teradataml 20.0.0.8__py3-none-any.whl

teradataml/store/feature_store/constants.py

@@ -0,0 +1,658 @@
+"""
+Copyright (c) 2024 by Teradata Corporation. All rights reserved.
+TERADATA CORPORATION CONFIDENTIAL AND TRADE SECRET
+
+Primary Owner: pradeep.garre@teradata.com
+Secondary Owner: adithya.avvaru@teradata.com
+
+This file implements constants required for Teradata Enterprise Feature Store.
+"""
+
+from teradatasqlalchemy.types import *
+from enum import Enum
+
+# Template for creating the triggers on
+# corresponding tables.
+
+# Tables for storing the data domains.
+EFS_DATA_DOMAINS="""
+    CREATE MULTISET TABLE {0}.{1}
+        (
+        name VARCHAR(255) CHARACTER SET LATIN NOT CASESPECIFIC NOT NULL,
+        created_time TIMESTAMP(6)
+        )
+    UNIQUE PRIMARY INDEX (name);
+"""
+
+
+# Tables for storing the features.
+EFS_FEATURES = """
+    CREATE MULTISET TABLE {0}.{1}
+        (
+        id INTEGER,
+        name VARCHAR(255) CHARACTER SET LATIN NOT CASESPECIFIC NOT NULL,
+        data_domain VARCHAR(255) CHARACTER SET LATIN NOT CASESPECIFIC NOT NULL,
+        column_name VARCHAR(200) CHARACTER SET LATIN NOT CASESPECIFIC,
+        description VARCHAR(1024) CHARACTER SET LATIN NOT CASESPECIFIC,
+        tags VARCHAR(2000) CHARACTER SET LATIN NOT CASESPECIFIC,
+        data_type VARCHAR(1024) CHARACTER SET LATIN NOT CASESPECIFIC,
+        feature_type VARCHAR(100) CHARACTER SET LATIN NOT CASESPECIFIC,
+        status VARCHAR(100) CHARACTER SET LATIN NOT CASESPECIFIC,
+        creation_time TIMESTAMP(6),
+        modified_time TIMESTAMP(6),
+        CONSTRAINT data_domain_fk FOREIGN KEY (data_domain) REFERENCES _efs_data_domains (name)
+        )
+    UNIQUE PRIMARY INDEX (name, data_domain)
+    UNIQUE INDEX (id);
+"""
+
+EFS_FEATURES_STAGING="""
+    CREATE MULTISET TABLE {0}.{1}
+        (
+        id INTEGER,
+        name VARCHAR(255) CHARACTER SET LATIN NOT CASESPECIFIC NOT NULL,
+        data_domain VARCHAR(255) CHARACTER SET LATIN NOT CASESPECIFIC NOT NULL,
+        column_name VARCHAR(200) CHARACTER SET LATIN NOT CASESPECIFIC,
+        description VARCHAR(1024) CHARACTER SET LATIN NOT CASESPECIFIC,
+        tags VARCHAR(2000) CHARACTER SET LATIN NOT CASESPECIFIC,
+        data_type VARCHAR(1024) CHARACTER SET LATIN NOT CASESPECIFIC,
+        feature_type VARCHAR(100) CHARACTER SET LATIN NOT CASESPECIFIC,
+        status VARCHAR(100) CHARACTER SET LATIN NOT CASESPECIFIC,
+        creation_time TIMESTAMP(6),
+        modified_time TIMESTAMP(6),
+        archived_time TIMESTAMP(6)
+        )
+    NO PRIMARY INDEX ;
+"""
+
+EFS_FEATURES_TRG="""
+    CREATE TRIGGER {0}.{1}
+    AFTER DELETE ON {0}.{2}
+    REFERENCING OLD AS DeletedRow
+    FOR EACH ROW
+        INSERT INTO {3}
+        VALUES (DeletedRow.id, DeletedRow.name, DeletedRow.data_domain, DeletedRow.column_name, DeletedRow.description, DeletedRow.tags, DeletedRow.data_type, DeletedRow.feature_type, DeletedRow.status, DeletedRow.creation_time, DeletedRow.modified_time, 
+                current_timestamp(6)
+                );
+"""
+
+EFS_GROUP_FEATURES = """
