Untitled
unknown
plain_text
a year ago
22 kB
3
Indexable
import pandas as pd
import json
from base64 import b64decode as base_b64decode
import logging
from pandas import read_sql as pd_read_sql
import sys
from sqlalchemy import create_engine
import argparse
from io import TextIOBase as io_TextIOBase
from json import load as json_load
import os
import datetime
# define Python user-defined exceptions
class Error(Exception):
"""Base class for other exceptions"""
pass
# define Python user-defined exceptions
class ETL_PPM_REPLICATION_MASTER_ERROR(Error):
pass
class DB_CONNECTION_ERROR(Error):
pass
def is_datetime(value):
try:
# Attempt to parse the value as a datetime
datetime.datetime.strptime(str(value), '%Y-%m-%d %H:%M:%S')
return True
except ValueError:
return False
# Function to write SQL query to the log file
def write_sql(sql_log_file, query_info):
sql_log_file.write(query_info)
sql_log_file.write('\n')
def setDbConnection(logging, json_data, serverInfo, encoding):
from sqlalchemy import create_engine as sqlalchemy_create_engine
from base64 import b64decode as base_b64decode
import mysql.connector
try:
cnx = cursor = schema = db_type = None
encrypt = json_data.get(serverInfo, {}).get('ENCRYPT')
host = json_data.get(serverInfo, {}).get('DB_HOST')
port = json_data.get(serverInfo, {}).get('DB_PORT')
user = json_data.get(serverInfo, {}).get('DB_USER')
db_type = json_data.get(serverInfo, {}).get('DB_TYPE')
schema = json_data.get(serverInfo, {}).get('DB_SCHEMA')
if encrypt == 'Y':
password = base_b64decode(json_data.get(serverInfo, {}).get('DB_PASSWORD')).decode('utf-8')
else:
password = json_data.get(serverInfo, {}).get('DB_PASSWORD')
if db_type in ('MYSQL', 'MARIA'):
connection_text = ('mysql+mysqlconnector://%s:%s@%s:%s/%s' % (user
, password
, host
, port
, schema))
elif db_type == 'ORACLE':
connection_text = ('oracle://%s:%s@%s:%s/%s' % (user
, password
, host
, port
, json_data.get(serverInfo, {}).get('DB_SID')
))
cnx = sqlalchemy_create_engine(connection_text
, encoding=encoding)
# ,fast_executemany=True
# , connect_args={'connect_timeout': 600})
cursor = cnx.connect()
logging.info(f"Connected to database server {serverInfo}: {host}:{port}/{schema}")
# except mysql.connector.Error as dberr:
# logging.error("DATABASE CONNECTION ERROR")
# logging.error("Error - {} . Line No - {} ".format(str(dberr), str(sys.exc_info()[-1].tb_lineno)))
# cnx = cursor = schema = None
except Exception as dbexp:
logging.error("DATABASE CONNECTION EXCEPTION")
logging.error("Error - {} . Line No - {} ".format(str(dbexp), str(sys.exc_info()[-1].tb_lineno)))
cnx = cursor = schema = None
return cnx, cursor, schema, db_type
# def main(args, json_data, log_file, logging, encoding):
def process_changes_only(pdb_ppm_sc, pdb_source, pdb_target, ps_rep_env, ps_op_id, ps_bu_id, ps_rep_job_id, pl_logging):
try:
pl_logging.info("Started")
replicationJobId = args.replicationJobId
return_flag = True
STATUS = "Success"
STATUS_MESSAGE = "Service Catalogue Replication Successful."
