Implement database-level deduplication using ROW_NUMBER() window function
to remove duplicate material records based on MaterialCode field.
Changes:
- Add query_all_distinct_by_material_code() to deduplicate full table results
- Add query_by_source_numbers_distinct() to deduplicate filtered results
- Update validator to use new deduplication methods with statistics logging
- Retention strategy: keep earliest record (CreateDate ASC, SequenceNumber ASC)
Benefits:
- Reduces data redundancy from 54,358 to 1,679 records (96.91% reduction)
- Improves validation result readability by removing duplicate materials
- Maintains backward compatibility with existing query methods
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
Fix parameter mismatch error when saving discrete material plan data.
Changed placeholder count from 28 to 29 to match the 29 columns in the
INSERT statement (Factory through BOMVersion).
Resolves error: "The SQL contains 28 parameter markers, but 29 parameters were supplied"
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
This commit enhances the material validation functionality to support
database-driven workflows alongside the existing Excel-based approach.
## New Features
### DAO Layer
- Add ProductionContractDataDAO for querying production contract data
- Add MaterialsToBeDeletedDAO with full CRUD operations
- Enhance DiscreteMaterialPlanDAO with query_all(), query_by_source_numbers(),
and get_unique_material_names() methods
### Validation Modes
Support for 4 validation modes in MaterialValidationTab:
1. Database Full - Query all materials from DiscreteMaterialPlanData
2. Database Filtered - Query by ProductionID.txt file
3. Excel Existing - Validate from existing Excel file
4. Excel Full - Complete workflow with ERP extraction
### Configuration
- Add ValidationConfig dataclass with data_source, batch_size, match_mode,
enable_crud_operations, and default_manager fields
- Update ConfigLoader to support validation configuration
- Add validation settings section in SettingsTab GUI
### Query Chain
Implementation of full query chain:
ProductionID.txt (总排号) → productionContractData (生产订单号) →
DiscreteMaterialPlanData (SourceNumber) → MaterialName →
MaterialsToBeDeleted comparison
## Backward Compatibility
All existing Excel-based validation methods remain unchanged, ensuring
no breaking changes for existing workflows.
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
Add optional database persistence feature that automatically saves extracted
discrete material plan data to SQL Server. Users can enable this feature in
the settings tab.
Changes:
- Add enable_db_persistence flag to ExtractionConfig (default: disabled)
- Create DiscreteMaterialPlanDAO for database operations with REPLACE pattern
- Update progress tracking to include database persistence stage (90-100%)
- Add database persistence checkbox in settings UI
- Remove verbose logging checkbox from data extraction UI (config-only now)
- Update extraction workflow to save merged DataFrame to database
Progress weights adjusted:
- download: 65% -> 60%
- database: 10% (new stage)
- Other stages adjusted accordingly
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
