Implemented TypeORM, Prisma and Enums on a schema

TODO.md

@@ -2,21 +2,21 @@
 
 
 ## Input Readers / Writers
-- [x] **Database Inspector**
+- [✔️] **Database Inspector**
-  - [x] PostgreSQL driver
+  - [✔️] PostgreSQL driver
   - [ ] MySQL driver
   - [ ] SQLite driver
   - [ ] MSSQL driver
-  - [x] Foreign key detection
+  - [✔️] Foreign key detection
-  - [x] Index extraction
+  - [✔️] Index extraction
   - [ ] .sql file generation with sequence and priority
-- [*] .dbml: Database Markup Language (DBML) for textual schema representation.
+- [✔️] .dbml: Database Markup Language (DBML) for textual schema representation.
-- [ ] Prisma schema support (PSL format) .prisma
+- [✔️] Prisma schema support (PSL format) .prisma
-- [ ] Entity Framework (.NET) model .edmx
+- [☠️] Entity Framework (.NET) model .edmx  (Fuck no, EDMX files were bloated, verbose XML nightmares—hard to merge, error-prone, and a pain in teams. Microsoft wisely ditched them in EF Core for code-first. Classic overkill from old MS era.)
-- [ ] TypeORM support
+- [✔️] TypeORM support
-- [ ] .hbm.xml / schema.xml: Hibernate/Propel mappings (Java/PHP)
+- [] .hbm.xml / schema.xml: Hibernate/Propel mappings (Java/PHP)  (💲 Someone can do this, not me)
-- [ ] Django models.py (Python classes), Sequelize migrations (JS)
+- [ ] Django models.py (Python classes), Sequelize migrations (JS) (💲 Someone can do this, not me)
-- [ ] .avsc: Avro schema (JSON format for data serialization)
+- [] .avsc: Avro schema (JSON format for data serialization) (💲 Someone can do this, not me)
 
 
 

cmd/relspec/convert.go

@@ -15,6 +15,8 @@ import (
 	"git.warky.dev/wdevs/relspecgo/pkg/readers/gorm"
 	"git.warky.dev/wdevs/relspecgo/pkg/readers/json"
 	"git.warky.dev/wdevs/relspecgo/pkg/readers/pgsql"
+	"git.warky.dev/wdevs/relspecgo/pkg/readers/prisma"
+	"git.warky.dev/wdevs/relspecgo/pkg/readers/typeorm"
 	"git.warky.dev/wdevs/relspecgo/pkg/readers/yaml"
 	"git.warky.dev/wdevs/relspecgo/pkg/writers"
 	wbun "git.warky.dev/wdevs/relspecgo/pkg/writers/bun"
@@ -24,6 +26,8 @@ import (
 	wgorm "git.warky.dev/wdevs/relspecgo/pkg/writers/gorm"
 	wjson "git.warky.dev/wdevs/relspecgo/pkg/writers/json"
 	wpgsql "git.warky.dev/wdevs/relspecgo/pkg/writers/pgsql"
+	wprisma "git.warky.dev/wdevs/relspecgo/pkg/writers/prisma"
+	wtypeorm "git.warky.dev/wdevs/relspecgo/pkg/writers/typeorm"
 	wyaml "git.warky.dev/wdevs/relspecgo/pkg/writers/yaml"
 	"github.com/spf13/cobra"
 )
@@ -55,6 +59,8 @@ Input formats:
   - yaml: YAML database schema
   - gorm: GORM model files (Go, file or directory)
   - bun: Bun model files (Go, file or directory)
+  - prisma: Prisma schema files (.prisma)
+  - typeorm: TypeORM entity files (TypeScript)
   - pgsql: PostgreSQL database (live connection)
 
 Output formats:
@@ -65,6 +71,8 @@ Output formats:
   - yaml: YAML database schema
   - gorm: GORM model files (Go)
   - bun: Bun model files (Go)
+  - prisma: Prisma schema files (.prisma)
+  - typeorm: TypeORM entity files (TypeScript)
   - pgsql: PostgreSQL SQL schema
 
 PostgreSQL Connection String Examples:
@@ -123,11 +131,11 @@ Examples:
 }
 
 func init() {
-	convertCmd.Flags().StringVar(&convertSourceType, "from", "", "Source format (dbml, dctx, drawdb, json, yaml, gorm, bun, pgsql)")
+	convertCmd.Flags().StringVar(&convertSourceType, "from", "", "Source format (dbml, dctx, drawdb, json, yaml, gorm, bun, prisma, typeorm, pgsql)")
 	convertCmd.Flags().StringVar(&convertSourcePath, "from-path", "", "Source file path (for file-based formats)")
 	convertCmd.Flags().StringVar(&convertSourceConn, "from-conn", "", "Source connection string (for database formats)")
 
-	convertCmd.Flags().StringVar(&convertTargetType, "to", "", "Target format (dbml, dctx, drawdb, json, yaml, gorm, bun, pgsql)")
+	convertCmd.Flags().StringVar(&convertTargetType, "to", "", "Target format (dbml, dctx, drawdb, json, yaml, gorm, bun, prisma, typeorm, pgsql)")
 	convertCmd.Flags().StringVar(&convertTargetPath, "to-path", "", "Target output path (file or directory)")
 	convertCmd.Flags().StringVar(&convertPackageName, "package", "", "Package name (for code generation formats like gorm/bun)")
 	convertCmd.Flags().StringVar(&convertSchemaFilter, "schema", "", "Filter to a specific schema by name (required for formats like dctx that only support single schemas)")
@@ -239,6 +247,18 @@ func readDatabaseForConvert(dbType, filePath, connString string) (*models.Databa
 		}
 		reader = bun.NewReader(&readers.ReaderOptions{FilePath: filePath})
 
+	case "prisma":
+		if filePath == "" {
+			return nil, fmt.Errorf("file path is required for Prisma format")
+		}
+		reader = prisma.NewReader(&readers.ReaderOptions{FilePath: filePath})
+
+	case "typeorm":
+		if filePath == "" {
+			return nil, fmt.Errorf("file path is required for TypeORM format")
+		}
+		reader = typeorm.NewReader(&readers.ReaderOptions{FilePath: filePath})
+
 	default:
 		return nil, fmt.Errorf("unsupported source format: %s", dbType)
 	}
@@ -290,6 +310,12 @@ func writeDatabase(db *models.Database, dbType, outputPath, packageName, schemaF
 	case "pgsql", "postgres", "postgresql", "sql":
 		writer = wpgsql.NewWriter(writerOpts)
 
+	case "prisma":
+		writer = wprisma.NewWriter(writerOpts)
+
+	case "typeorm":
+		writer = wtypeorm.NewWriter(writerOpts)
+
 	default:
 		return fmt.Errorf("unsupported target format: %s", dbType)
 	}

pkg/models/models.go

@@ -40,6 +40,7 @@ type Schema struct {
 	Sequence    uint              `json:"sequence,omitempty" yaml:"sequence,omitempty" xml:"sequence,omitempty"`
 	RefDatabase *Database         `json:"-" yaml:"-" xml:"-"` // Excluded to prevent circular references
 	Relations   []*Relationship   `json:"relations,omitempty" yaml:"relations,omitempty" xml:"-"`
+	Enums       []*Enum           `json:"enums,omitempty" yaml:"enums,omitempty" xml:"enums"`
 }
 
 // SQLName returns the schema name in lowercase
@@ -225,6 +226,16 @@ func (d *Constraint) SQLName() string {
 
 type ConstraintType string
 
+type Enum struct {
+	Name   string   `json:"name" yaml:"name" xml:"name"`
+	Values []string `json:"values" yaml:"values" xml:"values"`
+	Schema string   `json:"schema,omitempty" yaml:"schema,omitempty" xml:"schema,omitempty"`
+}
+
+func (d *Enum) SQLName() string {
+	return strings.ToLower(d.Name)
+}
+
 const (
 	PrimaryKeyConstraint ConstraintType = "primary_key"
 	ForeignKeyConstraint ConstraintType = "foreign_key"