+    CREATE MULTISET TABLE {0}.{1}
+        (
+        feature_name VARCHAR(255) CHARACTER SET LATIN NOT CASESPECIFIC,
+        feature_data_domain VARCHAR(255) CHARACTER SET LATIN NOT CASESPECIFIC,
+        group_name VARCHAR(200) CHARACTER SET LATIN NOT CASESPECIFIC,
+        group_data_domain VARCHAR(255) CHARACTER SET LATIN NOT CASESPECIFIC,
+        creation_time TIMESTAMP(6),
+        modified_time TIMESTAMP(6),	
+    CONSTRAINT feature_name_fk FOREIGN KEY (feature_name, feature_data_domain) REFERENCES {0}._efs_features (name, data_domain),
+    CONSTRAINT group_name_fk FOREIGN KEY (group_name, group_data_domain) REFERENCES {0}._efs_feature_group (name, data_domain),
+    CONSTRAINT data_domain_fk1 FOREIGN KEY (feature_data_domain) REFERENCES {0}._efs_data_domains (name),
+    CONSTRAINT data_domain_fk2 FOREIGN KEY (group_data_domain) REFERENCES {0}._efs_data_domains (name)
+        )
+    UNIQUE PRIMARY INDEX (feature_name, feature_data_domain, group_name, group_data_domain);
+"""
+
+EFS_GROUP_FEATURES_STAGING = """
+    CREATE MULTISET TABLE {0}.{1}
+        (
+        feature_name VARCHAR(255),
+        feature_data_domain VARCHAR(255),
+        group_name VARCHAR(200) CHARACTER SET LATIN NOT CASESPECIFIC,
+        group_data_domain VARCHAR(200) CHARACTER SET LATIN NOT CASESPECIFIC,
+        creation_time TIMESTAMP(6),
+        modified_time TIMESTAMP(6),
+        archived_time TIMESTAMP(6)
+        )
+    NO PRIMARY INDEX ;  
+"""
+
+EFS_GROUP_FEATURES_TRG = """
+    CREATE TRIGGER {0}.{1}
+    AFTER DELETE ON {0}.{2}
+    REFERENCING OLD AS DeletedRow
+    FOR EACH ROW
+        INSERT INTO {3}
+        VALUES (DeletedRow.feature_name, DeletedRow.feature_data_domain, DeletedRow.group_name, DeletedRow.group_data_domain, DeletedRow.creation_time, DeletedRow.modified_time, 
+                current_timestamp(6)
+                );   
+"""  
+
+# Tables for Entities.
+
+EFS_ENTITY = """
+    CREATE MULTISET TABLE {0}.{1}
+        (
+        name VARCHAR(200) CHARACTER SET LATIN NOT CASESPECIFIC,
+        data_domain VARCHAR(255) CHARACTER SET LATIN NOT CASESPECIFIC,
+        description VARCHAR(200) CHARACTER SET LATIN NOT CASESPECIFIC,
+        creation_time TIMESTAMP(6),
+        modified_time TIMESTAMP(6),
+        CONSTRAINT data_domain_fk FOREIGN KEY (data_domain) REFERENCES {0}._efs_data_domains (name)
+        )
+    UNIQUE PRIMARY INDEX (name, data_domain);
+"""
+
+EFS_ENTITY_STAGING= """
+    CREATE MULTISET TABLE {0}.{1}
+        (
+        name VARCHAR(200) CHARACTER SET LATIN NOT CASESPECIFIC,
+        data_domain VARCHAR(255) CHARACTER SET LATIN NOT CASESPECIFIC,      
+        description VARCHAR(200) CHARACTER SET LATIN NOT CASESPECIFIC,
+        creation_time TIMESTAMP(6),
+        modified_time TIMESTAMP(6),
+        archived_time TIMESTAMP(6))
+    NO PRIMARY INDEX ;
+"""
+
+EFS_ENTITY_TRG = """
+    CREATE TRIGGER {0}.{1}
+    AFTER DELETE ON {0}.{2}
+    REFERENCING OLD AS DeletedRow
+    FOR EACH ROW
+        INSERT INTO {3}
+        VALUES (DeletedRow.name, DeletedRow.data_domain, DeletedRow.description, DeletedRow.creation_time, DeletedRow.modified_time, 
