Fixes for not found records

pkg/common/adapters/database/bun.go

@@ -691,6 +691,11 @@ func (b *BunSelectQuery) Order(order string) common.SelectQuery {
 	return b
 }
 
+func (b *BunSelectQuery) OrderExpr(order string, args ...interface{}) common.SelectQuery {
+	b.query = b.query.OrderExpr(order, args...)
+	return b
+}
+
 func (b *BunSelectQuery) Limit(n int) common.SelectQuery {
 	b.query = b.query.Limit(n)
 	return b

pkg/common/adapters/database/gorm.go

@@ -386,6 +386,12 @@ func (g *GormSelectQuery) Order(order string) common.SelectQuery {
 	return g
 }
 
+func (g *GormSelectQuery) OrderExpr(order string, args ...interface{}) common.SelectQuery {
+	// GORM's Order can handle expressions directly
+	g.db = g.db.Order(gorm.Expr(order, args...))
+	return g
+}
+
 func (g *GormSelectQuery) Limit(n int) common.SelectQuery {
 	g.db = g.db.Limit(n)
 	return g

pkg/common/adapters/database/pgsql.go

@@ -281,6 +281,13 @@ func (p *PgSQLSelectQuery) Order(order string) common.SelectQuery {
 	return p
 }
 
+func (p *PgSQLSelectQuery) OrderExpr(order string, args ...interface{}) common.SelectQuery {
+	// For PgSQL, expressions are passed directly without quoting
+	// If there are args, we would need to format them, but for now just append the expression
+	p.orderBy = append(p.orderBy, order)
+	return p
+}
+
 func (p *PgSQLSelectQuery) Limit(n int) common.SelectQuery {
 	p.limit = n
 	return p

pkg/common/interfaces.go

@@ -46,6 +46,7 @@ type SelectQuery interface {
 	PreloadRelation(relation string, apply ...func(SelectQuery) SelectQuery) SelectQuery
 	JoinRelation(relation string, apply ...func(SelectQuery) SelectQuery) SelectQuery
 	Order(order string) SelectQuery
+	OrderExpr(order string, args ...interface{}) SelectQuery
 	Limit(n int) SelectQuery
 	Offset(n int) SelectQuery
 	Group(group string) SelectQuery

pkg/common/sql_helpers.go

@@ -2,6 +2,7 @@ package common
 
 import (
 	"fmt"
+	"regexp"
 	"strings"
 
 	"github.com/bitechdev/ResolveSpec/pkg/logger"
@@ -207,6 +208,20 @@ func SanitizeWhereClause(where string, tableName string, options ...*RequestOpti
 					}
 				}
 			}
+		} else if tableName != "" && !hasTablePrefix(condToCheck) {
+			// If tableName is provided and the condition DOESN'T have a table prefix,
+			// qualify unambiguous column references to prevent "ambiguous column" errors
+			// when there are multiple joins on the same table (e.g., recursive preloads)
+			columnName := extractUnqualifiedColumnName(condToCheck)
+			if columnName != "" && (validColumns == nil || isValidColumn(columnName, validColumns)) {
+				// Qualify the column with the table name
+				// Be careful to only replace the column name, not other occurrences of the string
+				oldRef := columnName
+				newRef := tableName + "." + columnName
+				// Use word boundary matching to avoid replacing partial matches
+				cond = qualifyColumnInCondition(cond, oldRef, newRef)
+				logger.Debug("Qualified unqualified column in condition: '%s' added table prefix '%s'", oldRef, tableName)
+			}
 		}
 
 		validConditions = append(validConditions, cond)
@@ -483,6 +498,86 @@ func extractTableAndColumn(cond string) (table string, column string) {
 	return "", ""
 }
 
+// extractUnqualifiedColumnName extracts the column name from an unqualified condition
+// For example: "rid_parentmastertaskitem is null" returns "rid_parentmastertaskitem"
+// "status = 'active'" returns "status"
+func extractUnqualifiedColumnName(cond string) string {
+	// Common SQL operators
+	operators := []string{" = ", " != ", " <> ", " > ", " >= ", " < ", " <= ", " LIKE ", " like ", " IN ", " in ", " IS ", " is ", " NOT ", " not "}
+
+	// Find the column reference (left side of the operator)
+	minIdx := -1
+	for _, op := range operators {
+		idx := strings.Index(cond, op)
+		if idx > 0 && (minIdx == -1 || idx < minIdx) {
+			minIdx = idx
+		}
+	}
+
+	var columnRef string
+	if minIdx > 0 {
+		columnRef = strings.TrimSpace(cond[:minIdx])
+	} else {
+		// No operator found, might be a single column reference
+		parts := strings.Fields(cond)
+		if len(parts) > 0 {
+			columnRef = parts[0]
+		}
+	}
+
+	if columnRef == "" {
+		return ""
+	}
+
+	// Remove any quotes
+	columnRef = strings.Trim(columnRef, "`\"'")
+
+	// Return empty if it contains a dot (already qualified) or function call
+	if strings.Contains(columnRef, ".") || strings.Contains(columnRef, "(") {
+		return ""
+	}
+
+	return columnRef
+}
+
+// qualifyColumnInCondition replaces an unqualified column name with a qualified one in a condition
+// Uses word boundaries to avoid partial matches
+// For example: qualifyColumnInCondition("rid_item is null", "rid_item", "table.rid_item")
+// returns "table.rid_item is null"
+func qualifyColumnInCondition(cond, oldRef, newRef string) string {
+	// Use word boundary matching with Go's supported regex syntax
+	// \b matches word boundaries
+	escapedOld := regexp.QuoteMeta(oldRef)
+	pattern := `\b` + escapedOld + `\b`
+
+	re, err := regexp.Compile(pattern)
+	if err != nil {
+		// If regex fails, fall back to simple string replacement
+		logger.Debug("Failed to compile regex for column qualification, using simple replace: %v", err)
+		return strings.Replace(cond, oldRef, newRef, 1)
+	}
+
+	// Only replace if the match is not preceded by a dot (to avoid replacing already qualified columns)
+	result := cond
+	matches := re.FindAllStringIndex(cond, -1)
+
+	// Process matches in reverse order to maintain correct indices
+	for i := len(matches) - 1; i >= 0; i-- {
+		match := matches[i]
+		start := match[0]
+
+		// Check if preceded by a dot (already qualified)
+		if start > 0 && cond[start-1] == '.' {
+			continue
+		}
+
+		// Replace this occurrence
+		result = result[:start] + newRef + result[match[1]:]
+	}
+
+	return result
+}
+
 // findOperatorOutsideParentheses finds the first occurrence of an operator outside of parentheses
 // Returns the index of the operator, or -1 if not found or only found inside parentheses
 func findOperatorOutsideParentheses(s string, operator string) int {

pkg/common/sql_helpers_test.go

@@ -33,16 +33,16 @@ func TestSanitizeWhereClause(t *testing.T) {
 			expected:  "",
 		},
 		{
-			name:      "valid condition with parentheses - no prefix added",
+			name:      "valid condition with parentheses - prefix added to prevent ambiguity",
 			where:     "(status = 'active')",
 			tableName: "users",
-			expected:  "status = 'active'",
+			expected:  "users.status = 'active'",
 		},
 		{
-			name:      "mixed trivial and valid conditions - no prefix added",
+			name:      "mixed trivial and valid conditions - prefix added",
 			where:     "true AND status = 'active' AND 1=1",
 			tableName: "users",
-			expected:  "status = 'active'",
+			expected:  "users.status = 'active'",
 		},
 		{
 			name:      "condition with correct table prefix - unchanged",
@@ -63,10 +63,10 @@ func TestSanitizeWhereClause(t *testing.T) {
 			expected:  "users.status = 'active' AND users.age > 18",
 		},
 		{
-			name:      "multiple valid conditions without prefix - no prefix added",
+			name:      "multiple valid conditions without prefix - prefixes added",
 			where:     "status = 'active' AND age > 18",
 			tableName: "users",
-			expected:  "status = 'active' AND age > 18",
+			expected:  "users.status = 'active' AND users.age > 18",
 		},
 		{
 			name:      "no table name provided",
@@ -90,13 +90,13 @@ func TestSanitizeWhereClause(t *testing.T) {
 			name:      "mixed case AND operators",
 			where:     "status = 'active' AND age > 18 and name = 'John'",
 			tableName: "users",
-			expected:  "status = 'active' AND age > 18 AND name = 'John'",
+			expected:  "users.status = 'active' AND users.age > 18 AND users.name = 'John'",
 		},
 		{
 			name:      "subquery with ORDER BY and LIMIT - allowed",
 			where:     "id IN (SELECT id FROM users WHERE status = 'active' ORDER BY created_at DESC LIMIT 10)",
 			tableName: "users",
-			expected:  "id IN (SELECT id FROM users WHERE status = 'active' ORDER BY created_at DESC LIMIT 10)",
+			expected:  "users.id IN (SELECT users.id FROM users WHERE status = 'active' ORDER BY created_at DESC LIMIT 10)",
 		},
 		{
 			name:      "dangerous DELETE keyword - blocked",

pkg/common/validation.go

@@ -237,6 +237,13 @@ func (v *ColumnValidator) FilterRequestOptions(options RequestOptions) RequestOp
 	for _, sort := range options.Sort {
 		if v.IsValidColumn(sort.Column) {
 			validSorts = append(validSorts, sort)
+		} else if strings.HasPrefix(sort.Column, "(") && strings.HasSuffix(sort.Column, ")") {
+			// Allow sort by expression/subquery, but validate for security
+			if IsSafeSortExpression(sort.Column) {
+				validSorts = append(validSorts, sort)
+			} else {
+				logger.Warn("Unsafe sort expression '%s' removed", sort.Column)
+			}
 		} else {
 			logger.Warn("Invalid column in sort '%s' removed", sort.Column)
 		}
@@ -262,6 +269,24 @@ func (v *ColumnValidator) FilterRequestOptions(options RequestOptions) RequestOp
 		}
 		filteredPreload.Filters = validPreloadFilters
 
