add sql file mode support to gendao command

This commit is contained in:
John Guo
2026-03-14 15:09:16 +08:00
parent 766579d868
commit 33ccec6d68
25 changed files with 3626 additions and 104 deletions

3
.gitignore vendored
View File

@ -25,4 +25,5 @@ node_modules
output
.example/
.golangci.bck.yml
*.exe
*.exe
.aiprompt.zh.md

View File

@ -1,5 +1,27 @@
SHELL := /bin/bash
# commit changes with AI-generated commit message
.PHONY: up
up:
@if git diff --quiet HEAD && git diff --cached --quiet && [ -z "$$(git ls-files --others --exclude-standard)" ]; then \
echo "No changes to commit"; \
exit 0; \
fi
@git add -A
@echo "Analyzing changes and generating commit message via AI..."
@set -e; \
MSG=$$(git diff --cached --stat && echo "---" && git diff --cached | head -2000 | \
claude -p "Analyze the git diff above and generate a concise commit message (single line, max 72 chars, lowercase, no quotes). Output only the commit message itself, nothing else." \
--model haiku) || { echo "Error: Claude command failed"; exit 1; }; \
COMMIT_MSG=$$(echo "$$MSG" | tail -1); \
if [ -z "$$COMMIT_MSG" ]; then \
echo "Error: Failed to generate commit message"; \
exit 1; \
fi; \
echo "Commit: $$COMMIT_MSG"; \
git commit -m "$$COMMIT_MSG" && \
git push origin
# execute "go mod tidy" on all folders that have go.mod file
.PHONY: tidy
tidy:

View File

@ -33,65 +33,88 @@ import (
)
type (
CGenDao struct{}
// CGenDao is the command handler struct for "gen dao" command.
CGenDao struct{}
// CGenDaoInput defines all input parameters for the "gen dao" command.
// It supports both command-line arguments and configuration file options.
CGenDaoInput struct {
g.Meta `name:"dao" config:"{CGenDaoConfig}" usage:"{CGenDaoUsage}" brief:"{CGenDaoBrief}" eg:"{CGenDaoEg}" ad:"{CGenDaoAd}"`
Path string `name:"path" short:"p" brief:"{CGenDaoBriefPath}" d:"internal"`
Link string `name:"link" short:"l" brief:"{CGenDaoBriefLink}"`
Tables string `name:"tables" short:"t" brief:"{CGenDaoBriefTables}"`
TablesEx string `name:"tablesEx" short:"x" brief:"{CGenDaoBriefTablesEx}"`
ShardingPattern []string `name:"shardingPattern" short:"sp" brief:"{CGenDaoBriefShardingPattern}"`
Group string `name:"group" short:"g" brief:"{CGenDaoBriefGroup}" d:"default"`
Prefix string `name:"prefix" short:"f" brief:"{CGenDaoBriefPrefix}"`
RemovePrefix string `name:"removePrefix" short:"r" brief:"{CGenDaoBriefRemovePrefix}"`
RemoveFieldPrefix string `name:"removeFieldPrefix" short:"rf" brief:"{CGenDaoBriefRemoveFieldPrefix}"`
JsonCase string `name:"jsonCase" short:"j" brief:"{CGenDaoBriefJsonCase}" d:"CamelLower"`
ImportPrefix string `name:"importPrefix" short:"i" brief:"{CGenDaoBriefImportPrefix}"`
DaoPath string `name:"daoPath" short:"d" brief:"{CGenDaoBriefDaoPath}" d:"dao"`
TablePath string `name:"tablePath" short:"tp" brief:"{CGenDaoBriefTablePath}" d:"table"`
DoPath string `name:"doPath" short:"o" brief:"{CGenDaoBriefDoPath}" d:"model/do"`
EntityPath string `name:"entityPath" short:"e" brief:"{CGenDaoBriefEntityPath}" d:"model/entity"`
TplDaoTablePath string `name:"tplDaoTablePath" short:"t0" brief:"{CGenDaoBriefTplDaoTablePath}"`
TplDaoIndexPath string `name:"tplDaoIndexPath" short:"t1" brief:"{CGenDaoBriefTplDaoIndexPath}"`
TplDaoInternalPath string `name:"tplDaoInternalPath" short:"t2" brief:"{CGenDaoBriefTplDaoInternalPath}"`
TplDaoDoPath string `name:"tplDaoDoPath" short:"t3" brief:"{CGenDaoBriefTplDaoDoPathPath}"`
TplDaoEntityPath string `name:"tplDaoEntityPath" short:"t4" brief:"{CGenDaoBriefTplDaoEntityPath}"`
StdTime bool `name:"stdTime" short:"s" brief:"{CGenDaoBriefStdTime}" orphan:"true"`
WithTime bool `name:"withTime" short:"w" brief:"{CGenDaoBriefWithTime}" orphan:"true"`
GJsonSupport bool `name:"gJsonSupport" short:"n" brief:"{CGenDaoBriefGJsonSupport}" orphan:"true"`
OverwriteDao bool `name:"overwriteDao" short:"v" brief:"{CGenDaoBriefOverwriteDao}" orphan:"true"`
DescriptionTag bool `name:"descriptionTag" short:"c" brief:"{CGenDaoBriefDescriptionTag}" orphan:"true"`
NoJsonTag bool `name:"noJsonTag" short:"k" brief:"{CGenDaoBriefNoJsonTag}" orphan:"true"`
NoModelComment bool `name:"noModelComment" short:"m" brief:"{CGenDaoBriefNoModelComment}" orphan:"true"`
Clear bool `name:"clear" short:"a" brief:"{CGenDaoBriefClear}" orphan:"true"`
GenTable bool `name:"genTable" short:"gt" brief:"{CGenDaoBriefGenTable}" orphan:"true"`
Path string `name:"path" short:"p" brief:"{CGenDaoBriefPath}" d:"internal"` // Base directory path for generated files.
Link string `name:"link" short:"l" brief:"{CGenDaoBriefLink}"` // Database connection string (e.g., "mysql:root:pass@tcp(127.0.0.1:3306)/db").
Tables string `name:"tables" short:"t" brief:"{CGenDaoBriefTables}"` // Comma-separated table names or wildcard patterns to include.
TablesEx string `name:"tablesEx" short:"x" brief:"{CGenDaoBriefTablesEx}"` // Comma-separated table names or wildcard patterns to exclude.
ShardingPattern []string `name:"shardingPattern" short:"sp" brief:"{CGenDaoBriefShardingPattern}"` // Patterns for sharding tables (e.g., "users_?" merges users_001, users_002 into one dao).
Group string `name:"group" short:"g" brief:"{CGenDaoBriefGroup}" d:"default"` // Database configuration group name for ORM instance.
Prefix string `name:"prefix" short:"f" brief:"{CGenDaoBriefPrefix}"` // Prefix to add to all generated table names.
RemovePrefix string `name:"removePrefix" short:"r" brief:"{CGenDaoBriefRemovePrefix}"` // Comma-separated prefixes to remove from table names.
RemoveFieldPrefix string `name:"removeFieldPrefix" short:"rf" brief:"{CGenDaoBriefRemoveFieldPrefix}"` // Comma-separated prefixes to remove from field names.
JsonCase string `name:"jsonCase" short:"j" brief:"{CGenDaoBriefJsonCase}" d:"CamelLower"` // Naming convention for JSON tags (e.g., CamelLower, Snake).
ImportPrefix string `name:"importPrefix" short:"i" brief:"{CGenDaoBriefImportPrefix}"` // Custom Go import path prefix for generated files.
DaoPath string `name:"daoPath" short:"d" brief:"{CGenDaoBriefDaoPath}" d:"dao"` // Sub-directory under Path for dao files.
TablePath string `name:"tablePath" short:"tp" brief:"{CGenDaoBriefTablePath}" d:"table"` // Sub-directory under Path for table field definition files.
DoPath string `name:"doPath" short:"o" brief:"{CGenDaoBriefDoPath}" d:"model/do"` // Sub-directory under Path for DO (Data Object) files.
EntityPath string `name:"entityPath" short:"e" brief:"{CGenDaoBriefEntityPath}" d:"model/entity"` // Sub-directory under Path for entity struct files.
TplDaoTablePath string `name:"tplDaoTablePath" short:"t0" brief:"{CGenDaoBriefTplDaoTablePath}"` // Custom template file for dao table generation.
TplDaoIndexPath string `name:"tplDaoIndexPath" short:"t1" brief:"{CGenDaoBriefTplDaoIndexPath}"` // Custom template file for dao index generation.
TplDaoInternalPath string `name:"tplDaoInternalPath" short:"t2" brief:"{CGenDaoBriefTplDaoInternalPath}"` // Custom template file for dao internal generation.
TplDaoDoPath string `name:"tplDaoDoPath" short:"t3" brief:"{CGenDaoBriefTplDaoDoPathPath}"` // Custom template file for DO generation.
TplDaoEntityPath string `name:"tplDaoEntityPath" short:"t4" brief:"{CGenDaoBriefTplDaoEntityPath}"` // Custom template file for entity generation.
StdTime bool `name:"stdTime" short:"s" brief:"{CGenDaoBriefStdTime}" orphan:"true"` // Use stdlib time.Time instead of gtime.Time for time fields.
WithTime bool `name:"withTime" short:"w" brief:"{CGenDaoBriefWithTime}" orphan:"true"` // Add creation timestamp to generated file headers.
GJsonSupport bool `name:"gJsonSupport" short:"n" brief:"{CGenDaoBriefGJsonSupport}" orphan:"true"` // Use *gjson.Json instead of string for JSON fields.
OverwriteDao bool `name:"overwriteDao" short:"v" brief:"{CGenDaoBriefOverwriteDao}" orphan:"true"` // Overwrite existing dao files (both index and internal).
DescriptionTag bool `name:"descriptionTag" short:"c" brief:"{CGenDaoBriefDescriptionTag}" orphan:"true"`// Add description struct tag with field comment.
NoJsonTag bool `name:"noJsonTag" short:"k" brief:"{CGenDaoBriefNoJsonTag}" orphan:"true"` // Omit json struct tags from generated structs.
NoModelComment bool `name:"noModelComment" short:"m" brief:"{CGenDaoBriefNoModelComment}" orphan:"true"`// Omit inline comments from generated struct fields.
Clear bool `name:"clear" short:"a" brief:"{CGenDaoBriefClear}" orphan:"true"` // Delete generated files that no longer correspond to database tables.
GenTable bool `name:"genTable" short:"gt" brief:"{CGenDaoBriefGenTable}" orphan:"true"` // Enable generation of table field definition files.
SqlDir string `name:"sqlDir" short:"sd" brief:"{CGenDaoBriefSqlDir}"` // Directory of SQL DDL files for offline generation (no DB connection needed).
SqlType string `name:"sqlType" short:"st" brief:"{CGenDaoBriefSqlType}" d:"mysql"` // SQL dialect when using SqlDir (mysql, pgsql, mssql, oracle, sqlite).
TypeMapping map[DBFieldTypeName]CustomAttributeType `name:"typeMapping" short:"y" brief:"{CGenDaoBriefTypeMapping}" orphan:"true"`
// TypeMapping maps database field type names to custom Go types.
// For example, mapping "decimal" to "float64" or "uuid" to "uuid.UUID".
TypeMapping map[DBFieldTypeName]CustomAttributeType `name:"typeMapping" short:"y" brief:"{CGenDaoBriefTypeMapping}" orphan:"true"`
// FieldMapping maps specific table.field combinations to custom Go types.
// For example, mapping "user.balance" to "decimal.Decimal".
FieldMapping map[DBTableFieldName]CustomAttributeType `name:"fieldMapping" short:"fm" brief:"{CGenDaoBriefFieldMapping}" orphan:"true"`
// internal usage purpose.
// genItems tracks all generated file paths and directories for cleanup purposes.
genItems *CGenDaoInternalGenItems
}
// CGenDaoOutput is the output of the "gen dao" command (currently empty).
CGenDaoOutput struct{}
// CGenDaoInternalInput extends CGenDaoInput with runtime-resolved fields
// used during the actual generation process.
CGenDaoInternalInput struct {
CGenDaoInput
DB gdb.DB
TableNames []string
NewTableNames []string
ShardingTableSet *gset.StrSet
DB gdb.DB // Database connection instance (nil in SQL file mode).
TableNames []string // Original table names from database or SQL files.
NewTableNames []string // Processed table names after prefix removal and sharding.
ShardingTableSet *gset.StrSet // Set of table names identified as sharding tables.
// TableFieldsMap stores pre-parsed table fields from SQL files.
// When this is set (SQL file mode), DB may be nil.
TableFieldsMap map[string]map[string]*gdb.TableField
}
DBTableFieldName = string
DBFieldTypeName = string
// DBTableFieldName is the fully-qualified field name in "table.field" format.
DBTableFieldName = string
// DBFieldTypeName is the database column type name (e.g., "varchar", "decimal").
DBFieldTypeName = string
// CustomAttributeType defines a custom Go type mapping with its import path.
CustomAttributeType struct {
Type string `brief:"custom attribute type name"`
Import string `brief:"custom import for this type"`
Type string `brief:"custom attribute type name"` // Go type name (e.g., "decimal.Decimal").
Import string `brief:"custom import for this type"` // Go import path (e.g., "github.com/shopspring/decimal").
}
)
var (
createdAt = gtime.Now()
tplView = gview.New()
createdAt = gtime.Now() // Timestamp captured at program start, used in generated file headers.
tplView = gview.New() // Shared template view instance for rendering all Go file templates.
// defaultTypeMapping provides built-in type mappings from database types to Go types.
// User-provided TypeMapping takes precedence over these defaults.
defaultTypeMapping = map[DBFieldTypeName]CustomAttributeType{
"decimal": {
Type: "float64",
@ -111,7 +134,8 @@ var (
},
}
// tablewriter Options
// twRenderer configures the tablewriter to render without borders or separators,
// producing clean aligned text output for generated Go source code.
twRenderer = tablewriter.WithRenderer(renderer.NewBlueprint(tw.Rendition{
Borders: tw.Border{Top: tw.Off, Bottom: tw.Off, Left: tw.Off, Right: tw.Off},
Settings: tw.Settings{
@ -126,9 +150,17 @@ var (
})
)
// Dao is the main entry point for the "gen dao" command.
// It dispatches to the appropriate generation mode based on input:
// - SQL file mode (SqlDir is set): generates from DDL files without database connection.
// - Link mode (Link is set): uses a direct database connection string.
// - Config mode: reads database configuration from the application config file.
func (c CGenDao) Dao(ctx context.Context, in CGenDaoInput) (out *CGenDaoOutput, err error) {
in.genItems = newCGenDaoInternalGenItems()
if in.Link != "" {
if in.SqlDir != "" {
// SQL file mode: generate from SQL DDL files without database connection.
doGenDaoFromSQLFiles(ctx, in)
} else if in.Link != "" {
doGenDaoForArray(ctx, -1, in)
} else if g.Cfg().Available(ctx) {
v := g.Cfg().MustGet(ctx, CGenDaoConfig)
@ -147,7 +179,11 @@ func (c CGenDao) Dao(ctx context.Context, in CGenDaoInput) (out *CGenDaoOutput,
return
}
// doGenDaoForArray implements the "gen dao" command for configuration array.
// doGenDaoForArray implements the "gen dao" command for a single configuration entry.
// When index >= 0, it reads configuration from the array at that index.
// When index < 0, it uses the input as-is (for Link mode or single config mode).
// It performs the full generation pipeline: connect to DB, resolve tables,
// apply sharding patterns, and generate dao/table/do/entity files.
func doGenDaoForArray(ctx context.Context, index int, in CGenDaoInput) {
var (
err error
@ -332,6 +368,10 @@ func doGenDaoForArray(ctx context.Context, index int, in CGenDaoInput) {
in.genItems.SetClear(in.Clear)
}
// getImportPartContent analyzes the generated Go source code and builds the import block.
// It automatically detects usage of gtime.Time, time.Time, and gjson.Json in the source,
// and includes the corresponding import paths. Additional custom imports (from TypeMapping
// or FieldMapping) are appended and their dependencies are resolved via "go get" if needed.
func getImportPartContent(ctx context.Context, source string, isDo bool, appendImports []string) string {
var packageImportsArray = garray.NewStrArray()
if isDo {
@ -385,6 +425,9 @@ func getImportPartContent(ctx context.Context, source string, isDo bool, appendI
return packageImportsStr
}
// assignDefaultVar sets the default template variables for datetime strings
// used in generated file headers. The creation timestamp is only included
// when WithTime is enabled in the input configuration.
func assignDefaultVar(view *gview.View, in CGenDaoInternalInput) {
var (
tplCreatedAtDatetimeStr string
@ -399,6 +442,8 @@ func assignDefaultVar(view *gview.View, in CGenDaoInternalInput) {
})
}
// sortFieldKeyForDao returns field names sorted by their Index in the TableField map.
// This preserves the original column order as defined in the database table schema.
func sortFieldKeyForDao(fieldMap map[string]*gdb.TableField) []string {
names := make(map[int]string)
for _, field := range fieldMap {
@ -423,6 +468,20 @@ func sortFieldKeyForDao(fieldMap map[string]*gdb.TableField) []string {
return result
}
// getTableFields retrieves table fields either from the pre-parsed TableFieldsMap (SQL file mode)
// or from the database connection. This abstracts the data source for generation functions.
func getTableFields(ctx context.Context, in CGenDaoInternalInput, tableName string) (map[string]*gdb.TableField, error) {
if in.TableFieldsMap != nil {
if fields, ok := in.TableFieldsMap[tableName]; ok {
return fields, nil
}
return nil, fmt.Errorf("table '%s' not found in SQL files", tableName)
}
return in.DB.TableFields(ctx, tableName)
}
// getTemplateFromPathOrDefault returns the template content from the given file path.
// If the file path is empty or the file has no content, it falls back to the default template.
func getTemplateFromPathOrDefault(filePath string, def string) string {
if filePath != "" {
if contents := gfile.GetContents(filePath); contents != "" {
@ -489,3 +548,130 @@ func filterTablesByPatterns(allTables []string, patterns []string) []string {
}
return result
}
// doGenDaoFromSQLFiles implements the "gen dao" command for SQL file mode.
// It parses DDL SQL files to obtain table structures without requiring a database connection.
func doGenDaoFromSQLFiles(ctx context.Context, in CGenDaoInput) {
if dirRealPath := gfile.RealPath(in.Path); dirRealPath == "" {
mlog.Fatalf(`path "%s" does not exist`, in.Path)
}
if dirRealPath := gfile.RealPath(in.SqlDir); dirRealPath == "" {
mlog.Fatalf(`SQL directory "%s" does not exist`, in.SqlDir)
}
dialect := SQLDialect(strings.ToLower(in.SqlType))
tableNames, tableFieldsMap := ParseSQLFilesFromDir(in.SqlDir, dialect)
removePrefixArray := gstr.SplitAndTrim(in.RemovePrefix, ",")
// Table filtering by name patterns.
if in.Tables != "" {
inputTables := gstr.SplitAndTrim(in.Tables, ",")
var hasPattern bool
for _, t := range inputTables {
if containsWildcard(t) {
hasPattern = true
break
}
}
if hasPattern {
tableNames = filterTablesByPatterns(tableNames, inputTables)
} else {
tableNames = inputTables
}
}
// Table excluding.
if in.TablesEx != "" {
array := garray.NewStrArrayFrom(tableNames)
for _, p := range gstr.SplitAndTrim(in.TablesEx, ",") {
if containsWildcard(p) {
regPattern := "^" + patternToRegex(p) + "$"
for _, v := range array.Clone().Slice() {
if gregex.IsMatchString(regPattern, v) {
array.RemoveValue(v)
}
}
} else {
array.RemoveValue(p)
}
}
tableNames = array.Slice()
}
// merge default typeMapping.
if in.TypeMapping == nil {
in.TypeMapping = defaultTypeMapping
} else {
for key, typeMapping := range defaultTypeMapping {
if _, ok := in.TypeMapping[key]; !ok {
in.TypeMapping[key] = typeMapping
}
}
}
// Process table names (prefix removal, sharding, etc.)
var (
newTableNames = make([]string, len(tableNames))
shardingNewTableSet = gset.NewStrSet()
)
sortedShardingPatterns := make([]string, len(in.ShardingPattern))
copy(sortedShardingPatterns, in.ShardingPattern)
sort.Slice(sortedShardingPatterns, func(i, j int) bool {
return len(sortedShardingPatterns[i]) > len(sortedShardingPatterns[j])
})
for i, tableName := range tableNames {
newTableName := tableName
for _, v := range removePrefixArray {
newTableName = gstr.TrimLeftStr(newTableName, v, 1)
}
if len(sortedShardingPatterns) > 0 {
for _, pattern := range sortedShardingPatterns {
var (
match []string
regPattern = gstr.Replace(pattern, "?", `(.+)`)
err error
)
match, err = gregex.MatchString(regPattern, newTableName)
if err != nil {
mlog.Fatalf(`invalid sharding pattern "%s": %+v`, pattern, err)
}
if len(match) < 2 {
continue
}
newTableName = gstr.Replace(pattern, "?", "")
newTableName = gstr.Trim(newTableName, `_.-`)
if shardingNewTableSet.Contains(newTableName) {
tableNames[i] = ""
break
}
shardingNewTableSet.Add(in.Prefix + newTableName)
break
}
}
newTableName = in.Prefix + newTableName
if tableNames[i] != "" {
newTableNames[i] = newTableName
}
}
tableNames = garray.NewStrArrayFrom(tableNames).FilterEmpty().Slice()
newTableNames = garray.NewStrArrayFrom(newTableNames).FilterEmpty().Slice()
in.genItems.Scale()
internalInput := CGenDaoInternalInput{
CGenDaoInput: in,
DB: nil,
TableNames: tableNames,
NewTableNames: newTableNames,
ShardingTableSet: shardingNewTableSet,
TableFieldsMap: tableFieldsMap,
}
// Generate all files using the same flow as database mode.
generateDao(ctx, internalInput)
generateTable(ctx, internalInput)
generateDo(ctx, internalInput)
generateEntity(ctx, internalInput)
in.genItems.SetClear(in.Clear)
}