+                current_timestamp(6)
+                );
+"""
+
+EFS_ENTITY_XREF= """
+    CREATE MULTISET TABLE {0}.{1}
+        (
+        entity_name VARCHAR(200) CHARACTER SET LATIN NOT CASESPECIFIC,
+        data_domain VARCHAR(255) CHARACTER SET LATIN NOT CASESPECIFIC, 
+        entity_column VARCHAR(200) CHARACTER SET LATIN NOT CASESPECIFIC, 
+    CONSTRAINT entity_xref_fk FOREIGN KEY (entity_name, data_domain) REFERENCES {0}._efs_entity (name, data_domain),
+    CONSTRAINT data_domain_fk FOREIGN KEY (data_domain) REFERENCES {0}._efs_data_domains (name)
+        )
+    UNIQUE PRIMARY INDEX (entity_name, data_domain, entity_column);
+"""
+
+EFS_ENTITY_XREF_STAGING = """
+    CREATE MULTISET TABLE {0}.{1}
+        (
+        entity_name VARCHAR(200) CHARACTER SET LATIN NOT CASESPECIFIC,
+        data_domain VARCHAR(255) CHARACTER SET LATIN NOT CASESPECIFIC,
+        entity_column VARCHAR(200) CHARACTER SET LATIN NOT CASESPECIFIC,      
+        archived_time TIMESTAMP(6)
+        )
+    NO PRIMARY INDEX ;
+"""
+
+EFS_ENTITY_XREF_TRG = """
+    CREATE TRIGGER {0}.{1}
+    AFTER DELETE ON {0}.{2}
+    REFERENCING OLD AS DeletedRow
+    FOR EACH ROW
+        INSERT INTO {3}
+        VALUES (DeletedRow.entity_name, DeletedRow.data_domain, DeletedRow.entity_column, 
+                current_timestamp(6)
+                );       
+"""
+
+# Table for Data sources.
+
+EFS_DATA_SOURCE = """
+    CREATE MULTISET TABLE {0}.{1}
+        (
+        name VARCHAR(200) CHARACTER SET LATIN NOT CASESPECIFIC,
+        data_domain VARCHAR(255) CHARACTER SET LATIN NOT CASESPECIFIC,
+        description VARCHAR(1024) CHARACTER SET LATIN NOT CASESPECIFIC,
+        timestamp_column VARCHAR(50) CHARACTER SET LATIN NOT CASESPECIFIC,
+        source VARCHAR(5000) CHARACTER SET LATIN NOT CASESPECIFIC,
+        creation_time TIMESTAMP(6),
+        modified_time TIMESTAMP(6),
+        CONSTRAINT data_domain_fk FOREIGN KEY (data_domain) REFERENCES {0}._efs_data_domains (name)
+        )
+    UNIQUE PRIMARY INDEX (name, data_domain);
+"""
+
+EFS_DATA_SOURCE_STAGING = """
+    CREATE MULTISET TABLE {0}.{1}
+        (
+        name VARCHAR(200) CHARACTER SET LATIN NOT CASESPECIFIC,
+        data_domain VARCHAR(200) CHARACTER SET LATIN NOT CASESPECIFIC,
+        description VARCHAR(1024) CHARACTER SET LATIN NOT CASESPECIFIC,
+        timestamp_column VARCHAR(50) CHARACTER SET LATIN NOT CASESPECIFIC,
+        source VARCHAR(5000) CHARACTER SET LATIN NOT CASESPECIFIC,
+        creation_time TIMESTAMP(6),
+        modified_time TIMESTAMP(6),
+        archived_time TIMESTAMP(6))
+    NO PRIMARY INDEX;
+"""
+
+EFS_DATA_SOURCE_TRG = """
+    CREATE TRIGGER {0}.{1}
+    AFTER DELETE ON {0}.{2}
+    REFERENCING OLD AS DeletedRow
+    FOR EACH ROW
+        INSERT INTO {3}
+        VALUES (DeletedRow.name, DeletedRow.data_domain, DeletedRow.description, DeletedRow.timestamp_column, DeletedRow.source, DeletedRow.creation_time, DeletedRow.modified_time, 
+                current_timestamp(6)
+                );
+"""
+
+# Table for Feature groups.