pkg/readers/prisma/reader.go

@@ -0,0 +1,823 @@
+package prisma
+
+import (
+	"bufio"
+	"fmt"
+	"os"
+	"regexp"
+	"strings"
+
+	"git.warky.dev/wdevs/relspecgo/pkg/models"
+	"git.warky.dev/wdevs/relspecgo/pkg/readers"
+)
+
+// Reader implements the readers.Reader interface for Prisma schema format
+type Reader struct {
+	options *readers.ReaderOptions
+}
+
+// NewReader creates a new Prisma reader with the given options
+func NewReader(options *readers.ReaderOptions) *Reader {
+	return &Reader{
+		options: options,
+	}
+}
+
+// ReadDatabase reads and parses Prisma schema input, returning a Database model
+func (r *Reader) ReadDatabase() (*models.Database, error) {
+	if r.options.FilePath == "" {
+		return nil, fmt.Errorf("file path is required for Prisma reader")
+	}
+
+	content, err := os.ReadFile(r.options.FilePath)
+	if err != nil {
+		return nil, fmt.Errorf("failed to read file: %w", err)
+	}
+
+	return r.parsePrisma(string(content))
+}
+
+// ReadSchema reads and parses Prisma schema input, returning a Schema model
+func (r *Reader) ReadSchema() (*models.Schema, error) {
+	db, err := r.ReadDatabase()
+	if err != nil {
+		return nil, err
+	}
+
+	if len(db.Schemas) == 0 {
+		return nil, fmt.Errorf("no schemas found in Prisma schema")
+	}
+
+	// Return the first schema
+	return db.Schemas[0], nil
+}
+
+// ReadTable reads and parses Prisma schema input, returning a Table model
+func (r *Reader) ReadTable() (*models.Table, error) {
+	schema, err := r.ReadSchema()
+	if err != nil {
+		return nil, err
+	}
+
+	if len(schema.Tables) == 0 {
+		return nil, fmt.Errorf("no tables found in Prisma schema")
+	}
+
+	// Return the first table
+	return schema.Tables[0], nil
+}
+
+// stripQuotes removes surrounding quotes from an identifier
+func stripQuotes(s string) string {
+	s = strings.TrimSpace(s)
+	if len(s) >= 2 && ((s[0] == '"' && s[len(s)-1] == '"') || (s[0] == '\'' && s[len(s)-1] == '\'')) {
+		return s[1 : len(s)-1]
+	}
+	return s
+}
+
+// parsePrisma parses Prisma schema content and returns a Database model
+func (r *Reader) parsePrisma(content string) (*models.Database, error) {
+	db := models.InitDatabase("database")
+
+	if r.options.Metadata != nil {
+		if name, ok := r.options.Metadata["name"].(string); ok {
+			db.Name = name
+		}
+	}
+
+	// Default schema for Prisma (doesn't have explicit schema concept in most cases)
+	schema := models.InitSchema("public")
+	schema.Enums = make([]*models.Enum, 0)
+
+	scanner := bufio.NewScanner(strings.NewReader(content))
+
+	// State tracking
+	var currentBlock string // "datasource", "generator", "model", "enum"
+	var currentTable *models.Table
+	var currentEnum *models.Enum
+	var blockContent []string
+
+	// Regex patterns
+	datasourceRegex := regexp.MustCompile(`^datasource\s+\w+\s*{`)
+	generatorRegex := regexp.MustCompile(`^generator\s+\w+\s*{`)
+	modelRegex := regexp.MustCompile(`^model\s+(\w+)\s*{`)
+	enumRegex := regexp.MustCompile(`^enum\s+(\w+)\s*{`)
+
+	for scanner.Scan() {
+		line := scanner.Text()
+		trimmed := strings.TrimSpace(line)
+
+		// Skip empty lines and comments
+		if trimmed == "" || strings.HasPrefix(trimmed, "//") {
+			continue
+		}
+
+		// Check for block start
+		if matches := datasourceRegex.FindStringSubmatch(trimmed); matches != nil {
+			currentBlock = "datasource"
+			blockContent = []string{}
+			continue
+		}
+
+		if matches := generatorRegex.FindStringSubmatch(trimmed); matches != nil {
+			currentBlock = "generator"
+			blockContent = []string{}
+			continue
+		}
+
+		if matches := modelRegex.FindStringSubmatch(trimmed); matches != nil {
+			currentBlock = "model"
+			tableName := matches[1]
+			currentTable = models.InitTable(tableName, "public")
+			blockContent = []string{}
+			continue
+		}
+
+		if matches := enumRegex.FindStringSubmatch(trimmed); matches != nil {
+			currentBlock = "enum"
+			enumName := matches[1]
+			currentEnum = &models.Enum{
+				Name:   enumName,
+				Schema: "public",
+				Values: make([]string, 0),
+			}
+			blockContent = []string{}
+			continue
+		}
+
+		// Check for block end
+		if trimmed == "}" {
+			switch currentBlock {
+			case "datasource":
+				r.parseDatasource(blockContent, db)
+			case "generator":
+				// We don't need to do anything with generator blocks
+			case "model":
+				if currentTable != nil {
+					r.parseModelFields(blockContent, currentTable)
+					schema.Tables = append(schema.Tables, currentTable)
+					currentTable = nil
+				}
+			case "enum":
+				if currentEnum != nil {
+					schema.Enums = append(schema.Enums, currentEnum)
+					currentEnum = nil
+				}
+			}
+			currentBlock = ""
+			blockContent = []string{}
+			continue
+		}
+
+		// Accumulate block content
+		if currentBlock != "" {
+			if currentBlock == "enum" && currentEnum != nil {
+				// For enums, just add the trimmed value
+				if trimmed != "" {
+					currentEnum.Values = append(currentEnum.Values, trimmed)
+				}
+			} else {
+				blockContent = append(blockContent, line)
+			}
+		}
+	}
+
+	// Second pass: resolve relationships
+	r.resolveRelationships(schema)
+
+	db.Schemas = append(db.Schemas, schema)
+	return db, nil
+}
+
+// parseDatasource extracts database type from datasource block
+func (r *Reader) parseDatasource(lines []string, db *models.Database) {
+	providerRegex := regexp.MustCompile(`provider\s*=\s*"?(\w+)"?`)
+
+	for _, line := range lines {
+		if matches := providerRegex.FindStringSubmatch(line); matches != nil {
+			provider := matches[1]
+			switch provider {
+			case "postgresql", "postgres":
+				db.DatabaseType = models.PostgresqlDatabaseType
+			case "mysql":
+				db.DatabaseType = "mysql"
+			case "sqlite":
+				db.DatabaseType = models.SqlLiteDatabaseType
+			case "sqlserver":
+				db.DatabaseType = models.MSSQLDatabaseType
+			default:
+				db.DatabaseType = models.PostgresqlDatabaseType
+			}
+			break
+		}
+	}
+}
+
+// parseModelFields parses model field definitions
+func (r *Reader) parseModelFields(lines []string, table *models.Table) {
+	fieldRegex := regexp.MustCompile(`^(\w+)\s+(\w+)(\?|\[\])?\s*(@.+)?`)
+	blockAttrRegex := regexp.MustCompile(`^@@(\w+)\((.*?)\)`)
+
+	for _, line := range lines {
+		trimmed := strings.TrimSpace(line)
+
+		// Skip empty lines and comments
+		if trimmed == "" || strings.HasPrefix(trimmed, "//") {
+			continue
+		}
+
+		// Check for block attributes (@@id, @@unique, @@index)
+		if matches := blockAttrRegex.FindStringSubmatch(trimmed); matches != nil {
+			attrName := matches[1]
+			attrContent := matches[2]
+			r.parseBlockAttribute(attrName, attrContent, table)
+			continue
+		}
+
+		// Parse field definition
+		if matches := fieldRegex.FindStringSubmatch(trimmed); matches != nil {
+			fieldName := matches[1]
+			fieldType := matches[2]
+			modifier := matches[3]      // ? or []
+			attributes := matches[4]    // @... part
+
+			column := r.parseField(fieldName, fieldType, modifier, attributes, table)
+			if column != nil {
+				table.Columns[column.Name] = column
+			}
+		}
+	}
+}
+
+// parseField parses a single field definition
+func (r *Reader) parseField(name, fieldType, modifier, attributes string, table *models.Table) *models.Column {
+	// Check if this is a relation field (array or references another model)
+	if modifier == "[]" {
+		// Array field - this is a relation field, not a column
+		// We'll handle this in relationship resolution
+		return nil
+	}
+
+	// Check if this is a non-primitive type (relation field)
+	// Note: We need to allow enum types through as they're like primitives
+	if !r.isPrimitiveType(fieldType) && !r.isEnumType(fieldType, table) {
+		// This is a relation field (e.g., user User), not a scalar column
+		// Only process this if it has @relation attribute (which means it's the owning side with FK)
+		// Otherwise skip it as it's just the inverse relation field
+		if attributes == "" || !strings.Contains(attributes, "@relation") {
+			return nil
+		}
+		// If it has @relation, we still don't create a column for it
+		// The actual FK column will be in the fields: [...] part of @relation
+		return nil
+	}
+
+	column := models.InitColumn(name, table.Name, table.Schema)
+
+	// Map Prisma type to SQL type
+	column.Type = r.prismaTypeToSQL(fieldType)
+
+	// Handle modifiers
+	if modifier == "?" {
+		column.NotNull = false
+	} else {
+		// Default: required fields are NOT NULL
+		column.NotNull = true
+	}
+
+	// Parse field attributes
+	if attributes != "" {
+		r.parseFieldAttributes(attributes, column, table)
+	}
+
+	return column
+}
+
+// prismaTypeToSQL converts Prisma types to SQL types
+func (r *Reader) prismaTypeToSQL(prismaType string) string {
+	typeMap := map[string]string{
+		"String":   "text",
+		"Boolean":  "boolean",
+		"Int":      "integer",
+		"BigInt":   "bigint",
+		"Float":    "double precision",
+		"Decimal":  "decimal",
+		"DateTime": "timestamp",
+		"Json":     "jsonb",
+		"Bytes":    "bytea",
+	}
+
+	if sqlType, ok := typeMap[prismaType]; ok {
+		return sqlType
+	}
+
+	// If not a built-in type, it might be an enum or model reference
+	// For enums, we'll use the enum name directly
+	return prismaType
+}
+
+// parseFieldAttributes parses field attributes like @id, @unique, @default
+func (r *Reader) parseFieldAttributes(attributes string, column *models.Column, table *models.Table) {
+	// @id attribute
+	if strings.Contains(attributes, "@id") {
+		column.IsPrimaryKey = true
+		column.NotNull = true
+	}
+
+	// @unique attribute
+	if regexp.MustCompile(`@unique\b`).MatchString(attributes) {
+		uniqueConstraint := models.InitConstraint(
+			fmt.Sprintf("uq_%s", column.Name),
+			models.UniqueConstraint,
+		)
+		uniqueConstraint.Schema = table.Schema
+		uniqueConstraint.Table = table.Name
+		uniqueConstraint.Columns = []string{column.Name}
+		table.Constraints[uniqueConstraint.Name] = uniqueConstraint
+	}
+
+	// @default attribute - extract value with balanced parentheses
+	if strings.Contains(attributes, "@default(") {
+		defaultValue := r.extractDefaultValue(attributes)
+		if defaultValue != "" {
+			r.parseDefaultValue(defaultValue, column)
+		}
+	}
+
+	// @updatedAt attribute - store in comment for now
+	if strings.Contains(attributes, "@updatedAt") {
+		if column.Comment != "" {
+			column.Comment += "; @updatedAt"
+		} else {
+			column.Comment = "@updatedAt"
+		}
+	}
+
+	// @relation attribute - we'll handle this in relationship resolution
+	// For now, just note that this field is part of a relation
+}
+
+// extractDefaultValue extracts the default value from @default(...) handling nested parentheses
+func (r *Reader) extractDefaultValue(attributes string) string {
+	idx := strings.Index(attributes, "@default(")
+	if idx == -1 {
+		return ""
+	}
+
+	start := idx + len("@default(")
+	depth := 1
+	i := start
+
+	for i < len(attributes) && depth > 0 {
+		if attributes[i] == '(' {
+			depth++
+		} else if attributes[i] == ')' {
+			depth--
+		}
+		i++
+	}
+
+	if depth == 0 {
+		return attributes[start : i-1]
+	}
+
+	return ""
+}
+
+// parseDefaultValue parses Prisma default value expressions
+func (r *Reader) parseDefaultValue(defaultExpr string, column *models.Column) {
+	defaultExpr = strings.TrimSpace(defaultExpr)
+
+	switch defaultExpr {
+	case "autoincrement()":
+		column.AutoIncrement = true
+	case "now()":
+		column.Default = "now()"
+	case "uuid()":
+		column.Default = "gen_random_uuid()"
+	case "cuid()":
+		// CUID is Prisma-specific, store in comment
+		if column.Comment != "" {
+			column.Comment += "; default(cuid())"
+		} else {
+			column.Comment = "default(cuid())"
+		}
+	case "true":
+		column.Default = true
+	case "false":
+		column.Default = false
+	default:
+		// Check if it's a string literal
+		if strings.HasPrefix(defaultExpr, "\"") && strings.HasSuffix(defaultExpr, "\"") {
+			column.Default = defaultExpr[1 : len(defaultExpr)-1]
+		} else if strings.HasPrefix(defaultExpr, "'") && strings.HasSuffix(defaultExpr, "'") {
+			column.Default = defaultExpr[1 : len(defaultExpr)-1]
+		} else {
+			// Try to parse as number or enum value
+			column.Default = defaultExpr
+		}
+	}
+}
+
+// parseBlockAttribute parses block-level attributes like @@id, @@unique, @@index