+		// Filter preload sort columns
+		validPreloadSorts := make([]SortOption, 0, len(preload.Sort))
+		for _, sort := range preload.Sort {
+			if v.IsValidColumn(sort.Column) {
+				validPreloadSorts = append(validPreloadSorts, sort)
+			} else if strings.HasPrefix(sort.Column, "(") && strings.HasSuffix(sort.Column, ")") {
+				// Allow sort by expression/subquery, but validate for security
+				if IsSafeSortExpression(sort.Column) {
+					validPreloadSorts = append(validPreloadSorts, sort)
+				} else {
+					logger.Warn("Unsafe sort expression in preload '%s' removed: '%s'", preload.Relation, sort.Column)
+				}
+			} else {
+				logger.Warn("Invalid column in preload '%s' sort '%s' removed", preload.Relation, sort.Column)
+			}
+		}
+		filteredPreload.Sort = validPreloadSorts
+
 		validPreloads = append(validPreloads, filteredPreload)
 	}
 	filtered.Preload = validPreloads
@@ -269,6 +294,56 @@ func (v *ColumnValidator) FilterRequestOptions(options RequestOptions) RequestOp
 	return filtered
 }
 
+// IsSafeSortExpression validates that a sort expression (enclosed in brackets) is safe
+// and doesn't contain SQL injection attempts or dangerous commands
+func IsSafeSortExpression(expr string) bool {
+	if expr == "" {
+		return false
+	}
+
+	// Expression must be enclosed in brackets
+	expr = strings.TrimSpace(expr)
+	if !strings.HasPrefix(expr, "(") || !strings.HasSuffix(expr, ")") {
+		return false
+	}
+
+	// Remove outer brackets for content validation
+	expr = expr[1 : len(expr)-1]
+	expr = strings.TrimSpace(expr)
+
+	// Convert to lowercase for checking dangerous keywords
+	exprLower := strings.ToLower(expr)
+
+	// Check for dangerous SQL commands that should never be in a sort expression
+	dangerousKeywords := []string{
+		"drop ", "delete ", "insert ", "update ", "alter ", "create ",
+		"truncate ", "exec ", "execute ", "grant ", "revoke ",
+		"into ", "values ", "set ", "shutdown", "xp_",
+	}
+
+	for _, keyword := range dangerousKeywords {
+		if strings.Contains(exprLower, keyword) {
+			logger.Warn("Dangerous SQL keyword '%s' detected in sort expression: %s", keyword, expr)
+			return false
+		}
+	}
+
+	// Check for SQL comment attempts
+	if strings.Contains(expr, "--") || strings.Contains(expr, "/*") || strings.Contains(expr, "*/") {
+		logger.Warn("SQL comment detected in sort expression: %s", expr)
+		return false
+	}
+
+	// Check for semicolon (command separator)
+	if strings.Contains(expr, ";") {
+		logger.Warn("Command separator (;) detected in sort expression: %s", expr)
+		return false
+	}
+
+	// Expression appears safe
+	return true
+}
+
 // GetValidColumns returns a list of all valid column names for debugging purposes
 func (v *ColumnValidator) GetValidColumns() []string {
 	columns := make([]string, 0, len(v.validColumns))

pkg/common/validation_test.go

@@ -361,3 +361,83 @@ func TestFilterRequestOptions(t *testing.T) {
 		t.Errorf("Expected sort column 'id', got %s", filtered.Sort[0].Column)
 	}
 }
+
+func TestIsSafeSortExpression(t *testing.T) {
+	tests := []struct {
+		name       string
+		expression string
+		shouldPass bool
+	}{
+		// Safe expressions
+		{"Valid subquery", "(SELECT MAX(price) FROM products)", true},
+		{"Valid CASE expression", "(CASE WHEN status = 'active' THEN 1 ELSE 0 END)", true},
+		{"Valid aggregate", "(COUNT(*) OVER (PARTITION BY category))", true},
+		{"Valid function", "(COALESCE(discount, 0))", true},
+
+		// Dangerous expressions - SQL injection attempts
+		{"DROP TABLE attempt", "(id); DROP TABLE users; --", false},
+		{"DELETE attempt", "(id WHERE 1=1); DELETE FROM users; --", false},
+		{"INSERT attempt", "(id); INSERT INTO admin VALUES ('hacker'); --", false},
+		{"UPDATE attempt", "(id); UPDATE users SET role='admin'; --", false},
+		{"EXEC attempt", "(id); EXEC sp_executesql 'DROP TABLE users'; --", false},
+		{"XP_ stored proc", "(id); xp_cmdshell 'dir'; --", false},
+
+		// Comment injection
+		{"SQL comment dash", "(id) -- malicious comment", false},
+		{"SQL comment block start", "(id) /* comment", false},
+		{"SQL comment block end", "(id) comment */", false},
+
+		// Semicolon attempts
+		{"Semicolon separator", "(id); SELECT * FROM passwords", false},
+
+		// Empty/invalid
+		{"Empty string", "", false},
+		{"Just brackets", "()", true}, // Empty but technically valid structure
+		{"No brackets", "id", false},  // Must have brackets for expressions
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			result := IsSafeSortExpression(tt.expression)
+			if result != tt.shouldPass {
+				t.Errorf("IsSafeSortExpression(%q) = %v, want %v", tt.expression, result, tt.shouldPass)
+			}
+		})
+	}
+}
+
+func TestFilterRequestOptions_WithSortExpressions(t *testing.T) {
+	model := TestModel{}
+	validator := NewColumnValidator(model)
+
+	options := RequestOptions{
+		Sort: []SortOption{
+			{Column: "id", Direction: "ASC"},                                    // Valid column
+			{Column: "(SELECT MAX(age) FROM users)", Direction: "DESC"},         // Safe expression
+			{Column: "name", Direction: "ASC"},                                  // Valid column
+			{Column: "(id); DROP TABLE users; --", Direction: "DESC"},          // Dangerous expression
+			{Column: "invalid_col", Direction: "ASC"},                           // Invalid column
+			{Column: "(CASE WHEN age > 18 THEN 1 ELSE 0 END)", Direction: "ASC"}, // Safe expression
+		},
+	}
+
+	filtered := validator.FilterRequestOptions(options)
+
+	// Should keep: id, safe expression, name, another safe expression
+	// Should remove: dangerous expression, invalid column
+	expectedCount := 4
+	if len(filtered.Sort) != expectedCount {
+		t.Errorf("Expected %d sort options, got %d", expectedCount, len(filtered.Sort))
+	}
+
+	// Verify the kept options
+	if filtered.Sort[0].Column != "id" {
+		t.Errorf("Expected first sort to be 'id', got '%s'", filtered.Sort[0].Column)
+	}
+	if filtered.Sort[1].Column != "(SELECT MAX(age) FROM users)" {
+		t.Errorf("Expected second sort to be safe expression, got '%s'", filtered.Sort[1].Column)
+	}
+	if filtered.Sort[2].Column != "name" {
+		t.Errorf("Expected third sort to be 'name', got '%s'", filtered.Sort[2].Column)
+	}
+}

pkg/modelregistry/model_registry.go

@@ -6,15 +6,37 @@ import (
 	"sync"
 )
 
+// ModelRules defines the permissions and security settings for a model
+type ModelRules struct {
+	CanRead          bool // Whether the model can be read (GET operations)
+	CanUpdate        bool // Whether the model can be updated (PUT/PATCH operations)
+	CanCreate        bool // Whether the model can be created (POST operations)
+	CanDelete        bool // Whether the model can be deleted (DELETE operations)
+	SecurityDisabled bool // Whether security checks are disabled for this model
+}
+
+// DefaultModelRules returns the default rules for a model (all operations allowed, security enabled)
+func DefaultModelRules() ModelRules {
+	return ModelRules{
+		CanRead:          true,
+		CanUpdate:        true,
+		CanCreate:        true,
+		CanDelete:        true,
+		SecurityDisabled: false,
+	}
+}
+
 // DefaultModelRegistry implements ModelRegistry interface
 type DefaultModelRegistry struct {
 	models map[string]interface{}
+	rules  map[string]ModelRules
 	mutex  sync.RWMutex
 }
 