+
+EFS_FEATURE_GROUP = """
+    CREATE MULTISET TABLE {0}.{1}
+        (
+        name VARCHAR(200) CHARACTER SET LATIN NOT CASESPECIFIC,
+        data_domain VARCHAR(255) CHARACTER SET LATIN NOT CASESPECIFIC,      
+        description VARCHAR(200) CHARACTER SET LATIN NOT CASESPECIFIC,
+        data_source_name VARCHAR(200) CHARACTER SET LATIN NOT CASESPECIFIC,
+        entity_name VARCHAR(200) CHARACTER SET LATIN NOT CASESPECIFIC,
+        creation_time TIMESTAMP(6),
+        modified_time TIMESTAMP(6), 
+    CONSTRAINT data_source_name_fk FOREIGN KEY (data_source_name, data_domain) REFERENCES {0}._efs_data_source (name, data_domain),
+    CONSTRAINT entity_fk FOREIGN KEY (entity_name, data_domain) REFERENCES {0}._efs_entity (name, data_domain),
+    CONSTRAINT data_domain_fk FOREIGN KEY (data_domain) REFERENCES {0}._efs_data_domains (name)
+        )
+    UNIQUE PRIMARY INDEX (name, data_domain);
+"""
+
+EFS_FEATURE_GROUP_STAGING = """
+    CREATE MULTISET TABLE {0}.{1}
+        (
+        name VARCHAR(200) CHARACTER SET LATIN NOT CASESPECIFIC,
+        data_domain VARCHAR(255) CHARACTER SET LATIN NOT CASESPECIFIC,      
+        description VARCHAR(200) CHARACTER SET LATIN NOT CASESPECIFIC,
+        data_source_name VARCHAR(200) CHARACTER SET LATIN NOT CASESPECIFIC,
+        entity_name VARCHAR(200) CHARACTER SET LATIN NOT CASESPECIFIC,
+        creation_time TIMESTAMP(6),
+        modified_time TIMESTAMP(6),
+        archived_time TIMESTAMP(6))
+    NO PRIMARY INDEX ;
+"""
+
+EFS_FEATURE_GROUP_TRG = """
+    CREATE TRIGGER {0}.{1}
+    AFTER DELETE ON {0}.{2}
+    REFERENCING OLD AS DeletedRow
+    FOR EACH ROW
+        INSERT INTO {3}
+        VALUES (DeletedRow.name, DeletedRow.data_domain, DeletedRow.description, DeletedRow.data_source_name, DeletedRow.entity_name, DeletedRow.creation_time, DeletedRow.modified_time, 
+                current_timestamp(6)
+                );
+"""
+
+# Table for feature process.
+EFS_FEATURE_PROCESS = """
+    CREATE MULTISET TABLE {0}.{1}
+        (
+        process_id VARCHAR(200) CHARACTER SET LATIN NOT CASESPECIFIC,
+        description VARCHAR(2000) CHARACTER SET LATIN CASESPECIFIC,
+        data_domain VARCHAR(255) CHARACTER SET LATIN NOT CASESPECIFIC,
+        process_type VARCHAR(255) CHARACTER SET LATIN NOT CASESPECIFIC,
+        data_source VARCHAR(255) CHARACTER SET LATIN NOT CASESPECIFIC,
+        entity_id VARCHAR(255) CHARACTER SET LATIN CASESPECIFIC,
+        feature_names VARCHAR(2000) CHARACTER SET LATIN CASESPECIFIC,
+        feature_ids VARCHAR(2000) CHARACTER SET LATIN CASESPECIFIC,
+        valid_start TIMESTAMP(6) WITH TIME ZONE NOT NULL,
+        valid_end TIMESTAMP(6) WITH TIME ZONE NOT NULL,
+        PERIOD FOR ValidPeriod  (valid_start, valid_end) AS VALIDTIME)
+    PRIMARY INDEX (process_id);
+"""
+
+
+EFS_FEATURE_RUNS = """
+CREATE MULTISET TABLE {0}.{1}
+        (
+        run_id BIGINT GENERATED ALWAYS AS IDENTITY (START WITH 1 INCREMENT BY 1 MINVALUE 1 NO MAXVALUE NO CYCLE) NOT NULL,
+        process_id VARCHAR(200) CHARACTER SET LATIN NOT CASESPECIFIC,
+        data_domain VARCHAR(255) CHARACTER SET LATIN NOT CASESPECIFIC,
+        start_time TIMESTAMP(6),
+        end_time TIMESTAMP(6),
+        status VARCHAR(20) CHARACTER SET LATIN NOT CASESPECIFIC,
+        filter VARCHAR(2000) CHARACTER SET LATIN NOT CASESPECIFIC,
+        as_of_start TIMESTAMP(6) WITH TIME ZONE,
+        as_of_end TIMESTAMP(6) WITH TIME ZONE,
+        failure_reason VARCHAR(2000) CHARACTER SET LATIN CASESPECIFIC)
+    UNIQUE PRIMARY INDEX (run_id);
+"""
+
+# Table for storing the features metadata.