View File

@ -13,6 +13,10 @@ import (
"github.com/gogf/gf/cmd/gf/v2/internal/utility/mlog"
)
// doClear performs cleanup of stale generated files across all generation items.
// It collects all generated file paths from all items, then for each item with
// Clear enabled, removes any .go files in its directories that are NOT in the
// generated file list. This ensures files for dropped/removed tables are cleaned up.
func doClear(items *CGenDaoInternalGenItems) {
var allGeneratedFilePaths = make([]string, 0)
for _, item := range items.Items {
@ -29,6 +33,10 @@ func doClear(items *CGenDaoInternalGenItems) {
}
}
// doClearItem removes stale .go files for a single generation item.
// It scans all storage directories for .go files and deletes any file
// that is not in the allGeneratedFilePaths list (i.e., no longer corresponds
// to an existing database table).
func doClearItem(item CGenDaoInternalGenItem, allGeneratedFilePaths []string) {
var generatedFilePaths = make([]string, 0)
for _, dirPath := range item.StorageDirPaths {

View File

@ -26,6 +26,9 @@ import (
"github.com/gogf/gf/cmd/gf/v2/internal/utility/utils"
)
// generateDao generates dao files (index + internal) for all tables in the input.
// It creates the dao directory structure and iterates over each table to generate
// individual dao files via generateDaoSingle.
func generateDao(ctx context.Context, in CGenDaoInternalInput) {
var (
dirPathDao = gfile.Join(in.Path, in.DaoPath)
@ -48,21 +51,20 @@ func generateDao(ctx context.Context, in CGenDaoInternalInput) {
}
}
// generateDaoSingleInput holds all parameters needed to generate dao files for a single table.
type generateDaoSingleInput struct {
CGenDaoInternalInput
// TableName specifies the table name of the table.
TableName string
// NewTableName specifies the prefix-stripped or custom edited name of the table.
NewTableName string
DirPathDao string
DirPathDaoInternal string
IsSharding bool
TableName string // Original table name as it exists in the database.
NewTableName string // Processed table name after prefix removal and sharding.
DirPathDao string // Directory path for the dao index files.
DirPathDaoInternal string // Directory path for the dao internal implementation files.
IsSharding bool // Whether this table is a sharding table (merged from multiple physical tables).
}
// generateDaoSingle generates the dao and model content of given table.
func generateDaoSingle(ctx context.Context, in generateDaoSingleInput) {
// Generating table data preparing.
fieldMap, err := in.DB.TableFields(ctx, in.TableName)
fieldMap, err := getTableFields(ctx, in.CGenDaoInternalInput, in.TableName)
if err != nil {
mlog.Fatalf(`fetching tables fields failed for table "%s": %+v`, in.TableName, err)
}
@ -105,14 +107,21 @@ func generateDaoSingle(ctx context.Context, in generateDaoSingleInput) {
})
}
// generateDaoIndexInput holds parameters for generating the dao index file.
// The index file provides the public API (exported struct and constructor)
// for accessing the DAO, delegating to the internal implementation.
type generateDaoIndexInput struct {
generateDaoSingleInput
TableNameCamelCase string
TableNameCamelLowerCase string
ImportPrefix string
FileName string
TableNameCamelCase string // CamelCase version of the table name (e.g., "UserDetail").
TableNameCamelLowerCase string // camelCase version of the table name (e.g., "userDetail").
ImportPrefix string // Go import path prefix for the dao package.
FileName string // Output file name (without extension).
}
// generateDaoIndex generates the dao index file for a single table.
// The index file is the public-facing dao file that users import directly.
// It will NOT overwrite an existing file unless OverwriteDao is enabled,
// allowing users to customize the index file without losing changes.
func generateDaoIndex(in generateDaoIndexInput) {
path := filepath.FromSlash(gfile.Join(in.DirPathDao, in.FileName+".go"))
// It should add path to result slice whenever it would generate the path file or not.
@ -147,15 +156,21 @@ func generateDaoIndex(in generateDaoIndexInput) {
}
}
// generateDaoInternalInput holds parameters for generating the dao internal file.
// The internal file contains the actual DAO implementation with column definitions
// and is always overwritten on regeneration.
type generateDaoInternalInput struct {
generateDaoSingleInput
TableNameCamelCase string
TableNameCamelLowerCase string
ImportPrefix string
FileName string
FieldMap map[string]*gdb.TableField
TableNameCamelCase string // CamelCase version of the table name.
TableNameCamelLowerCase string // camelCase version of the table name.
ImportPrefix string // Go import path prefix for the dao package.
FileName string // Output file name (without extension).
FieldMap map[string]*gdb.TableField // Map of column name to field metadata.
}
// generateDaoInternal generates the dao internal implementation file for a single table.
// This file is always regenerated (overwritten) and contains the Columns struct definition
// with column name constants and their string value assignments.
func generateDaoInternal(in generateDaoInternalInput) {
var (
ctx = context.Background()

View File

@ -22,6 +22,10 @@ import (
"github.com/gogf/gf/cmd/gf/v2/internal/utility/utils"
)
// generateDo generates DO (Data Object) files for all tables.
// DO structs use "any" type for all scalar fields (replacing concrete types),
// enabling flexible query building with the g.Meta `orm:"do:true"` tag.
// Pointer, slice, and map types are preserved as-is.
func generateDo(ctx context.Context, in CGenDaoInternalInput) {
var dirPathDo = filepath.FromSlash(gfile.Join(in.Path, in.DoPath))
in.genItems.AppendDirPath(dirPathDo)
@ -30,7 +34,7 @@ func generateDo(ctx context.Context, in CGenDaoInternalInput) {
in.NoModelComment = false
// Model content.
for i, tableName := range in.TableNames {
fieldMap, err := in.DB.TableFields(ctx, tableName)
fieldMap, err := getTableFields(ctx, in, tableName)
if err != nil {
mlog.Fatalf("fetching tables fields failed for table '%s':\n%v", tableName, err)
}
@ -75,6 +79,9 @@ func generateDo(ctx context.Context, in CGenDaoInternalInput) {
}
}
// generateDoContent renders the DO file content using the template engine.
// It assembles template variables including package imports, struct definition,
// and metadata, then parses the DO template to produce the final file content.
func generateDoContent(
ctx context.Context, in CGenDaoInternalInput, tableName, tableNameCamelCase, structDefine string,
) string {

View File

@ -20,12 +20,15 @@ import (
"github.com/gogf/gf/cmd/gf/v2/internal/utility/utils"
)
// generateEntity generates entity struct files for all tables.
// Entity structs represent database table rows with concrete Go types,
// including orm tags for field-to-column mapping and json tags for serialization.
func generateEntity(ctx context.Context, in CGenDaoInternalInput) {
var dirPathEntity = gfile.Join(in.Path, in.EntityPath)
in.genItems.AppendDirPath(dirPathEntity)
// Model content.
for i, tableName := range in.TableNames {
fieldMap, err := in.DB.TableFields(ctx, tableName)
fieldMap, err := getTableFields(ctx, in, tableName)
if err != nil {
mlog.Fatalf("fetching tables fields failed for table '%s':\n%v", tableName, err)
}
@ -60,6 +63,9 @@ func generateEntity(ctx context.Context, in CGenDaoInternalInput) {
}
}
// generateEntityContent renders the entity file content using the template engine.
// It assembles template variables and parses the entity template to produce
// the final Go source file content with proper imports and struct definition.
func generateEntityContent(
ctx context.Context, in CGenDaoInternalInput, tableName, tableNameCamelCase, structDefine string, appendImports []string,
) string {

View File

@ -7,17 +7,25 @@
package gendao
type (
// CGenDaoInternalGenItems tracks generation state across multiple configuration entries.
// Each configuration entry (e.g., different database links in the config array)
// gets its own CGenDaoInternalGenItem via Scale(). The index field points to the
// current active item.
CGenDaoInternalGenItems struct {
index int
Items []CGenDaoInternalGenItem
index int // Index of the current active generation item.
Items []CGenDaoInternalGenItem // List of all generation items, one per config entry.
}
// CGenDaoInternalGenItem tracks generated files and directories for a single
// configuration entry. Used by the Clear feature to identify and remove stale files.
CGenDaoInternalGenItem struct {
Clear bool
StorageDirPaths []string
GeneratedFilePaths []string
Clear bool // Whether to clear stale files for this item.
StorageDirPaths []string // Directories where generated files are stored (dao, do, entity, table).
GeneratedFilePaths []string // All file paths generated in this run.
}
)
// newCGenDaoInternalGenItems creates a new generation items tracker with an empty item list.
func newCGenDaoInternalGenItems() *CGenDaoInternalGenItems {
return &CGenDaoInternalGenItems{
index: -1,
@ -25,6 +33,8 @@ func newCGenDaoInternalGenItems() *CGenDaoInternalGenItems {
}
}
// Scale adds a new generation item and advances the index to it.
// Must be called once per configuration entry before generating files.
func (i *CGenDaoInternalGenItems) Scale() {
i.Items = append(i.Items, CGenDaoInternalGenItem{
StorageDirPaths: make([]string, 0),
@ -34,10 +44,12 @@ func (i *CGenDaoInternalGenItems) Scale() {
i.index++
}
// SetClear enables or disables the clear (stale file removal) flag for the current item.
func (i *CGenDaoInternalGenItems) SetClear(clear bool) {
i.Items[i.index].Clear = clear
}
// AppendDirPath records a directory path used for storing generated files in the current item.
func (i *CGenDaoInternalGenItems) AppendDirPath(storageDirPath string) {
i.Items[i.index].StorageDirPaths = append(
i.Items[i.index].StorageDirPaths,
@ -45,6 +57,7 @@ func (i *CGenDaoInternalGenItems) AppendDirPath(storageDirPath string) {
)
}
// AppendGeneratedFilePath records a file path that was generated in the current item.
func (i *CGenDaoInternalGenItems) AppendGeneratedFilePath(generatedFilePath string) {
i.Items[i.index].GeneratedFilePaths = append(
i.Items[i.index].GeneratedFilePaths,