+func (r *Reader) parseBlockAttribute(attrName, content string, table *models.Table) {
+	// Extract column list from brackets [col1, col2]
+	colListRegex := regexp.MustCompile(`\[(.*?)\]`)
+	matches := colListRegex.FindStringSubmatch(content)
+	if matches == nil {
+		return
+	}
+
+	columnList := strings.Split(matches[1], ",")
+	columns := make([]string, 0)
+	for _, col := range columnList {
+		columns = append(columns, strings.TrimSpace(col))
+	}
+
+	switch attrName {
+	case "id":
+		// Composite primary key
+		for _, colName := range columns {
+			if col, exists := table.Columns[colName]; exists {
+				col.IsPrimaryKey = true
+				col.NotNull = true
+			}
+		}
+		// Also create a PK constraint
+		pkConstraint := models.InitConstraint(
+			fmt.Sprintf("pk_%s", table.Name),
+			models.PrimaryKeyConstraint,
+		)
+		pkConstraint.Schema = table.Schema
+		pkConstraint.Table = table.Name
+		pkConstraint.Columns = columns
+		table.Constraints[pkConstraint.Name] = pkConstraint
+
+	case "unique":
+		// Multi-column unique constraint
+		uniqueConstraint := models.InitConstraint(
+			fmt.Sprintf("uq_%s_%s", table.Name, strings.Join(columns, "_")),
+			models.UniqueConstraint,
+		)
+		uniqueConstraint.Schema = table.Schema
+		uniqueConstraint.Table = table.Name
+		uniqueConstraint.Columns = columns
+		table.Constraints[uniqueConstraint.Name] = uniqueConstraint
+
+	case "index":
+		// Index
+		index := models.InitIndex(
+			fmt.Sprintf("idx_%s_%s", table.Name, strings.Join(columns, "_")),
+			table.Name,
+			table.Schema,
+		)
+		index.Columns = columns
+		table.Indexes[index.Name] = index
+	}
+}
+
+// relationField stores information about a relation field for second-pass processing
+type relationField struct {
+	tableName      string
+	fieldName      string
+	relatedModel   string
+	isArray        bool
+	relationAttr   string
+}
+
+// resolveRelationships performs a second pass to resolve @relation attributes
+func (r *Reader) resolveRelationships(schema *models.Schema) {
+	// Build a map of table names for quick lookup
+	tableMap := make(map[string]*models.Table)
+	for _, table := range schema.Tables {
+		tableMap[table.Name] = table
+	}
+
+	// First, we need to re-parse to find relation fields
+	// We'll re-read the file to extract relation information
+	if r.options.FilePath == "" {
+		return
+	}
+
+	content, err := os.ReadFile(r.options.FilePath)
+	if err != nil {
+		return
+	}
+
+	relations := r.extractRelationFields(string(content))
+
+	// Process explicit @relation attributes to create FK constraints
+	for _, rel := range relations {
+		if rel.relationAttr != "" {
+			r.createConstraintFromRelation(rel, tableMap, schema)
+		}
+	}
+
+	// Detect implicit many-to-many relationships
+	r.detectImplicitManyToMany(relations, tableMap, schema)
+}
+
+// extractRelationFields extracts relation field information from the schema
+func (r *Reader) extractRelationFields(content string) []relationField {
+	relations := make([]relationField, 0)
+	scanner := bufio.NewScanner(strings.NewReader(content))
+
+	modelRegex := regexp.MustCompile(`^model\s+(\w+)\s*{`)
+	fieldRegex := regexp.MustCompile(`^(\w+)\s+(\w+)(\?|\[\])?\s*(@.+)?`)
+
+	var currentModel string
+	inModel := false
+
+	for scanner.Scan() {
+		line := scanner.Text()
+		trimmed := strings.TrimSpace(line)
+
+		if trimmed == "" || strings.HasPrefix(trimmed, "//") {
+			continue
+		}
+
+		if matches := modelRegex.FindStringSubmatch(trimmed); matches != nil {
+			currentModel = matches[1]
+			inModel = true
+			continue
+		}
+
+		if trimmed == "}" {
+			inModel = false
+			currentModel = ""
+			continue
+		}
+
+		if inModel && currentModel != "" {
+			if matches := fieldRegex.FindStringSubmatch(trimmed); matches != nil {
+				fieldName := matches[1]
+				fieldType := matches[2]
+				modifier := matches[3]
+				attributes := matches[4]
+
+				// Check if this is a relation field (references another model or is an array)
+				isPotentialRelation := modifier == "[]" || !r.isPrimitiveType(fieldType)
+
+				if isPotentialRelation {
+					rel := relationField{
+						tableName:    currentModel,
+						fieldName:    fieldName,
+						relatedModel: fieldType,
+						isArray:      modifier == "[]",
+						relationAttr: attributes,
+					}
+					relations = append(relations, rel)
+				}
+			}
+		}
+	}
+
+	return relations
+}
+
+// isPrimitiveType checks if a type is a Prisma primitive type
+func (r *Reader) isPrimitiveType(typeName string) bool {
+	primitives := []string{"String", "Boolean", "Int", "BigInt", "Float", "Decimal", "DateTime", "Json", "Bytes"}
+	for _, p := range primitives {
+		if typeName == p {
+			return true
+		}
+	}
+	return false
+}
+
+// isEnumType checks if a type name might be an enum
+// Note: We can't definitively check against schema.Enums at parse time
+// because enums might be defined after the model, so we just check
+// if it starts with uppercase (Prisma convention for enums)
+func (r *Reader) isEnumType(typeName string, table *models.Table) bool {
+	// Simple heuristic: enum types start with uppercase letter
+	// and are not known model names (though we can't check that yet)
+	if len(typeName) > 0 && typeName[0] >= 'A' && typeName[0] <= 'Z' {
+		// Additional check: primitive types are already handled above
+		// So if it's uppercase and not primitive, it's likely an enum or model
+		// We'll assume it's an enum if it's a single word
+		return !strings.Contains(typeName, "_")
+	}
+	return false
+}
+
+// createConstraintFromRelation creates a FK constraint from a @relation attribute
+func (r *Reader) createConstraintFromRelation(rel relationField, tableMap map[string]*models.Table, schema *models.Schema) {
+	// Skip array fields (they are the inverse side of the relation)
+	if rel.isArray {
+		return
+	}
+
+	if rel.relationAttr == "" {
+		return
+	}
+
+	// Parse @relation attribute
+	relationRegex := regexp.MustCompile(`@relation\((.*?)\)`)
+	matches := relationRegex.FindStringSubmatch(rel.relationAttr)
+	if matches == nil {
+		return
+	}
+
+	relationContent := matches[1]
+
+	// Extract fields and references
+	fieldsRegex := regexp.MustCompile(`fields:\s*\[(.*?)\]`)
+	referencesRegex := regexp.MustCompile(`references:\s*\[(.*?)\]`)
+	nameRegex := regexp.MustCompile(`name:\s*"([^"]+)"`)
+	onDeleteRegex := regexp.MustCompile(`onDelete:\s*(\w+)`)
+	onUpdateRegex := regexp.MustCompile(`onUpdate:\s*(\w+)`)
+
+	fieldsMatch := fieldsRegex.FindStringSubmatch(relationContent)
+	referencesMatch := referencesRegex.FindStringSubmatch(relationContent)
+
+	if fieldsMatch == nil || referencesMatch == nil {
+		return
+	}
+
+	// Parse field and reference column lists
+	fieldCols := r.parseColumnList(fieldsMatch[1])
+	refCols := r.parseColumnList(referencesMatch[1])
+
+	if len(fieldCols) == 0 || len(refCols) == 0 {
+		return
+	}
+
+	// Create FK constraint
+	constraintName := fmt.Sprintf("fk_%s_%s", rel.tableName, fieldCols[0])
+
+	// Check for custom name
+	if nameMatch := nameRegex.FindStringSubmatch(relationContent); nameMatch != nil {
+		constraintName = nameMatch[1]
+	}
+
+	constraint := models.InitConstraint(constraintName, models.ForeignKeyConstraint)
+	constraint.Schema = "public"
+	constraint.Table = rel.tableName
+	constraint.Columns = fieldCols
+	constraint.ReferencedSchema = "public"
+	constraint.ReferencedTable = rel.relatedModel
+	constraint.ReferencedColumns = refCols
+
+	// Parse referential actions
+	if onDeleteMatch := onDeleteRegex.FindStringSubmatch(relationContent); onDeleteMatch != nil {
+		constraint.OnDelete = onDeleteMatch[1]
+	}
+
+	if onUpdateMatch := onUpdateRegex.FindStringSubmatch(relationContent); onUpdateMatch != nil {
+		constraint.OnUpdate = onUpdateMatch[1]
+	}
+
+	// Add constraint to table
+	if table, exists := tableMap[rel.tableName]; exists {
+		table.Constraints[constraint.Name] = constraint
+	}
+}
+
+// parseColumnList parses a comma-separated list of column names
+func (r *Reader) parseColumnList(list string) []string {
+	parts := strings.Split(list, ",")
+	result := make([]string, 0)
+	for _, part := range parts {
+		trimmed := strings.TrimSpace(part)
+		if trimmed != "" {
+			result = append(result, trimmed)
+		}
+	}
+	return result
+}
+
+// detectImplicitManyToMany detects implicit M2M relationships and creates join tables
+func (r *Reader) detectImplicitManyToMany(relations []relationField, tableMap map[string]*models.Table, schema *models.Schema) {
+	// Group relations by model pairs
+	type modelPair struct {
+		model1 string
+		model2 string
+	}
+
+	pairMap := make(map[modelPair][]relationField)
+
+	for _, rel := range relations {
+		if !rel.isArray || rel.relationAttr != "" {
+			// Skip non-array fields and explicit relations
+			continue
+		}
+
+		// Create a normalized pair (alphabetically sorted to avoid duplicates)
+		pair := modelPair{}
+		if rel.tableName < rel.relatedModel {
+			pair.model1 = rel.tableName
+			pair.model2 = rel.relatedModel
+		} else {
+			pair.model1 = rel.relatedModel
+			pair.model2 = rel.tableName
+		}
+
+		pairMap[pair] = append(pairMap[pair], rel)
+	}
+
+	// Check for pairs with arrays on both sides (implicit M2M)
+	for pair, rels := range pairMap {
+		if len(rels) >= 2 {
+			// This is an implicit many-to-many relationship
+			r.createImplicitJoinTable(pair.model1, pair.model2, tableMap, schema)
+		}
+	}
+}
+
+// createImplicitJoinTable creates a virtual join table for implicit M2M relations
+func (r *Reader) createImplicitJoinTable(model1, model2 string, tableMap map[string]*models.Table, schema *models.Schema) {
+	// Prisma naming convention: _Model1ToModel2 (alphabetically sorted)
+	joinTableName := fmt.Sprintf("_%sTo%s", model1, model2)
+
+	// Check if join table already exists
+	if _, exists := tableMap[joinTableName]; exists {
+		return
+	}
+
+	// Create join table
+	joinTable := models.InitTable(joinTableName, "public")
+
+	// Get primary keys from both tables
+	pk1 := r.getPrimaryKeyColumn(tableMap[model1])
+	pk2 := r.getPrimaryKeyColumn(tableMap[model2])
+
+	if pk1 == nil || pk2 == nil {
+		return // Can't create join table without PKs
+	}
+
+	// Create FK columns in join table
+	fkCol1Name := fmt.Sprintf("%sId", model1)
+	fkCol1 := models.InitColumn(fkCol1Name, joinTableName, "public")
+	fkCol1.Type = pk1.Type
+	fkCol1.NotNull = true
+	joinTable.Columns[fkCol1Name] = fkCol1
+
+	fkCol2Name := fmt.Sprintf("%sId", model2)
+	fkCol2 := models.InitColumn(fkCol2Name, joinTableName, "public")
+	fkCol2.Type = pk2.Type
+	fkCol2.NotNull = true
+	joinTable.Columns[fkCol2Name] = fkCol2
+
+	// Create composite primary key
+	pkConstraint := models.InitConstraint(
+		fmt.Sprintf("pk_%s", joinTableName),
+		models.PrimaryKeyConstraint,
+	)
+	pkConstraint.Schema = "public"
+	pkConstraint.Table = joinTableName
+	pkConstraint.Columns = []string{fkCol1Name, fkCol2Name}
+	joinTable.Constraints[pkConstraint.Name] = pkConstraint
+
+	// Mark columns as PK
+	fkCol1.IsPrimaryKey = true
+	fkCol2.IsPrimaryKey = true
+
+	// Create FK constraints
+	fk1 := models.InitConstraint(
+		fmt.Sprintf("fk_%s_%s", joinTableName, model1),
+		models.ForeignKeyConstraint,
+	)
+	fk1.Schema = "public"
+	fk1.Table = joinTableName
+	fk1.Columns = []string{fkCol1Name}
+	fk1.ReferencedSchema = "public"
+	fk1.ReferencedTable = model1
+	fk1.ReferencedColumns = []string{pk1.Name}
+	fk1.OnDelete = "Cascade"
+	joinTable.Constraints[fk1.Name] = fk1
+
+	fk2 := models.InitConstraint(
+		fmt.Sprintf("fk_%s_%s", joinTableName, model2),
+		models.ForeignKeyConstraint,
+	)
+	fk2.Schema = "public"
+	fk2.Table = joinTableName
+	fk2.Columns = []string{fkCol2Name}
+	fk2.ReferencedSchema = "public"
+	fk2.ReferencedTable = model2
+	fk2.ReferencedColumns = []string{pk2.Name}
+	fk2.OnDelete = "Cascade"
+	joinTable.Constraints[fk2.Name] = fk2
+
+	// Add join table to schema
+	schema.Tables = append(schema.Tables, joinTable)
+	tableMap[joinTableName] = joinTable
+}
+
+// getPrimaryKeyColumn returns the primary key column of a table
+func (r *Reader) getPrimaryKeyColumn(table *models.Table) *models.Column {
+	if table == nil {
+		return nil
+	}
+
+	for _, col := range table.Columns {
+		if col.IsPrimaryKey {
+			return col
+		}
+	}
+
+	return nil
+}

