try:

    from datetime import datetime, timezone, timedelta
    from airflow import DAG
    
    from airflow.operators.python_operator import PythonOperator
    from airflow.operators.http_operator import SimpleHttpOperator
    from datetime import datetime
    from pandas.io.json import json_normalize
    from airflow.operators.postgres_operator import PostgresOperator

    import pandas as pd
    import json
    import requests
    import numpy as np

    import psycopg2
    from sqlalchemy import create_engine

except Exception as e:
    print("Error {} ".format(e))

dRoW_api_end_url = "https://drow.cloud"

def getDrowToken(**context):
    response = requests.post(
    url=f"{dRoW_api_end_url}/api/auth/authenticate",
    data={
    "username": "icwp2@drow.cloud",
    "password": "dGVzdDAxQHRlc3QuY29t"
    }
    ).json()
    context["ti"].xcom_push(key="token", value=response['token'])

def getSheetData(token , sheetId):
    response = requests.get(
    url=f"{dRoW_api_end_url}/api/sheets/{sheetId}?with_records=true&fields=",
    headers={
    "x-access-token": f"Bearer {token}",
    }
    )
    sheet = json.loads(response.text)
    headers = sheet['header']
    record = sheet['record']
    dataToExtract=[]
    for d in record: 
        objectToPush = {}
        for v in d['values']:
            for c in headers:
                colNameToExtract = c['colName']
                if v['colName'] == colNameToExtract:
                    # # print(v)
                    if v.get('multValue') != None:
                        if v['multValue'] == True:
                            if v['colType'] == 'Table':
                                tObjectArray = []
                                for t in v['tableValue']:
                                    tObjectToPush = {}
                                    for s in t['subValues']:
                                        tObjectToPush[s['colName']] = s.value
                                    tObjectArray.push(tObjectToPush)
                            else:
                                objectToPush[v['colName']] = v['valueArray']
                        else:
                            if v.get('value') != None:
                                if v.get('value') == 'NA':
                                    objectToPush[v['colName']] = None
                                else:
                                    objectToPush[v['colName']] = v['value']
                            else:
                                objectToPush[v['colName']] = None
                    else:
                        if v.get('value') != None:
                            if v.get('value') == 'NA':
                                objectToPush[v['colName']] = None
                            else:
                                objectToPush[v['colName']] = v['value']
                        else:
                            objectToPush[v['colName']] = None
        dataToExtract.append(objectToPush)
    return dataToExtract

def getWorkflowData(token , workflowId):
    response = requests.get(
    url=f"{dRoW_api_end_url}/api/module/document-export/airflow/workflow/{workflowId}?export_type=0",
    headers={
    "x-access-token": f"Bearer {token}",
    }
    )
    return json.loads(response.text)

def changeTimeFormat(date_string):
    if date_string is None:
        return datetime.now() 
    if "GMT" in date_string:
        format_string = "%a %b %d %Y %H:%M:%S GMT%z"
        date_parts = date_string.split(' (')
        date_object = datetime.strptime(date_parts[0], format_string)
        date_object = date_object.replace(tzinfo=timezone.utc)  # make date_object timezone-aware
        date64_object = np.datetime64(date_object, utc=True)  # convert date_object to a datetime64 object
        return date64_object
        return dt
    else:
        return datetime.strptime(date_string, "%d-%b-%Y")

def getdrowPSQLConnectionString():
    host                  = 'drowdatewarehouse.crlwwhgepgi7.ap-east-1.rds.amazonaws.com'  

    # User name of the database server
    dbUserName            = 'dRowAdmin'  

    # Password for the database user
    dbUserPassword        = 'drowsuper'  

    # Name of the database 
    database              = 'drowDateWareHouse'

    # Character set
    charSet               = "utf8mb4"  

    port                  = "5432"

    conn_string = ('postgres://' +
                           dbUserName + ':' + 
                           dbUserPassword +
                           '@' + host + ':' + port +
                           '/' + database)
    return conn_string

def pipelineProcess(**context):
    token = context.get("ti").xcom_pull(key="token")
    Data = getSheetData(token, "63fec465a01bde0cac754b4d")
    Data2 = getSheetData(token, "63fec436df8d4c0cb6f328bf")
    
    conn_string = getdrowPSQLConnectionString()
    db = create_engine(conn_string)
    conn = db.connect()

    df = pd.DataFrame()
    _df = pd.DataFrame()
    with conn as conn:
        for x in Data:
            df_nested_list = json_normalize(x)
            df2 = df_nested_list
            df = df.append(df2)        
        df['starting date']=df['starting date'].apply(pd.to_datetime)
        df['ori comp date']=df['ori comp date'].apply(pd.to_datetime)
        df.columns = df.columns.str.replace(' ', '_').str.replace('.', '').str.replace('(', '_').str.replace(')', '').str.replace('%', 'percent').str.replace('/', '_')
        df.to_sql('c1_nec_section_of_work', con=conn, if_exists='replace', index= False)
        
