Mẫu dữ liệu cho quy trình mua hàng đến thanh toán: yêu cầu mua hàng

REPORT CONFIGURATION for SAP Ariba Analytical Reporting

REPORT TYPE: Analytical Report

PRIMARY DATA SOURCE (FACT):
*   [Requisition Fact]

JOINED DATA SOURCES (FACTS):
*   [Approval History Fact] on Requisition ID
*   [Purchase Order Fact] on Requisition ID
*   [Requisition Line Item Fact] on Requisition ID

REPORT FIELDS (COLUMNS TO EXPORT for subsequent transformation):

// Case and General Attributes
*   Requisition.Requisition ID (as PurchaseRequisitionId)
*   Requisition.Status (as RequisitionStatus)
*   Requisition.Requester.Department (as RequesterDepartment)
*   Requisition.Total Cost.Amount (as TotalRequisitionAmount)
*   Requisition.Line Items.Category (as ItemCategory)
*   Requisition.Is Urgent (as UrgencyLevel)
*   Requisition.Approval Flow Name (as ApprovalWorkflowPath)

// Fields for 'Requisition Created'
*   Requisition.Created Date (Timestamp)
*   Requisition.Requester.User ID (User)

// Fields for 'Requisition Submitted'
*   Requisition.Submitted Date (Timestamp)
*   Requisition.Submitter.User ID (User)

// Fields for 'Requisition Amended'
*   Requisition.Last Modified Date (Timestamp) - Note: Requires logic to isolate true amendment events
*   Requisition.Last Modified By.User ID (User)

// Fields for 'Approval Step' activities
*   Approval History.Approval Request Date (Timestamp for 'Approval Step Started')
*   Approval History.Action Date (Timestamp for 'Approval Step Approved'/'Approval Step Rejected')
*   Approval History.Action (Value: 'Approve'/'Deny')
*   Approval History.Approver.User ID (User)
*   Approval History.Comments (for RejectionReason)

// Fields for final status activities
*   Requisition.Approved Date (Timestamp for 'Requisition Approved')
*   Requisition.Denied Date (Timestamp for 'Requisition Denied')
*   Requisition.Withdrawn Date (Timestamp for 'Requisition Withdrawn')
*   Requisition.Closed Date (Timestamp for 'Requisition Closed')

// Fields for downstream activities
*   Requisition.Sourcing Status Date (Timestamp for 'Requisition Sent To Sourcing')
*   Purchase Order.Created Date (Timestamp for 'Purchase Order Created')
*   Purchase Order.Creator.User ID (User for 'Purchase Order Created')
*   Requisition Line Item.Ordered Date (Timestamp for 'Requisition Line Item Ordered')

FILTERS:
*   Requisition.Created Date BETWEEN [Start Date] AND [End Date]
*   Company Code IN ([Your Company Codes])

import requests
import json
import pandas as pd
import time
import zipfile
import io

class=class="hl-string">"hl-comment"># --- Configuration --- 
class=class="hl-string">"hl-comment"># Replace with your SAP Ariba realm, datacenter-specific URL, and credentials
ARIBA_REALM = class="hl-string">"[Your Ariba Realm]"
BASE_URL = fclass="hl-string">"https://{ARIBA_REALM}.api.ariba.com/api/operationalreporting/v2/"
TOKEN_URL = class="hl-string">"https://api.ariba.com/v2/oauth/token"
API_KEY = class="hl-string">"[Your Application Key]"
BASE64_CREDS = class="hl-string">"[Your Base64-encoded ClientID:ClientSecret]"

class=class="hl-string">"hl-comment"># --- Report Parameters ---
class=class="hl-string">"hl-comment"># Set the date range for the extraction
START_DATE = class="hl-string">"2023-01-01T00:00:00Z"
END_DATE = class="hl-string">"2023-06-30T23:59:59Z"

class=class="hl-string">"hl-comment"># The main function to orchestrate the extraction process
def main():
    print(class="hl-string">"1. Acquiring access token...")
    access_token = get_access_token()
    if not access_token:
        return

    print(class="hl-string">"2. Creating asynchronous report job...")
    job_id = create_report_job(access_token)
    if not job_id:
        return