 // Global default registry instance
 var defaultRegistry = &DefaultModelRegistry{
 	models: make(map[string]interface{}),
+	rules:  make(map[string]ModelRules),
 }
 
 // Global list of registries (searched in order)
@@ -25,6 +47,7 @@ var registriesMutex sync.RWMutex
 func NewModelRegistry() *DefaultModelRegistry {
 	return &DefaultModelRegistry{
 		models: make(map[string]interface{}),
+		rules:  make(map[string]ModelRules),
 	}
 }
 
@@ -98,6 +121,10 @@ func (r *DefaultModelRegistry) RegisterModel(name string, model interface{}) err
 	}
 
 	r.models[name] = model
+	// Initialize with default rules if not already set
+	if _, exists := r.rules[name]; !exists {
+		r.rules[name] = DefaultModelRules()
+	}
 	return nil
 }
 
@@ -135,6 +162,54 @@ func (r *DefaultModelRegistry) GetModelByEntity(schema, entity string) (interfac
 	return r.GetModel(entity)
 }
 
+// SetModelRules sets the rules for a specific model
+func (r *DefaultModelRegistry) SetModelRules(name string, rules ModelRules) error {
+	r.mutex.Lock()
+	defer r.mutex.Unlock()
+
+	// Check if model exists
+	if _, exists := r.models[name]; !exists {
+		return fmt.Errorf("model %s not found", name)
+	}
+
+	r.rules[name] = rules
+	return nil
+}
+
+// GetModelRules retrieves the rules for a specific model
+// Returns default rules if model exists but rules are not set
+func (r *DefaultModelRegistry) GetModelRules(name string) (ModelRules, error) {
+	r.mutex.RLock()
+	defer r.mutex.RUnlock()
+
+	// Check if model exists
+	if _, exists := r.models[name]; !exists {
+		return ModelRules{}, fmt.Errorf("model %s not found", name)
+	}
+
+	// Return rules if set, otherwise return default rules
+	if rules, exists := r.rules[name]; exists {
+		return rules, nil
+	}
+
+	return DefaultModelRules(), nil
+}
+
+// RegisterModelWithRules registers a model with specific rules
+func (r *DefaultModelRegistry) RegisterModelWithRules(name string, model interface{}, rules ModelRules) error {
+	// First register the model
+	if err := r.RegisterModel(name, model); err != nil {
+		return err
+	}
+
+	// Then set the rules (we need to lock again for rules)
+	r.mutex.Lock()
+	defer r.mutex.Unlock()
+	r.rules[name] = rules
+
+	return nil
+}
+
 // Global convenience functions using the default registry
 
 // RegisterModel registers a model with the default global registry
@@ -190,3 +265,34 @@ func GetModels() []interface{} {
 
 	return models
 }
+
+// SetModelRules sets the rules for a specific model in the default registry
+func SetModelRules(name string, rules ModelRules) error {
+	return defaultRegistry.SetModelRules(name, rules)
+}
+
+// GetModelRules retrieves the rules for a specific model from the default registry
+func GetModelRules(name string) (ModelRules, error) {
+	return defaultRegistry.GetModelRules(name)
+}
+
+// GetModelRulesByName retrieves the rules for a model by searching through all registries in order
+// Returns the first match found
+func GetModelRulesByName(name string) (ModelRules, error) {
+	registriesMutex.RLock()
+	defer registriesMutex.RUnlock()
+
+	for _, registry := range registries {
+		if _, err := registry.GetModel(name); err == nil {
+			// Model found in this registry, get its rules
+			return registry.GetModelRules(name)
+		}
+	}
+
+	return ModelRules{}, fmt.Errorf("model %s not found in any registry", name)
+}
+
+// RegisterModelWithRules registers a model with specific rules in the default registry
+func RegisterModelWithRules(model interface{}, name string, rules ModelRules) error {
+	return defaultRegistry.RegisterModelWithRules(name, model, rules)
+}

pkg/reflection/model_utils.go

@@ -1,10 +1,12 @@
 package reflection
 
 import (
+	"encoding/json"
 	"fmt"
 	"reflect"
 	"strconv"
 	"strings"
+	"time"
 
 	"github.com/bitechdev/ResolveSpec/pkg/modelregistry"
 )
@@ -897,6 +899,368 @@ func GetRelationModel(model interface{}, fieldName string) interface{} {
 	return currentModel
 }
 