pkg/readers/typeorm/reader.go

@@ -0,0 +1,785 @@
+package typeorm
+
+import (
+	"bufio"
+	"fmt"
+	"os"
+	"regexp"
+	"strings"
+
+	"git.warky.dev/wdevs/relspecgo/pkg/models"
+	"git.warky.dev/wdevs/relspecgo/pkg/readers"
+)
+
+// Reader implements the readers.Reader interface for TypeORM entity files
+type Reader struct {
+	options *readers.ReaderOptions
+}
+
+// NewReader creates a new TypeORM reader with the given options
+func NewReader(options *readers.ReaderOptions) *Reader {
+	return &Reader{
+		options: options,
+	}
+}
+
+// ReadDatabase reads and parses TypeORM entity files, returning a Database model
+func (r *Reader) ReadDatabase() (*models.Database, error) {
+	if r.options.FilePath == "" {
+		return nil, fmt.Errorf("file path is required for TypeORM reader")
+	}
+
+	content, err := os.ReadFile(r.options.FilePath)
+	if err != nil {
+		return nil, fmt.Errorf("failed to read file: %w", err)
+	}
+
+	return r.parseTypeORM(string(content))
+}
+
+// ReadSchema reads and parses TypeORM entity files, returning a Schema model
+func (r *Reader) ReadSchema() (*models.Schema, error) {
+	db, err := r.ReadDatabase()
+	if err != nil {
+		return nil, err
+	}
+
+	if len(db.Schemas) == 0 {
+		return nil, fmt.Errorf("no schemas found in TypeORM entities")
+	}
+
+	return db.Schemas[0], nil
+}
+
+// ReadTable reads and parses TypeORM entity files, returning a Table model
+func (r *Reader) ReadTable() (*models.Table, error) {
+	schema, err := r.ReadSchema()
+	if err != nil {
+		return nil, err
+	}
+
+	if len(schema.Tables) == 0 {
+		return nil, fmt.Errorf("no tables found in TypeORM entities")
+	}
+
+	return schema.Tables[0], nil
+}
+
+// entityInfo stores information about an entity during parsing
+type entityInfo struct {
+	name       string
+	fields     []fieldInfo
+	decorators []string
+}
+
+// fieldInfo stores information about a field during parsing
+type fieldInfo struct {
+	name       string
+	typeName   string
+	decorators []string
+}
+
+// parseTypeORM parses TypeORM entity content and returns a Database model
+func (r *Reader) parseTypeORM(content string) (*models.Database, error) {
+	db := models.InitDatabase("database")
+	schema := models.InitSchema("public")
+
+	// Parse entities
+	entities := r.extractEntities(content)
+
+	// Convert entities to tables and views
+	tableMap := make(map[string]*models.Table)
+	for _, entity := range entities {
+		// Check if this is a view
+		isView := false
+		for _, decorator := range entity.decorators {
+			if strings.HasPrefix(decorator, "@ViewEntity") {
+				isView = true
+				break
+			}
+		}
+
+		if isView {
+			view := r.entityToView(entity)
+			schema.Views = append(schema.Views, view)
+		} else {
+			table := r.entityToTable(entity)
+			schema.Tables = append(schema.Tables, table)
+			tableMap[table.Name] = table
+		}
+	}
+
+	// Second pass: resolve relationships
+	r.resolveRelationships(entities, tableMap, schema)
+
+	db.Schemas = append(db.Schemas, schema)
+	return db, nil
+}
+
+// extractEntities extracts entity and view definitions from TypeORM content
+func (r *Reader) extractEntities(content string) []entityInfo {
+	entities := make([]entityInfo, 0)
+
+	// First, extract decorators properly (handling multi-line)
+	content = r.normalizeDecorators(content)
+
+	scanner := bufio.NewScanner(strings.NewReader(content))
+
+	entityRegex := regexp.MustCompile(`^export\s+class\s+(\w+)`)
+	decoratorRegex := regexp.MustCompile(`^\s*@(\w+)(\([^)]*\))?`)
+	fieldRegex := regexp.MustCompile(`^\s*(\w+):\s*([^;]+);`)
+
+	var currentEntity *entityInfo
+	var pendingDecorators []string
+	inClass := false
+
+	for scanner.Scan() {
+		line := scanner.Text()
+		trimmed := strings.TrimSpace(line)
+
+		// Skip empty lines and comments
+		if trimmed == "" || strings.HasPrefix(trimmed, "//") || strings.HasPrefix(trimmed, "import ") {
+			continue
+		}
+
+		// Check for decorator
+		if matches := decoratorRegex.FindStringSubmatch(trimmed); matches != nil {
+			decorator := matches[0]
+			pendingDecorators = append(pendingDecorators, decorator)
+			continue
+		}
+
+		// Check for entity/view class
+		if matches := entityRegex.FindStringSubmatch(trimmed); matches != nil {
+			// Save previous entity if exists
+			if currentEntity != nil {
+				entities = append(entities, *currentEntity)
+			}
+
+			currentEntity = &entityInfo{
+				name:       matches[1],
+				fields:     make([]fieldInfo, 0),
+				decorators: pendingDecorators,
+			}
+			pendingDecorators = []string{}
+			inClass = true
+			continue
+		}
+
+		// Check for class end
+		if inClass && trimmed == "}" {
+			if currentEntity != nil {
+				entities = append(entities, *currentEntity)
+				currentEntity = nil
+			}
+			inClass = false
+			pendingDecorators = []string{}
+			continue
+		}
+
+		// Check for field definition
+		if inClass && currentEntity != nil {
+			if matches := fieldRegex.FindStringSubmatch(trimmed); matches != nil {
+				fieldName := matches[1]
+				fieldType := strings.TrimSpace(matches[2])
+
+				field := fieldInfo{
+					name:       fieldName,
+					typeName:   fieldType,
+					decorators: pendingDecorators,
+				}
+				currentEntity.fields = append(currentEntity.fields, field)
+				pendingDecorators = []string{}
+			}
+		}
+	}
+
+	// Save last entity
+	if currentEntity != nil {
+		entities = append(entities, *currentEntity)
+	}
+
+	return entities
+}
+
+// normalizeDecorators combines multi-line decorators into single lines
+func (r *Reader) normalizeDecorators(content string) string {
+	// Replace multi-line decorators with single-line versions
+	// Match @Decorator({ ... }) across multiple lines
+	decoratorRegex := regexp.MustCompile(`@(\w+)\s*\(\s*\{([^}]*)\}\s*\)`)
+
+	return decoratorRegex.ReplaceAllStringFunc(content, func(match string) string {
+		// Remove newlines and extra spaces from decorator
+		match = strings.ReplaceAll(match, "\n", " ")
+		match = strings.ReplaceAll(match, "\r", " ")
+		// Normalize multiple spaces
+		spaceRegex := regexp.MustCompile(`\s+`)
+		match = spaceRegex.ReplaceAllString(match, " ")
+		return match
+	})
+}
+
+// entityToView converts a view entity to a view
+func (r *Reader) entityToView(entity entityInfo) *models.View {
+	// Parse @ViewEntity decorator options
+	viewName := entity.name
+	schemaName := "public"
+	var expression string
+
+	for _, decorator := range entity.decorators {
+		if strings.HasPrefix(decorator, "@ViewEntity") {
+			// Extract options from @ViewEntity({ ... })
+			options := r.parseViewEntityOptions(decorator)
+
+			// Check for custom view name
+			if name, ok := options["name"]; ok {
+				viewName = name
+			}
+
+			// Check for schema
+			if schema, ok := options["schema"]; ok {
+				schemaName = schema
+			}
+
+			// Check for expression (SQL definition)
+			if expr, ok := options["expression"]; ok {
+				expression = expr
+			}
+			break
+		}
+	}
+
+	view := models.InitView(viewName, schemaName)
+	view.Definition = expression
+
+	// Add columns from fields (if any are defined in the view class)
+	for _, field := range entity.fields {
+		column := models.InitColumn(field.name, viewName, schemaName)
+		column.Type = r.typeScriptTypeToSQL(field.typeName)
+		view.Columns[column.Name] = column
+	}
+
+	return view
+}
+
+// parseViewEntityOptions parses @ViewEntity decorator options
+func (r *Reader) parseViewEntityOptions(decorator string) map[string]string {
+	options := make(map[string]string)
+
+	// Extract content between parentheses
+	start := strings.Index(decorator, "(")
+	end := strings.LastIndex(decorator, ")")
+
+	if start == -1 || end == -1 || start >= end {
+		return options
+	}
+
+	content := decorator[start+1 : end]
+
+	// Skip if empty @ViewEntity()
+	if strings.TrimSpace(content) == "" {
+		return options
+	}
+
+	// Parse name: "value"
+	nameRegex := regexp.MustCompile(`name:\s*["']([^"']+)["']`)
+	if matches := nameRegex.FindStringSubmatch(content); matches != nil {
+		options["name"] = matches[1]
+	}
+
+	// Parse schema: "value"
+	schemaRegex := regexp.MustCompile(`schema:\s*["']([^"']+)["']`)
+	if matches := schemaRegex.FindStringSubmatch(content); matches != nil {
+		options["schema"] = matches[1]
+	}
+
+	// Parse expression: ` ... ` (can be multi-line, captured as single line after normalization)
+	// Look for expression followed by backtick or quote
+	expressionRegex := regexp.MustCompile(`expression:\s*` + "`" + `([^` + "`" + `]+)` + "`")
+	if matches := expressionRegex.FindStringSubmatch(content); matches != nil {
+		options["expression"] = strings.TrimSpace(matches[1])
+	} else {
+		// Try with regular quotes
+		expressionRegex = regexp.MustCompile(`expression:\s*["']([^"']+)["']`)
+		if matches := expressionRegex.FindStringSubmatch(content); matches != nil {
+			options["expression"] = strings.TrimSpace(matches[1])
+		}
+	}
+
+	return options
+}
+
+// entityToTable converts an entity to a table
+func (r *Reader) entityToTable(entity entityInfo) *models.Table {
+	// Parse @Entity decorator options
+	tableName := entity.name
+	schemaName := "public"
+	var entityOptions map[string]string
+
+	for _, decorator := range entity.decorators {
+		if strings.HasPrefix(decorator, "@Entity") {
+			// Extract options from @Entity({ ... })
+			entityOptions = r.parseEntityOptions(decorator)
+
+			// Check for custom table name
+			if name, ok := entityOptions["name"]; ok {
+				tableName = name
+			}
+
+			// Check for schema
+			if schema, ok := entityOptions["schema"]; ok {
+				schemaName = schema
+			}
+			break
+		}
+	}
+
+	table := models.InitTable(tableName, schemaName)
+
+	// Store additional metadata from @Entity options
+	if entityOptions != nil {
+		// Store database name in metadata
+		if database, ok := entityOptions["database"]; ok {
+			if table.Metadata == nil {
+				table.Metadata = make(map[string]any)
+			}
+			table.Metadata["database"] = database
+		}
+
+		// Store engine in metadata
+		if engine, ok := entityOptions["engine"]; ok {
+			if table.Metadata == nil {
+				table.Metadata = make(map[string]any)
+			}
+			table.Metadata["engine"] = engine
+		}
+
+		// Store original class name if different from table name
+		if entity.name != tableName {
+			if table.Metadata == nil {
+				table.Metadata = make(map[string]any)
+			}
+			table.Metadata["class_name"] = entity.name
+		}
+	}
+
+	for _, field := range entity.fields {
+		// Skip relation fields (they'll be handled in relationship resolution)
+		if r.isRelationField(field) {
+			continue
+		}
+
+		column := r.fieldToColumn(field, table)
+		if column != nil {
+			table.Columns[column.Name] = column
+		}
+	}
+
+	return table
+}
+
+// parseEntityOptions parses @Entity decorator options
+func (r *Reader) parseEntityOptions(decorator string) map[string]string {
+	options := make(map[string]string)
+
+	// Extract content between parentheses
+	start := strings.Index(decorator, "(")
+	end := strings.LastIndex(decorator, ")")
+
+	if start == -1 || end == -1 || start >= end {
+		return options
+	}
+
+	content := decorator[start+1 : end]
+
+	// Skip if empty @Entity()
+	if strings.TrimSpace(content) == "" {
+		return options
+	}
+
+	// Parse name: "value" or name: 'value'
+	nameRegex := regexp.MustCompile(`name:\s*["']([^"']+)["']`)
+	if matches := nameRegex.FindStringSubmatch(content); matches != nil {
+		options["name"] = matches[1]
+	}
+
+	// Parse schema: "value"