failed_entities = []
opId=ps_op_id
buId=ps_bu_id
pl_logging.info("Variable declaration")
total_events_logged = 0
source_read_count = 0
target_read_count = 0
inserted_count = 0
updated_count = 0
insert_failed_count = 0
update_failed_count = 0
query_count=0
pl_logging.info("Master query")
sql_log_file = open(f"/app/scripts/PPM_Release_Management/Service_Catalog_ETL/logs/{replicationJobId}_ppm_sc_changesonly.sql","w")
primary_query = f"SELECT * FROM {schema_ppm}.etl_ppm_replication_master WHERE eprm_catalog='SC' AND eprm_enabled_flg='Y'"
write_sql(sql_log_file,f"-- ++++++++++++++++++++++++++++++++ Primary Query")
write_sql(sql_log_file,primary_query)
primary_df = pd_read_sql(primary_query, con=pdb_ppm_sc)
for index, row in primary_df.iterrows():
# Fetch primary key column name and table name
table_name = row['eprm_table_name'].lower()
eprm_table_col_pk = row['eprm_table_col_pk']
pk = eprm_table_col_pk.lower()
where_clause = ((row['eprm_join_cols_reim'] % (ps_op_id, ps_bu_id)))
query_count+=1
table_alias=row['eprm_table_alias']
write_sql(sql_log_file,
f"-- ++++++++++++++++++++++++++++++++ Seq# {query_count}| entity# {pk} | {table_name}")
try:
source_query = f"SELECT /*+ PARALLEL(2) */ * FROM {schema_source}.{table_name} {table_alias} where {where_clause}"
write_sql(sql_log_file,source_query)
pl_logging.info("Reading data from source")
source_df = pd.read_sql(source_query, connection_source)
pl_logging.info("source_df: %s" % len(source_df))
write_sql(sql_log_file,f"source: {table_name}: {source_read_count}")
source_read_count+=len(source_df)
target_query = f"SELECT /*+ PARALLEL(2) */ * FROM {schema_ext}.{table_name} {table_alias} where {where_clause}"
write_sql(sql_log_file,target_query)
pl_logging.info("Reading data from target")
target_df = pd.read_sql(target_query, connection_target)
write_sql(sql_log_file, f"target: {table_name}: {target_read_count}")
target_read_count+=len(target_df)
for index, source_row in source_df.iterrows():
pk_value = source_row[pk]
# Check if the primary key exists in the target DataFrame
if pk_value not in target_df[pk].values:
# Replace 'None' and 'NaT' with None in source_row
for column_name, source_val in source_row.items():
if source_val == 'None' or source_val == 'NaT' or source_val == 'nan':
source_row[column_name] = NULL
# Generate an INSERT query dynamically
insert_query = f"INSERT INTO {schema_ext}.{table_name} ("
insert_columns = []
insert_values = []
for column_name, source_val in source_row.items():
if source_val is not None:
if isinstance(source_val, str) and source_val.startswith('TO_DATE'):
# If it already starts with TO_DATE, don't add TO_DATE again
insert_values.append(source_val)
elif is_datetime(source_val):
# Format datetime values using the appropriate function for the database type
if db_type_ext == 'ORACLE':
insert_values.append(f"TO_DATE('{source_val}', 'YYYY-MM-DD HH24:MI:SS')")
elif db_type_ext in ('MYSQL', 'MARIA'):
# For MariaDB, use STR_TO_DATE
insert_values.append(f"STR_TO_DATE('{source_val}', '%Y-%m-%d %H:%i:%s')")
else:
# Enclose other values in single quotes
insert_values.append(f"'{source_val}'")
insert_columns.append(column_name) # Add the column name
elif str(source_val) == 'NaT':
# Replace 'NaT' with NULL without single quotes
insert_values.append('NULL')
insert_columns.append(column_name) # Add the column name
elif column_name == 'extended_rule_code':
parts = source_val.split('==')
if len(parts) == 2:
extended_rule_code = f"'{parts[0]}=='{parts[1]}''"
insert_values.append(extended_rule_code)
insert_columns.append(column_name)
elif str(source_val) == 'nan':
insert_values.append('NULL')
insert_columns.append(column_name)
else:
# Enclose other values in single quotes
source_val = str(source_val).replace("'","''")
insert_values.append(f"'{source_val}'")
insert_columns.append(column_name) # Add the column name
else:
insert_values.append('NULL') # Insert a true NULL
insert_columns.append(column_name) # Add the column name
# Construct the INSERT query with column names
insert_query = f"INSERT INTO {schema_ext}.{table_name} ({', '.join(insert_columns)}) VALUES ({', '.join(insert_values)})"