+// MapToStruct populates a struct from a map while preserving custom types
+// It uses reflection to set struct fields based on map keys, matching by:
+// 1. Bun tag column name
+// 2. Gorm tag column name
+// 3. JSON tag name
+// 4. Field name (case-insensitive)
+// This preserves custom types that implement driver.Valuer like SqlJSONB
+func MapToStruct(dataMap map[string]interface{}, target interface{}) error {
+	if dataMap == nil || target == nil {
+		return fmt.Errorf("dataMap and target cannot be nil")
+	}
+
+	targetValue := reflect.ValueOf(target)
+	if targetValue.Kind() != reflect.Ptr {
+		return fmt.Errorf("target must be a pointer to a struct")
+	}
+
+	targetValue = targetValue.Elem()
+	if targetValue.Kind() != reflect.Struct {
+		return fmt.Errorf("target must be a pointer to a struct")
+	}
+
+	targetType := targetValue.Type()
+
+	// Create a map of column names to field indices for faster lookup
+	columnToField := make(map[string]int)
+	for i := 0; i < targetType.NumField(); i++ {
+		field := targetType.Field(i)
+
+		// Skip unexported fields
+		if !field.IsExported() {
+			continue
+		}
+
+		// Build list of possible column names for this field
+		var columnNames []string
+
+		// 1. Bun tag
+		if bunTag := field.Tag.Get("bun"); bunTag != "" && bunTag != "-" {
+			if colName := ExtractColumnFromBunTag(bunTag); colName != "" {
+				columnNames = append(columnNames, colName)
+			}
+		}
+
+		// 2. Gorm tag
+		if gormTag := field.Tag.Get("gorm"); gormTag != "" && gormTag != "-" {
+			if colName := ExtractColumnFromGormTag(gormTag); colName != "" {
+				columnNames = append(columnNames, colName)
+			}
+		}
+
+		// 3. JSON tag
+		if jsonTag := field.Tag.Get("json"); jsonTag != "" && jsonTag != "-" {
+			parts := strings.Split(jsonTag, ",")
+			if len(parts) > 0 && parts[0] != "" {
+				columnNames = append(columnNames, parts[0])
+			}
+		}
+
+		// 4. Field name variations
+		columnNames = append(columnNames, field.Name)
+		columnNames = append(columnNames, strings.ToLower(field.Name))
+		columnNames = append(columnNames, ToSnakeCase(field.Name))
+
+		// Map all column name variations to this field index
+		for _, colName := range columnNames {
+			columnToField[strings.ToLower(colName)] = i
+		}
+	}
+
+	// Iterate through the map and set struct fields
+	for key, value := range dataMap {
+		// Find the field index for this key
+		fieldIndex, found := columnToField[strings.ToLower(key)]
+		if !found {
+			// Skip keys that don't map to any field
+			continue
+		}
+
+		field := targetValue.Field(fieldIndex)
+		if !field.CanSet() {
+			continue
+		}
+
+		// Set the value, preserving custom types
+		if err := setFieldValue(field, value); err != nil {
+			return fmt.Errorf("failed to set field %s: %w", targetType.Field(fieldIndex).Name, err)
+		}
+	}
+
+	return nil
+}
+
+// setFieldValue sets a reflect.Value from an interface{} value, handling type conversions
+func setFieldValue(field reflect.Value, value interface{}) error {
+	if value == nil {
+		// Set zero value for nil
+		field.Set(reflect.Zero(field.Type()))
+		return nil
+	}
+
+	valueReflect := reflect.ValueOf(value)
+
+	// If types match exactly, just set it
+	if valueReflect.Type().AssignableTo(field.Type()) {
+		field.Set(valueReflect)
+		return nil
+	}
+
+	// Handle pointer fields
+	if field.Kind() == reflect.Ptr {
+		if valueReflect.Kind() != reflect.Ptr {
+			// Create a new pointer and set its value
+			newPtr := reflect.New(field.Type().Elem())
+			if err := setFieldValue(newPtr.Elem(), value); err != nil {
+				return err
+			}
+			field.Set(newPtr)
+			return nil
+		}
+	}
+
+	// Handle conversions for basic types
+	switch field.Kind() {
+	case reflect.String:
+		if str, ok := value.(string); ok {
+			field.SetString(str)
+			return nil
+		}
+	case reflect.Int, reflect.Int8, reflect.Int16, reflect.Int32, reflect.Int64:
+		if num, ok := convertToInt64(value); ok {
+			if field.OverflowInt(num) {
+				return fmt.Errorf("integer overflow")
+			}
+			field.SetInt(num)
+			return nil
+		}
+	case reflect.Uint, reflect.Uint8, reflect.Uint16, reflect.Uint32, reflect.Uint64:
+		if num, ok := convertToUint64(value); ok {
+			if field.OverflowUint(num) {
+				return fmt.Errorf("unsigned integer overflow")
+			}
+			field.SetUint(num)
+			return nil
+		}
+	case reflect.Float32, reflect.Float64:
+		if num, ok := convertToFloat64(value); ok {
+			if field.OverflowFloat(num) {
+				return fmt.Errorf("float overflow")
+			}
+			field.SetFloat(num)
+			return nil
+		}
+	case reflect.Bool:
+		if b, ok := value.(bool); ok {
+			field.SetBool(b)
+			return nil
+		}
+	case reflect.Slice:
+		// Handle []byte specially (for types like SqlJSONB)
+		if field.Type().Elem().Kind() == reflect.Uint8 {
+			switch v := value.(type) {
+			case []byte:
+				field.SetBytes(v)
+				return nil
+			case string:
+				field.SetBytes([]byte(v))
+				return nil
+			case map[string]interface{}, []interface{}:
+				// Marshal complex types to JSON for SqlJSONB fields
+				jsonBytes, err := json.Marshal(v)
+				if err != nil {
+					return fmt.Errorf("failed to marshal value to JSON: %w", err)
+				}
+				field.SetBytes(jsonBytes)
+				return nil
+			}
+		}
+	}
+
+	// Handle struct types (like SqlTimeStamp, SqlDate, SqlTime which wrap SqlNull[time.Time])
+	if field.Kind() == reflect.Struct {
+
+		// Handle datatypes.SqlNull[T] and wrapped types (SqlTimeStamp, SqlDate, SqlTime)
+		// Check if the type has a Scan method (sql.Scanner interface)
+		if field.CanAddr() {
+			scanMethod := field.Addr().MethodByName("Scan")
+			if scanMethod.IsValid() {
+				// Call the Scan method with the value
+				results := scanMethod.Call([]reflect.Value{reflect.ValueOf(value)})
+				if len(results) > 0 {
+					// Check if there was an error
+					if err, ok := results[0].Interface().(error); ok && err != nil {
+						return err
+					}
+					return nil
+				}
+			}
+		}
+
+		// Handle time.Time with ISO string fallback
+		if field.Type() == reflect.TypeOf(time.Time{}) {
+			switch v := value.(type) {
+			case time.Time:
+				field.Set(reflect.ValueOf(v))
+				return nil
+			case string:
+				// Try parsing as ISO 8601 / RFC3339
+				if t, err := time.Parse(time.RFC3339, v); err == nil {
+					field.Set(reflect.ValueOf(t))
+					return nil
+				}
+				// Try other common formats
+				formats := []string{
+					"2006-01-02T15:04:05.000-0700",
+					"2006-01-02T15:04:05.000",
+					"2006-01-02T15:04:05",
+					"2006-01-02 15:04:05",
+					"2006-01-02",
+				}
+				for _, format := range formats {
+					if t, err := time.Parse(format, v); err == nil {
+						field.Set(reflect.ValueOf(t))
+						return nil
+					}
+				}
+				return fmt.Errorf("cannot parse time string: %s", v)
+			}
+		}
+
+		// Fallback: Try to find a "Val" field (for SqlNull types) and set it directly
+		valField := field.FieldByName("Val")
+		if valField.IsValid() && valField.CanSet() {
+			// Also set Valid field to true
+			validField := field.FieldByName("Valid")
+			if validField.IsValid() && validField.CanSet() && validField.Kind() == reflect.Bool {
+				// Set the Val field
+				if err := setFieldValue(valField, value); err != nil {
+					return err
+				}
+				// Set Valid to true
+				validField.SetBool(true)
+				return nil
+			}
+		}
+
+	}
+
+	// If we can convert the type, do it
+	if valueReflect.Type().ConvertibleTo(field.Type()) {
+		field.Set(valueReflect.Convert(field.Type()))
+		return nil
+	}
+
+	return fmt.Errorf("cannot convert %v to %v", valueReflect.Type(), field.Type())
+}
+
+// convertToInt64 attempts to convert various types to int64
+func convertToInt64(value interface{}) (int64, bool) {
+	switch v := value.(type) {
+	case int:
+		return int64(v), true
+	case int8:
+		return int64(v), true
+	case int16:
+		return int64(v), true
+	case int32:
+		return int64(v), true
+	case int64:
+		return v, true
+	case uint:
+		return int64(v), true
+	case uint8:
+		return int64(v), true
+	case uint16:
+		return int64(v), true
+	case uint32:
+		return int64(v), true
+	case uint64:
+		return int64(v), true
+	case float32:
+		return int64(v), true
+	case float64:
+		return int64(v), true
+	case string:
+		if num, err := strconv.ParseInt(v, 10, 64); err == nil {
+			return num, true
+		}
+	}
+	return 0, false
+}
+
+// convertToUint64 attempts to convert various types to uint64
+func convertToUint64(value interface{}) (uint64, bool) {
+	switch v := value.(type) {
+	case int:
+		return uint64(v), true
+	case int8:
+		return uint64(v), true
+	case int16:
+		return uint64(v), true
+	case int32:
+		return uint64(v), true
+	case int64:
+		return uint64(v), true
+	case uint:
+		return uint64(v), true
+	case uint8:
+		return uint64(v), true
+	case uint16:
+		return uint64(v), true
+	case uint32:
+		return uint64(v), true
+	case uint64:
+		return v, true
+	case float32:
+		return uint64(v), true
+	case float64:
+		return uint64(v), true
+	case string:
+		if num, err := strconv.ParseUint(v, 10, 64); err == nil {
+			return num, true
+		}
+	}
+	return 0, false
+}
+
+// convertToFloat64 attempts to convert various types to float64
+func convertToFloat64(value interface{}) (float64, bool) {
+	switch v := value.(type) {
+	case int:
+		return float64(v), true
+	case int8:
+		return float64(v), true
+	case int16:
+		return float64(v), true
+	case int32:
+		return float64(v), true
+	case int64:
+		return float64(v), true
+	case uint:
+		return float64(v), true
+	case uint8:
+		return float64(v), true
+	case uint16:
+		return float64(v), true
+	case uint32:
+		return float64(v), true
+	case uint64:
+		return float64(v), true
+	case float32:
+		return float64(v), true
+	case float64:
+		return v, true
+	case string:
+		if num, err := strconv.ParseFloat(v, 64); err == nil {
+			return num, true
+		}
+	}
+	return 0, false
+}
+
 // getRelationModelSingleLevel gets the model type for a single level field (non-recursive)
 // This is a helper function used by GetRelationModel to handle one level at a time
 func getRelationModelSingleLevel(model interface{}, fieldName string) interface{} {