+EFS_FEATURES_METADATA = """
+    CREATE MULTISET TABLE {0}.{1}
+        (
+        entity_name VARCHAR(255) CHARACTER SET LATIN NOT CASESPECIFIC NOT NULL,
+        data_domain VARCHAR(200) CHARACTER SET LATIN NOT CASESPECIFIC,
+        feature_id BIGINT NOT NULL,
+        table_name VARCHAR(255) CHARACTER SET LATIN NOT CASESPECIFIC,
+        valid_start TIMESTAMP(6) WITH TIME ZONE NOT NULL,
+        valid_end TIMESTAMP(6) WITH TIME ZONE NOT NULL,
+        PERIOD FOR ValidPeriod  (valid_start, valid_end) AS VALIDTIME)
+    PRIMARY INDEX (entity_name);
+"""
+
+EFS_DATASET_CATALOG = """
+    CREATE MULTISET TABLE {0}.{1}
+        (
+        id VARCHAR(36) CHARACTER SET LATIN NOT CASESPECIFIC NOT NULL,
+        data_domain VARCHAR(200) CHARACTER SET LATIN NOT CASESPECIFIC,
+        name VARCHAR(255) CHARACTER SET LATIN NOT CASESPECIFIC NOT NULL,
+        entity_name VARCHAR(255) CHARACTER SET LATIN NOT CASESPECIFIC NOT NULL,
+        database_name VARCHAR(255) CHARACTER SET LATIN NOT CASESPECIFIC NOT NULL,
+        description VARCHAR(2000) CHARACTER SET LATIN NOT CASESPECIFIC,
+        valid_start TIMESTAMP(6) WITH TIME ZONE NOT NULL,
+        valid_end TIMESTAMP(6) WITH TIME ZONE NOT NULL,
+        PERIOD FOR ValidPeriod  (valid_start, valid_end) AS VALIDTIME)
+    PRIMARY INDEX (id);
+"""
+
+EFS_DATASET_FEATURES = """
+    CREATE MULTISET TABLE {0}.{1}
+        (
+        dataset_id VARCHAR(36) CHARACTER SET LATIN NOT CASESPECIFIC NOT NULL,
+        data_domain VARCHAR(200) CHARACTER SET LATIN NOT CASESPECIFIC,
+        feature_id BIGINT,
+        feature_name VARCHAR(255) CHARACTER SET LATIN NOT CASESPECIFIC NOT NULL,
+        feature_version VARCHAR(255) CHARACTER SET LATIN NOT CASESPECIFIC NOT NULL,
+        feature_repo VARCHAR(255) CHARACTER SET LATIN NOT CASESPECIFIC NOT NULL,
+        feature_view VARCHAR(255) CHARACTER SET LATIN NOT CASESPECIFIC NOT NULL,
+        valid_start TIMESTAMP(6) WITH TIME ZONE NOT NULL,
+        valid_end TIMESTAMP(6) WITH TIME ZONE NOT NULL,
+        PERIOD FOR ValidPeriod  (valid_start, valid_end) AS VALIDTIME)
+    PRIMARY INDEX (dataset_id);
+"""
+
+EFS_FEATURE_VERSION = """
+CREATE VIEW {}.{} AS 
+LOCK ROW FOR ACCESS
+SELECT
+    data_domain,
+    entity_id,        
+    trim(NGRAM) AS feature_name,
+    PROCESS_ID as feature_version
+FROM NGramSplitter (
+    ON (
+        SELECT * FROM {}.{}
+        ) as paragraphs_input
+        USING
+            TextColumn ('FEATURE_NAMES')
+            ConvertToLowerCase ('false')
+            Grams ('1')
+            Delimiter(',')
+    ) AS dt; 
+"""
+
+# Select the archived records.