File diff suppressed because it is too large Load Diff

View File

@ -0,0 +1,211 @@
// Copyright GoFrame gf Author(https://goframe.org). All Rights Reserved.
//
// This Source Code Form is subject to the terms of the MIT License.
// If a copy of the MIT was not distributed with this file,
// You can obtain one at https://github.com/gogf/gf.
package gendao
import (
"fmt"
"strings"
"github.com/gogf/gf/v2/database/gdb"
)
// MSSQLParser implements SQLParser for SQL Server (T-SQL) DDL.
type MSSQLParser struct{}
// ParseCreateTable parses a single MSSQL CREATE TABLE statement.
func (p *MSSQLParser) ParseCreateTable(stmt string) (string, map[string]*gdb.TableField, error) {
body, _, ok := extractBodyAndTrailing(stmt)
if !ok {
return "", nil, nil
}
parenIdx := strings.Index(stmt, "(")
header := stmt[:parenIdx]
tableName := extractTableName(header)
if tableName == "" {
return "", nil, fmt.Errorf("cannot extract table name from: %s", header)
}
columnDefs := splitColumns(body)
fields := make(map[string]*gdb.TableField)
pkColumns := findPrimaryKeysFromConstraints(columnDefs)
fieldIndex := 0
for _, def := range columnDefs {
def = strings.TrimSpace(def)
if def == "" {
continue
}
firstWord := strings.ToUpper(strings.Fields(def)[0])
if isConstraintKeyword(firstWord) {
continue
}
field, err := p.parseColumnDef(def, fieldIndex)
if err != nil {
continue
}
if field != nil {
fields[field.Name] = field
fieldIndex++
}
}
for _, pkCol := range pkColumns {
if f, ok := fields[pkCol]; ok {
f.Key = "PRI"
}
}
return tableName, fields, nil
}
// ParseAlterTable parses MSSQL ALTER TABLE statements.
func (p *MSSQLParser) ParseAlterTable(stmt string, tables map[string]map[string]*gdb.TableField) error {
return parseAlterTableCommon(stmt, tables, p.parseColumnDef)
}
// ParseComment parses EXEC sp_addextendedproperty to extract column comments.
func (p *MSSQLParser) ParseComment(stmt string, tables map[string]map[string]*gdb.TableField) {
upper := strings.ToUpper(strings.TrimSpace(stmt))
if !strings.Contains(upper, "SP_ADDEXTENDEDPROPERTY") ||
!strings.Contains(upper, "MS_DESCRIPTION") {
return
}
// Extract quoted string values
var values []string
inQuote := false
var current strings.Builder
for i := 0; i < len(stmt); i++ {
ch := stmt[i]
if ch == '\'' {
if inQuote {
if i+1 < len(stmt) && stmt[i+1] == '\'' {
current.WriteByte('\'')
i++
continue
}
values = append(values, current.String())
current.Reset()
inQuote = false
} else {
inQuote = true
}
} else if inQuote {
current.WriteByte(ch)
}
}
if len(values) < 8 {
return
}
var (
comment string
tableName string
columnName string
)
for i := 0; i < len(values)-1; i++ {
switch strings.ToUpper(values[i]) {
case "MS_DESCRIPTION":
comment = values[i+1]
case "TABLE":
tableName = values[i+1]
case "COLUMN":
columnName = values[i+1]
}
}
if tableName != "" && columnName != "" && comment != "" {
if fields, ok := tables[tableName]; ok {
if field, ok := fields[columnName]; ok {
field.Comment = comment
}
}
}
}
// parseColumnDef parses a single MSSQL column definition string into a TableField.
// It handles MSSQL-specific syntax including bracket-quoted identifiers and
// type parameters like varchar(max).
func (p *MSSQLParser) parseColumnDef(def string, index int) (*gdb.TableField, error) {
tokens := mysqlTokenize(def)
if len(tokens) < 2 {
return nil, fmt.Errorf("invalid column definition: %s", def)
}
field := &gdb.TableField{
Index: index,
Name: unquoteIdentifier(tokens[0]),
Null: true,
}
field.Type = tokens[1]
rest := ""
if len(tokens) > 2 {
rest = strings.Join(tokens[2:], " ")
}
if !strings.Contains(field.Type, "(") && strings.HasPrefix(strings.TrimSpace(rest), "(") {
end := strings.Index(rest, ")")
if end >= 0 {
field.Type += rest[:end+1]
rest = strings.TrimSpace(rest[end+1:])
}
}
p.parseColumnAttributes(field, rest)
return field, nil
}
// parseColumnAttributes parses MSSQL column constraint keywords including
// NOT NULL, NULL, PRIMARY KEY, UNIQUE, IDENTITY (auto-increment), and DEFAULT.
func (p *MSSQLParser) parseColumnAttributes(field *gdb.TableField, attrs string) {
words := strings.Fields(attrs)
upperWords := strings.Fields(strings.ToUpper(attrs))
for i := 0; i < len(upperWords); i++ {
switch upperWords[i] {
case "NOT":
if i+1 < len(upperWords) && upperWords[i+1] == "NULL" {
field.Null = false
i++
}
case "NULL":
field.Null = true
case "PRIMARY":
if i+1 < len(upperWords) && upperWords[i+1] == "KEY" {
field.Key = "PRI"
i++
}
case "UNIQUE":
if field.Key == "" {
field.Key = "UNI"
}
case "IDENTITY":
field.Extra = "auto_increment"
if i+1 < len(words) && strings.HasPrefix(words[i+1], "(") {
i++
}
default:
if strings.HasPrefix(upperWords[i], "IDENTITY(") || strings.HasPrefix(upperWords[i], "IDENTITY (") {
field.Extra = "auto_increment"
}
case "DEFAULT":
if i+1 < len(words) {
defaultVal, _ := extractDefaultValue("DEFAULT " + strings.Join(words[i+1:], " "))
field.Default = defaultVal
if defaultVal != nil {
i++
}
}
}
}
}

View File

@ -0,0 +1,72 @@
// Copyright GoFrame gf Author(https://goframe.org). All Rights Reserved.
//
// This Source Code Form is subject to the terms of the MIT License.
// If a copy of the MIT was not distributed with this file,
// You can obtain one at https://github.com/gogf/gf.
package gendao
import (
"testing"
"github.com/gogf/gf/v2/database/gdb"
"github.com/gogf/gf/v2/test/gtest"
)
func Test_MSSQL_CreateTable_Basic(t *testing.T) {
gtest.C(t, func(t *gtest.T) {
parser := &MSSQLParser{}
sql := `
CREATE TABLE [dbo].[users] (
[id] INT IDENTITY(1,1) NOT NULL,
[name] NVARCHAR(100) NOT NULL,
[email] NVARCHAR(200) NULL,
[balance] DECIMAL(18,2) DEFAULT 0,
[created_at] DATETIME2 NOT NULL DEFAULT GETDATE(),
CONSTRAINT [PK_users] PRIMARY KEY CLUSTERED ([id])
);
EXEC sp_addextendedproperty 'MS_Description', 'User ID', 'SCHEMA', 'dbo', 'TABLE', 'users', 'COLUMN', 'id';
EXEC sp_addextendedproperty 'MS_Description', 'User name', 'SCHEMA', 'dbo', 'TABLE', 'users', 'COLUMN', 'name';
`
tables := make(map[string]map[string]*gdb.TableField)
err := processSQL(parser, sql, tables)
t.AssertNil(err)
fields := tables["users"]
t.Assert(len(fields), 5)
t.Assert(fields["id"].Extra, "auto_increment")
t.Assert(fields["id"].Null, false)
t.Assert(fields["id"].Key, "PRI")
t.Assert(fields["id"].Comment, "User ID")
t.Assert(fields["name"].Comment, "User name")
t.Assert(fields["name"].Null, false)
t.Assert(fields["email"].Null, true)
})
}
func Test_MSSQL_AlterTable(t *testing.T) {
gtest.C(t, func(t *gtest.T) {
parser := &MSSQLParser{}
sql := `
CREATE TABLE users (
id INT IDENTITY(1,1) NOT NULL,
name NVARCHAR(100) NOT NULL,
CONSTRAINT PK_users PRIMARY KEY (id)
);
ALTER TABLE users ADD email NVARCHAR(200) NULL;
ALTER TABLE users DROP COLUMN name;
`
tables := make(map[string]map[string]*gdb.TableField)
err := processSQL(parser, sql, tables)
t.AssertNil(err)
fields := tables["users"]
t.Assert(len(fields), 2) // id, email
_, ok := fields["name"]
t.Assert(ok, false)
t.Assert(fields["email"].Null, true)
})
}

View File

@ -0,0 +1,199 @@
// Copyright GoFrame gf Author(https://goframe.org). All Rights Reserved.
//
// This Source Code Form is subject to the terms of the MIT License.
// If a copy of the MIT was not distributed with this file,
// You can obtain one at https://github.com/gogf/gf.
package gendao
import (
"fmt"
"strings"
"github.com/gogf/gf/v2/database/gdb"
)
// MySQLParser implements SQLParser for MySQL/MariaDB/TiDB DDL.
type MySQLParser struct{}
// ParseCreateTable parses a single MySQL CREATE TABLE statement.
func (p *MySQLParser) ParseCreateTable(stmt string) (string, map[string]*gdb.TableField, error) {
body, trailing, ok := extractBodyAndTrailing(stmt)
if !ok {
return "", nil, nil
}
parenIdx := strings.Index(stmt, "(")
header := stmt[:parenIdx]
tableName := extractTableName(header)
if tableName == "" {
return "", nil, fmt.Errorf("cannot extract table name from: %s", header)
}
columnDefs := splitColumns(body)
fields := make(map[string]*gdb.TableField)
pkColumns := findPrimaryKeysFromConstraints(columnDefs)
fieldIndex := 0
for _, def := range columnDefs {
def = strings.TrimSpace(def)
if def == "" {
continue
}
firstWord := strings.ToUpper(strings.Fields(def)[0])
if isConstraintKeyword(firstWord) {
continue
}
field, err := p.parseColumnDef(def, fieldIndex)
if err != nil {
continue
}
if field != nil {
fields[field.Name] = field
fieldIndex++
}
}
for _, pkCol := range pkColumns {
if f, ok := fields[pkCol]; ok {
f.Key = "PRI"
}
}
// Extract inline comments from trailing table options (not used for field generation)
_ = trailing
return tableName, fields, nil
}
// ParseAlterTable parses MySQL ALTER TABLE statements.
func (p *MySQLParser) ParseAlterTable(stmt string, tables map[string]map[string]*gdb.TableField) error {
return parseAlterTableCommon(stmt, tables, p.parseColumnDef)
}
// ParseComment handles MySQL-style comments (inline COMMENT keyword is handled in parseColumnDef).
func (p *MySQLParser) ParseComment(stmt string, tables map[string]map[string]*gdb.TableField) {
// MySQL uses inline COMMENT 'xxx' in column definitions,
// which is already handled by parseColumnDef. No separate COMMENT ON statement.
}
// parseColumnDef parses a single MySQL column definition string into a TableField.
// It extracts the column name, data type (including UNSIGNED modifier), and delegates
// attribute parsing (NULL, DEFAULT, PRIMARY KEY, COMMENT, etc.) to parseColumnAttributes.
func (p *MySQLParser) parseColumnDef(def string, index int) (*gdb.TableField, error) {
tokens := mysqlTokenize(def)
if len(tokens) < 2 {
return nil, fmt.Errorf("invalid column definition: %s", def)
}
field := &gdb.TableField{
Index: index,
Name: unquoteIdentifier(tokens[0]),
Null: true,
}
typeStr := tokens[1]
rest := ""
if len(tokens) > 2 {
rest = strings.Join(tokens[2:], " ")
}
// Check if rest starts with '(' meaning the type params are in rest
if !strings.Contains(typeStr, "(") && strings.HasPrefix(strings.TrimSpace(rest), "(") {
endParen := strings.Index(rest, ")")
if endParen >= 0 {
typeStr += rest[:endParen+1]
rest = strings.TrimSpace(rest[endParen+1:])
}
}
field.Type = typeStr
// Handle UNSIGNED
upperRest := strings.ToUpper(rest)
if strings.HasPrefix(upperRest, "UNSIGNED") {
field.Type += " unsigned"
rest = strings.TrimSpace(rest[8:])
}
p.parseColumnAttributes(field, rest)
return field, nil
}
// parseColumnAttributes parses MySQL column constraint keywords from the attribute string
// following the column type. It handles NOT NULL, NULL, PRIMARY KEY, UNIQUE, AUTO_INCREMENT,
// DEFAULT, COMMENT, and ON UPDATE clauses.
func (p *MySQLParser) parseColumnAttributes(field *gdb.TableField, attrs string) {
words := strings.Fields(attrs)
upperWords := strings.Fields(strings.ToUpper(attrs))
for i := 0; i < len(upperWords); i++ {
switch upperWords[i] {
case "NOT":
if i+1 < len(upperWords) && upperWords[i+1] == "NULL" {
field.Null = false
i++
}
case "NULL":
field.Null = true
case "PRIMARY":
if i+1 < len(upperWords) && upperWords[i+1] == "KEY" {
field.Key = "PRI"
i++
}
case "UNIQUE":
if field.Key == "" {
field.Key = "UNI"
}
if i+1 < len(upperWords) && upperWords[i+1] == "KEY" {
i++
}
case "KEY":
if field.Key == "" {
field.Key = "MUL"
}
case "AUTO_INCREMENT":
field.Extra = "auto_increment"
case "DEFAULT":
if i+1 < len(words) {
defaultVal, _ := extractDefaultValue("DEFAULT " + strings.Join(words[i+1:], " "))
field.Default = defaultVal
if defaultVal != nil {
if strings.HasPrefix(words[i+1], "'") {
for j := i + 1; j < len(words); j++ {
if strings.HasSuffix(words[j], "'") {
i = j
break
}
}
} else {
i++
}
}
}
case "COMMENT":
if i+1 < len(words) {
comment := strings.Join(words[i+1:], " ")
comment = strings.TrimSpace(comment)
if len(comment) >= 2 && comment[0] == '\'' && comment[len(comment)-1] == '\'' {
comment = comment[1 : len(comment)-1]
comment = strings.ReplaceAll(comment, "''", "'")
}
field.Comment = comment
return
}
case "ON":
if i+1 < len(upperWords) && upperWords[i+1] == "UPDATE" {
if i+2 < len(upperWords) {
if field.Extra != "" {
field.Extra += ", "
}
field.Extra += "on update " + strings.ToLower(words[i+2])
i += 2
}
}
}
}
}