    print(fclass="hl-string">"Job created with ID: {job_id}. Polling for completion...")
    status = poll_job_status(access_token, job_id)

    if status == class="hl-string">'COMPLETED':
        print(class="hl-string">"3. Job completed. Downloading report data...")
        raw_data = download_report(access_token, job_id)
        if raw_data:
            print(class="hl-string">"4. Processing raw data into event log...")
            event_log = process_data(raw_data)
            output_filename = class="hl-string">'ariba_requisition_event_log.csv'
            event_log.to_csv(output_filename, index=False)
            print(fclass="hl-string">"5. Success! Event log saved to {output_filename}")
    else:
        print(fclass="hl-string">"Job failed or timed out with status: {status}")

def get_access_token():
    headers = {
        class="hl-string">'Authorization': fclass="hl-string">'Basic {BASE64_CREDS}',
        class="hl-string">'Content-Type': class="hl-string">'application/x-www-form-urlencoded',
    }
    data = {class="hl-string">'grant_type': class="hl-string">'client_credentials'}
    try:
        response = requests.post(TOKEN_URL, headers=headers, data=data)
        response.raise_for_status()
        return response.json().get(class="hl-string">'access_token')
    except requests.exceptions.RequestException as e:
        print(fclass="hl-string">"Error getting access token: {e}")
        return None

def create_report_job(token):
    headers = {
        class="hl-string">'Authorization': fclass="hl-string">'Bearer {token}',
        class="hl-string">'apiKey': API_KEY,
        class="hl-string">'Content-Type': class="hl-string">'application/json'
    }
    report_payload = {
        class="hl-string">"name": class="hl-string">"ProcessMind Requisition Extraction",
        class="hl-string">"viewTemplateName": class="hl-string">"RequisitionFactSystemView",
        class="hl-string">"outputFormat": class="hl-string">"CSV",
        class="hl-string">"filters": {
            class="hl-string">"creationDate": {
                class="hl-string">"from": START_DATE,
                class="hl-string">"to": END_DATE
            }
        },
        class="hl-string">"select": [
            class="hl-string">"RequisitionId", class="hl-string">"StatusString", class="hl-string">"SubmitDate", class="hl-string">"ApprovedState.TimeUpdated",
            class="hl-string">"DeniedState.TimeUpdated", class="hl-string">"SourcingState.TimeUpdated", class="hl-string">"ClosedState.TimeUpdated",
            class="hl-string">"WithdrawnState.TimeUpdated", class="hl-string">"Requester.Name", class="hl-string">"Requester.Department",
            class="hl-string">"TotalCost.Amount", class="hl-string">"LineItems.Category.Name", class="hl-string">"IsUrgent", class="hl-string">"ApprovalFlow.Name",
            class="hl-string">"RequisitionVersion", class="hl-string">"TimeCreated", class="hl-string">"LineItems.OrderedState.TimeUpdated", class="hl-string">"POInfo.POId"
        ]
    }
    try:
        response = requests.post(fclass="hl-string">"{BASE_URL}jobs", headers=headers, json=report_payload)
        response.raise_for_status()
        return response.json().get(class="hl-string">'jobId')
    except requests.exceptions.RequestException as e:
        print(fclass="hl-string">"Error creating report job: {e}")
        return None

def poll_job_status(token, job_id):
    headers = {class="hl-string">'Authorization': fclass="hl-string">'Bearer {token}', class="hl-string">'apiKey': API_KEY}
    while True:
        try:
            response = requests.get(fclass="hl-string">"{BASE_URL}jobs('{job_id}')", headers=headers)
            response.raise_for_status()
            status = response.json().get(class="hl-string">'status')
            print(fclass="hl-string">"Current job status: {status}")
            if status in [class="hl-string">'COMPLETED', class="hl-string">'FAILED', class="hl-string">'CANCELLED']:
                return status
            time.sleep(60) class=class="hl-string">"hl-comment"># Wait 60 seconds before polling again
        except requests.exceptions.RequestException as e:
            print(fclass="hl-string">"Error polling job status: {e}")
            return class="hl-string">'FAILED'