+	schemaRegex := regexp.MustCompile(`schema:\s*["']([^"']+)["']`)
+	if matches := schemaRegex.FindStringSubmatch(content); matches != nil {
+		options["schema"] = matches[1]
+	}
+
+	// Parse database: "value"
+	databaseRegex := regexp.MustCompile(`database:\s*["']([^"']+)["']`)
+	if matches := databaseRegex.FindStringSubmatch(content); matches != nil {
+		options["database"] = matches[1]
+	}
+
+	// Parse engine: "value"
+	engineRegex := regexp.MustCompile(`engine:\s*["']([^"']+)["']`)
+	if matches := engineRegex.FindStringSubmatch(content); matches != nil {
+		options["engine"] = matches[1]
+	}
+
+	return options
+}
+
+// isRelationField checks if a field is a relation field
+func (r *Reader) isRelationField(field fieldInfo) bool {
+	for _, decorator := range field.decorators {
+		if strings.Contains(decorator, "@ManyToOne") ||
+			strings.Contains(decorator, "@OneToMany") ||
+			strings.Contains(decorator, "@ManyToMany") ||
+			strings.Contains(decorator, "@OneToOne") {
+			return true
+		}
+	}
+	return false
+}
+
+// fieldToColumn converts a field to a column
+func (r *Reader) fieldToColumn(field fieldInfo, table *models.Table) *models.Column {
+	column := models.InitColumn(field.name, table.Name, table.Schema)
+
+	// Map TypeScript type to SQL type
+	column.Type = r.typeScriptTypeToSQL(field.typeName)
+
+	// Default to NOT NULL
+	column.NotNull = true
+
+	// Parse decorators
+	for _, decorator := range field.decorators {
+		r.parseColumnDecorator(decorator, column, table)
+	}
+
+	return column
+}
+
+// typeScriptTypeToSQL converts TypeScript types to SQL types
+func (r *Reader) typeScriptTypeToSQL(tsType string) string {
+	// Remove array brackets and optional markers
+	tsType = strings.TrimSuffix(tsType, "[]")
+	tsType = strings.TrimSuffix(tsType, " | null")
+
+	typeMap := map[string]string{
+		"string":  "text",
+		"number":  "integer",
+		"boolean": "boolean",
+		"Date":    "timestamp",
+		"any":     "jsonb",
+	}
+
+	for tsPattern, sqlType := range typeMap {
+		if strings.Contains(tsType, tsPattern) {
+			return sqlType
+		}
+	}
+
+	// Default to text
+	return "text"
+}
+
+// parseColumnDecorator parses a column decorator
+func (r *Reader) parseColumnDecorator(decorator string, column *models.Column, table *models.Table) {
+	// @PrimaryGeneratedColumn
+	if strings.HasPrefix(decorator, "@PrimaryGeneratedColumn") {
+		column.IsPrimaryKey = true
+		column.NotNull = true
+
+		if strings.Contains(decorator, "'uuid'") {
+			column.Type = "uuid"
+			column.Default = "gen_random_uuid()"
+		} else if strings.Contains(decorator, "'increment'") || strings.Contains(decorator, "()") {
+			column.AutoIncrement = true
+		}
+		return
+	}
+
+	// @Column
+	if strings.HasPrefix(decorator, "@Column") {
+		r.parseColumnOptions(decorator, column, table)
+		return
+	}
+
+	// @CreateDateColumn
+	if strings.HasPrefix(decorator, "@CreateDateColumn") {
+		column.Type = "timestamp"
+		column.Default = "now()"
+		column.NotNull = true
+		return
+	}
+
+	// @UpdateDateColumn
+	if strings.HasPrefix(decorator, "@UpdateDateColumn") {
+		column.Type = "timestamp"
+		column.NotNull = true
+		if column.Comment != "" {
+			column.Comment += "; auto-update"
+		} else {
+			column.Comment = "auto-update"
+		}
+		return
+	}
+}
+
+// parseColumnOptions parses @Column decorator options
+func (r *Reader) parseColumnOptions(decorator string, column *models.Column, table *models.Table) {
+	// Extract content between parentheses
+	start := strings.Index(decorator, "(")
+	end := strings.LastIndex(decorator, ")")
+
+	if start == -1 || end == -1 || start >= end {
+		return
+	}
+
+	content := decorator[start+1 : end]
+
+	// Check for shorthand type: @Column('text')
+	if strings.HasPrefix(content, "'") || strings.HasPrefix(content, "\"") {
+		typeStr := strings.Trim(content, "'\"`")
+		column.Type = typeStr
+		return
+	}
+
+	// Parse options object
+	if strings.Contains(content, "type:") {
+		typeRegex := regexp.MustCompile(`type:\s*['"]([^'"]+)['"]`)
+		if matches := typeRegex.FindStringSubmatch(content); matches != nil {
+			column.Type = matches[1]
+		}
+	}
+
+	if strings.Contains(content, "nullable: true") || strings.Contains(content, "nullable:true") {
+		column.NotNull = false
+	}
+
+	if strings.Contains(content, "unique: true") || strings.Contains(content, "unique:true") {
+		uniqueConstraint := models.InitConstraint(
+			fmt.Sprintf("uq_%s", column.Name),
+			models.UniqueConstraint,
+		)
+		uniqueConstraint.Schema = table.Schema
+		uniqueConstraint.Table = table.Name
+		uniqueConstraint.Columns = []string{column.Name}
+		table.Constraints[uniqueConstraint.Name] = uniqueConstraint
+	}
+
+	if strings.Contains(content, "default:") {
+		defaultRegex := regexp.MustCompile(`default:\s*['"]?([^,}'"]+)['"]?`)
+		if matches := defaultRegex.FindStringSubmatch(content); matches != nil {
+			defaultValue := strings.TrimSpace(matches[1])
+			defaultValue = strings.Trim(defaultValue, "'\"")
+			column.Default = defaultValue
+		}
+	}
+}
+
+// resolveRelationships resolves TypeORM relationships
+func (r *Reader) resolveRelationships(entities []entityInfo, tableMap map[string]*models.Table, schema *models.Schema) {
+	// Track M2M relations that need join tables
+	type m2mRelation struct {
+		ownerEntity string
+		targetEntity string
+		ownerField string
+	}
+	m2mRelations := make([]m2mRelation, 0)
+
+	for _, entity := range entities {
+		table := tableMap[entity.name]
+		if table == nil {
+			continue
+		}
+
+		for _, field := range entity.fields {
+			// Handle @ManyToOne relations
+			if r.hasDecorator(field, "@ManyToOne") {
+				r.createManyToOneConstraint(field, entity.name, table, tableMap)
+			}
+
+			// Track @ManyToMany relations with @JoinTable
+			if r.hasDecorator(field, "@ManyToMany") && r.hasDecorator(field, "@JoinTable") {
+				targetEntity := r.extractRelationTarget(field)
+				if targetEntity != "" {
+					m2mRelations = append(m2mRelations, m2mRelation{
+						ownerEntity:  entity.name,
+						targetEntity: targetEntity,
+						ownerField:   field.name,
+					})
+				}
+			}
+		}
+	}
+
+	// Create join tables for M2M relations
+	for _, rel := range m2mRelations {
+		r.createManyToManyJoinTable(rel.ownerEntity, rel.targetEntity, tableMap, schema)
+	}
+}
+
+// hasDecorator checks if a field has a specific decorator
+func (r *Reader) hasDecorator(field fieldInfo, decoratorName string) bool {
+	for _, decorator := range field.decorators {
+		if strings.HasPrefix(decorator, decoratorName) {
+			return true
+		}
+	}
+	return false
+}
+
+// extractRelationTarget extracts the target entity from a relation decorator
+func (r *Reader) extractRelationTarget(field fieldInfo) string {
+	// Remove array brackets from type
+	targetType := strings.TrimSuffix(field.typeName, "[]")
+	targetType = strings.TrimSpace(targetType)
+	return targetType
+}
+
+// createManyToOneConstraint creates a foreign key constraint for @ManyToOne
+func (r *Reader) createManyToOneConstraint(field fieldInfo, entityName string, table *models.Table, tableMap map[string]*models.Table) {
+	targetEntity := r.extractRelationTarget(field)
+	if targetEntity == "" {
+		return
+	}
+
+	// Get target table to find its PK
+	targetTable := tableMap[targetEntity]
+	if targetTable == nil {
+		return
+	}
+
+	targetPK := r.getPrimaryKeyColumn(targetTable)
+	if targetPK == nil {
+		return
+	}
+
+	// Create FK column
+	fkColumnName := fmt.Sprintf("%sId", field.name)
+	fkColumn := models.InitColumn(fkColumnName, table.Name, table.Schema)
+	fkColumn.Type = targetPK.Type
+
+	// Check if nullable option is set in @ManyToOne decorator
+	isNullable := false
+	for _, decorator := range field.decorators {
+		if strings.Contains(decorator, "nullable: true") || strings.Contains(decorator, "nullable:true") {
+			isNullable = true
+			break
+		}
+	}
+	fkColumn.NotNull = !isNullable
+
+	table.Columns[fkColumnName] = fkColumn
+
+	// Create FK constraint
+	constraint := models.InitConstraint(
+		fmt.Sprintf("fk_%s_%s", entityName, field.name),
+		models.ForeignKeyConstraint,
+	)
+	constraint.Schema = table.Schema
+	constraint.Table = table.Name
+	constraint.Columns = []string{fkColumnName}
+	constraint.ReferencedSchema = "public"
+	constraint.ReferencedTable = targetEntity
+	constraint.ReferencedColumns = []string{targetPK.Name}
+	constraint.OnDelete = "CASCADE"
+
+	table.Constraints[constraint.Name] = constraint
+}
+
+// createManyToManyJoinTable creates a join table for M2M relations
+func (r *Reader) createManyToManyJoinTable(entity1, entity2 string, tableMap map[string]*models.Table, schema *models.Schema) {
+	// TypeORM naming convention: entity1_entity2_entity1field
+	// We'll simplify to entity1_entity2
+	joinTableName := fmt.Sprintf("%s_%s", strings.ToLower(entity1), strings.ToLower(entity2))
+
+	// Check if join table already exists
+	if _, exists := tableMap[joinTableName]; exists {
+		return
+	}
+
+	// Get PKs from both tables
+	table1 := tableMap[entity1]
+	table2 := tableMap[entity2]
+	if table1 == nil || table2 == nil {
+		return
+	}
+
+	pk1 := r.getPrimaryKeyColumn(table1)
+	pk2 := r.getPrimaryKeyColumn(table2)
+	if pk1 == nil || pk2 == nil {
+		return
+	}
+
+	// Create join table
+	joinTable := models.InitTable(joinTableName, "public")
+
+	// Create FK columns
+	fkCol1Name := fmt.Sprintf("%sId", strings.ToLower(entity1))
+	fkCol1 := models.InitColumn(fkCol1Name, joinTableName, "public")
+	fkCol1.Type = pk1.Type
+	fkCol1.NotNull = true
+	fkCol1.IsPrimaryKey = true
+	joinTable.Columns[fkCol1Name] = fkCol1
+
+	fkCol2Name := fmt.Sprintf("%sId", strings.ToLower(entity2))
+	fkCol2 := models.InitColumn(fkCol2Name, joinTableName, "public")
+	fkCol2.Type = pk2.Type
+	fkCol2.NotNull = true
+	fkCol2.IsPrimaryKey = true
+	joinTable.Columns[fkCol2Name] = fkCol2
+
+	// Create composite PK constraint
+	pkConstraint := models.InitConstraint(
+		fmt.Sprintf("pk_%s", joinTableName),
+		models.PrimaryKeyConstraint,
+	)
+	pkConstraint.Schema = "public"
+	pkConstraint.Table = joinTableName
+	pkConstraint.Columns = []string{fkCol1Name, fkCol2Name}
+	joinTable.Constraints[pkConstraint.Name] = pkConstraint
+
+	// Create FK constraints
+	fk1 := models.InitConstraint(
+		fmt.Sprintf("fk_%s_%s", joinTableName, entity1),
+		models.ForeignKeyConstraint,
+	)
+	fk1.Schema = "public"
+	fk1.Table = joinTableName
+	fk1.Columns = []string{fkCol1Name}
+	fk1.ReferencedSchema = "public"
+	fk1.ReferencedTable = entity1
+	fk1.ReferencedColumns = []string{pk1.Name}
+	fk1.OnDelete = "CASCADE"
+	joinTable.Constraints[fk1.Name] = fk1
+
+	fk2 := models.InitConstraint(
+		fmt.Sprintf("fk_%s_%s", joinTableName, entity2),
+		models.ForeignKeyConstraint,
+	)
+	fk2.Schema = "public"
+	fk2.Table = joinTableName
+	fk2.Columns = []string{fkCol2Name}
+	fk2.ReferencedSchema = "public"
+	fk2.ReferencedTable = entity2
+	fk2.ReferencedColumns = []string{pk2.Name}
+	fk2.OnDelete = "CASCADE"
+	joinTable.Constraints[fk2.Name] = fk2
+
+	// Add join table to schema
+	schema.Tables = append(schema.Tables, joinTable)
+	tableMap[joinTableName] = joinTable
+}
+
+// getPrimaryKeyColumn returns the primary key column of a table
+func (r *Reader) getPrimaryKeyColumn(table *models.Table) *models.Column {
+	if table == nil {
+		return nil
+	}
+
+	for _, col := range table.Columns {
+		if col.IsPrimaryKey {
+			return col
+		}
+	}
+
+	return nil
+}