View File

@ -0,0 +1,300 @@
// Copyright GoFrame gf Author(https://goframe.org). All Rights Reserved.
//
// This Source Code Form is subject to the terms of the MIT License.
// If a copy of the MIT was not distributed with this file,
// You can obtain one at https://github.com/gogf/gf.
package gendao
import (
"testing"
"github.com/gogf/gf/v2/database/gdb"
"github.com/gogf/gf/v2/test/gtest"
)
func Test_MySQL_CreateTable_Basic(t *testing.T) {
gtest.C(t, func(t *gtest.T) {
parser := &MySQLParser{}
sql := `
CREATE TABLE users (
id BIGINT(20) UNSIGNED NOT NULL AUTO_INCREMENT COMMENT 'User ID',
name VARCHAR(100) NOT NULL DEFAULT '' COMMENT 'User name',
email VARCHAR(200) NULL COMMENT 'Email address',
age INT(11) DEFAULT 0,
score DECIMAL(10,2) DEFAULT 0.00,
status TINYINT(1) NOT NULL DEFAULT 1,
created_at DATETIME NOT NULL DEFAULT CURRENT_TIMESTAMP,
updated_at DATETIME NULL ON UPDATE CURRENT_TIMESTAMP,
PRIMARY KEY (id),
UNIQUE KEY uk_email (email)
) ENGINE=InnoDB DEFAULT CHARSET=utf8mb4 COMMENT='User table';
`
tables := make(map[string]map[string]*gdb.TableField)
err := processSQL(parser, sql, tables)
t.AssertNil(err)
t.Assert(len(tables), 1)
fields := tables["users"]
t.Assert(len(fields), 8)
// Check id field
t.Assert(fields["id"].Name, "id")
t.Assert(fields["id"].Type, "BIGINT(20) unsigned")
t.Assert(fields["id"].Null, false)
t.Assert(fields["id"].Key, "PRI")
t.Assert(fields["id"].Extra, "auto_increment")
t.Assert(fields["id"].Comment, "User ID")
t.Assert(fields["id"].Index, 0)
// Check name field
t.Assert(fields["name"].Name, "name")
t.Assert(fields["name"].Null, false)
t.Assert(fields["name"].Comment, "User name")
// Check email field
t.Assert(fields["email"].Null, true)
// Check created_at
t.Assert(fields["created_at"].Null, false)
})
}
func Test_MySQL_AlterTable_AddColumn(t *testing.T) {
gtest.C(t, func(t *gtest.T) {
parser := &MySQLParser{}
sql := `
CREATE TABLE users (
id INT NOT NULL AUTO_INCREMENT,
name VARCHAR(100) NOT NULL,
PRIMARY KEY (id)
);
ALTER TABLE users ADD COLUMN email VARCHAR(200) NULL COMMENT 'Email';
ALTER TABLE users ADD COLUMN age INT DEFAULT 0;
`
tables := make(map[string]map[string]*gdb.TableField)
err := processSQL(parser, sql, tables)
t.AssertNil(err)
fields := tables["users"]
t.Assert(len(fields), 4)
t.Assert(fields["email"].Name, "email")
t.Assert(fields["email"].Null, true)
t.Assert(fields["email"].Comment, "Email")
t.Assert(fields["age"].Name, "age")
})
}
func Test_MySQL_AlterTable_DropColumn(t *testing.T) {
gtest.C(t, func(t *gtest.T) {
parser := &MySQLParser{}
sql := `
CREATE TABLE users (
id INT NOT NULL AUTO_INCREMENT,
name VARCHAR(100),
old_field VARCHAR(50),
PRIMARY KEY (id)
);
ALTER TABLE users DROP COLUMN old_field;
`
tables := make(map[string]map[string]*gdb.TableField)
err := processSQL(parser, sql, tables)
t.AssertNil(err)
fields := tables["users"]
t.Assert(len(fields), 2)
_, ok := fields["old_field"]
t.Assert(ok, false)
})
}
func Test_MySQL_AlterTable_ModifyColumn(t *testing.T) {
gtest.C(t, func(t *gtest.T) {
parser := &MySQLParser{}
sql := `
CREATE TABLE users (
id INT NOT NULL AUTO_INCREMENT,
name VARCHAR(100),
PRIMARY KEY (id)
);
ALTER TABLE users MODIFY COLUMN name VARCHAR(200) NOT NULL COMMENT 'Full name';
`
tables := make(map[string]map[string]*gdb.TableField)
err := processSQL(parser, sql, tables)
t.AssertNil(err)
fields := tables["users"]
t.Assert(fields["name"].Type, "VARCHAR(200)")
t.Assert(fields["name"].Null, false)
t.Assert(fields["name"].Comment, "Full name")
})
}
func Test_MySQL_AlterTable_ChangeColumn(t *testing.T) {
gtest.C(t, func(t *gtest.T) {
parser := &MySQLParser{}
sql := `
CREATE TABLE users (
id INT NOT NULL AUTO_INCREMENT,
old_name VARCHAR(100),
PRIMARY KEY (id)
);
ALTER TABLE users CHANGE COLUMN old_name new_name VARCHAR(200) NOT NULL;
`
tables := make(map[string]map[string]*gdb.TableField)
err := processSQL(parser, sql, tables)
t.AssertNil(err)
fields := tables["users"]
_, ok := fields["old_name"]
t.Assert(ok, false)
t.Assert(fields["new_name"].Name, "new_name")
t.Assert(fields["new_name"].Type, "VARCHAR(200)")
})
}
func Test_MySQL_AlterTable_AddPrimaryKey(t *testing.T) {
gtest.C(t, func(t *gtest.T) {
parser := &MySQLParser{}
sql := `
CREATE TABLE users (
id INT NOT NULL,
name VARCHAR(100)
);
ALTER TABLE users ADD PRIMARY KEY (id);
`
tables := make(map[string]map[string]*gdb.TableField)
err := processSQL(parser, sql, tables)
t.AssertNil(err)
t.Assert(tables["users"]["id"].Key, "PRI")
})
}
func Test_MySQL_DropTable(t *testing.T) {
gtest.C(t, func(t *gtest.T) {
parser := &MySQLParser{}
sql := `
CREATE TABLE temp_log (id INT, msg TEXT);
CREATE TABLE users (id INT, name VARCHAR(100));
DROP TABLE IF EXISTS temp_log;
`
tables := make(map[string]map[string]*gdb.TableField)
err := processSQL(parser, sql, tables)
t.AssertNil(err)
t.Assert(len(tables), 1)
_, ok := tables["temp_log"]
t.Assert(ok, false)
_, ok = tables["users"]
t.Assert(ok, true)
})
}
func Test_MySQL_MultipleMigrations(t *testing.T) {
gtest.C(t, func(t *gtest.T) {
parser := &MySQLParser{}
// Simulate V1: initial schema
tables := make(map[string]map[string]*gdb.TableField)
err := processSQL(parser, `
CREATE TABLE users (
id BIGINT UNSIGNED NOT NULL AUTO_INCREMENT,
name VARCHAR(50) NOT NULL,
PRIMARY KEY (id)
);
`, tables)
t.AssertNil(err)
// Simulate V2: add columns
err = processSQL(parser, `
ALTER TABLE users ADD COLUMN email VARCHAR(200) NULL;
ALTER TABLE users ADD COLUMN phone VARCHAR(20) NULL;
`, tables)
t.AssertNil(err)
// Simulate V3: modify + drop
err = processSQL(parser, `
ALTER TABLE users MODIFY COLUMN name VARCHAR(100) NOT NULL COMMENT 'Full name';
ALTER TABLE users DROP COLUMN phone;
`, tables)
t.AssertNil(err)
fields := tables["users"]
t.Assert(len(fields), 3) // id, name, email
t.Assert(fields["name"].Type, "VARCHAR(100)")
t.Assert(fields["name"].Comment, "Full name")
_, ok := fields["phone"]
t.Assert(ok, false)
t.Assert(fields["email"].Null, true)
})
}
func Test_MySQL_FullMigrationScenario(t *testing.T) {
gtest.C(t, func(t *gtest.T) {
parser := &MySQLParser{}
tables := make(map[string]map[string]*gdb.TableField)
// V001: Initial tables
err := processSQL(parser, `
CREATE TABLE IF NOT EXISTS users (
id BIGINT UNSIGNED NOT NULL AUTO_INCREMENT COMMENT 'Primary key',
username VARCHAR(50) NOT NULL COMMENT 'Username',
password VARCHAR(128) NOT NULL COMMENT 'Hashed password',
created_at DATETIME NOT NULL DEFAULT CURRENT_TIMESTAMP,
PRIMARY KEY (id),
UNIQUE KEY uk_username (username)
);
CREATE TABLE IF NOT EXISTS orders (
id BIGINT UNSIGNED NOT NULL AUTO_INCREMENT,
user_id BIGINT UNSIGNED NOT NULL,
amount DECIMAL(10,2) NOT NULL DEFAULT 0.00,
PRIMARY KEY (id)
);
`, tables)
t.AssertNil(err)
t.Assert(len(tables), 2)
// V002: Add email, phone
err = processSQL(parser, `
ALTER TABLE users ADD COLUMN email VARCHAR(200) NULL COMMENT 'User email';
ALTER TABLE users ADD COLUMN phone VARCHAR(20) NULL COMMENT 'Phone number';
`, tables)
t.AssertNil(err)
t.Assert(len(tables["users"]), 6)
// V003: Modify, rename, drop
err = processSQL(parser, `
ALTER TABLE users MODIFY COLUMN username VARCHAR(100) NOT NULL COMMENT 'Login name';
ALTER TABLE users CHANGE COLUMN phone mobile VARCHAR(20) NULL COMMENT 'Mobile number';
ALTER TABLE users DROP COLUMN password;
ALTER TABLE orders ADD COLUMN status TINYINT(1) NOT NULL DEFAULT 0 COMMENT 'Order status';
`, tables)
t.AssertNil(err)
userFields := tables["users"]
t.Assert(len(userFields), 5) // id, username, email, mobile, created_at
t.Assert(userFields["username"].Type, "VARCHAR(100)")
t.Assert(userFields["username"].Comment, "Login name")
_, ok := userFields["password"]
t.Assert(ok, false)
_, ok = userFields["phone"]
t.Assert(ok, false)
t.Assert(userFields["mobile"].Name, "mobile")
t.Assert(userFields["mobile"].Comment, "Mobile number")
orderFields := tables["orders"]
t.Assert(len(orderFields), 4)
t.Assert(orderFields["status"].Default, "0")
// V004: Drop table
err = processSQL(parser, `
DROP TABLE IF EXISTS orders;
`, tables)
t.AssertNil(err)
t.Assert(len(tables), 1)
_, ok = tables["orders"]
t.Assert(ok, false)
})
}