pkg/reflection/model_utils_sqltypes_test.go

@@ -0,0 +1,266 @@
+package reflection_test
+
+import (
+	"testing"
+	"time"
+
+	"github.com/bitechdev/ResolveSpec/pkg/reflection"
+	"github.com/bitechdev/ResolveSpec/pkg/spectypes"
+)
+
+func TestMapToStruct_SqlJSONB_PreservesDriverValuer(t *testing.T) {
+	// Test that SqlJSONB type preserves driver.Valuer interface
+	type TestModel struct {
+		ID   int64           `bun:"id,pk" json:"id"`
+		Meta spectypes.SqlJSONB `bun:"meta" json:"meta"`
+	}
+
+	dataMap := map[string]interface{}{
+		"id": int64(123),
+		"meta": map[string]interface{}{
+			"key": "value",
+			"num": 42,
+		},
+	}
+
+	var result TestModel
+	err := reflection.MapToStruct(dataMap, &result)
+	if err != nil {
+		t.Fatalf("MapToStruct() error = %v", err)
+	}
+
+	// Verify the field was set
+	if result.ID != 123 {
+		t.Errorf("ID = %v, want 123", result.ID)
+	}
+
+	// Verify SqlJSONB was populated
+	if len(result.Meta) == 0 {
+		t.Error("Meta is empty, want non-empty")
+	}
+
+	// Most importantly: verify driver.Valuer interface works
+	value, err := result.Meta.Value()
+	if err != nil {
+		t.Errorf("Meta.Value() error = %v, want nil", err)
+	}
+
+	// Value should return a string representation of the JSON
+	if value == nil {
+		t.Error("Meta.Value() returned nil, want non-nil")
+	}
+
+	// Check it's a valid JSON string
+	if str, ok := value.(string); ok {
+		if len(str) == 0 {
+			t.Error("Meta.Value() returned empty string, want valid JSON")
+		}
+		t.Logf("SqlJSONB.Value() returned: %s", str)
+	} else {
+		t.Errorf("Meta.Value() returned type %T, want string", value)
+	}
+}
+
+func TestMapToStruct_SqlJSONB_FromBytes(t *testing.T) {
+	// Test that SqlJSONB can be set from []byte directly
+	type TestModel struct {
+		ID   int64           `bun:"id,pk" json:"id"`
+		Meta spectypes.SqlJSONB `bun:"meta" json:"meta"`
+	}
+
+	jsonBytes := []byte(`{"direct":"bytes"}`)
+	dataMap := map[string]interface{}{
+		"id":   int64(456),
+		"meta": jsonBytes,
+	}
+
+	var result TestModel
+	err := reflection.MapToStruct(dataMap, &result)
+	if err != nil {
+		t.Fatalf("MapToStruct() error = %v", err)
+	}
+
+	if result.ID != 456 {
+		t.Errorf("ID = %v, want 456", result.ID)
+	}
+
+	if string(result.Meta) != string(jsonBytes) {
+		t.Errorf("Meta = %s, want %s", string(result.Meta), string(jsonBytes))
+	}
+
+	// Verify driver.Valuer works
+	value, err := result.Meta.Value()
+	if err != nil {
+		t.Errorf("Meta.Value() error = %v", err)
+	}
+	if value == nil {
+		t.Error("Meta.Value() returned nil")
+	}
+}
+
+func TestMapToStruct_AllSqlTypes(t *testing.T) {
+	// Test model with all SQL custom types
+	type TestModel struct {
+		ID        int64               `bun:"id,pk" json:"id"`
+		Name      string              `bun:"name" json:"name"`
+		CreatedAt spectypes.SqlTimeStamp `bun:"created_at" json:"created_at"`
+		BirthDate spectypes.SqlDate      `bun:"birth_date" json:"birth_date"`
+		LoginTime spectypes.SqlTime      `bun:"login_time" json:"login_time"`
+		Meta      spectypes.SqlJSONB     `bun:"meta" json:"meta"`
+		Tags      spectypes.SqlJSONB     `bun:"tags" json:"tags"`
+	}
+
+	now := time.Now()
+	birthDate := time.Date(1990, 1, 15, 0, 0, 0, 0, time.UTC)
+	loginTime := time.Date(0, 1, 1, 14, 30, 0, 0, time.UTC)
+
+	dataMap := map[string]interface{}{
+		"id":         int64(100),
+		"name":       "Test User",
+		"created_at": now,
+		"birth_date": birthDate,
+		"login_time": loginTime,
+		"meta": map[string]interface{}{
+			"role":   "admin",
+			"active": true,
+		},
+		"tags": []interface{}{"golang", "testing", "sql"},
+	}
+
+	var result TestModel
+	err := reflection.MapToStruct(dataMap, &result)
+	if err != nil {
+		t.Fatalf("MapToStruct() error = %v", err)
+	}
+
+	// Verify basic fields
+	if result.ID != 100 {
+		t.Errorf("ID = %v, want 100", result.ID)
+	}
+	if result.Name != "Test User" {
+		t.Errorf("Name = %v, want 'Test User'", result.Name)
+	}
+
+	// Verify SqlTimeStamp
+	if !result.CreatedAt.Valid {
+		t.Error("CreatedAt.Valid = false, want true")
+	}
+	if !result.CreatedAt.Val.Equal(now) {
+		t.Errorf("CreatedAt.Val = %v, want %v", result.CreatedAt.Val, now)
+	}
+
+	// Verify driver.Valuer for SqlTimeStamp
+	tsValue, err := result.CreatedAt.Value()
+	if err != nil {
+		t.Errorf("CreatedAt.Value() error = %v", err)
+	}
+	if tsValue == nil {
+		t.Error("CreatedAt.Value() returned nil")
+	}
+
+	// Verify SqlDate
+	if !result.BirthDate.Valid {
+		t.Error("BirthDate.Valid = false, want true")
+	}
+	if !result.BirthDate.Val.Equal(birthDate) {
+		t.Errorf("BirthDate.Val = %v, want %v", result.BirthDate.Val, birthDate)
+	}
+
+	// Verify driver.Valuer for SqlDate
+	dateValue, err := result.BirthDate.Value()
+	if err != nil {
+		t.Errorf("BirthDate.Value() error = %v", err)
+	}
+	if dateValue == nil {
+		t.Error("BirthDate.Value() returned nil")
+	}
+
+	// Verify SqlTime
+	if !result.LoginTime.Valid {
+		t.Error("LoginTime.Valid = false, want true")
+	}
+
+	// Verify driver.Valuer for SqlTime
+	timeValue, err := result.LoginTime.Value()
+	if err != nil {
+		t.Errorf("LoginTime.Value() error = %v", err)
+	}
+	if timeValue == nil {
+		t.Error("LoginTime.Value() returned nil")
+	}
+
+	// Verify SqlJSONB for Meta
+	if len(result.Meta) == 0 {
+		t.Error("Meta is empty")
+	}
+	metaValue, err := result.Meta.Value()
+	if err != nil {
+		t.Errorf("Meta.Value() error = %v", err)
+	}
+	if metaValue == nil {
+		t.Error("Meta.Value() returned nil")
+	}
+
+	// Verify SqlJSONB for Tags
+	if len(result.Tags) == 0 {
+		t.Error("Tags is empty")
+	}
+	tagsValue, err := result.Tags.Value()
+	if err != nil {
+		t.Errorf("Tags.Value() error = %v", err)
+	}
+	if tagsValue == nil {
+		t.Error("Tags.Value() returned nil")
+	}
+
+	t.Logf("All SQL types successfully preserved driver.Valuer interface:")
+	t.Logf("  - SqlTimeStamp: %v", tsValue)
+	t.Logf("  - SqlDate: %v", dateValue)
+	t.Logf("  - SqlTime: %v", timeValue)
+	t.Logf("  - SqlJSONB (Meta): %v", metaValue)
+	t.Logf("  - SqlJSONB (Tags): %v", tagsValue)
+}
+
+func TestMapToStruct_SqlNull_NilValues(t *testing.T) {
+	// Test that SqlNull types handle nil values correctly
+	type TestModel struct {
+		ID        int64               `bun:"id,pk" json:"id"`
+		UpdatedAt spectypes.SqlTimeStamp `bun:"updated_at" json:"updated_at"`
+		DeletedAt spectypes.SqlTimeStamp `bun:"deleted_at" json:"deleted_at"`
+	}
+
+	now := time.Now()
+	dataMap := map[string]interface{}{
+		"id":         int64(200),
+		"updated_at": now,
+		"deleted_at": nil, // Explicitly nil
+	}
+
+	var result TestModel
+	err := reflection.MapToStruct(dataMap, &result)
+	if err != nil {
+		t.Fatalf("MapToStruct() error = %v", err)
+	}
+
+	// UpdatedAt should be valid
+	if !result.UpdatedAt.Valid {
+		t.Error("UpdatedAt.Valid = false, want true")
+	}
+	if !result.UpdatedAt.Val.Equal(now) {
+		t.Errorf("UpdatedAt.Val = %v, want %v", result.UpdatedAt.Val, now)
+	}
+
+	// DeletedAt should be invalid (null)
+	if result.DeletedAt.Valid {
+		t.Error("DeletedAt.Valid = true, want false (null)")
+	}
+
+	// Verify driver.Valuer for null SqlTimeStamp
+	deletedValue, err := result.DeletedAt.Value()
+	if err != nil {
+		t.Errorf("DeletedAt.Value() error = %v", err)
+	}
+	if deletedValue != nil {
+		t.Errorf("DeletedAt.Value() = %v, want nil", deletedValue)
+	}
+}