        for x in Data2:
            df_nested_list = json_normalize(x)
            df2=df_nested_list
            _df = _df.append(df2)
        _df['Starting Date']=_df['Starting Date'].apply(pd.to_datetime)
        _df['Original completion dates']=_df['Original completion dates'].apply(pd.to_datetime)
        _df.columns = _df.columns.str.replace(' ', '_').str.replace('.', '').str.replace('(', '_').str.replace(')', '').str.replace('%', 'percent').str.replace('/', '_')
        _df.to_sql('c1_nec_section_of_work_key_date', con=conn, if_exists='replace', index= False)
        
    Data = getSheetData(token, "63fec4f938237d0c7931f18c")
    df = pd.DataFrame.from_dict(Data)
    numerics = df.select_dtypes(include="number").columns
    df=df.apply(pd.to_numeric, errors='ignore')
    df[numerics]=df[numerics].apply(lambda x: np.round(x, decimals=5))
    df['IP No.']=df['IP No.'].astype(str)
    df['Month - Year']=df['Month - Year'].apply(pd.to_datetime)
    df.columns = df.columns.str.replace(' ', '_').str.replace('.', '_').str.replace('(', '_').str.replace(')', '').str.replace('%', 'percent')

    db = create_engine(conn_string)
    conn = db.connect()
    with conn as conn:
        df.to_sql('c1_finance_data', con=conn, if_exists='replace')
    conn.close()

    data = getSheetData(token, "63fec367a01bde0cac754807")
    df = pd.DataFrame.from_dict(data)
    numerics = df.select_dtypes(include="number").columns
    df=df.apply(pd.to_numeric, errors='ignore')
    df[numerics]=df[numerics].apply(lambda x: np.round(x, decimals=5))
    df['Month - Year']=df['Month - Year'].apply(pd.to_datetime)
    df.columns = df.columns.str.replace(' ', '_').str.replace('.', '_').str.replace('(', '_').str.replace(')', '').str.replace('%', 'percent')
    db = create_engine(conn_string)
    conn = db.connect()
    with conn as conn:
        df.to_sql('c1_finance_status_data', con=conn, if_exists='replace')
    conn.close()

    data = getSheetData(token, "63fc9ef84243400ca9af7c70")
    df = pd.DataFrame.from_dict(data)
    numerics = df.select_dtypes(include="number").columns
    df=df.apply(pd.to_numeric, errors='ignore')
    df[numerics]=df[numerics].apply(lambda x: np.round(x, decimals=5))
    df['Month - Year']=df['Month - Year'].apply(pd.to_datetime)
    df.columns = df.columns.str.replace(' ', '_').str.replace('.', '_').str.replace('(', '_').str.replace(')', '').str.replace('%', 'percent')
    db = create_engine(conn_string)
    conn = db.connect()
    with conn as conn:
        df.to_sql('c1_eot_data', con=conn, if_exists='replace')
    conn.close()

    data= getSheetData(token, "63fec3fca01bde0cac7549a8")
    if data:
        df = pd.DataFrame.from_dict(data)
        df['Submission Date']=df['Submission Date'].apply(changeTimeFormat)
        df['Acceptance Date']=df['Acceptance Date'].apply(changeTimeFormat)
        df.columns = df.columns.str.replace(' ', '_').str.replace('.', '_').str.replace('(', '_').str.replace(')', '').str.replace('%', 'percent')
        db = create_engine(conn_string)
        conn = db.connect()
        with conn as conn:
            df.to_sql('c1_programme_data', con=conn, if_exists='replace')
        conn.close()

    data = getSheetData(token, "63fec4a9df8d4c0cb6f32adf")
    df = pd.DataFrame.from_dict(data)
    df['Planned Completion Date(PCD)']=df['Planned Completion Date(PCD)'].apply(pd.to_datetime)
    df.columns = df.columns.str.replace(' ', '_').str.replace('.', '_').str.replace('(', '_').str.replace(')', '').str.replace('%', 'percent')
    db = create_engine(conn_string)
    conn = db.connect()
    with conn as conn:
        df.to_sql('c1_key_date_data', con=conn, if_exists='replace')
    conn.close()
    