View File

@ -0,0 +1,209 @@
// Copyright GoFrame gf Author(https://goframe.org). All Rights Reserved.
//
// This Source Code Form is subject to the terms of the MIT License.
// If a copy of the MIT was not distributed with this file,
// You can obtain one at https://github.com/gogf/gf.
package gendao
import (
"fmt"
"strings"
"github.com/gogf/gf/v2/database/gdb"
)
// OracleParser implements SQLParser for Oracle/DM DDL.
type OracleParser struct{}
// ParseCreateTable parses a single Oracle CREATE TABLE statement.
func (p *OracleParser) ParseCreateTable(stmt string) (string, map[string]*gdb.TableField, error) {
body, _, ok := extractBodyAndTrailing(stmt)
if !ok {
return "", nil, nil
}
parenIdx := strings.Index(stmt, "(")
header := stmt[:parenIdx]
tableName := extractTableName(header)
if tableName == "" {
return "", nil, fmt.Errorf("cannot extract table name from: %s", header)
}
columnDefs := splitColumns(body)
fields := make(map[string]*gdb.TableField)
pkColumns := findPrimaryKeysFromConstraints(columnDefs)
fieldIndex := 0
for _, def := range columnDefs {
def = strings.TrimSpace(def)
if def == "" {
continue
}
firstWord := strings.ToUpper(strings.Fields(def)[0])
if isConstraintKeyword(firstWord) {
continue
}
field, err := p.parseColumnDef(def, fieldIndex)
if err != nil {
continue
}
if field != nil {
fields[field.Name] = field
fieldIndex++
}
}
for _, pkCol := range pkColumns {
if f, ok := fields[pkCol]; ok {
f.Key = "PRI"
}
upperPk := strings.ToUpper(pkCol)
if f, ok := fields[upperPk]; ok {
f.Key = "PRI"
}
}
return tableName, fields, nil
}
// ParseAlterTable parses Oracle ALTER TABLE statements.
func (p *OracleParser) ParseAlterTable(stmt string, tables map[string]map[string]*gdb.TableField) error {
return parseAlterTableCommon(stmt, tables, p.parseColumnDef)
}
// ParseComment parses COMMENT ON COLUMN table.column IS 'comment'.
func (p *OracleParser) ParseComment(stmt string, tables map[string]map[string]*gdb.TableField) {
upper := strings.ToUpper(strings.TrimSpace(stmt))
if !strings.HasPrefix(upper, "COMMENT ON COLUMN") {
return
}
rest := strings.TrimSpace(stmt[len("COMMENT ON COLUMN"):])
isIdx := strings.Index(strings.ToUpper(rest), " IS ")
if isIdx < 0 {
return
}
ref := strings.TrimSpace(rest[:isIdx])
comment := strings.TrimSpace(rest[isIdx+4:])
if len(comment) >= 2 && comment[0] == '\'' && comment[len(comment)-1] == '\'' {
comment = comment[1 : len(comment)-1]
comment = strings.ReplaceAll(comment, "''", "'")
}
parts := strings.Split(ref, ".")
var tableName, columnName string
switch len(parts) {
case 2:
tableName = unquoteIdentifier(parts[0])
columnName = unquoteIdentifier(parts[1])
case 3:
tableName = unquoteIdentifier(parts[1])
columnName = unquoteIdentifier(parts[2])
default:
return
}
if fields, ok := tables[tableName]; ok {
if field, ok := fields[columnName]; ok {
field.Comment = comment
}
}
}
// parseColumnDef parses a single Oracle column definition string into a TableField.
// It handles Oracle-specific types including TIMESTAMP WITH TIME ZONE and
// TIMESTAMP WITH LOCAL TIME ZONE.
func (p *OracleParser) parseColumnDef(def string, index int) (*gdb.TableField, error) {
tokens := mysqlTokenize(def)
if len(tokens) < 2 {
return nil, fmt.Errorf("invalid column definition: %s", def)
}
field := &gdb.TableField{
Index: index,
Name: unquoteIdentifier(tokens[0]),
Null: true,
}
field.Type = tokens[1]
rest := ""
if len(tokens) > 2 {
rest = strings.Join(tokens[2:], " ")
}
if !strings.Contains(field.Type, "(") && strings.HasPrefix(strings.TrimSpace(rest), "(") {
end := strings.Index(rest, ")")
if end >= 0 {
field.Type += rest[:end+1]
rest = strings.TrimSpace(rest[end+1:])
}
}
// Handle TIMESTAMP WITH TIME ZONE / WITH LOCAL TIME ZONE
upperType := strings.ToUpper(field.Type)
upperRest := strings.ToUpper(rest)
if upperType == "TIMESTAMP" {
if strings.HasPrefix(upperRest, "WITH LOCAL TIME ZONE") {
field.Type = "timestamp with local time zone"
rest = strings.TrimSpace(rest[len("WITH LOCAL TIME ZONE"):])
} else if strings.HasPrefix(upperRest, "WITH TIME ZONE") {
field.Type = "timestamp with time zone"
rest = strings.TrimSpace(rest[len("WITH TIME ZONE"):])
}
}
p.parseColumnAttributes(field, rest)
return field, nil
}
// parseColumnAttributes parses Oracle column constraint keywords including
// NOT NULL, NULL, PRIMARY KEY, UNIQUE, DEFAULT, and GENERATED ... AS IDENTITY.
func (p *OracleParser) parseColumnAttributes(field *gdb.TableField, attrs string) {
words := strings.Fields(attrs)
upperWords := strings.Fields(strings.ToUpper(attrs))
for i := 0; i < len(upperWords); i++ {
switch upperWords[i] {
case "NOT":
if i+1 < len(upperWords) && upperWords[i+1] == "NULL" {
field.Null = false
i++
}
case "NULL":
field.Null = true
case "PRIMARY":
if i+1 < len(upperWords) && upperWords[i+1] == "KEY" {
field.Key = "PRI"
i++
}
case "UNIQUE":
if field.Key == "" {
field.Key = "UNI"
}
case "DEFAULT":
if i+1 < len(words) {
defaultVal, _ := extractDefaultValue("DEFAULT " + strings.Join(words[i+1:], " "))
field.Default = defaultVal
if defaultVal != nil {
i++
}
}
case "GENERATED":
rest := strings.Join(upperWords[i:], " ")
if strings.Contains(rest, "AS IDENTITY") {
field.Extra = "auto_increment"
for j := i + 1; j < len(upperWords); j++ {
if upperWords[j] == "IDENTITY" {
i = j
break
}
}
}
}
}
}

View File

@ -0,0 +1,97 @@
// Copyright GoFrame gf Author(https://goframe.org). All Rights Reserved.
//
// This Source Code Form is subject to the terms of the MIT License.
// If a copy of the MIT was not distributed with this file,
// You can obtain one at https://github.com/gogf/gf.
package gendao
import (
"testing"
"github.com/gogf/gf/v2/database/gdb"
"github.com/gogf/gf/v2/test/gtest"
)
func Test_Oracle_CreateTable_Basic(t *testing.T) {
gtest.C(t, func(t *gtest.T) {
parser := &OracleParser{}
sql := `
CREATE TABLE users (
ID NUMBER(10) NOT NULL,
NAME VARCHAR2(100) NOT NULL,
EMAIL VARCHAR2(200),
CREATED_AT TIMESTAMP WITH TIME ZONE DEFAULT SYSTIMESTAMP,
CONSTRAINT PK_USERS PRIMARY KEY (ID)
);
COMMENT ON COLUMN users.ID IS 'User ID';
COMMENT ON COLUMN users.NAME IS 'User name';
`
tables := make(map[string]map[string]*gdb.TableField)
err := processSQL(parser, sql, tables)
t.AssertNil(err)
fields := tables["users"]
t.Assert(len(fields), 4)
t.Assert(fields["ID"].Key, "PRI")
t.Assert(fields["ID"].Null, false)
t.Assert(fields["ID"].Comment, "User ID")
t.Assert(fields["NAME"].Null, false)
t.Assert(fields["NAME"].Comment, "User name")
t.Assert(fields["CREATED_AT"].Type, "timestamp with time zone")
})
}
func Test_Oracle_AlterTable(t *testing.T) {
gtest.C(t, func(t *gtest.T) {
parser := &OracleParser{}
sql := `
CREATE TABLE users (
ID NUMBER(10) NOT NULL,
NAME VARCHAR2(100),
CONSTRAINT PK_USERS PRIMARY KEY (ID)
);
ALTER TABLE users ADD EMAIL VARCHAR2(200);
ALTER TABLE users MODIFY NAME VARCHAR2(200) NOT NULL;
COMMENT ON COLUMN users.EMAIL IS 'Email address';
`
tables := make(map[string]map[string]*gdb.TableField)
err := processSQL(parser, sql, tables)
t.AssertNil(err)
fields := tables["users"]
t.Assert(len(fields), 3)
t.Assert(fields["EMAIL"].Comment, "Email address")
t.Assert(fields["NAME"].Type, "VARCHAR2(200)")
t.Assert(fields["NAME"].Null, false)
})
}
func Test_Oracle_AlterTable_DropColumn(t *testing.T) {
gtest.C(t, func(t *gtest.T) {
parser := &OracleParser{}
sql := `
CREATE TABLE users (
ID NUMBER(10) NOT NULL,
NAME VARCHAR2(100) NOT NULL,
OLD_COL VARCHAR2(50),
EMAIL VARCHAR2(200),
CONSTRAINT PK_USERS PRIMARY KEY (ID)
);
ALTER TABLE users DROP COLUMN OLD_COL;
`
tables := make(map[string]map[string]*gdb.TableField)
err := processSQL(parser, sql, tables)
t.AssertNil(err)
fields := tables["users"]
t.Assert(len(fields), 3)
_, ok := fields["OLD_COL"]
t.Assert(ok, false)
t.Assert(fields["NAME"].Name, "NAME")
t.Assert(fields["EMAIL"].Name, "EMAIL")
})
}

