feat: Rebuild complete CSV import system for legacy data migration

PROBLEM SOLVED:
- Completely removed broken import functionality
- Built new robust, modular CSV import system from scratch
- Provides reliable data migration path for legacy .sc files

NEW IMPORT SYSTEM FEATURES:
✅ Modular CSV parsers for all 5 tables (ROLODEX, PHONE, FILES, LEDGER, QDROS)
✅ RESTful API endpoints with background processing (/api/admin/import/*)
✅ Admin web interface at /admin/import for file uploads
✅ Comprehensive validation and error handling
✅ Real-time progress tracking and status monitoring
✅ Detailed logging with import session tracking
✅ Transaction rollback on failures
✅ Batch import with dependency ordering
✅ Foreign key validation and duplicate detection

TECHNICAL IMPLEMENTATION:
- Clean /app/import_export/ module structure with base classes
- Enhanced logging system with import-specific logs
- Background task processing with FastAPI BackgroundTasks
- Auto-detection of CSV delimiters and encoding
- Field validation with proper data type conversion
- Admin authentication integration
- Console logging for debugging support

IMPORT WORKFLOW:
1. Admin selects table type and uploads CSV file
2. System validates headers and data structure
3. Background processing with real-time status updates
4. Detailed error reporting and success metrics
5. Import logs stored in logs/imports/ directory

SUPPORTED TABLES:
- ROLODEX (contacts/people) - 19 fields, requires: id, last
- PHONE (phone numbers) - 3 fields, requires: rolodex_id, phone
- FILES (case files) - 29 fields, requires: file_no, id, empl_num, file_type, opened, status, rate_per_hour
- LEDGER (transactions) - 12 fields, requires: file_no, date, t_code, t_type, empl_num, amount
- QDROS (documents) - 31 fields, requires: file_no

REMOVED FILES:
- app/api/unified_import_api.py
- app/services/unified_import.py
- app/api/flexible.py
- app/models/flexible.py
- templates/unified_import.html
- templates/flexible.html
- static/js/flexible.js
- All legacy import routes and references

TESTING COMPLETED:
✅ Schema validation for all table types
✅ CSV header validation
✅ Single file import functionality
✅ Multi-table dependency validation
✅ Error handling and logging
✅ API endpoint integration

READY FOR PRODUCTION: System tested and validated with sample data.
Administrators can now reliably import CSV files converted from legacy .sc files.

🤖 Generated with [Claude Code](https://claude.ai/code)

Co-Authored-By: Claude <noreply@anthropic.com>

app/import_export/rolodex_importer.py

@@ -0,0 +1,93 @@
+"""
+ROLODEX CSV Importer
+"""
+from typing import Dict, List, Any
+from datetime import date
+from sqlalchemy.orm import Session
+
+from .base import BaseCSVImporter, ImportValidationError
+from app.models.rolodex import Rolodex
+
+
+class RolodexCSVImporter(BaseCSVImporter):
+    """CSV importer for ROLODEX table"""
+    
+    @property
+    def table_name(self) -> str:
+        return "rolodex"
+        
+    @property
+    def required_fields(self) -> List[str]:
+        return ["id", "last"]  # Only ID and last name are required
+        
+    @property
+    def field_mapping(self) -> Dict[str, str]:
+        """Map CSV headers to database field names"""
+        return {
+            "id": "id",
+            "last": "last",
+            "first": "first", 
+            "middle": "middle",
+            "prefix": "prefix",
+            "suffix": "suffix",
+            "title": "title",
+            "group": "group",
+            "a1": "a1",
+            "a2": "a2", 
+            "a3": "a3",
+            "city": "city",
+            "abrev": "abrev",
+            "zip": "zip",
+            "email": "email",
+            "dob": "dob",
+            "ss_number": "ss_number",
+            "legal_status": "legal_status",
+            "memo": "memo"
+        }
+        
+    def create_model_instance(self, row_data: Dict[str, Any]) -> Rolodex:
+        """Create a Rolodex instance from processed row data"""
+        
+        # Validate required fields
+        if not row_data.get("id"):
+            raise ImportValidationError("ID is required")
+        if not row_data.get("last"):
+            raise ImportValidationError("Last name is required")
+            
+        # Check for duplicate ID
+        existing = self.db_session.query(Rolodex).filter_by(id=row_data["id"]).first()
+        if existing:
+            raise ImportValidationError(f"Rolodex ID '{row_data['id']}' already exists")
+            
+        # Parse date of birth
+        dob = None
+        if row_data.get("dob"):
+            try:
+                dob = self.parse_date(row_data["dob"])
+            except ValueError as e:
+                raise ImportValidationError(f"Invalid date of birth: {e}")
+                
+        # Create instance with field length validation
+        rolodex = Rolodex(
+            id=self.normalize_string(row_data["id"], 80),
+            last=self.normalize_string(row_data["last"], 80),
+            first=self.normalize_string(row_data.get("first", ""), 45),
+            middle=self.normalize_string(row_data.get("middle", ""), 45),
+            prefix=self.normalize_string(row_data.get("prefix", ""), 45),
+            suffix=self.normalize_string(row_data.get("suffix", ""), 45),
+            title=self.normalize_string(row_data.get("title", ""), 45),
+            group=self.normalize_string(row_data.get("group", ""), 45),
+            a1=self.normalize_string(row_data.get("a1", ""), 45),
+            a2=self.normalize_string(row_data.get("a2", ""), 45),
+            a3=self.normalize_string(row_data.get("a3", ""), 45),
+            city=self.normalize_string(row_data.get("city", ""), 80),
+            abrev=self.normalize_string(row_data.get("abrev", ""), 45),
+            zip=self.normalize_string(row_data.get("zip", ""), 45),
+            email=self.normalize_string(row_data.get("email", ""), 100),
+            dob=dob,
+            ss_number=self.normalize_string(row_data.get("ss_number", ""), 20),
+            legal_status=self.normalize_string(row_data.get("legal_status", ""), 45),
+            memo=row_data.get("memo", "")  # Text field, no length limit
+        )
+        
+        return rolodex