pkg/reflection/model_utils_test.go

@@ -1687,3 +1687,201 @@ func TestGetRelationModel_WithTags(t *testing.T) {
 		})
 	}
 }
+
+func TestMapToStruct(t *testing.T) {
+	// Test model with various field types
+	type TestModel struct {
+		ID       int64   `bun:"id,pk" json:"id"`
+		Name     string  `bun:"name" json:"name"`
+		Age      int     `bun:"age" json:"age"`
+		Active   bool    `bun:"active" json:"active"`
+		Score    float64 `bun:"score" json:"score"`
+		Data     []byte  `bun:"data" json:"data"`
+		MetaJSON []byte  `bun:"meta_json" json:"meta_json"`
+	}
+
+	tests := []struct {
+		name     string
+		dataMap  map[string]interface{}
+		expected TestModel
+		wantErr  bool
+	}{
+		{
+			name: "Basic types conversion",
+			dataMap: map[string]interface{}{
+				"id":     int64(123),
+				"name":   "Test User",
+				"age":    30,
+				"active": true,
+				"score":  95.5,
+			},
+			expected: TestModel{
+				ID:     123,
+				Name:   "Test User",
+				Age:    30,
+				Active: true,
+				Score:  95.5,
+			},
+			wantErr: false,
+		},
+		{
+			name: "Byte slice (SqlJSONB-like) from []byte",
+			dataMap: map[string]interface{}{
+				"id":   int64(456),
+				"name": "JSON Test",
+				"data": []byte(`{"key":"value"}`),
+			},
+			expected: TestModel{
+				ID:   456,
+				Name: "JSON Test",
+				Data: []byte(`{"key":"value"}`),
+			},
+			wantErr: false,
+		},
+		{
+			name: "Byte slice from string",
+			dataMap: map[string]interface{}{
+				"id":   int64(789),
+				"data": "string data",
+			},
+			expected: TestModel{
+				ID:   789,
+				Data: []byte("string data"),
+			},
+			wantErr: false,
+		},
+		{
+			name: "Byte slice from map (JSON marshal)",
+			dataMap: map[string]interface{}{
+				"id": int64(999),
+				"meta_json": map[string]interface{}{
+					"field1": "value1",
+					"field2": 42,
+				},
+			},
+			expected: TestModel{
+				ID:       999,
+				MetaJSON: []byte(`{"field1":"value1","field2":42}`),
+			},
+			wantErr: false,
+		},
+		{
+			name: "Byte slice from slice (JSON marshal)",
+			dataMap: map[string]interface{}{
+				"id":        int64(111),
+				"meta_json": []interface{}{"item1", "item2", 3},
+			},
+			expected: TestModel{
+				ID:       111,
+				MetaJSON: []byte(`["item1","item2",3]`),
+			},
+			wantErr: false,
+		},
+		{
+			name: "Field matching by bun tag",
+			dataMap: map[string]interface{}{
+				"id":   int64(222),
+				"name": "Tagged Field",
+			},
+			expected: TestModel{
+				ID:   222,
+				Name: "Tagged Field",
+			},
+			wantErr: false,
+		},
+		{
+			name: "Nil values",
+			dataMap: map[string]interface{}{
+				"id":   int64(333),
+				"data": nil,
+			},
+			expected: TestModel{
+				ID:   333,
+				Data: nil,
+			},
+			wantErr: false,
+		},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			var result TestModel
+			err := MapToStruct(tt.dataMap, &result)
+
+			if (err != nil) != tt.wantErr {
+				t.Errorf("MapToStruct() error = %v, wantErr %v", err, tt.wantErr)
+				return
+			}
+
+			// Compare fields individually for better error messages
+			if result.ID != tt.expected.ID {
+				t.Errorf("ID = %v, want %v", result.ID, tt.expected.ID)
+			}
+			if result.Name != tt.expected.Name {
+				t.Errorf("Name = %v, want %v", result.Name, tt.expected.Name)
+			}
+			if result.Age != tt.expected.Age {
+				t.Errorf("Age = %v, want %v", result.Age, tt.expected.Age)
+			}
+			if result.Active != tt.expected.Active {
+				t.Errorf("Active = %v, want %v", result.Active, tt.expected.Active)
+			}
+			if result.Score != tt.expected.Score {
+				t.Errorf("Score = %v, want %v", result.Score, tt.expected.Score)
+			}
+
+			// For byte slices, compare as strings for JSON data
+			if tt.expected.Data != nil {
+				if string(result.Data) != string(tt.expected.Data) {
+					t.Errorf("Data = %s, want %s", string(result.Data), string(tt.expected.Data))
+				}
+			}
+			if tt.expected.MetaJSON != nil {
+				if string(result.MetaJSON) != string(tt.expected.MetaJSON) {
+					t.Errorf("MetaJSON = %s, want %s", string(result.MetaJSON), string(tt.expected.MetaJSON))
+				}
+			}
+		})
+	}
+}
+
+func TestMapToStruct_Errors(t *testing.T) {
+	type TestModel struct {
+		ID int `bun:"id" json:"id"`
+	}
+
+	tests := []struct {
+		name    string
+		dataMap map[string]interface{}
+		target  interface{}
+		wantErr bool
+	}{
+		{
+			name:    "Nil dataMap",
+			dataMap: nil,
+			target:  &TestModel{},
+			wantErr: true,
+		},
+		{
+			name:    "Nil target",
+			dataMap: map[string]interface{}{"id": 1},
+			target:  nil,
+			wantErr: true,
+		},
+		{
+			name:    "Non-pointer target",
+			dataMap: map[string]interface{}{"id": 1},
+			target:  TestModel{},
+			wantErr: true,
+		},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			err := MapToStruct(tt.dataMap, tt.target)
+			if (err != nil) != tt.wantErr {
+				t.Errorf("MapToStruct() error = %v, wantErr %v", err, tt.wantErr)
+			}
+		})
+	}
+}

pkg/resolvespec/handler.go

@@ -2,6 +2,7 @@ package resolvespec
 
 import (
 	"context"
+	"database/sql"
 	"encoding/json"
 	"fmt"
 	"net/http"
@@ -957,7 +958,29 @@ func (h *Handler) handleDelete(ctx context.Context, w common.ResponseWriter, id
 		return
 	}
 
-	query := h.db.NewDelete().Table(tableName).Where(fmt.Sprintf("%s = ?", common.QuoteIdent(reflection.GetPrimaryKeyName(model))), id)
+	// Get primary key name
+	pkName := reflection.GetPrimaryKeyName(model)
+
+	// First, fetch the record that will be deleted
+	modelType := reflect.TypeOf(model)
+	if modelType.Kind() == reflect.Ptr {
+		modelType = modelType.Elem()
+	}
+	recordToDelete := reflect.New(modelType).Interface()
+
+	selectQuery := h.db.NewSelect().Model(recordToDelete).Where(fmt.Sprintf("%s = ?", common.QuoteIdent(pkName)), id)
+	if err := selectQuery.ScanModel(ctx); err != nil {
+		if err == sql.ErrNoRows {
+			logger.Warn("Record not found for delete: %s = %s", pkName, id)
+			h.sendError(w, http.StatusNotFound, "not_found", "Record not found", err)
+			return
+		}
+		logger.Error("Error fetching record for delete: %v", err)
+		h.sendError(w, http.StatusInternalServerError, "fetch_error", "Error fetching record", err)
+		return
+	}
+
+	query := h.db.NewDelete().Table(tableName).Where(fmt.Sprintf("%s = ?", common.QuoteIdent(pkName)), id)
 
 	result, err := query.Exec(ctx)
 	if err != nil {
@@ -966,14 +989,16 @@ func (h *Handler) handleDelete(ctx context.Context, w common.ResponseWriter, id
 		return
 	}
 
+	// Check if the record was actually deleted
 	if result.RowsAffected() == 0 {
-		logger.Warn("No record found to delete with ID: %s", id)
+		logger.Warn("No rows deleted for ID: %s", id)
-		h.sendError(w, http.StatusNotFound, "not_found", "Record not found", nil)
+		h.sendError(w, http.StatusNotFound, "not_found", "Record not found or already deleted", nil)
 		return
 	}
 
 	logger.Info("Successfully deleted record with ID: %s", id)
-	h.sendResponse(w, nil, nil)
+	// Return the deleted record data
+	h.sendResponse(w, recordToDelete, nil)
 }
 
 func (h *Handler) applyFilter(query common.SelectQuery, filter common.FilterOption) common.SelectQuery {

pkg/restheadspec/handler.go

@@ -2,6 +2,7 @@ package restheadspec
 
 import (
 	"context"
+	"database/sql"
 	"encoding/json"
 	"fmt"
 	"net/http"
@@ -513,7 +514,15 @@ func (h *Handler) handleRead(ctx context.Context, w common.ResponseWriter, id st
 			direction = "DESC"
 		}
 		logger.Debug("Applying sort: %s %s", sort.Column, direction)
-		query = query.Order(fmt.Sprintf("%s %s", sort.Column, direction))
+
+		// Check if it's an expression (enclosed in brackets) - use directly without quoting
+		if strings.HasPrefix(sort.Column, "(") && strings.HasSuffix(sort.Column, ")") {
+			// For expressions, pass as raw SQL to prevent auto-quoting
+			query = query.OrderExpr(fmt.Sprintf("%s %s", sort.Column, direction))
+		} else {
+			// Regular column - let Bun handle quoting
+			query = query.Order(fmt.Sprintf("%s %s", sort.Column, direction))
+		}
 	}
 
 	// Get total count before pagination (unless skip count is requested)
@@ -653,6 +662,14 @@ func (h *Handler) handleRead(ctx context.Context, w common.ResponseWriter, id st
 		return
 	}
 