def download_report(token, job_id):
    headers = {class="hl-string">'Authorization': fclass="hl-string">'Bearer {token}', class="hl-string">'apiKey': API_KEY}
    try:
        response = requests.get(fclass="hl-string">"{BASE_URL}jobs('{job_id}')/result", headers=headers, stream=True)
        response.raise_for_status()
        return response.content
    except requests.exceptions.RequestException as e:
        print(fclass="hl-string">"Error downloading report: {e}")
        return None

def process_data(zip_content):
    try:
        with zipfile.ZipFile(io.BytesIO(zip_content)) as z:
            class=class="hl-string">"hl-comment"># Assuming the first file in the zip is the CSV we need
            csv_filename = z.namelist()[0]
            with z.open(csv_filename) as f:
                df = pd.read_csv(f)
    except Exception as e:
        print(fclass="hl-string">"Error reading zipped CSV: {e}")
        return pd.DataFrame()

    class=class="hl-string">"hl-comment"># Rename columns for clarity
    df.rename(columns={
        class="hl-string">'RequisitionId': class="hl-string">'PurchaseRequisitionId',
        class="hl-string">'TimeCreated': class="hl-string">'Timestamp_Created',
        class="hl-string">'SubmitDate': class="hl-string">'Timestamp_Submitted',
        class="hl-string">'ApprovedState.TimeUpdated': class="hl-string">'Timestamp_Approved',
        class="hl-string">'DeniedState.TimeUpdated': class="hl-string">'Timestamp_Denied',
        class="hl-string">'SourcingState.TimeUpdated': class="hl-string">'Timestamp_SentToSourcing',
        class="hl-string">'ClosedState.TimeUpdated': class="hl-string">'Timestamp_Closed',
        class="hl-string">'WithdrawnState.TimeUpdated': class="hl-string">'Timestamp_Withdrawn',
        class="hl-string">'LineItems.OrderedState.TimeUpdated': class="hl-string">'Timestamp_LineItemOrdered',
        class="hl-string">'POInfo.POId': class="hl-string">'POId',
        class="hl-string">'Requester.Name': class="hl-string">'EventUser',
        class="hl-string">'Requester.Department': class="hl-string">'RequesterDepartment',
        class="hl-string">'TotalCost.Amount': class="hl-string">'TotalRequisitionAmount',
        class="hl-string">'LineItems.Category.Name': class="hl-string">'ItemCategory',
        class="hl-string">'IsUrgent': class="hl-string">'UrgencyLevel',
        class="hl-string">'ApprovalFlow.Name': class="hl-string">'ApprovalWorkflowPath',
        class="hl-string">'StatusString': class="hl-string">'RequisitionStatus'
    }, inplace=True)
    
    class=class="hl-string">"hl-comment"># Create a base dataframe for melting
    id_vars = [
        class="hl-string">'PurchaseRequisitionId', class="hl-string">'EventUser', class="hl-string">'RequesterDepartment', class="hl-string">'TotalRequisitionAmount',
        class="hl-string">'ItemCategory', class="hl-string">'UrgencyLevel', class="hl-string">'ApprovalWorkflowPath', class="hl-string">'RequisitionStatus', class="hl-string">'RequisitionVersion', class="hl-string">'POId'
    ]
    
    timestamp_cols = [col for col in df.columns if col.startswith(class="hl-string">'Timestamp_')]
    
    class=class="hl-string">"hl-comment"># Melt the dataframe to turn timestamp columns into rows
    event_log = pd.melt(df, id_vars=id_vars, value_vars=timestamp_cols, var_name=class="hl-string">'ActivityName', value_name=class="hl-string">'EventTimestamp')
    
    class=class="hl-string">"hl-comment"># Clean up and format
    event_log.dropna(subset=[class="hl-string">'EventTimestamp'], inplace=True)
    event_log[class="hl-string">'ActivityName'] = event_log[class="hl-string">'ActivityName'].str.replace(class="hl-string">'Timestamp_', class="hl-string">'Requisition ')
    event_log[class="hl-string">'EventTimestamp'] = pd.to_datetime(event_log[class="hl-string">'EventTimestamp'])