+EFS_ARCHIVED_RECORDS = """
+SELECT {}, 
+CASE WHEN valid_end < current_timestamp then 1 else 0 end as is_archived 
+FROM {} 
+WHERE {}"""
+
+# Table to store the version of feature store. This is very important.
+# When teradataml incrementally adds functionality for feature store, this
+# version will be deciding factor whether teradataml should automatically
+# update metadata or not.
+
+EFS_VERSION = """
+    CREATE MULTISET TABLE {0}.{1} (
+        version VARCHAR(20) CHARACTER SET LATIN NOT CASESPECIFIC,
+        creation_time TIMESTAMP(6)
+    );
+"""
+
+EFS_VERSION_ = "2.0.0"
+
+EFS_DB_COMPONENTS = {
+    "data_domain": "_efs_data_domains",
+    "feature": "_efs_features",
+    "feature_staging": "_efs_features_staging",
+    "feature_trg": "_efs_features_trg",
+    "group_features": "_efs_group_features",
+    "group_features_staging": "_efs_group_features_staging",
+    "group_features_trg": "_efs_group_features_trg",
+    "entity": "_efs_entity",
+    "entity_staging": "_efs_entity_staging",
+    "entity_trg": "_efs_entity_trg",
+    "entity_xref": "_efs_entity_xref",
+    "entity_staging_xref": "_efs_entity_xref_staging",
+    "entity_xref_trg": "_efs_entity_xref_trg",
+    "data_source": "_efs_data_source",
+    "data_source_staging": "_efs_data_source_staging",
+    "data_source_trg": "_efs_data_source_trg",
+    "feature_group": "_efs_feature_group",
+    "feature_group_staging": "_efs_feature_group_staging",
+    "feature_group_trg": "_efs_feature_group_trg",
+    "feature_process": "_efs_feature_process",
+    "feature_runs": "_efs_feature_runs",
+    "feature_metadata": "_efs_features_metadata",
+    "dataset_catalog": "_efs_dataset_catalog",
+    "dataset_features": "_efs_dataset_features",
+    "feature_version": "_efs_feature_version",
+    "version": "_efs_version"
+}
+
+
+EFS_TABLES = {
+    EFS_DATA_DOMAINS: "_efs_data_domains",
+    EFS_FEATURES: "_efs_features",
+    EFS_FEATURES_STAGING: "_efs_features_staging",
+    EFS_GROUP_FEATURES: "_efs_group_features",
+    EFS_GROUP_FEATURES_STAGING: "_efs_group_features_staging",
+    EFS_ENTITY: "_efs_entity",
+    EFS_ENTITY_STAGING: "_efs_entity_staging",
+    EFS_ENTITY_XREF: "_efs_entity_xref",
+    EFS_ENTITY_XREF_STAGING: "_efs_entity_xref_staging",
+    EFS_DATA_SOURCE: "_efs_data_source",
+    EFS_DATA_SOURCE_STAGING: "_efs_data_source_staging",
+    EFS_FEATURE_GROUP: "_efs_feature_group",
+    EFS_FEATURE_RUNS: "_efs_feature_runs",
+    EFS_FEATURE_GROUP_STAGING: "_efs_feature_group_staging",
+    EFS_FEATURE_PROCESS: "_efs_feature_process",
+    EFS_FEATURES_METADATA: "_efs_features_metadata",
+    EFS_DATASET_CATALOG: "_efs_dataset_catalog",
+    EFS_DATASET_FEATURES: "_efs_dataset_features",
+    EFS_VERSION: "_efs_version"
+}
+
+EFS_TRIGGERS = {
+    EFS_FEATURES_TRG: "_efs_features_trg",
+    EFS_GROUP_FEATURES_TRG: "_efs_group_features_trg",
+    EFS_ENTITY_TRG: "_efs_entity_trg",
+    EFS_ENTITY_XREF_TRG: "_efs_entity_xref_trg",
+    EFS_DATA_SOURCE_TRG: "_efs_data_source_trg",
+    EFS_FEATURE_GROUP_TRG: "_efs_feature_group_trg"
+}
+
+class FeatureStatus(Enum):
+    ACTIVE = 1
+    INACTIVE = 2
+
+
+class FeatureType(Enum):
+    CONTINUOUS = 1
+    CATEGORICAL = 2
+    NUMERICAL = 3
+
+class ProcessType(Enum):
+    DENORMALIZED_VIEW = 'denormalized view'
+    FEATURE_GROUP = 'feature group'
+    NEW = 'new'
+    EXISTING = 'existing'
+
+
+class ProcessStatus(Enum):
+    NOT_STARTED = 'not started'
+    RUNNING = 'running'
+    COMPLETED = 'completed'
+    FAILED = 'failed'
+
+
+class _FeatureStoreDFContainer:
+    """
+    Utility class for FeatureStore DataFrame operations.