    _Data = getWorkflowData(token, "637c7d22b38f8ca02f5c49ac")
    Mapping= {
            "Original Doc No.": "Original_Doc_No",
            "NEC Doc Type": "NEC_Doc_Type",
            "NEC Event No.": "NEC_Event_No",
            "Doc Ver.": "Doc_Ver",
            "Doc Date": "Doc_Date",
            "Subject": "Subject",
            "From": "From",
            "To": "To",
            "CE/PMI Amount": "CE_PMI_Amount",
            "CE Increase / Decrease": "CE_Increase_Decrease",
            "Quotation Status": "Quotation_Status",
            "NEC Clause": "NEC_Clause"
    }

    conn_string = getdrowPSQLConnectionString()
    db = create_engine(conn_string)
    conn = db.connect()
    df = pd.DataFrame()
    with conn as conn:
        for x in _Data:
            try:
                # print(x['data'])
                if len(x['data'].keys()) == 0:
                    continue
                df_nested_list = json_normalize(x['data'])
                df2 = df_nested_list.reindex(columns=Mapping.keys())
                df2['NEC Doc Title']=x['data']['NEC Doc Type']+x['data']['NEC Event No.']
                df2['Doc Org Ver']= x['data']['Doc Ver.']
                y=0
                if x['data']['Doc Ver.'] == None:
                    df2['Doc Ver.'] = y
                elif x['data']['Doc Ver.'].startswith('Rev. '):
                    y = x['data']['Doc Ver.'].replace('Rev. ', '')
                    y = int(y)
                else:
                    y = x['data']['Doc Ver.'].replace('-', '').replace('r','')
                    if y!='' and not y.isnumeric():
                        y = int(ord(y)) - int(ord('A')) + 1
                    elif y =='':
                        y = 0
                    else :
                        y = int(y)
                    df2['Doc Ver.'] = y
                if (not df2['NEC Doc Title'].empty and 'NEC Doc Title' in df.columns):
                    check_ver_df = df.loc[(df['NEC Doc Title'] == x['data']['NEC Doc Type']+x['data']['NEC Event No.'])]
                    if check_ver_df.empty:
                        df2['is_latest'] = 'Yes'
                    else :
                        check_ver_df2 = check_ver_df.loc[(check_ver_df['Doc Ver.'] > y)]
                        if not check_ver_df2.empty:
                            df2['is_latest'] = "No"
                        else:
                            df.loc[(df['NEC Doc Title'] == x['data']['NEC Doc Type']+x['data']['NEC Event No.']) & ( df['Doc Ver.']<y), 'is_latest'] = 'No'
                            df2['is_latest'] = 'Yes'
                else:
                    df2['is_latest'] = 'Yes'

                if (x['data'].get('NEC Doc Type') or '').strip().upper() == 'PMN-' and y==0 and (x['Status'] == 'Receipt by Contractor' or x['Status'] == 'Closed'):
                    df2['From_Status'] = '1. CE notified'
                elif (x['data'].get('NEC Doc Type') or '').strip().upper() == 'CSQ-' and y==0:
                    df2['From_Status'] = '2. Quotation Submitted'
                elif (x['data'].get('NEC Doc Type') or '').strip().upper() == 'QA-'  and (x['Status'] == 'Receipt by Contractor' or x['Status'] == 'Closed'):
                    df2['From_Status'] = '3. CE implemented'
                elif (x['data'].get('NEC Doc Type') or '').strip().upper() == 'PMIQ-' and y==0:
                    df['From_Status'] = '4. Quotation to be Submitted'
                else:
                    df2['From_Status'] = None

                df2['NEC Doc Title With Version']=x['data']['NEC Doc Type']+x['data']['NEC Event No.']+'-'+str(y)
                if len(x['data']['Change to Time'])>0 and x['data']['NEC Doc Type']!='EW':
                    df4=pd.DataFrame()
                    for change_to_time_table in x['data']['Change to Time']:
                        df3=df2.copy()
                        if 'Key Date' in change_to_time_table:
                            key_date_value = change_to_time_table['Key Date']
                            extracted_value = key_date_value.split(' (')[0]  # This assumes the format is consistent as 'Section 18 (Subject to excision)'
                            df3['Key Date'] = extracted_value
                        if 'Extension in days' in change_to_time_table:
                            df3['Extension in days'] = change_to_time_table['Extension in days']
                        if 'Ori Completion Date' in change_to_time_table:
                            df3['Ori Completion Date'] = change_to_time_table['Ori Completion Date']
                        if 'Revised Completion Date' in change_to_time_table:
                            df3['Revised Completion Date'] = change_to_time_table['Revised Completion Date']
                        df3['Status']=x['Status']
                        df4 = df4.append(df3)
                    df2 = df2.iloc[0:0]
                    df2=df2.append(df4)
                df2['Status']=x['Status']
                df = df.append(df2)
            except Exception as e:
                print(e)
                continue
        df.rename(columns=Mapping, inplace=True)
        df['Doc_Date']=df['Doc_Date'].apply(pd.to_datetime)
        df['Doc_Date'] = df['Doc_Date'] - pd.Timedelta(hours=8)
        df['Ori Completion Date']=df['Ori Completion Date'].apply(pd.to_datetime)
        # df['Ori Completion Date'] = df['Ori Completion Date'] - pd.Timedelta(hours=8)
        df['Revised Completion Date']=df['Revised Completion Date'].apply(pd.to_datetime)
        # df['Revised Completion Date'] = df['Revised Completion Date'] - pd.Timedelta(hours=8)
        df.columns = df.columns.str.replace(' ', '_').str.replace('.', '').str.replace('(', '_').str.replace(')', '').str.replace('%', 'percent').str.replace('/', '_')
        df.to_sql('c1_nec_cas', con=conn, if_exists='replace', index= False)
    conn.close()
    