    class=class="hl-string">"hl-comment"># --- Handle specific/derived activities ---
    class=class="hl-string">"hl-comment"># Requisition Amended
    amended_df = df[df[class="hl-string">'RequisitionVersion'] > 1].copy()
    if not amended_df.empty:
        amended_df[class="hl-string">'ActivityName'] = class="hl-string">'Requisition Amended'
        amended_df[class="hl-string">'EventTimestamp'] = pd.to_datetime(amended_df[class="hl-string">'Timestamp_Submitted'])
        event_log = pd.concat([event_log, amended_df[event_log.columns]], ignore_index=True)

    class=class="hl-string">"hl-comment"># Purchase Order Created (use the timestamp of the first line item ordered as a proxy)
    po_created_df = df[df[class="hl-string">'POId'].notna() & df[class="hl-string">'Timestamp_LineItemOrdered'].notna()].copy()
    if not po_created_df.empty:
        po_created_df[class="hl-string">'ActivityName'] = class="hl-string">'Purchase Order Created'
        po_created_df[class="hl-string">'EventTimestamp'] = pd.to_datetime(po_created_df[class="hl-string">'Timestamp_LineItemOrdered'])
        event_log = pd.concat([event_log, po_created_df[event_log.columns]], ignore_index=True)
    
    class=class="hl-string">"hl-comment"># Placeholder for Approval Steps - requires parsing ApprovalRecords not included in this basic script
    class=class="hl-string">"hl-comment"># For now, we only have the final class="hl-string">'Approved' status. This logic can be expanded.
    class=class="hl-string">"hl-comment"># Creating dummy class="hl-string">'Approval Step Started/Approved' for demonstration.
    approval_df = df[df[class="hl-string">'Timestamp_Approved'].notna()].copy()
    if not approval_df.empty:
        approval_started = approval_df.copy()
        approval_started[class="hl-string">'ActivityName'] = class="hl-string">'Approval Step Started'
        class=class="hl-string">"hl-comment"># Approximate start time as submit time
        approval_started[class="hl-string">'EventTimestamp'] = pd.to_datetime(approval_started[class="hl-string">'Timestamp_Submitted'])
        
        approval_approved = approval_df.copy()
        approval_approved[class="hl-string">'ActivityName'] = class="hl-string">'Approval Step Approved'
        approval_approved[class="hl-string">'EventTimestamp'] = pd.to_datetime(approval_approved[class="hl-string">'Timestamp_Approved'])
        approval_approved[class="hl-string">'RejectionReason'] = None class=class="hl-string">"hl-comment"># Add column for consistency
        
        event_log = pd.concat([event_log, approval_started[event_log.columns], approval_approved[event_log.columns.tolist() + [class="hl-string">'RejectionReason']]], ignore_index=True)

    class=class="hl-string">"hl-comment"># Add placeholder for RejectionReason, as it's not directly in this view
    if class="hl-string">'RejectionReason' not in event_log.columns:
        event_log[class="hl-string">'RejectionReason'] = None

    class=class="hl-string">"hl-comment"># Final sorting and cleanup
    event_log = event_log.sort_values(by=[class="hl-string">'PurchaseRequisitionId', class="hl-string">'EventTimestamp']).reset_index(drop=True)
    
    class=class="hl-string">"hl-comment"># Rename class="hl-string">'Requisition LineItemOrdered' to class="hl-string">'Requisition Line Item Ordered'
    event_log[class="hl-string">'ActivityName'] = event_log[class="hl-string">'ActivityName'].replace({class="hl-string">'Requisition LineItemOrdered': class="hl-string">'Requisition Line Item Ordered'})

    class=class="hl-string">"hl-comment"># Final column selection
    final_columns = [
        class="hl-string">'PurchaseRequisitionId', class="hl-string">'ActivityName', class="hl-string">'EventTimestamp', class="hl-string">'EventUser',
        class="hl-string">'RequesterDepartment', class="hl-string">'TotalRequisitionAmount', class="hl-string">'ItemCategory', class="hl-string">'UrgencyLevel',
        class="hl-string">'ApprovalWorkflowPath', class="hl-string">'RejectionReason', class="hl-string">'RequisitionStatus'
    ]
    event_log = event_log[final_columns]

    return event_log

if __name__ == class="hl-string">"__main__":
    main()