+	// Check if a specific ID was requested but no record was found
+	resultCount := reflection.Len(modelPtr)
+	if id != "" && resultCount == 0 {
+		logger.Warn("Record not found for ID: %s", id)
+		h.sendError(w, http.StatusNotFound, "not_found", "Record not found", nil)
+		return
+	}
+
 	limit := 0
 	if options.Limit != nil {
 		limit = *options.Limit
@@ -667,7 +684,7 @@ func (h *Handler) handleRead(ctx context.Context, w common.ResponseWriter, id st
 
 	metadata := &common.Metadata{
 		Total:    int64(total),
-		Count:    int64(reflection.Len(modelPtr)),
+		Count:    int64(resultCount),
 		Filtered: int64(total),
 		Limit:    limit,
 		Offset:   offset,
@@ -746,9 +763,42 @@ func (h *Handler) applyPreloadWithRecursion(query common.SelectQuery, preload co
 
 			// Apply ComputedQL fields if any
 			if len(preload.ComputedQL) > 0 {
+				// Get the base table name from the related model
+				baseTableName := getTableNameFromModel(relatedModel)
+
+				// Convert the preload relation path to the appropriate alias format
+				// This is ORM-specific. Currently we only support Bun's format.
+				// TODO: Add support for other ORMs if needed
+				preloadAlias := ""
+				if h.db.GetUnderlyingDB() != nil {
+					// Check if we're using Bun by checking the type name
+					underlyingType := fmt.Sprintf("%T", h.db.GetUnderlyingDB())
+					if strings.Contains(underlyingType, "bun.DB") {
+						// Use Bun's alias format: lowercase with double underscores
+						preloadAlias = relationPathToBunAlias(preload.Relation)
+					}
+					// For GORM: GORM doesn't use the same alias format, and this fix
+					// may not be needed since GORM handles preloads differently
+				}
+
+				logger.Debug("Applying computed columns to preload %s (alias: %s, base table: %s)",
+					preload.Relation, preloadAlias, baseTableName)
+
 				for colName, colExpr := range preload.ComputedQL {
+					// Replace table references in the expression with the preload alias
+					// This fixes the ambiguous column reference issue when there are multiple
+					// levels of recursive/nested preloads
+					adjustedExpr := colExpr
+					if baseTableName != "" && preloadAlias != "" {
+						adjustedExpr = replaceTableReferencesInSQL(colExpr, baseTableName, preloadAlias)
+						if adjustedExpr != colExpr {
+							logger.Debug("Adjusted computed column expression for %s: '%s' -> '%s'",
+								colName, colExpr, adjustedExpr)
+						}
+					}
+
 					logger.Debug("Applying computed column to preload %s: %s", preload.Relation, colName)
-					sq = sq.ColumnExpr(fmt.Sprintf("(%s) AS %s", colExpr, colName))
+					sq = sq.ColumnExpr(fmt.Sprintf("(%s) AS %s", adjustedExpr, colName))
 					// Remove the computed column from selected columns to avoid duplication
 					for colIndex := range preload.Columns {
 						if preload.Columns[colIndex] == colName {
@@ -794,7 +844,14 @@ func (h *Handler) applyPreloadWithRecursion(query common.SelectQuery, preload co
 		// Apply sorting
 		if len(preload.Sort) > 0 {
 			for _, sort := range preload.Sort {
-				sq = sq.Order(fmt.Sprintf("%s %s", sort.Column, sort.Direction))
+				// Check if it's an expression (enclosed in brackets) - use directly without quoting
+				if strings.HasPrefix(sort.Column, "(") && strings.HasSuffix(sort.Column, ")") {
+					// For expressions, pass as raw SQL to prevent auto-quoting
+					sq = sq.OrderExpr(fmt.Sprintf("%s %s", sort.Column, sort.Direction))
+				} else {
+					// Regular column - let ORM handle quoting
+					sq = sq.Order(fmt.Sprintf("%s %s", sort.Column, sort.Direction))
+				}
 			}
 		}
 
@@ -841,6 +898,73 @@ func (h *Handler) applyPreloadWithRecursion(query common.SelectQuery, preload co
 	return query
 }
 
+// relationPathToBunAlias converts a relation path like "MAL.MAL.DEF" to the Bun alias format "mal__mal__def"
+// Bun generates aliases for nested relations by lowercasing and replacing dots with double underscores
+func relationPathToBunAlias(relationPath string) string {
+	if relationPath == "" {
+		return ""
+	}
+	// Convert to lowercase and replace dots with double underscores
+	alias := strings.ToLower(relationPath)
+	alias = strings.ReplaceAll(alias, ".", "__")
+	return alias
+}
+
+// replaceTableReferencesInSQL replaces references to a base table name in a SQL expression
+// with the appropriate alias for the current preload level
+// For example, if baseTableName is "mastertaskitem" and targetAlias is "mal__mal",
+// it will replace "mastertaskitem.rid_mastertaskitem" with "mal__mal.rid_mastertaskitem"
+func replaceTableReferencesInSQL(sqlExpr, baseTableName, targetAlias string) string {
+	if sqlExpr == "" || baseTableName == "" || targetAlias == "" {
+		return sqlExpr
+	}
+
+	// Replace both quoted and unquoted table references
+	// Handle patterns like: tablename.column, "tablename".column, tablename."column", "tablename"."column"
+
+	// Pattern 1: tablename.column (unquoted)
+	result := strings.ReplaceAll(sqlExpr, baseTableName+".", targetAlias+".")
+
+	// Pattern 2: "tablename".column or "tablename"."column" (quoted table name)
+	result = strings.ReplaceAll(result, "\""+baseTableName+"\".", "\""+targetAlias+"\".")
+
+	return result
+}
+
+// getTableNameFromModel extracts the table name from a model
+// It checks the bun tag first, then falls back to converting the struct name to snake_case
+func getTableNameFromModel(model interface{}) string {
+	if model == nil {
+		return ""
+	}
+
+	modelType := reflect.TypeOf(model)
+
+	// Unwrap pointers
+	for modelType != nil && modelType.Kind() == reflect.Ptr {
+		modelType = modelType.Elem()
+	}
+
+	if modelType == nil || modelType.Kind() != reflect.Struct {
+		return ""
+	}
+
+	// Look for bun tag on embedded BaseModel
+	for i := 0; i < modelType.NumField(); i++ {
+		field := modelType.Field(i)
+		if field.Anonymous {
+			bunTag := field.Tag.Get("bun")
+			if strings.HasPrefix(bunTag, "table:") {
+				return strings.TrimPrefix(bunTag, "table:")
+			}
+		}
+	}
+
+	// Fallback: convert struct name to lowercase (simple heuristic)
+	// This handles cases like "MasterTaskItem" -> "mastertaskitem"
+	return strings.ToLower(modelType.Name())
+}
+
 func (h *Handler) handleCreate(ctx context.Context, w common.ResponseWriter, data interface{}, options ExtendedRequestOptions) {
 	// Capture panics and return error response
 	defer func() {
@@ -1120,8 +1244,19 @@ func (h *Handler) handleUpdate(ctx context.Context, w common.ResponseWriter, id
 		// Ensure ID is in the data map for the update
 		dataMap[pkName] = targetID
 
-		// Create update query
+		// Populate model instance from dataMap to preserve custom types (like SqlJSONB)
-		query := tx.NewUpdate().Table(tableName).SetMap(dataMap)
+		// Get the type of the model, handling both pointer and non-pointer types
+		modelType := reflect.TypeOf(model)
+		if modelType.Kind() == reflect.Ptr {
+			modelType = modelType.Elem()
+		}
+		modelInstance := reflect.New(modelType).Interface()
+		if err := reflection.MapToStruct(dataMap, modelInstance); err != nil {
+			return fmt.Errorf("failed to populate model from data: %w", err)
+		}
+
+		// Create update query using Model() to preserve custom types and driver.Valuer interfaces
+		query := tx.NewUpdate().Model(modelInstance)
 		query = query.Where(fmt.Sprintf("%s = ?", common.QuoteIdent(pkName)), targetID)
 
 		// Execute BeforeScan hooks - pass query chain so hooks can modify it
@@ -1388,7 +1523,34 @@ func (h *Handler) handleDelete(ctx context.Context, w common.ResponseWriter, id
 	}
 