View File

@ -0,0 +1,268 @@
// Copyright GoFrame gf Author(https://goframe.org). All Rights Reserved.
//
// This Source Code Form is subject to the terms of the MIT License.
// If a copy of the MIT was not distributed with this file,
// You can obtain one at https://github.com/gogf/gf.
package gendao
import (
"fmt"
"strings"
"github.com/gogf/gf/v2/database/gdb"
)
// PgSQLParser implements SQLParser for PostgreSQL DDL.
type PgSQLParser struct{}
// ParseCreateTable parses a single PostgreSQL CREATE TABLE statement.
func (p *PgSQLParser) ParseCreateTable(stmt string) (string, map[string]*gdb.TableField, error) {
body, _, ok := extractBodyAndTrailing(stmt)
if !ok {
return "", nil, nil
}
parenIdx := strings.Index(stmt, "(")
header := stmt[:parenIdx]
tableName := extractTableName(header)
if tableName == "" {
return "", nil, fmt.Errorf("cannot extract table name from: %s", header)
}
columnDefs := splitColumns(body)
fields := make(map[string]*gdb.TableField)
pkColumns := findPrimaryKeysFromConstraints(columnDefs)
fieldIndex := 0
for _, def := range columnDefs {
def = strings.TrimSpace(def)
if def == "" {
continue
}
firstWord := strings.ToUpper(strings.Fields(def)[0])
if isConstraintKeyword(firstWord) {
continue
}
field, err := p.parseColumnDef(def, fieldIndex)
if err != nil {
continue
}
if field != nil {
fields[field.Name] = field
fieldIndex++
}
}
for _, pkCol := range pkColumns {
if f, ok := fields[pkCol]; ok {
f.Key = "PRI"
}
}
return tableName, fields, nil
}
// ParseAlterTable parses PostgreSQL ALTER TABLE statements.
func (p *PgSQLParser) ParseAlterTable(stmt string, tables map[string]map[string]*gdb.TableField) error {
return parseAlterTableCommon(stmt, tables, p.parseColumnDef)
}
// ParseComment parses COMMENT ON COLUMN schema.table.column IS 'comment' statements.
func (p *PgSQLParser) ParseComment(stmt string, tables map[string]map[string]*gdb.TableField) {
upper := strings.ToUpper(strings.TrimSpace(stmt))
if !strings.HasPrefix(upper, "COMMENT ON COLUMN") {
return
}
rest := strings.TrimSpace(stmt[len("COMMENT ON COLUMN"):])
isIdx := strings.Index(strings.ToUpper(rest), " IS ")
if isIdx < 0 {
return
}
ref := strings.TrimSpace(rest[:isIdx])
comment := strings.TrimSpace(rest[isIdx+4:])
if len(comment) >= 2 && comment[0] == '\'' && comment[len(comment)-1] == '\'' {
comment = comment[1 : len(comment)-1]
comment = strings.ReplaceAll(comment, "''", "'")
}
parts := strings.Split(ref, ".")
var tableName, columnName string
switch len(parts) {
case 2:
tableName = unquoteIdentifier(parts[0])
columnName = unquoteIdentifier(parts[1])
case 3:
tableName = unquoteIdentifier(parts[1])
columnName = unquoteIdentifier(parts[2])
default:
return
}
if fields, ok := tables[tableName]; ok {
if field, ok := fields[columnName]; ok {
field.Comment = comment
}
}
}
// parseColumnDef parses a single PostgreSQL column definition string into a TableField.
// It handles PostgreSQL-specific types like SERIAL/BIGSERIAL (auto-increment shorthand),
// CHARACTER VARYING, DOUBLE PRECISION, TIMESTAMP WITH TIME ZONE, and array types.
func (p *PgSQLParser) parseColumnDef(def string, index int) (*gdb.TableField, error) {
tokens := mysqlTokenize(def)
if len(tokens) < 2 {
return nil, fmt.Errorf("invalid column definition: %s", def)
}
field := &gdb.TableField{
Index: index,
Name: unquoteIdentifier(tokens[0]),
Null: true,
}
// Handle SERIAL types
typeToken := strings.ToUpper(tokens[1])
switch typeToken {
case "SERIAL":
field.Type = "int"
field.Extra = "auto_increment"
field.Null = false
case "BIGSERIAL":
field.Type = "bigint"
field.Extra = "auto_increment"
field.Null = false
case "SMALLSERIAL":
field.Type = "smallint"
field.Extra = "auto_increment"
field.Null = false
default:
field.Type = tokens[1]
}
rest := ""
if len(tokens) > 2 {
rest = strings.Join(tokens[2:], " ")
}
upperType := strings.ToUpper(field.Type)
upperRest := strings.ToUpper(rest)
switch {
case upperType == "CHARACTER" && strings.HasPrefix(upperRest, "VARYING"):
rest = strings.TrimSpace(rest[len("VARYING"):])
if strings.HasPrefix(rest, "(") {
end := strings.Index(rest, ")")
if end >= 0 {
field.Type = "character varying" + rest[:end+1]
rest = strings.TrimSpace(rest[end+1:])
}
} else {
field.Type = "character varying"
}
case upperType == "DOUBLE" && strings.HasPrefix(upperRest, "PRECISION"):
field.Type = "double precision"
rest = strings.TrimSpace(rest[len("PRECISION"):])
case (upperType == "TIMESTAMP" || upperType == "TIME") &&
(strings.HasPrefix(upperRest, "WITH TIME ZONE") || strings.HasPrefix(upperRest, "WITHOUT TIME ZONE")):
if strings.HasPrefix(upperRest, "WITH TIME ZONE") {
if upperType == "TIMESTAMP" {
field.Type = "timestamptz"
} else {
field.Type = "time with time zone"
}
rest = strings.TrimSpace(rest[len("WITH TIME ZONE"):])
} else {
field.Type = strings.ToLower(upperType)
rest = strings.TrimSpace(rest[len("WITHOUT TIME ZONE"):])
}
case !strings.Contains(field.Type, "(") && strings.HasPrefix(strings.TrimSpace(rest), "("):
end := strings.Index(rest, ")")
if end >= 0 {
field.Type += rest[:end+1]
rest = strings.TrimSpace(rest[end+1:])
}
}
// Handle array types
if strings.HasPrefix(rest, "[]") {
field.Type += "[]"
rest = strings.TrimSpace(rest[2:])
} else if strings.HasPrefix(strings.ToUpper(rest), "ARRAY") {
field.Type += "[]"
rest = strings.TrimSpace(rest[5:])
}
p.parseColumnAttributes(field, rest)
return field, nil
}
// parseColumnAttributes parses PostgreSQL column constraint keywords including
// NOT NULL, NULL, PRIMARY KEY, UNIQUE, DEFAULT, GENERATED ... AS IDENTITY, and REFERENCES.
func (p *PgSQLParser) parseColumnAttributes(field *gdb.TableField, attrs string) {
words := strings.Fields(attrs)
upperWords := strings.Fields(strings.ToUpper(attrs))
for i := 0; i < len(upperWords); i++ {
switch upperWords[i] {
case "NOT":
if i+1 < len(upperWords) && upperWords[i+1] == "NULL" {
field.Null = false
i++
}
case "NULL":
field.Null = true
case "PRIMARY":
if i+1 < len(upperWords) && upperWords[i+1] == "KEY" {
field.Key = "PRI"
i++
}
case "UNIQUE":
if field.Key == "" {
field.Key = "UNI"
}
case "DEFAULT":
if i+1 < len(words) {
defaultVal, _ := extractDefaultValue("DEFAULT " + strings.Join(words[i+1:], " "))
field.Default = defaultVal
if defaultVal != nil {
i++
}
}
case "GENERATED":
if containsSequence(upperWords[i:], "ALWAYS", "AS", "IDENTITY") ||
containsSequence(upperWords[i:], "BY", "DEFAULT", "AS", "IDENTITY") {
field.Extra = "auto_increment"
for j := i + 1; j < len(upperWords); j++ {
if upperWords[j] == "IDENTITY" {
i = j
break
}
}
}
case "REFERENCES":
for j := i + 1; j < len(upperWords); j++ {
i = j
if strings.Contains(words[j], ")") {
break
}
}
}
}
}
// containsSequence checks if words slice contains the given word sequence starting from index 1.
func containsSequence(words []string, seq ...string) bool {
if len(words) < len(seq)+1 {
return false
}
for i, s := range seq {
if words[i+1] != s {
return false
}
}
return true
}