# Execute the INSERT query
try:
inserted_count += 1
cursor_target.execute(insert_query)
write_sql(sql_log_file, "INSERTION" +insert_query + ";\n")
except Exception as e:
print(insert_query)
pl_logging.error(
"Error - {} . Line No - {} ".format(str(e), str(sys.exc_info()[-1].tb_lineno)))
write_sql(sql_log_file, f"INSERT FAILED: {insert_query}")
insert_failed_count += 1
else:
# Fetch the corresponding row from the target DataFrame based on the primary key
target_row = target_df[target_df[pk] == pk_value].iloc[0]
if not source_row.equals(target_row):
columns_to_update = []
for column_name, source_val in source_row.items():
target_val = target_row[column_name]
if source_val != target_val:
if is_datetime(source_val):
# Format datetime values using the appropriate function for the database type
if db_type_ext == 'ORACLE':
update_value = f"TO_DATE('{source_val}', 'YYYY-MM-DD HH24:MI:SS')"
elif db_type_ext in ('MYSQL', 'MARIA'):
# For MariaDB and MySQL, use STR_TO_DATE
update_value = f"STR_TO_DATE('{source_val}', '%Y-%m-%d %H:%i:%s')"
else:
# Enclose other values in single quotes
update_value = f"'{source_val}'"
elif str(source_val) == 'NaT':
# Replace 'NaT' with NULL without single quotes
update_value = 'NULL'
elif str(source_val) == 'nan':
# Replace 'NaT' with NULL without single quotes
update_value = 'NULL'
elif str(source_val) == 'None':
update_value = 'NULL'
elif column_name == 'extended_rule_code':
parts = source_val.split('==')
if len(parts) == 2:
extended_rule_code = f"'{parts[0]}=='{parts[1]}'"
update_value = extended_rule_code
else:
if column_name == 'created_by':
update_value = f"'{source_val}'"
else:
# Handle non-datetime columns (e.g., strings, numbers) here
update_value = f"'{source_val}'"
# Add the column name and formatted value to the update statement
columns_to_update.append(f"{column_name} = {update_value}")
# Generate an update query dynamically
if columns_to_update:
update_query = f"UPDATE {schema_ext}.{table_name} SET "
update_query += ", ".join(columns_to_update)
update_query += f" WHERE {eprm_table_col_pk} = '{pk_value}' AND OP_ID='{opId}' AND BU_ID='{buId}'"
try:
print(update_query)
cursor_target.execute(update_query)
updated_count+=1
write_sql(sql_log_file, f"UPDATE: {update_query}")
except Exception as e:
pl_logging.error(
"Error - {} . Line No - {} ".format(str(e), str(sys.exc_info()[-1].tb_lineno)))
write_sql(sql_log_file, f"UPDATE FAILED: {update_query}")
update_failed_count += 1
except Exception as e:
error_msg=f"Error querying table {schema_ext}.{table_name}: {str(e)} skipping the table and proceeding for next table"
pl_logging.error(error_msg)
failed_entities.append((table_name,error_msg))
continue
# Calculate the inserted and updated counts
total_events_logged=inserted_count+updated_count
pl_logging.info(f"Total Events Logged: {total_events_logged}")
pl_logging.info(f"Source Read Count: {source_read_count}")
pl_logging.info(f"Target Read Count: {target_read_count}")
pl_logging.info(f"Inserted Count: {inserted_count}")
pl_logging.info(f"Updated Count: {updated_count}")
pl_logging.info(f"Failed Inserted Count: {insert_failed_count}")
pl_logging.info(f"Failed Updated Count: {update_failed_count}")
pl_logging.info(f"-- ++++++++++++++++++++++++++++++++++ BACK UPDATE STATUS FOR UI ++++++++++++++++++ \n")
if len(failed_entities) > 0:
return_flag = False
STATUS = "Error"
STATUS_MESSAGE = str(failed_entities).replace("'", '').replace('"', '')
pl_logging.info("STATUS: %s" % STATUS)
pl_logging.info("STATUS_MESSAGE: %s" % STATUS_MESSAGE)
query_update = f"UPDATE {schema_source}.ppm_replication_status SET status='" + STATUS + "'" \
+ ", status_message='" + STATUS_MESSAGE + "'" \
+ ", error_description=NULL" \
+ ", updated_by='" + replicationJobId + "' /**/" \
+ " WHERE replication_job_id='"+replicationJobId+"'"
if db_type_source == 'ORACLE':
query_update = query_update.replace('/**/', ', updated_date=SYSDATE')
elif db_type_source in ('MARIA', 'MYSQL'):
query_update = query_update.replace('/**/', ', updated_date=NOW()')
