f"NestedGlossaryTest.{random_name}")
+
+        for _ in range(500):
+            random_name = self.fake.first_name()
+            yield Either(
+                right=CreateGlossaryTermRequest(
+                    glossary="NestedGlossaryTest",
+                    name=random_name,
+                    displayName=random_name,
+                    description="Test glossary term 1",
+                )
+            )
+
     def generate_name(self):
         return f"Sample-@!3_(%t3st@)%_^{self.fake.name()}"