 	// Single delete with URL ID
-	// Execute BeforeDelete hooks
+	if id == "" {
+		h.sendError(w, http.StatusBadRequest, "missing_id", "ID is required for delete", nil)
+		return
+	}
+
+	// Get primary key name
+	pkName := reflection.GetPrimaryKeyName(model)
+
+	// First, fetch the record that will be deleted
+	modelType := reflect.TypeOf(model)
+	if modelType.Kind() == reflect.Ptr {
+		modelType = modelType.Elem()
+	}
+	recordToDelete := reflect.New(modelType).Interface()
+
+	selectQuery := h.db.NewSelect().Model(recordToDelete).Where(fmt.Sprintf("%s = ?", common.QuoteIdent(pkName)), id)
+	if err := selectQuery.ScanModel(ctx); err != nil {
+		if err == sql.ErrNoRows {
+			logger.Warn("Record not found for delete: %s = %s", pkName, id)
+			h.sendError(w, http.StatusNotFound, "not_found", "Record not found", err)
+			return
+		}
+		logger.Error("Error fetching record for delete: %v", err)
+		h.sendError(w, http.StatusInternalServerError, "fetch_error", "Error fetching record", err)
+		return
+	}
+
+	// Execute BeforeDelete hooks with the record data
 	hookCtx := &HookContext{
 		Context:   ctx,
 		Handler:   h,
@@ -1399,6 +1561,7 @@ func (h *Handler) handleDelete(ctx context.Context, w common.ResponseWriter, id
 		ID:        id,
 		Writer:    w,
 		Tx:        h.db,
+		Data:      recordToDelete,
 	}
 
 	if err := h.hooks.Execute(BeforeDelete, hookCtx); err != nil {
@@ -1408,13 +1571,7 @@ func (h *Handler) handleDelete(ctx context.Context, w common.ResponseWriter, id
 	}
 
 	query := h.db.NewDelete().Table(tableName)
-
+	query = query.Where(fmt.Sprintf("%s = ?", common.QuoteIdent(pkName)), id)
-	if id == "" {
-		h.sendError(w, http.StatusBadRequest, "missing_id", "ID is required for delete", nil)
-		return
-	}
-
-	query = query.Where(fmt.Sprintf("%s = ?", common.QuoteIdent(reflection.GetPrimaryKeyName(model))), id)
 
 	// Execute BeforeScan hooks - pass query chain so hooks can modify it
 	hookCtx.Query = query
@@ -1436,11 +1593,15 @@ func (h *Handler) handleDelete(ctx context.Context, w common.ResponseWriter, id
 		return
 	}
 
-	// Execute AfterDelete hooks
+	// Check if the record was actually deleted
-	responseData := map[string]interface{}{
+	if result.RowsAffected() == 0 {
-		"deleted": result.RowsAffected(),
+		logger.Warn("No rows deleted for ID: %s", id)
+		h.sendError(w, http.StatusNotFound, "not_found", "Record not found or already deleted", nil)
+		return
 	}
-	hookCtx.Result = responseData
+
+	// Execute AfterDelete hooks with the deleted record data
+	hookCtx.Result = recordToDelete
 	hookCtx.Error = nil
 
 	if err := h.hooks.Execute(AfterDelete, hookCtx); err != nil {
@@ -1449,7 +1610,8 @@ func (h *Handler) handleDelete(ctx context.Context, w common.ResponseWriter, id
 		return
 	}
 
-	h.sendResponse(w, responseData, nil)
+	// Return the deleted record data
+	h.sendResponse(w, recordToDelete, nil)
 }
 
 // mergeRecordWithRequest merges a database record with the original request data
@@ -1962,7 +2124,7 @@ func (h *Handler) sendResponseWithOptions(w common.ResponseWriter, data interfac
 // Returns the single element if data is a slice/array with exactly one element, otherwise returns data unchanged
 func (h *Handler) normalizeResultArray(data interface{}) interface{} {
 	if data == nil {
-		return nil
+		return map[string]interface{}{}
 	}
 
 	// Use reflection to check if data is a slice or array
@@ -1971,10 +2133,15 @@ func (h *Handler) normalizeResultArray(data interface{}) interface{} {
 		dataValue = dataValue.Elem()
 	}
 
-	// Check if it's a slice or array with exactly one element
+	// Check if it's a slice or array
-	if (dataValue.Kind() == reflect.Slice || dataValue.Kind() == reflect.Array) && dataValue.Len() == 1 {
+	if dataValue.Kind() == reflect.Slice || dataValue.Kind() == reflect.Array {
-		// Return the single element
+		if dataValue.Len() == 1 {
-		return dataValue.Index(0).Interface()
+			// Return the single element
+			return dataValue.Index(0).Interface()
+		} else if dataValue.Len() == 0 {
+			// Return empty object instead of empty array
+			return map[string]interface{}{}
+		}
 	}
 
 	return data
@@ -2082,7 +2249,14 @@ func (h *Handler) FetchRowNumber(ctx context.Context, tableName string, pkName s
 			if strings.EqualFold(sort.Direction, "desc") {
 				direction = "DESC"
 			}
-			sortParts = append(sortParts, fmt.Sprintf("%s.%s %s", tableName, sort.Column, direction))
+
+			// Check if it's an expression (enclosed in brackets) - use directly without table prefix
+			if strings.HasPrefix(sort.Column, "(") && strings.HasSuffix(sort.Column, ")") {
+				sortParts = append(sortParts, fmt.Sprintf("%s %s", sort.Column, direction))
+			} else {
+				// Regular column - add table prefix
+				sortParts = append(sortParts, fmt.Sprintf("%s.%s %s", tableName, sort.Column, direction))
+			}
 		}
 		sortSQL = strings.Join(sortParts, ", ")
 	} else {
@@ -2291,6 +2465,55 @@ func (h *Handler) filterExtendedOptions(validator *common.ColumnValidator, optio
 			expandValidator := common.NewColumnValidator(relInfo.relatedModel)
 			// Filter columns using the related model's validator
 			filteredExpand.Columns = expandValidator.FilterValidColumns(expand.Columns)
+
+			// Filter sort columns in the expand Sort string
+			if expand.Sort != "" {
+				sortFields := strings.Split(expand.Sort, ",")
+				validSortFields := make([]string, 0, len(sortFields))
+				for _, sortField := range sortFields {
+					sortField = strings.TrimSpace(sortField)
+					if sortField == "" {
+						continue
+					}
+
+					// Extract column name (remove direction prefixes/suffixes)
+					colName := sortField
+					direction := ""
+
+					if strings.HasPrefix(sortField, "-") {
+						direction = "-"
+						colName = strings.TrimPrefix(sortField, "-")
+					} else if strings.HasPrefix(sortField, "+") {
+						direction = "+"
+						colName = strings.TrimPrefix(sortField, "+")
+					}
+
+					if strings.HasSuffix(strings.ToLower(colName), " desc") {
+						direction = " desc"
+						colName = strings.TrimSuffix(strings.ToLower(colName), " desc")
+					} else if strings.HasSuffix(strings.ToLower(colName), " asc") {
+						direction = " asc"
+						colName = strings.TrimSuffix(strings.ToLower(colName), " asc")
+					}
+
+					colName = strings.TrimSpace(colName)
+
+					// Validate the column name
+					if expandValidator.IsValidColumn(colName) {
+						validSortFields = append(validSortFields, direction+colName)
+					} else if strings.HasPrefix(colName, "(") && strings.HasSuffix(colName, ")") {
+						// Allow sort by expression/subquery, but validate for security
+						if common.IsSafeSortExpression(colName) {
+							validSortFields = append(validSortFields, direction+colName)
+						} else {
+							logger.Warn("Unsafe sort expression in expand '%s' removed: '%s'", expand.Relation, colName)
+						}
+					} else {
+						logger.Warn("Invalid column in expand '%s' sort '%s' removed", expand.Relation, colName)
+					}
+				}
+				filteredExpand.Sort = strings.Join(validSortFields, ",")
+			}
 		} else {
 			// If we can't find the relationship, log a warning and skip column filtering
 			logger.Warn("Cannot validate columns for unknown relation: %s", expand.Relation)

pkg/restheadspec/headers.go

@@ -529,19 +529,47 @@ func (h *Handler) parseSorting(options *ExtendedRequestOptions, value string) {
 }
 
 // parseCommaSeparated parses comma-separated values and trims whitespace
+// It respects bracket nesting and only splits on commas outside of parentheses
 func (h *Handler) parseCommaSeparated(value string) []string {
 	if value == "" {
 		return nil
 	}
 
-	parts := strings.Split(value, ",")
+	result := make([]string, 0)
-	result := make([]string, 0, len(parts))
+	var current strings.Builder
-	for _, part := range parts {
+	nestingLevel := 0
-		part = strings.TrimSpace(part)
+
-		if part != "" {
+	for _, char := range value {
-			result = append(result, part)
+		switch char {
+		case '(':
+			nestingLevel++
+			current.WriteRune(char)
+		case ')':
+			nestingLevel--
+			current.WriteRune(char)
+		case ',':
+			if nestingLevel == 0 {
+				// We're outside all brackets, so split here
+				part := strings.TrimSpace(current.String())
+				if part != "" {
+					result = append(result, part)
+				}
+				current.Reset()
+			} else {
+				// Inside brackets, keep the comma
+				current.WriteRune(char)
+			}
+		default:
+			current.WriteRune(char)
 		}
 	}
+
+	// Add the last part
+	part := strings.TrimSpace(current.String())
+	if part != "" {
+		result = append(result, part)
+	}
+
 	return result
 }
 

pkg/spectypes/sql_types.go

@@ -1,5 +1,5 @@
-// Package common provides nullable SQL types with automatic casting and conversion methods.
+// Package spectypes provides nullable SQL types with automatic casting and conversion methods.
-package common
+package spectypes
 
 import (
 	"database/sql"

pkg/spectypes/sql_types_test.go

@@ -1,4 +1,4 @@
-package common
+package spectypes
 
 import (
 	"database/sql/driver"