pl_logging.info("-- + ppm_replication_status - UPDATE \n")
pl_logging.info(query_update + ";\n")
pl_logging.info(f"-- ++++++++++++++++++++++++++++++++++ FIN ++++++++++++++++++++++++++++++++++++++++ \n")
res = cursor_source.execute(query_update)
pl_logging.info("query_update: %s" % res)
except ETL_PPM_REPLICATION_MASTER_ERROR:
STATUS_MESSAGE = "NO RECORDS PRESENT IN etl_ppm_replication_master TABLE"
pl_logging.error("EXCEPTION:" + STATUS_MESSAGE)
return_flag = False
except Exception as e:
pl_logging.error("Error - {} . Line No - {} ".format(str(e), str(sys.exc_info()[-1].tb_lineno)))
return_flag = False
return return_flag
if __name__ == '__main__':
import logging
from configparser import ConfigParser as conf_ConfigParser
try:
# python3 ppm_sc_changesonly.py --operation ChangesOnly --replicationTarget SIT --opId HOB --buId DEFAULT --replicationJobId SC_JOB_999
parser = argparse.ArgumentParser(description="PPM Service Catalog Replication Script")
parser.add_argument('--operation', required=True, help="Operation Type")
parser.add_argument('--replicationTarget', required=True, help="Replication Target")
parser.add_argument('--opId', required=True, help="Operation ID")
parser.add_argument('--buId', required=True, help="Business Unit ID")
parser.add_argument('--replicationJobId', required=True, help="Replication Job ID")
args = parser.parse_args()
replicationJobId = args.replicationJobId
json_file_path = "/app/scripts/PPM_Release_Management/Service_Catalog_ETL/config/dna.json"
conf_file_path = "/app/scripts/PPM_Release_Management/Service_Catalog_ETL/config/ppm_sc_replication.conf"
sql_log_file = f"/app/scripts/PPM_Release_Management/Service_Catalog_ETL/logs/{replicationJobId}_ppm_sc_changesonly.sql"
log_file = f'/app/scripts/PPM_Release_Management/Service_Catalog_ETL/logs/{replicationJobId}_ppm_sc_changesonly.log'
args = parser.parse_args()
# Set up logging
CONFIG = conf_ConfigParser()
CONFIG.read(conf_file_path)
logging.basicConfig(filename=log_file
, level=CONFIG.get('CONFIG_LOGGING', 'LOG_LEVEL', raw=True)
, format=CONFIG.get('CONFIG_LOG_FORMAT', 'LOG_FORMAT_DISP', raw=True)
, datefmt=CONFIG.get('CONFIG_LOG_FORMAT', 'LOG_FORMAT_DATE', raw=True)
)
logging.info('LOGGER initiated')
encoding = CONFIG.get('CONFIG_GENERIC', 'DB_CHARSET', raw=True)
# Read JSON data from file
if not os.path.exists(json_file_path):
logging.error("CREDENTIAL FILE MISSING")
logging.error("CREDENTIAL FILE: %s" % json_file_path)
raise FileNotFoundError("CREDENTIAL FILE MISSING")
with open(json_file_path) as json_file:
json_data = json_load(json_file)
# Connect to PPM_PC database
connection_ppm, cursor_ppm, schema_ppm, db_type_ppm = setDbConnection(logging, json_data, 'PPM_PC', encoding)
# Connect to source database
connection_source, cursor_source, schema_source, db_type_source = setDbConnection(logging, json_data, 'SOURCE',
encoding)
# Connect to target_ext database
connection_target, cursor_target, schema_ext, db_type_ext = setDbConnection(logging, json_data,
args.replicationTarget,
encoding)
if not (connection_ppm and connection_source and connection_target):
raise DB_CONNECTION_ERROR
# if main(args, json_data, log_file, logging, encoding):
if process_changes_only(pdb_ppm_sc=connection_ppm
, pdb_source=connection_source
, pdb_target=connection_target
, ps_rep_env=args.replicationTarget
, ps_op_id=args.opId
, ps_bu_id=args.buId
, ps_rep_job_id=args.replicationJobId
, pl_logging=logging):
print("Update successful")
else:
print("Update FAILED")
except DB_CONNECTION_ERROR:
logging.error("EXCEPTION: DB CONNECTION ERROR")
print("EXCEPTION: DB CONNECTION ERROR")
except FileNotFoundError as ferr:
print("Error - {} . Line No - {} ".format(str(ferr), str(sys.exc_info()[-1].tb_lineno)))
except Exception as err:
print("Error - {} . Line No - {} ".format(str(err), str(sys.exc_info()[-1].tb_lineno)))
in the above insert statrement can you please update and optimize like if source pk_value is not present in target can you directly insert that record in target using to_sql and remove all the checks that we are doing