pkg/writers/prisma/writer.go

@@ -0,0 +1,551 @@
+package prisma
+
+import (
+	"fmt"
+	"os"
+	"sort"
+	"strings"
+
+	"git.warky.dev/wdevs/relspecgo/pkg/models"
+	"git.warky.dev/wdevs/relspecgo/pkg/writers"
+)
+
+// Writer implements the writers.Writer interface for Prisma schema format
+type Writer struct {
+	options *writers.WriterOptions
+}
+
+// NewWriter creates a new Prisma writer with the given options
+func NewWriter(options *writers.WriterOptions) *Writer {
+	return &Writer{
+		options: options,
+	}
+}
+
+// WriteDatabase writes a Database model to Prisma schema format
+func (w *Writer) WriteDatabase(db *models.Database) error {
+	content := w.databaseToPrisma(db)
+
+	if w.options.OutputPath != "" {
+		return os.WriteFile(w.options.OutputPath, []byte(content), 0644)
+	}
+
+	fmt.Print(content)
+	return nil
+}
+
+// WriteSchema writes a Schema model to Prisma schema format
+func (w *Writer) WriteSchema(schema *models.Schema) error {
+	// Create temporary database for schema
+	db := models.InitDatabase("database")
+	db.Schemas = []*models.Schema{schema}
+
+	return w.WriteDatabase(db)
+}
+
+// WriteTable writes a Table model to Prisma schema format
+func (w *Writer) WriteTable(table *models.Table) error {
+	// Create temporary schema and database for table
+	schema := models.InitSchema(table.Schema)
+	schema.Tables = []*models.Table{table}
+
+	return w.WriteSchema(schema)
+}
+
+// databaseToPrisma converts a Database to Prisma schema format string
+func (w *Writer) databaseToPrisma(db *models.Database) string {
+	var sb strings.Builder
+
+	// Write datasource block
+	sb.WriteString(w.generateDatasource(db))
+	sb.WriteString("\n")
+
+	// Write generator block
+	sb.WriteString(w.generateGenerator())
+	sb.WriteString("\n")
+
+	// Process all schemas (typically just one in Prisma)
+	for _, schema := range db.Schemas {
+		// Write enums
+		if len(schema.Enums) > 0 {
+			for _, enum := range schema.Enums {
+				sb.WriteString(w.enumToPrisma(enum))
+				sb.WriteString("\n")
+			}
+		}
+
+		// Identify join tables for implicit M2M
+		joinTables := w.identifyJoinTables(schema)
+
+		// Write models (excluding join tables)
+		for _, table := range schema.Tables {
+			if joinTables[table.Name] {
+				continue // Skip join tables
+			}
+			sb.WriteString(w.tableToPrisma(table, schema, joinTables))
+			sb.WriteString("\n")
+		}
+	}
+
+	return sb.String()
+}
+
+// generateDatasource generates the datasource block
+func (w *Writer) generateDatasource(db *models.Database) string {
+	provider := "postgresql"
+
+	// Map database type to Prisma provider
+	switch db.DatabaseType {
+	case models.PostgresqlDatabaseType:
+		provider = "postgresql"
+	case models.MSSQLDatabaseType:
+		provider = "sqlserver"
+	case models.SqlLiteDatabaseType:
+		provider = "sqlite"
+	case "mysql":
+		provider = "mysql"
+	}
+
+	return fmt.Sprintf(`datasource db {
+  provider = "%s"
+  url      = env("DATABASE_URL")
+}
+`, provider)
+}
+
+// generateGenerator generates the generator block
+func (w *Writer) generateGenerator() string {
+	return `generator client {
+  provider = "prisma-client-js"
+}
+`
+}
+
+// enumToPrisma converts an Enum to Prisma enum block
+func (w *Writer) enumToPrisma(enum *models.Enum) string {
+	var sb strings.Builder
+
+	sb.WriteString(fmt.Sprintf("enum %s {\n", enum.Name))
+	for _, value := range enum.Values {
+		sb.WriteString(fmt.Sprintf("  %s\n", value))
+	}
+	sb.WriteString("}\n")
+
+	return sb.String()
+}
+
+// identifyJoinTables identifies tables that are join tables for M2M relations
+func (w *Writer) identifyJoinTables(schema *models.Schema) map[string]bool {
+	joinTables := make(map[string]bool)
+
+	for _, table := range schema.Tables {
+		// Check if this is a join table:
+		// 1. Starts with _ (Prisma convention)
+		// 2. Has exactly 2 FK constraints
+		// 3. Has composite PK with those 2 columns
+		// 4. Has no other columns except the FK columns
+
+		if !strings.HasPrefix(table.Name, "_") {
+			continue
+		}
+
+		fks := table.GetForeignKeys()
+		if len(fks) != 2 {
+			continue
+		}
+
+		// Check if columns are only the FK columns
+		if len(table.Columns) != 2 {
+			continue
+		}
+
+		// Check if both FK columns are part of PK
+		pkCols := 0
+		for _, col := range table.Columns {
+			if col.IsPrimaryKey {
+				pkCols++
+			}
+		}
+
+		if pkCols == 2 {
+			joinTables[table.Name] = true
+		}
+	}
+
+	return joinTables
+}
+
+// tableToPrisma converts a Table to Prisma model block
+func (w *Writer) tableToPrisma(table *models.Table, schema *models.Schema, joinTables map[string]bool) string {
+	var sb strings.Builder
+
+	sb.WriteString(fmt.Sprintf("model %s {\n", table.Name))
+
+	// Collect columns to write
+	columns := make([]*models.Column, 0, len(table.Columns))
+	for _, col := range table.Columns {
+		columns = append(columns, col)
+	}
+
+	// Sort columns for consistent output
+	sort.Slice(columns, func(i, j int) bool {
+		return columns[i].Name < columns[j].Name
+	})
+
+	// Write scalar fields
+	for _, col := range columns {
+		// Skip if this column is part of a relation that will be output as array field
+		if w.isRelationColumn(col, table) {
+			// We'll output this with the relation field
+			continue
+		}
+
+		sb.WriteString(w.columnToField(col, table, schema))
+	}
+
+	// Write relation fields
+	sb.WriteString(w.generateRelationFields(table, schema, joinTables))
+
+	// Write block attributes (@@id, @@unique, @@index)
+	sb.WriteString(w.generateBlockAttributes(table))
+
+	sb.WriteString("}\n")
+
+	return sb.String()
+}
+
+// columnToField converts a Column to a Prisma field definition
+func (w *Writer) columnToField(col *models.Column, table *models.Table, schema *models.Schema) string {
+	var sb strings.Builder
+
+	// Field name
+	sb.WriteString(fmt.Sprintf("  %s", col.Name))
+
+	// Field type
+	prismaType := w.sqlTypeToPrisma(col.Type, schema)
+	sb.WriteString(fmt.Sprintf(" %s", prismaType))
+
+	// Optional modifier
+	if !col.NotNull && !col.IsPrimaryKey {
+		sb.WriteString("?")
+	}
+
+	// Field attributes
+	attributes := w.generateFieldAttributes(col, table)
+	if attributes != "" {
+		sb.WriteString(" ")
+		sb.WriteString(attributes)
+	}
+
+	sb.WriteString("\n")
+
+	return sb.String()
+}
+
+// sqlTypeToPrisma converts SQL types to Prisma types
+func (w *Writer) sqlTypeToPrisma(sqlType string, schema *models.Schema) string {
+	// Check if it's an enum
+	for _, enum := range schema.Enums {
+		if strings.EqualFold(sqlType, enum.Name) {
+			return enum.Name
+		}
+	}
+
+	// Standard type mapping
+	typeMap := map[string]string{
+		"text":              "String",
+		"varchar":           "String",
+		"character varying": "String",
+		"char":              "String",
+		"boolean":           "Boolean",
+		"bool":              "Boolean",
+		"integer":           "Int",
+		"int":               "Int",
+		"int4":              "Int",
+		"bigint":            "BigInt",
+		"int8":              "BigInt",
+		"double precision":  "Float",
+		"float":             "Float",
+		"float8":            "Float",
+		"decimal":           "Decimal",
+		"numeric":           "Decimal",
+		"timestamp":         "DateTime",
+		"timestamptz":       "DateTime",
+		"date":              "DateTime",
+		"jsonb":             "Json",
+		"json":              "Json",
+		"bytea":             "Bytes",
+	}
+
+	for sqlPattern, prismaType := range typeMap {
+		if strings.Contains(strings.ToLower(sqlType), sqlPattern) {
+			return prismaType
+		}
+	}
+
+	// Default to String for unknown types
+	return "String"
+}
+
+// generateFieldAttributes generates field attributes like @id, @unique, @default
+func (w *Writer) generateFieldAttributes(col *models.Column, table *models.Table) string {
+	attrs := make([]string, 0)
+
+	// @id
+	if col.IsPrimaryKey {
+		// Check if this is part of a composite key
+		pkCount := 0
+		for _, c := range table.Columns {
+			if c.IsPrimaryKey {
+				pkCount++
+			}
+		}
+		if pkCount == 1 {
+			attrs = append(attrs, "@id")
+		}
+	}
+
+	// @unique
+	if w.hasUniqueConstraint(col.Name, table) {
+		attrs = append(attrs, "@unique")
+	}
+
+	// @default
+	if col.AutoIncrement {
+		attrs = append(attrs, "@default(autoincrement())")
+	} else if col.Default != nil {
+		defaultAttr := w.formatDefaultValue(col.Default)
+		if defaultAttr != "" {
+			attrs = append(attrs, fmt.Sprintf("@default(%s)", defaultAttr))
+		}
+	}
+
+	// @updatedAt (check comment)
+	if strings.Contains(col.Comment, "@updatedAt") {
+		attrs = append(attrs, "@updatedAt")
+	}
+
+	return strings.Join(attrs, " ")
+}
+
+// formatDefaultValue formats a default value for Prisma
+func (w *Writer) formatDefaultValue(defaultValue any) string {
+	switch v := defaultValue.(type) {
+	case string:
+		if v == "now()" {
+			return "now()"
+		} else if v == "gen_random_uuid()" {
+			return "uuid()"
+		} else if strings.Contains(strings.ToLower(v), "uuid") {
+			return "uuid()"
+		} else {
+			// String literal
+			return fmt.Sprintf(`"%s"`, v)
+		}
+	case bool:
+		if v {
+			return "true"
+		}
+		return "false"
+	case int, int64, int32:
+		return fmt.Sprintf("%v", v)
+	default:
+		return fmt.Sprintf("%v", v)
+	}
+}
+
+// hasUniqueConstraint checks if a column has a unique constraint
+func (w *Writer) hasUniqueConstraint(colName string, table *models.Table) bool {
+	for _, constraint := range table.Constraints {
+		if constraint.Type == models.UniqueConstraint &&
+			len(constraint.Columns) == 1 &&
+			constraint.Columns[0] == colName {
+			return true
+		}
+	}
+	return false
+}
+
+// isRelationColumn checks if a column is a FK column
+func (w *Writer) isRelationColumn(col *models.Column, table *models.Table) bool {
+	for _, constraint := range table.Constraints {
+		if constraint.Type == models.ForeignKeyConstraint {
+			for _, fkCol := range constraint.Columns {
+				if fkCol == col.Name {
+					return true
+				}
+			}
+		}
+	}
+	return false
+}
+
+// generateRelationFields generates relation fields and their FK columns
+func (w *Writer) generateRelationFields(table *models.Table, schema *models.Schema, joinTables map[string]bool) string {
+	var sb strings.Builder
+
+	// Get all FK constraints
+	fks := table.GetForeignKeys()
+
+	for _, fk := range fks {
+		// Generate the FK scalar field
+		for _, fkCol := range fk.Columns {
+			if col, exists := table.Columns[fkCol]; exists {
+				sb.WriteString(w.columnToField(col, table, schema))
+			}
+		}
+
+		// Generate the relation field
+		relationType := fk.ReferencedTable
+		isOptional := false
+
+		// Check if FK column is nullable
+		for _, fkCol := range fk.Columns {
+			if col, exists := table.Columns[fkCol]; exists {
+				if !col.NotNull {
+					isOptional = true
+				}
+			}
+		}
+
+		relationName := relationType
+		if strings.HasSuffix(strings.ToLower(relationName), "s") {
+			relationName = relationName[:len(relationName)-1]
+		}
+
+		sb.WriteString(fmt.Sprintf("  %s %s", strings.ToLower(relationName), relationType))
+
+		if isOptional {
+			sb.WriteString("?")
+		}
+
+		// @relation attribute
+		relationAttr := w.generateRelationAttribute(fk)
+		if relationAttr != "" {
+			sb.WriteString(" ")
+			sb.WriteString(relationAttr)
+		}
+
+		sb.WriteString("\n")
+	}
+
+	// Generate inverse relations (arrays) for tables that reference this one
+	sb.WriteString(w.generateInverseRelations(table, schema, joinTables))
+
+	return sb.String()
+}
+
+// generateRelationAttribute generates the @relation(...) attribute
+func (w *Writer) generateRelationAttribute(fk *models.Constraint) string {
+	parts := make([]string, 0)
+
+	// fields
+	fieldsStr := strings.Join(fk.Columns, ", ")
+	parts = append(parts, fmt.Sprintf("fields: [%s]", fieldsStr))
+
+	// references
+	referencesStr := strings.Join(fk.ReferencedColumns, ", ")
+	parts = append(parts, fmt.Sprintf("references: [%s]", referencesStr))
+
+	// onDelete
+	if fk.OnDelete != "" {
+		parts = append(parts, fmt.Sprintf("onDelete: %s", fk.OnDelete))
+	}
+
+	// onUpdate
+	if fk.OnUpdate != "" {
+		parts = append(parts, fmt.Sprintf("onUpdate: %s", fk.OnUpdate))
+	}
+
+	return fmt.Sprintf("@relation(%s)", strings.Join(parts, ", "))
+}
+
+// generateInverseRelations generates array fields for reverse relationships
+func (w *Writer) generateInverseRelations(table *models.Table, schema *models.Schema, joinTables map[string]bool) string {
+	var sb strings.Builder
+
+	// Find all tables that have FKs pointing to this table
+	for _, otherTable := range schema.Tables {
+		if otherTable.Name == table.Name {
+			continue
+		}
+
+		// Check if this is a join table
+		if joinTables[otherTable.Name] {
+			// Handle implicit M2M
+			if w.isJoinTableFor(otherTable, table.Name) {
+				// Find the other side of the M2M
+				for _, fk := range otherTable.GetForeignKeys() {
+					if fk.ReferencedTable != table.Name {
+						// This is the other side
+						otherSide := fk.ReferencedTable
+						sb.WriteString(fmt.Sprintf("  %ss %s[]\n",
+							strings.ToLower(otherSide), otherSide))
+						break
+					}
+				}
+			}
+			continue
+		}
+
+		// Regular one-to-many inverse relation
+		for _, fk := range otherTable.GetForeignKeys() {
+			if fk.ReferencedTable == table.Name {
+				// This table is referenced by otherTable
+				pluralName := otherTable.Name
+				if !strings.HasSuffix(pluralName, "s") {
+					pluralName += "s"
+				}
+
+				sb.WriteString(fmt.Sprintf("  %s %s[]\n",
+					strings.ToLower(pluralName), otherTable.Name))
+			}
+		}
+	}
+
+	return sb.String()
+}
+
+// isJoinTableFor checks if a table is a join table involving the specified model
+func (w *Writer) isJoinTableFor(joinTable *models.Table, modelName string) bool {
+	for _, fk := range joinTable.GetForeignKeys() {
+		if fk.ReferencedTable == modelName {
+			return true
+		}
+	}
+	return false
+}
+
+// generateBlockAttributes generates block-level attributes like @@id, @@unique, @@index
+func (w *Writer) generateBlockAttributes(table *models.Table) string {
+	var sb strings.Builder
+
+	// @@id for composite primary key
+	pkCols := make([]string, 0)
+	for _, col := range table.Columns {
+		if col.IsPrimaryKey {
+			pkCols = append(pkCols, col.Name)
+		}
+	}
+
+	if len(pkCols) > 1 {
+		sort.Strings(pkCols)
+		sb.WriteString(fmt.Sprintf("  @@id([%s])\n", strings.Join(pkCols, ", ")))
+	}
+
+	// @@unique for multi-column unique constraints
+	for _, constraint := range table.Constraints {
+		if constraint.Type == models.UniqueConstraint && len(constraint.Columns) > 1 {
+			sb.WriteString(fmt.Sprintf("  @@unique([%s])\n", strings.Join(constraint.Columns, ", ")))
+		}
+	}
+
+	// @@index for indexes
+	for _, index := range table.Indexes {
+		if !index.Unique { // Unique indexes are handled by @@unique
+			sb.WriteString(fmt.Sprintf("  @@index([%s])\n", strings.Join(index.Columns, ", ")))
+		}
+	}
+
+	return sb.String()
+}