    resData = getWorkflowData(token, "638b3316a1faf60c870384e2")
    resData = getSheetData(token, "64170431b6c1cf0cd799149c")
    db = create_engine(conn_string)
    conn = db.connect()
    df = pd.DataFrame()
    with conn as conn:
        for x in resData:
            df_nested_list = json_normalize(x)
            df2 = df_nested_list
            if x['Date of Early Warning  (EW)'] == None:
                Date_of_Early_Warning = datetime.now(timezone.utc)
            else:
                Date_of_Early_Warning = datetime.strptime(x['Date of Early Warning  (EW)'], '%Y-%m-%dT%H:%M:%S.%f%z')
            if x['Date of Close of EW'] == None:
                Date_of_Close_of_EW = datetime.now(timezone.utc)
            else: 
                Date_of_Close_of_EW = datetime.strptime(x['Date of Close of EW'], '%Y-%m-%dT%H:%M:%S.%f%z')
            # print((Date_of_Close_of_EW - Date_of_Early_Warning))
            if (Date_of_Close_of_EW - Date_of_Early_Warning) > np.timedelta64(24, 'h'):
                df2['Elapsed_Time'] = ((Date_of_Close_of_EW - Date_of_Early_Warning))
            else:
                df2['Elapsed_Time'] = np.timedelta64(0, 'D')
            if (df2['Elapsed_Time'] >= np.timedelta64(365, 'D')).bool():
                df2['Elapsed_Time_more_then_1_year'] = True
            else:
                df2['Elapsed_Time_more_then_1_year'] = False
            df2['Elapsed_Time'] = df2['Elapsed_Time'] / 1000 / 1000 / 86400000
            df = df.append(df2)

        df['Date of Close of EW']=df['Date of Close of EW'].apply(pd.to_datetime)
        df['Date of Close of EW'] = df['Date of Close of EW'] - pd.Timedelta(hours=8)
        df['Date of Early Warning']=df['Date of Early Warning  (EW)'].apply(pd.to_datetime)
        df['Date of Early Warning'] = df['Date of Early Warning'] - pd.Timedelta(hours=8)
        # df['Action Party (CEDD / AECOM / BKREJV)']=df['Action Party (CEDD / AECOM / BKREJV)']
        df.columns = df.columns.str.replace(' ', '_').str.replace('.', '_').str.replace('(', '_').str.replace(')', '').str.replace('/', '_').str.replace('%', 'percent')
        # df['Action_Party___CEDD_/_AECOM_/_DCK_JV']=np.array(df['Action_Party___CEDD_/_AECOM_/_DCK_JV'].tolist())
        df.to_sql('c1_nec_risk_register', con=conn, if_exists='replace', index= False)



# */2 * * * * Execute every two minute 
with DAG(
        dag_id="c1_nec",
        schedule_interval="0 0,4,8,11,16 * * *",
        default_args={
            "owner": "airflow",
            "retries": 1,
            "retry_delay": timedelta(minutes=5),
            "start_date": datetime(2022, 10, 24)
        },
        catchup=False) as f:
    
    pipelineProcess = PythonOperator(
        task_id="pipelineProcess",
        python_callable=pipelineProcess,
        provide_context=True,
    )
    
    # getWorkflowRecords = PythonOperator(
    #     task_id="getWorkflowRecords",
    #     python_callable=getWorkflowRecords,
    #     provide_context=True,
    # )

    getDrowToken = PythonOperator(
        task_id="getDrowToken",
        python_callable=getDrowToken,
        provide_context=True,
        # op_kwargs={"name": "Dylan"}
    )
    

    # create_table = PostgresOperator(
    #     sql = create_table_sql_query,
    #     task_id = "create_table_task",
    #     postgres_conn_id = "postgres_rds",
    # )

    # insert_data = PostgresOperator(
    #     sql = insert_data_sql_query,
    #     task_id = "insertData_sql_query_task",
    #     postgres_conn_id = "postgres_rds",
    # )

# getDrowToken >> pipelineProcess >> getWorkflowRecords
getDrowToken >> pipelineProcess