+    
+    This class provides static methods for creating and managing DataFrames
+    used across different FeatureStore components, eliminating code duplication
+    and providing a centralized, efficient approach to DataFrame handling.
+    """
+    __df_container = {}
+
+    @staticmethod
+    def get_df(obj_type, repo, data_domain):
+        """
+        DESCRIPTION:
+            Generic static method to create and manage DataFrames for different object types
+            in FeatureStore. Handles joins and special object type processing.
+
+        PARAMETERS:
+            obj_type:
+                Required Argument.
+                Specifies the type of DataFrame to return.
+                Supported types: 'feature', 'feature_staging', 'entity', 'entity_staging',
+                'feature_wog', 'feature_info', 'feature_catalog', 'entity_info', and all
+                other types defined in EFS_DB_COMPONENTS.
+                Types: str
+
+            repo:
+                Required Argument.
+                Specifies the repository name.
+                Types: str
+
+            data_domain:
+                Required Argument.
+                Specifies the data domain for filtering operations.
+                Types: str
+
+        RETURNS:
+            teradataml DataFrame.
+
+        RAISES:
+            TeradataMlException
+
+        EXAMPLES:
+            >>> # Basic DataFrame retrieval
+            >>> df = _FeatureStoreDFContainer.get_df(
+            ...     obj_type='feature',
+            ...     repo='my_repo',
+            ...     data_domain='analytics'
+            ... )
+            
+            >>> # Complex join for feature info
+            >>> df = _FeatureStoreDFContainer.get_df(
+            ...     obj_type='feature_info',
+            ...     repo='my_repo',
+            ...     data_domain='analytics'
+            ... )
+        """
+        from teradataml.dataframe.dataframe import DataFrame, in_schema
+        repo_obj = repo + '.' + data_domain + '.' + obj_type
+
+        if repo_obj not in _FeatureStoreDFContainer.__df_container:
+
+            # Handle complex FeatureStore-specific patterns with joins
+            if obj_type in ["feature", "feature_staging"]:
+                # Join features with group_features for group name
+                map_ = {"feature": "group_features", "feature_staging": "group_features_staging"}
+                features = DataFrame(in_schema(repo, EFS_DB_COMPONENTS[obj_type]))
+                features_xref = DataFrame(in_schema(repo, EFS_DB_COMPONENTS[map_[obj_type]]))
+                features = features[features.data_domain == data_domain]
+                features_xref = features_xref[features_xref.feature_data_domain == data_domain].select(["feature_name", "group_name"])
+                df = features.join(features_xref, on=["name==feature_name"], how='left')
+                _FeatureStoreDFContainer.__df_container[repo_obj] = df.select(features.columns + ["group_name"])
+
+            elif obj_type in ["entity", "entity_staging"]:
+                # Join entity with entity_xref for entity columns
+                ent_df = DataFrame(in_schema(repo, EFS_DB_COMPONENTS[obj_type]))
+                xref_df = DataFrame(in_schema(repo, EFS_DB_COMPONENTS["{}_xref".format(obj_type)]))
+                ent_df = ent_df[ent_df.data_domain == data_domain]
+                xref_df = xref_df[xref_df.data_domain == data_domain].select(['entity_name', 'entity_column'])
+                df = ent_df.join(xref_df, on=["name==entity_name"], how="inner")
+                _FeatureStoreDFContainer.__df_container[repo_obj] = df.select(ent_df.columns + ["entity_column"])
+
+            elif obj_type == "feature_wog":
+                # Feature without group - direct access to feature table
+                _FeatureStoreDFContainer.__df_container[repo_obj] = DataFrame(in_schema(repo, EFS_DB_COMPONENTS["feature"]))
+
+            elif obj_type == "feature_info":
+                # join: features + metadata
+                # Use feature_wog (without group)
+                feature = _FeatureStoreDFContainer.get_df('feature_wog', repo, data_domain)
+
+                # Get metadata DataFrame