pkg/writers/typeorm/writer.go

@@ -0,0 +1,631 @@
+package typeorm
+
+import (
+	"fmt"
+	"os"
+	"sort"
+	"strings"
+
+	"git.warky.dev/wdevs/relspecgo/pkg/models"
+	"git.warky.dev/wdevs/relspecgo/pkg/writers"
+)
+
+// Writer implements the writers.Writer interface for TypeORM entity format
+type Writer struct {
+	options *writers.WriterOptions
+}
+
+// NewWriter creates a new TypeORM writer with the given options
+func NewWriter(options *writers.WriterOptions) *Writer {
+	return &Writer{
+		options: options,
+	}
+}
+
+// WriteDatabase writes a Database model to TypeORM entity format
+func (w *Writer) WriteDatabase(db *models.Database) error {
+	content := w.databaseToTypeORM(db)
+
+	if w.options.OutputPath != "" {
+		return os.WriteFile(w.options.OutputPath, []byte(content), 0644)
+	}
+
+	fmt.Print(content)
+	return nil
+}
+
+// WriteSchema writes a Schema model to TypeORM entity format
+func (w *Writer) WriteSchema(schema *models.Schema) error {
+	db := models.InitDatabase("database")
+	db.Schemas = []*models.Schema{schema}
+
+	return w.WriteDatabase(db)
+}
+
+// WriteTable writes a Table model to TypeORM entity format
+func (w *Writer) WriteTable(table *models.Table) error {
+	schema := models.InitSchema(table.Schema)
+	schema.Tables = []*models.Table{table}
+
+	return w.WriteSchema(schema)
+}
+
+// databaseToTypeORM converts a Database to TypeORM entity format string
+func (w *Writer) databaseToTypeORM(db *models.Database) string {
+	var sb strings.Builder
+
+	// Generate imports
+	sb.WriteString(w.generateImports(db))
+	sb.WriteString("\n")
+
+	// Process all schemas
+	for _, schema := range db.Schemas {
+		// Identify join tables
+		joinTables := w.identifyJoinTables(schema)
+
+		// Write entities (excluding join tables)
+		for _, table := range schema.Tables {
+			if joinTables[table.Name] {
+				continue
+			}
+			sb.WriteString(w.tableToEntity(table, schema, joinTables))
+			sb.WriteString("\n")
+		}
+
+		// Write view entities
+		for _, view := range schema.Views {
+			sb.WriteString(w.viewToEntity(view))
+			sb.WriteString("\n")
+		}
+	}
+
+	return sb.String()
+}
+
+// generateImports generates the TypeORM import statement
+func (w *Writer) generateImports(db *models.Database) string {
+	imports := make([]string, 0)
+
+	// Always include basic decorators
+	imports = append(imports, "Entity", "PrimaryGeneratedColumn", "Column")
+
+	// Check if we need relation decorators
+	needsManyToOne := false
+	needsOneToMany := false
+	needsManyToMany := false
+	needsJoinTable := false
+	needsCreateDate := false
+	needsUpdateDate := false
+	needsViewEntity := false
+
+	for _, schema := range db.Schemas {
+		// Check for views
+		if len(schema.Views) > 0 {
+			needsViewEntity = true
+		}
+
+		for _, table := range schema.Tables {
+			// Check for timestamp columns
+			for _, col := range table.Columns {
+				if col.Default == "now()" {
+					needsCreateDate = true
+				}
+				if strings.Contains(col.Comment, "auto-update") {
+					needsUpdateDate = true
+				}
+			}
+
+			// Check for relations
+			for _, constraint := range table.Constraints {
+				if constraint.Type == models.ForeignKeyConstraint {
+					needsManyToOne = true
+				}
+			}
+		}
+	}
+
+	// OneToMany is the inverse of ManyToOne
+	if needsManyToOne {
+		needsOneToMany = true
+	}
+
+	// Check for M2M (join tables indicate M2M relations)
+	joinTables := make(map[string]bool)
+	for _, schema := range db.Schemas {
+		jt := w.identifyJoinTables(schema)
+		for name := range jt {
+			joinTables[name] = true
+			needsManyToMany = true
+			needsJoinTable = true
+		}
+	}
+
+	if needsManyToOne {
+		imports = append(imports, "ManyToOne")
+	}
+	if needsOneToMany {
+		imports = append(imports, "OneToMany")
+	}
+	if needsManyToMany {
+		imports = append(imports, "ManyToMany")
+	}
+	if needsJoinTable {
+		imports = append(imports, "JoinTable")
+	}
+	if needsCreateDate {
+		imports = append(imports, "CreateDateColumn")
+	}
+	if needsUpdateDate {
+		imports = append(imports, "UpdateDateColumn")
+	}
+	if needsViewEntity {
+		imports = append(imports, "ViewEntity")
+	}
+
+	return fmt.Sprintf("import { %s } from 'typeorm';\n", strings.Join(imports, ", "))
+}
+
+// identifyJoinTables identifies tables that are join tables for M2M relations
+func (w *Writer) identifyJoinTables(schema *models.Schema) map[string]bool {
+	joinTables := make(map[string]bool)
+
+	for _, table := range schema.Tables {
+		// Check if this is a join table:
+		// 1. Has exactly 2 FK constraints
+		// 2. Has composite PK with those 2 columns
+		// 3. Has no other columns except the FK columns
+
+		fks := table.GetForeignKeys()
+		if len(fks) != 2 {
+			continue
+		}
+
+		// Check if columns are only the FK columns
+		if len(table.Columns) != 2 {
+			continue
+		}
+
+		// Check if both FK columns are part of PK
+		pkCols := 0
+		for _, col := range table.Columns {
+			if col.IsPrimaryKey {
+				pkCols++
+			}
+		}
+
+		if pkCols == 2 {
+			joinTables[table.Name] = true
+		}
+	}
+
+	return joinTables
+}
+
+// tableToEntity converts a Table to a TypeORM entity class
+func (w *Writer) tableToEntity(table *models.Table, schema *models.Schema, joinTables map[string]bool) string {
+	var sb strings.Builder
+
+	// Generate @Entity decorator with options
+	entityOptions := w.buildEntityOptions(table)
+	sb.WriteString(fmt.Sprintf("@Entity({\n%s\n})\n", entityOptions))
+
+	// Get class name (from metadata if different from table name)
+	className := table.Name
+	if table.Metadata != nil {
+		if classNameVal, ok := table.Metadata["class_name"]; ok {
+			if classNameStr, ok := classNameVal.(string); ok {
+				className = classNameStr
+			}
+		}
+	}
+
+	sb.WriteString(fmt.Sprintf("export class %s {\n", className))
+
+	// Collect and sort columns
+	columns := make([]*models.Column, 0, len(table.Columns))
+	for _, col := range table.Columns {
+		// Skip FK columns (they'll be represented as relations)
+		if w.isForeignKeyColumn(col, table) {
+			continue
+		}
+		columns = append(columns, col)
+	}
+
+	sort.Slice(columns, func(i, j int) bool {
+		// Put PK first, then alphabetical
+		if columns[i].IsPrimaryKey && !columns[j].IsPrimaryKey {
+			return true
+		}
+		if !columns[i].IsPrimaryKey && columns[j].IsPrimaryKey {
+			return false
+		}
+		return columns[i].Name < columns[j].Name
+	})
+
+	// Write scalar fields
+	for _, col := range columns {
+		sb.WriteString(w.columnToField(col, table))
+		sb.WriteString("\n")
+	}
+
+	// Write relation fields
+	sb.WriteString(w.generateRelationFields(table, schema, joinTables))
+
+	sb.WriteString("}\n")
+
+	return sb.String()
+}
+
+// viewToEntity converts a View to a TypeORM @ViewEntity class
+func (w *Writer) viewToEntity(view *models.View) string {
+	var sb strings.Builder
+
+	// Generate @ViewEntity decorator with expression
+	sb.WriteString("@ViewEntity({\n")
+	if view.Definition != "" {
+		// Format the SQL expression with proper indentation
+		sb.WriteString("  expression: `\n")
+		sb.WriteString("    ")
+		sb.WriteString(view.Definition)
+		sb.WriteString("\n  `,\n")
+	}
+	sb.WriteString("})\n")
+
+	// Generate class
+	sb.WriteString(fmt.Sprintf("export class %s {\n", view.Name))
+
+	// Generate field definitions (without decorators for view fields)
+	columns := make([]*models.Column, 0, len(view.Columns))
+	for _, col := range view.Columns {
+		columns = append(columns, col)
+	}
+	sort.Slice(columns, func(i, j int) bool {
+		return columns[i].Name < columns[j].Name
+	})
+
+	for _, col := range columns {
+		tsType := w.sqlTypeToTypeScript(col.Type)
+		sb.WriteString(fmt.Sprintf("  %s: %s;\n", col.Name, tsType))
+	}
+
+	sb.WriteString("}\n")
+
+	return sb.String()
+}
+
+// columnToField converts a Column to a TypeORM field
+func (w *Writer) columnToField(col *models.Column, table *models.Table) string {
+	var sb strings.Builder
+
+	// Generate decorator
+	if col.IsPrimaryKey {
+		if col.AutoIncrement {
+			sb.WriteString("  @PrimaryGeneratedColumn('increment')\n")
+		} else if col.Type == "uuid" || strings.Contains(fmt.Sprint(col.Default), "uuid") {
+			sb.WriteString("  @PrimaryGeneratedColumn('uuid')\n")
+		} else {
+			sb.WriteString("  @PrimaryGeneratedColumn()\n")
+		}
+	} else if col.Default == "now()" {
+		sb.WriteString("  @CreateDateColumn()\n")
+	} else if strings.Contains(col.Comment, "auto-update") {
+		sb.WriteString("  @UpdateDateColumn()\n")
+	} else {
+		// Regular @Column decorator
+		options := w.buildColumnOptions(col, table)
+		if options != "" {
+			sb.WriteString(fmt.Sprintf("  @Column({ %s })\n", options))
+		} else {
+			sb.WriteString("  @Column()\n")
+		}
+	}
+
+	// Generate field declaration
+	tsType := w.sqlTypeToTypeScript(col.Type)
+	nullable := ""
+	if !col.NotNull {
+		nullable = " | null"
+	}
+
+	sb.WriteString(fmt.Sprintf("  %s: %s%s;", col.Name, tsType, nullable))
+
+	return sb.String()
+}
+
+// buildColumnOptions builds the options object for @Column decorator
+func (w *Writer) buildColumnOptions(col *models.Column, table *models.Table) string {
+	options := make([]string, 0)
+
+	// Type (if not default)
+	if w.needsExplicitType(col.Type) {
+		options = append(options, fmt.Sprintf("type: '%s'", col.Type))
+	}
+
+	// Nullable
+	if !col.NotNull {
+		options = append(options, "nullable: true")
+	}
+
+	// Unique
+	if w.hasUniqueConstraint(col.Name, table) {
+		options = append(options, "unique: true")
+	}
+
+	// Default
+	if col.Default != nil && col.Default != "now()" {
+		defaultStr := fmt.Sprint(col.Default)
+		if defaultStr != "" {
+			options = append(options, fmt.Sprintf("default: '%s'", defaultStr))
+		}
+	}
+
+	return strings.Join(options, ", ")
+}
+
+// needsExplicitType checks if a SQL type needs explicit type declaration
+func (w *Writer) needsExplicitType(sqlType string) bool {
+	// Types that don't map cleanly to TypeScript types need explicit declaration
+	explicitTypes := []string{"text", "uuid", "jsonb", "bigint"}
+	for _, t := range explicitTypes {
+		if strings.Contains(sqlType, t) {
+			return true
+		}
+	}
+	return false
+}
+
+// hasUniqueConstraint checks if a column has a unique constraint
+func (w *Writer) hasUniqueConstraint(colName string, table *models.Table) bool {
+	for _, constraint := range table.Constraints {
+		if constraint.Type == models.UniqueConstraint &&
+			len(constraint.Columns) == 1 &&
+			constraint.Columns[0] == colName {
+			return true
+		}
+	}
+	return false
+}
+
+// sqlTypeToTypeScript converts SQL types to TypeScript types
+func (w *Writer) sqlTypeToTypeScript(sqlType string) string {
+	typeMap := map[string]string{
+		"text":              "string",
+		"varchar":           "string",
+		"character varying": "string",
+		"char":              "string",
+		"uuid":              "string",
+		"boolean":           "boolean",
+		"bool":              "boolean",
+		"integer":           "number",
+		"int":               "number",
+		"bigint":            "number",
+		"double precision":  "number",
+		"float":             "number",
+		"decimal":           "number",
+		"numeric":           "number",
+		"timestamp":         "Date",
+		"timestamptz":       "Date",
+		"date":              "Date",
+		"jsonb":             "any",
+		"json":              "any",
+	}
+
+	for sqlPattern, tsType := range typeMap {
+		if strings.Contains(strings.ToLower(sqlType), sqlPattern) {
+			return tsType
+		}
+	}
+
+	return "any"
+}
+
+// isForeignKeyColumn checks if a column is a FK column
+func (w *Writer) isForeignKeyColumn(col *models.Column, table *models.Table) bool {
+	for _, constraint := range table.Constraints {
+		if constraint.Type == models.ForeignKeyConstraint {
+			for _, fkCol := range constraint.Columns {
+				if fkCol == col.Name {
+					return true
+				}
+			}
+		}
+	}
+	return false
+}
+
+// generateRelationFields generates relation fields for a table
+func (w *Writer) generateRelationFields(table *models.Table, schema *models.Schema, joinTables map[string]bool) string {
+	var sb strings.Builder
+
+	// Get all FK constraints
+	fks := table.GetForeignKeys()
+
+	// Generate @ManyToOne fields
+	for _, fk := range fks {
+		relatedTable := fk.ReferencedTable
+		fieldName := strings.ToLower(relatedTable)
+
+		// Determine if nullable
+		isNullable := false
+		for _, fkCol := range fk.Columns {
+			if col, exists := table.Columns[fkCol]; exists {
+				if !col.NotNull {
+					isNullable = true
+				}
+			}
+		}
+
+		nullable := ""
+		if isNullable {
+			nullable = " | null"
+		}
+
+		// Find inverse field name if possible
+		inverseField := w.findInverseFieldName(table.Name, relatedTable, schema)
+
+		if inverseField != "" {
+			sb.WriteString(fmt.Sprintf("  @ManyToOne(() => %s, %s => %s.%s)\n",
+				relatedTable, strings.ToLower(relatedTable), strings.ToLower(relatedTable), inverseField))
+		} else {
+			if isNullable {
+				sb.WriteString(fmt.Sprintf("  @ManyToOne(() => %s, { nullable: true })\n", relatedTable))
+			} else {
+				sb.WriteString(fmt.Sprintf("  @ManyToOne(() => %s)\n", relatedTable))
+			}
+		}
+
+		sb.WriteString(fmt.Sprintf("  %s: %s%s;\n", fieldName, relatedTable, nullable))
+		sb.WriteString("\n")
+	}
+
+	// Generate @OneToMany fields (inverse of FKs pointing to this table)
+	w.generateInverseRelations(table, schema, joinTables, &sb)
+
+	// Generate @ManyToMany fields
+	w.generateManyToManyRelations(table, schema, joinTables, &sb)
+
+	return sb.String()
+}
+
+// findInverseFieldName finds the inverse field name for a relation
+func (w *Writer) findInverseFieldName(fromTable, toTable string, schema *models.Schema) string {
+	// Look for tables that have FKs pointing back to fromTable
+	for _, table := range schema.Tables {
+		if table.Name != toTable {
+			continue
+		}
+
+		for _, constraint := range table.Constraints {
+			if constraint.Type == models.ForeignKeyConstraint && constraint.ReferencedTable == fromTable {
+				// Found an inverse relation
+				// Use pluralized form of fromTable
+				return w.pluralize(strings.ToLower(fromTable))
+			}
+		}
+	}
+
+	return ""
+}
+
+// generateInverseRelations generates @OneToMany fields
+func (w *Writer) generateInverseRelations(table *models.Table, schema *models.Schema, joinTables map[string]bool, sb *strings.Builder) {
+	for _, otherTable := range schema.Tables {
+		if otherTable.Name == table.Name || joinTables[otherTable.Name] {
+			continue
+		}
+
+		for _, fk := range otherTable.GetForeignKeys() {
+			if fk.ReferencedTable == table.Name {
+				// This table is referenced by otherTable
+				fieldName := w.pluralize(strings.ToLower(otherTable.Name))
+				inverseName := strings.ToLower(table.Name)
+
+				sb.WriteString(fmt.Sprintf("  @OneToMany(() => %s, %s => %s.%s)\n",
+					otherTable.Name, strings.ToLower(otherTable.Name), strings.ToLower(otherTable.Name), inverseName))
+				sb.WriteString(fmt.Sprintf("  %s: %s[];\n", fieldName, otherTable.Name))
+				sb.WriteString("\n")
+			}
+		}
+	}
+}
+
+// generateManyToManyRelations generates @ManyToMany fields
+func (w *Writer) generateManyToManyRelations(table *models.Table, schema *models.Schema, joinTables map[string]bool, sb *strings.Builder) {
+	for joinTableName := range joinTables {
+		joinTable := w.findTable(joinTableName, schema)
+		if joinTable == nil {
+			continue
+		}
+
+		fks := joinTable.GetForeignKeys()
+		if len(fks) != 2 {
+			continue
+		}
+
+		// Check if this table is part of the M2M relation
+		var thisTableFK *models.Constraint
+		var otherTableFK *models.Constraint
+
+		for i, fk := range fks {
+			if fk.ReferencedTable == table.Name {
+				thisTableFK = fk
+				if i == 0 {
+					otherTableFK = fks[1]
+				} else {
+					otherTableFK = fks[0]
+				}
+			}
+		}
+
+		if thisTableFK == nil {
+			continue
+		}
+
+		// Determine which side owns the relation (has @JoinTable)
+		// We'll make the first entity alphabetically the owner
+		isOwner := table.Name < otherTableFK.ReferencedTable
+
+		otherTable := otherTableFK.ReferencedTable
+		fieldName := w.pluralize(strings.ToLower(otherTable))
+		inverseName := w.pluralize(strings.ToLower(table.Name))
+
+		if isOwner {
+			sb.WriteString(fmt.Sprintf("  @ManyToMany(() => %s, %s => %s.%s)\n",
+				otherTable, strings.ToLower(otherTable), strings.ToLower(otherTable), inverseName))
+			sb.WriteString("  @JoinTable()\n")
+		} else {
+			sb.WriteString(fmt.Sprintf("  @ManyToMany(() => %s, %s => %s.%s)\n",
+				otherTable, strings.ToLower(otherTable), strings.ToLower(otherTable), inverseName))
+		}
+
+		sb.WriteString(fmt.Sprintf("  %s: %s[];\n", fieldName, otherTable))
+		sb.WriteString("\n")
+	}
+}
+
+// findTable finds a table by name in a schema
+func (w *Writer) findTable(name string, schema *models.Schema) *models.Table {
+	for _, table := range schema.Tables {
+		if table.Name == name {
+			return table
+		}
+	}
+	return nil
+}
+
+// buildEntityOptions builds the options object for @Entity decorator
+func (w *Writer) buildEntityOptions(table *models.Table) string {
+	options := make([]string, 0)
+
+	// Always include table name
+	options = append(options, fmt.Sprintf("  name: \"%s\"", table.Name))
+
+	// Always include schema
+	options = append(options, fmt.Sprintf("  schema: \"%s\"", table.Schema))
+
+	// Database name from metadata
+	if table.Metadata != nil {
+		if database, ok := table.Metadata["database"]; ok {
+			if databaseStr, ok := database.(string); ok {
+				options = append(options, fmt.Sprintf("  database: \"%s\"", databaseStr))
+			}
+		}
+
+		// Engine from metadata
+		if engine, ok := table.Metadata["engine"]; ok {
+			if engineStr, ok := engine.(string); ok {
+				options = append(options, fmt.Sprintf("  engine: \"%s\"", engineStr))
+			}
+		}
+	}
+
+	return strings.Join(options, ",\n")
+}
+
+// pluralize adds 's' to make a word plural (simple version)
+func (w *Writer) pluralize(word string) string {
+	if strings.HasSuffix(word, "s") {
+		return word
+	}
+	return word + "s"
+}

