Skip to main content

Module 6: Transformations and Field Mapping

Learn how to transform data during replication - field mapping, calculations, filtering, and data enrichment.

1. Transformation Overview​

SLT supports three transformation types:

  • Field-level mapping (rename, reorder)
  • Value transformations (calculations, conversions)
  • Row-level filtering (conditional replication)

Transformation Architecture​

graph LR
    A[Source Data] --> B[Read]
    B --> C{Transformation<br/>Rules}
    C --> D[Field Mapping]
    C --> E[Calculations]
    C --> F[Filtering]
    D --> G[Target Data]
    E --> G
    F --> G

2. Field Mapping​

Simple Field Mapping​

Transaction: LTRC → Transformations

Source Table: MARA
Target Table: SLTREPL.MARA

Field Mappings:
MATNR     → MATERIAL_NUMBER  (rename)
MTART     → MATERIAL_TYPE
MATKL     → MATERIAL_GROUP
MEINS     → BASE_UOM
(skip NTGEW, BRGEW)  (exclude fields)

Mapping Types​

TypeExampleUse Case
DirectMATNR → MATNRNo change
RenameMATNR → MAT_NOField name change
ConstantNULL → 'DEFAULT'Add default values
ConcatenateVBELN+POSNR → ORDER_ITEMCombine fields
SplitKUNNR → CUST_ID, CUST_TYPEDivide field

3. Value Transformations​

Built-in Functions​

Transformation Expression Examples:

1. Date Conversion:
   ERDAT → TO_DATE(ERDAT, 'YYYYMMDD')

2. Unit Conversion:
   NTGEW → NTGEW * 1000  (KG to G)

3. Currency Conversion:
   NETWR → NETWR * EXCHANGE_RATE

4. String Manipulation:
   MATNR → LTRIM(MATNR, '0')  (remove leading zeros)

5. Conditional Logic:
   IF MTART = 'FERT' THEN 'FINISHED' ELSE 'OTHER'

Custom ABAP Routines​

" Create custom transformation routine
FUNCTION Z_TRANSFORM_MATERIAL.
  IMPORTING
    iv_matnr TYPE matnr
  EXPORTING
    ev_material_id TYPE string.
    
  " Remove leading zeros and add prefix
  ev_material_id = 'MAT-' && |{ iv_matnr ALPHA = OUT }|.
ENDFUNCTION.

4. Filtering Rules​

Row-Level Filtering​

-- Only replicate materials created this year
WHERE ERDAT >= '20260101'

-- Only specific material types
WHERE MTART IN ('FERT', 'HALB', 'ROH')

-- Complex conditions
WHERE (MTART = 'FERT' OR MTART = 'HALB')
  AND LVORM = ''
  AND ERDAT >= '20240101'

Performance Impact​

No Filter:    1,000,000 rows → 100% replicated
Date Filter:    500,000 rows →  50% replicated (50% savings)
Type Filter:    200,000 rows →  20% replicated (80% savings)

5. Configuration Steps​

Step 1: Access Transformation Designer​

Transaction: LTRC
Select MT_ID: ECC_HANA_01
Tab: Tables → Select MARA
Button: [Configure Transformations]

Step 2: Define Field Mappings​

Source Field → Target Field → Transformation

MATNR        → MATERIAL_NO   → LTRIM(MATNR, '0')
MTART        → MAT_TYPE      → <direct>
MATKL        → MAT_GROUP     → <direct>
MEINS        → UOM           → <direct>
CREATED_BY   → <NEW>         → CURRENT_USER (constant)

Step 3: Test Transformations​

-- Test with sample data
Test Input:
  MATNR: '000000000012345'
  MTART: 'FERT'
  
Expected Output:
  MATERIAL_NO: '12345'
  MAT_TYPE: 'FERT'
  CREATED_BY: 'SLTUSER'
  
[Execute Test] → ✅ Passed

6. Advanced Scenarios​

Scenario 1: Data Enrichment​

Enrich customer data with region lookup:

Source: KNA1 (Customer)
Lookup: T005 (Countries) 
Target: Enriched customer with region

Transformation:
1. Read KNA1.LAND1 (country)
2. Lookup T005.REGION for LAND1
3. Add REGION to target record

Scenario 2: Multi-Table Join​

-- Combine order header + item data
SELECT 
  VBAK.VBELN,
  VBAK.KUNNR,
  VBAP.POSNR,
  VBAP.MATNR,
  VBAP.KWMENG
FROM VBAK
INNER JOIN VBAP ON VBAK.VBELN = VBAP.VBELN
WHERE VBAK.ERDAT >= '20260101'

Scenario 3: Data Masking (Privacy)​

Mask sensitive data:

Original: KUNNR = 'CUST-123456'
Masked:   KUNNR = 'CUST-XXXXXX'

Original: EMAIL = 'john.doe@company.com'
Masked:   EMAIL = 'xxxx.xxx@company.com'

7. Performance Considerations​

Transformation Impact​

ComplexityPerformance ImpactLatency Increase
None0%+0 ms
Simple (rename)1-2%+10 ms
Calculations5-10%+50 ms
Lookups20-30%+200 ms
Complex ABAP50-100%+500 ms

Optimization Tips​

✅ Push transformations to target (HANA views) ✅ Minimize complex calculations ✅ Cache lookup tables ✅ Use parallel processing for heavy transforms ✅ Monitor transformation time per table

8. Troubleshooting​

Common Issues​

Issue 1: Transformation Syntax Error

Error: Invalid expression 'MATNR + WERKS'
Solution: Use concatenation: MATNR || WERKS

Issue 2: Data Type Mismatch

Error: Cannot convert CHAR to INT
Solution: Add explicit cast: TO_INTEGER(MATNR)

Issue 3: Lookup Table Not Found

Error: Table T005 not replicated
Solution: Add T005 to MT_ID or use alternative

Summary​

✅ Field mapping and renaming
✅ Value transformations and calculations
✅ Row-level filtering
✅ Custom ABAP routines
✅ Advanced scenarios (enrichment, joins, masking)
✅ Performance optimization
✅ Troubleshooting transformations

Next: Module 7 - Monitoring and Logging