+                feature_metadata = DataFrame(in_schema(repo, EFS_DB_COMPONENTS["feature_metadata"]))
+                
+                # Drop ValidPeriod column if it exists
+                if 'ValidPeriod' in feature_metadata.columns:
+                    feature_metadata = feature_metadata.drop(columns=["ValidPeriod"])
+
+                df = feature_metadata.join(feature,
+                                           how="inner",
+                                           on=[feature_metadata.feature_id == feature.id,
+                                               feature_metadata.data_domain == feature.data_domain,
+                                               feature_metadata.data_domain == data_domain],
+                                           lsuffix="_meta",
+                                           rsuffix="_feat")
+                _FeatureStoreDFContainer.__df_container[repo_obj] = df
+
+            elif obj_type == "feature_catalog":
+                # join: features + metadata + version
+                # Get the required DataFrames directly
+                fv = DataFrame(in_schema(repo, EFS_DB_COMPONENTS["feature_version"]))
+                f_ = _FeatureStoreDFContainer.get_df("feature", repo, data_domain)
+
+                # Feature can be mapped to more than one feature group. So, 'f_' can have duplicate rows
+                # which propagates these duplicates to final result.
+                f_ = f_.drop_duplicate(['id', 'data_domain', 'name'])
+                fm = DataFrame(in_schema(repo, EFS_DB_COMPONENTS["feature_metadata"]))
+
+                ndf = fm.select(['entity_name', 'data_domain', 'feature_id', 'table_name', 'valid_end'])
+                hdf = ndf.join(
+                    f_, on=((f_.id == ndf.feature_id) & (ndf.data_domain == f_.data_domain)),
+                    how='inner',
+                    lprefix='l'
+                ).select(['entity_name', 'data_domain', 'id', 'name', 'table_name', 'valid_end'])
+
+                vdf = hdf.join(fv,
+                               on=(
+                                       (hdf.data_domain == fv.data_domain) &
+                                       (hdf.entity_name == fv.entity_id) &
+                                       (fv.feature_name == hdf.name) &
+                                       (fv.data_domain == data_domain)
+                               ),
+                               how='inner',
+                               lprefix='l'
+                               )
+
+                _FeatureStoreDFContainer.__df_container[repo_obj] = vdf.select(
+                    ['entity_id', 'data_domain', 'id', 'name', 'table_name', 'feature_version', 'valid_end']
+                )
+                
+            elif obj_type == "entity_info":
+                # join: entity + entity_xref
+                entity_df = DataFrame(in_schema(repo, EFS_DB_COMPONENTS["entity"]))
+                entity_xref_df = DataFrame(in_schema(repo, EFS_DB_COMPONENTS["entity_xref"]))
+
+                # Build join conditions
+                join_conditions = [
+                    entity_df.name == entity_xref_df.entity_name,
+                    entity_df.data_domain == entity_xref_df.data_domain,
+                    entity_df.data_domain == data_domain
+                ]
+                
+                df = entity_df.join(
+                    other=entity_xref_df,
+                    on=join_conditions,
+                    lsuffix="l"
+                )
+
+                _FeatureStoreDFContainer.__df_container[repo_obj] = df.select(
+                    ['entity_name', 'data_domain', 'entity_column', 'description']
+                )
+
+            elif obj_type == 'data_domain':
+                _FeatureStoreDFContainer.__df_container[repo_obj] = DataFrame(in_schema(repo, EFS_DB_COMPONENTS["data_domain"]))
+
+            # Default case: simple DataFrame creation
+            else:
+                df = DataFrame(in_schema(repo, EFS_DB_COMPONENTS[obj_type]))
+                if 'data_domain' in df.columns:
+                    df = df[df.data_domain == data_domain]
+                
+                _FeatureStoreDFContainer.__df_container[repo_obj] = df
+
+        return _FeatureStoreDFContainer.__df_container[repo_obj]