tests/assets/prisma/example.prisma

@@ -0,0 +1,46 @@
+datasource db {
+  provider = "postgresql"
+}
+
+generator client {
+  provider = "prisma-client"
+  output   = "./generated"
+}
+
+model User {
+  id      Int      @id @default(autoincrement())
+  email   String   @unique
+  name    String?
+  role    Role     @default(USER)
+  posts   Post[]
+  profile Profile?
+}
+
+model Profile {
+  id     Int    @id @default(autoincrement())
+  bio    String
+  user   User   @relation(fields: [userId], references: [id])
+  userId Int    @unique
+}
+
+model Post {
+  id         Int        @id @default(autoincrement())
+  createdAt  DateTime   @default(now())
+  updatedAt  DateTime   @updatedAt
+  title      String
+  published  Boolean    @default(false)
+  author     User       @relation(fields: [authorId], references: [id])
+  authorId   Int
+  categories Category[]
+}
+
+model Category {
+  id    Int    @id @default(autoincrement())
+  name  String
+  posts Post[]
+}
+
+enum Role {
+  USER
+  ADMIN
+}

tests/assets/typeorm/example.ts

@@ -0,0 +1,115 @@
+//@ts-nocheck
+import { Entity, PrimaryGeneratedColumn, Column, ManyToOne, OneToMany, ManyToMany, JoinTable, CreateDateColumn, UpdateDateColumn } from 'typeorm';
+
+@Entity()
+export class User {
+  @PrimaryGeneratedColumn('uuid')
+  id: string;
+
+  @Column({ unique: true })
+  email: string;
+
+  @Column()
+  name: string;
+
+  @CreateDateColumn()
+  createdAt: Date;
+
+  @UpdateDateColumn()
+  updatedAt: Date;
+
+  @OneToMany(() => Project, project => project.owner)
+  ownedProjects: Project[];
+
+  @ManyToMany(() => Project, project => project.members)
+  @JoinTable()
+  projects: Project[];
+}
+
+@Entity()
+export class Project {
+  @PrimaryGeneratedColumn('uuid')
+  id: string;
+
+  @Column()
+  title: string;
+
+  @Column({ nullable: true })
+  description: string;
+
+  @Column({ default: 'active' })
+  status: string;
+
+  @ManyToOne(() => User, user => user.ownedProjects)
+  owner: User;
+
+  @ManyToMany(() => User, user => user.projects)
+  members: User[];
+
+  @OneToMany(() => Task, task => task.project)
+  tasks: Task[];
+
+  @CreateDateColumn()
+  createdAt: Date;
+}
+
+@Entity()
+export class Task {
+  @PrimaryGeneratedColumn('uuid')
+  id: string;
+
+  @Column()
+  title: string;
+
+  @Column({ type: 'text', nullable: true })
+  description: string;
+
+  @Column({ default: 'todo' })
+  status: string;
+
+  @Column({ nullable: true })
+  dueDate: Date;
+
+  @ManyToOne(() => Project, project => project.tasks)
+  project: Project;
+
+  @ManyToOne(() => User, { nullable: true })
+  assignee: User;
+
+  @OneToMany(() => Comment, comment => comment.task)
+  comments: Comment[];
+}
+
+@Entity()
+export class Comment {
+  @PrimaryGeneratedColumn('uuid')
+  id: string;
+
+  @Column('text')
+  content: string;
+
+  @ManyToOne(() => Task, task => task.comments)
+  task: Task;
+
+  @ManyToOne(() => User)
+  author: User;
+
+  @CreateDateColumn()
+  createdAt: Date;
+}
+
+@Entity()
+export class Tag {
+  @PrimaryGeneratedColumn('uuid')
+  id: string;
+
+  @Column({ unique: true })
+  name: string;
+
+  @Column()
+  color: string;
+
+  @ManyToMany(() => Task)
+  @JoinTable()
+  tasks: Task[];
+}