View File

@ -0,0 +1,232 @@
// Copyright GoFrame gf Author(https://goframe.org). All Rights Reserved.
//
// This Source Code Form is subject to the terms of the MIT License.
// If a copy of the MIT was not distributed with this file,
// You can obtain one at https://github.com/gogf/gf.
package gendao
import (
"testing"
"github.com/gogf/gf/v2/database/gdb"
"github.com/gogf/gf/v2/test/gtest"
)
func Test_PgSQL_CreateTable_Basic(t *testing.T) {
gtest.C(t, func(t *gtest.T) {
parser := &PgSQLParser{}
sql := `
CREATE TABLE users (
id BIGSERIAL PRIMARY KEY,
name VARCHAR(100) NOT NULL,
email CHARACTER VARYING(200),
score DOUBLE PRECISION DEFAULT 0.0,
metadata JSONB,
created_at TIMESTAMP WITH TIME ZONE DEFAULT CURRENT_TIMESTAMP,
is_active BOOLEAN NOT NULL DEFAULT TRUE
);
COMMENT ON COLUMN users.name IS 'User full name';
COMMENT ON COLUMN users.email IS 'Email address';
`
tables := make(map[string]map[string]*gdb.TableField)
err := processSQL(parser, sql, tables)
t.AssertNil(err)
fields := tables["users"]
t.Assert(len(fields), 7)
// BIGSERIAL should be auto_increment bigint
t.Assert(fields["id"].Type, "bigint")
t.Assert(fields["id"].Extra, "auto_increment")
t.Assert(fields["id"].Key, "PRI")
// CHARACTER VARYING
t.AssertNE(fields["email"], nil)
// DOUBLE PRECISION
t.Assert(fields["score"].Type, "double precision")
// JSONB
t.Assert(fields["metadata"].Type, "JSONB")
// TIMESTAMP WITH TIME ZONE
t.Assert(fields["created_at"].Type, "timestamptz")
// COMMENT ON COLUMN
t.Assert(fields["name"].Comment, "User full name")
t.Assert(fields["email"].Comment, "Email address")
})
}
func Test_PgSQL_AlterTable_AddColumn(t *testing.T) {
gtest.C(t, func(t *gtest.T) {
parser := &PgSQLParser{}
sql := `
CREATE TABLE users (
id SERIAL PRIMARY KEY,
name VARCHAR(100) NOT NULL
);
ALTER TABLE users ADD COLUMN email VARCHAR(200);
COMMENT ON COLUMN users.email IS 'User email';
`
tables := make(map[string]map[string]*gdb.TableField)
err := processSQL(parser, sql, tables)
t.AssertNil(err)
fields := tables["users"]
t.Assert(len(fields), 3)
t.Assert(fields["email"].Name, "email")
t.Assert(fields["email"].Comment, "User email")
})
}
func Test_PgSQL_AlterTable_AlterColumnType(t *testing.T) {
gtest.C(t, func(t *gtest.T) {
parser := &PgSQLParser{}
sql := `
CREATE TABLE users (
id SERIAL PRIMARY KEY,
name VARCHAR(100)
);
ALTER TABLE users ALTER COLUMN name TYPE VARCHAR(200);
ALTER TABLE users ALTER COLUMN name SET NOT NULL;
`
tables := make(map[string]map[string]*gdb.TableField)
err := processSQL(parser, sql, tables)
t.AssertNil(err)
fields := tables["users"]
t.Assert(fields["name"].Type, "VARCHAR(200)")
t.Assert(fields["name"].Null, false)
})
}
func Test_PgSQL_AlterTable_DropColumn(t *testing.T) {
gtest.C(t, func(t *gtest.T) {
parser := &PgSQLParser{}
sql := `
CREATE TABLE users (
id SERIAL PRIMARY KEY,
name VARCHAR(100),
old_col TEXT
);
ALTER TABLE users DROP COLUMN old_col;
`
tables := make(map[string]map[string]*gdb.TableField)
err := processSQL(parser, sql, tables)
t.AssertNil(err)
fields := tables["users"]
t.Assert(len(fields), 2)
_, ok := fields["old_col"]
t.Assert(ok, false)
})
}
func Test_PgSQL_AlterTable_RenameColumn(t *testing.T) {
gtest.C(t, func(t *gtest.T) {
parser := &PgSQLParser{}
sql := `
CREATE TABLE users (
id SERIAL PRIMARY KEY,
old_name VARCHAR(100)
);
ALTER TABLE users RENAME COLUMN old_name TO new_name;
`
tables := make(map[string]map[string]*gdb.TableField)
err := processSQL(parser, sql, tables)
t.AssertNil(err)
fields := tables["users"]
_, ok := fields["old_name"]
t.Assert(ok, false)
t.Assert(fields["new_name"].Name, "new_name")
})
}
func Test_PgSQL_MultipleMigrations(t *testing.T) {
gtest.C(t, func(t *gtest.T) {
parser := &PgSQLParser{}
tables := make(map[string]map[string]*gdb.TableField)
// V1
err := processSQL(parser, `
CREATE TABLE products (
id BIGSERIAL PRIMARY KEY,
name VARCHAR(100) NOT NULL,
price NUMERIC(10,2) DEFAULT 0.00
);
`, tables)
t.AssertNil(err)
// V2: add, alter, comment
err = processSQL(parser, `
ALTER TABLE products ADD COLUMN category VARCHAR(50);
ALTER TABLE products ALTER COLUMN name TYPE VARCHAR(200);
ALTER TABLE products ALTER COLUMN name SET NOT NULL;
COMMENT ON COLUMN products.category IS 'Product category';
`, tables)
t.AssertNil(err)
// V3: rename, drop
err = processSQL(parser, `
ALTER TABLE products RENAME COLUMN category TO product_category;
`, tables)
t.AssertNil(err)
fields := tables["products"]
t.Assert(len(fields), 4)
t.Assert(fields["name"].Type, "VARCHAR(200)")
t.Assert(fields["name"].Null, false)
_, ok := fields["category"]
t.Assert(ok, false)
t.Assert(fields["product_category"].Name, "product_category")
t.Assert(fields["product_category"].Comment, "Product category")
})
}
func Test_PgSQL_FullMigrationScenario(t *testing.T) {
gtest.C(t, func(t *gtest.T) {
parser := &PgSQLParser{}
tables := make(map[string]map[string]*gdb.TableField)
// V001: Initial
err := processSQL(parser, `
CREATE TABLE users (
id BIGSERIAL PRIMARY KEY,
name VARCHAR(100) NOT NULL,
email VARCHAR(200) UNIQUE
);
COMMENT ON COLUMN users.name IS 'User name';
`, tables)
t.AssertNil(err)
// V002: Add, alter type, set not null
err = processSQL(parser, `
ALTER TABLE users ADD COLUMN avatar TEXT;
ALTER TABLE users ALTER COLUMN name TYPE VARCHAR(200);
ALTER TABLE users ALTER COLUMN email SET NOT NULL;
COMMENT ON COLUMN users.avatar IS 'Avatar URL';
`, tables)
t.AssertNil(err)
fields := tables["users"]
t.Assert(len(fields), 4)
t.Assert(fields["name"].Type, "VARCHAR(200)")
t.Assert(fields["email"].Null, false)
t.Assert(fields["avatar"].Comment, "Avatar URL")
// V003: Rename column, drop not null
err = processSQL(parser, `
ALTER TABLE users RENAME COLUMN avatar TO profile_image;
ALTER TABLE users ALTER COLUMN email DROP NOT NULL;
`, tables)
t.AssertNil(err)
_, ok := fields["avatar"]
t.Assert(ok, false)
t.Assert(fields["profile_image"].Name, "profile_image")
t.Assert(fields["email"].Null, true)
})
}

View File

@ -0,0 +1,159 @@
// Copyright GoFrame gf Author(https://goframe.org). All Rights Reserved.
//
// This Source Code Form is subject to the terms of the MIT License.
// If a copy of the MIT was not distributed with this file,
// You can obtain one at https://github.com/gogf/gf.
package gendao
import (
"fmt"
"strings"
"github.com/gogf/gf/v2/database/gdb"
)
// SQLiteParser implements SQLParser for SQLite DDL.
type SQLiteParser struct{}
// ParseCreateTable parses a single SQLite CREATE TABLE statement.
func (p *SQLiteParser) ParseCreateTable(stmt string) (string, map[string]*gdb.TableField, error) {
body, _, ok := extractBodyAndTrailing(stmt)
if !ok {
return "", nil, nil
}
parenIdx := strings.Index(stmt, "(")
header := stmt[:parenIdx]
tableName := extractTableName(header)
if tableName == "" {
return "", nil, fmt.Errorf("cannot extract table name from: %s", header)
}
columnDefs := splitColumns(body)
fields := make(map[string]*gdb.TableField)
pkColumns := findPrimaryKeysFromConstraints(columnDefs)
fieldIndex := 0
for _, def := range columnDefs {
def = strings.TrimSpace(def)
if def == "" {
continue
}
firstWord := strings.ToUpper(strings.Fields(def)[0])
if isConstraintKeyword(firstWord) {
continue
}
field, err := p.parseColumnDef(def, fieldIndex)
if err != nil {
continue
}
if field != nil {
fields[field.Name] = field
fieldIndex++
}
}
for _, pkCol := range pkColumns {
if f, ok := fields[pkCol]; ok {
f.Key = "PRI"
}
}
return tableName, fields, nil
}
// ParseAlterTable parses SQLite ALTER TABLE statements.
// Note: SQLite only supports ADD COLUMN and RENAME COLUMN in ALTER TABLE.
func (p *SQLiteParser) ParseAlterTable(stmt string, tables map[string]map[string]*gdb.TableField) error {
return parseAlterTableCommon(stmt, tables, p.parseColumnDef)
}
// ParseComment is a no-op for SQLite as it doesn't support COMMENT ON statements.
func (p *SQLiteParser) ParseComment(stmt string, tables map[string]map[string]*gdb.TableField) {
// SQLite does not support comments on columns.
}
// parseColumnDef parses a single SQLite column definition string into a TableField.
// SQLite has flexible typing (type affinity), so columns may have no explicit type,
// in which case "text" is used as the default type.
func (p *SQLiteParser) parseColumnDef(def string, index int) (*gdb.TableField, error) {
tokens := mysqlTokenize(def)
if len(tokens) < 1 {
return nil, fmt.Errorf("invalid column definition: %s", def)
}
field := &gdb.TableField{
Index: index,
Name: unquoteIdentifier(tokens[0]),
Null: true,
}
if len(tokens) < 2 {
field.Type = "text"
return field, nil
}
field.Type = tokens[1]
rest := ""
if len(tokens) > 2 {
rest = strings.Join(tokens[2:], " ")
}
if !strings.Contains(field.Type, "(") && strings.HasPrefix(strings.TrimSpace(rest), "(") {
end := strings.Index(rest, ")")
if end >= 0 {
field.Type += rest[:end+1]
rest = strings.TrimSpace(rest[end+1:])
}
}
p.parseColumnAttributes(field, rest)
return field, nil
}
// parseColumnAttributes parses SQLite column constraint keywords including
// NOT NULL, NULL, PRIMARY KEY (with optional AUTOINCREMENT), UNIQUE, and DEFAULT.
func (p *SQLiteParser) parseColumnAttributes(field *gdb.TableField, attrs string) {
words := strings.Fields(attrs)
upperWords := strings.Fields(strings.ToUpper(attrs))
for i := 0; i < len(upperWords); i++ {
switch upperWords[i] {
case "NOT":
if i+1 < len(upperWords) && upperWords[i+1] == "NULL" {
field.Null = false
i++
}
case "NULL":
field.Null = true
case "PRIMARY":
if i+1 < len(upperWords) && upperWords[i+1] == "KEY" {
field.Key = "PRI"
field.Null = false
i++
if i+1 < len(upperWords) && upperWords[i+1] == "AUTOINCREMENT" {
field.Extra = "auto_increment"
i++
}
}
case "AUTOINCREMENT":
field.Extra = "auto_increment"
case "UNIQUE":
if field.Key == "" {
field.Key = "UNI"
}
case "DEFAULT":
if i+1 < len(words) {
defaultVal, _ := extractDefaultValue("DEFAULT " + strings.Join(words[i+1:], " "))
field.Default = defaultVal
if defaultVal != nil {
i++
}
}
}
}
}

View File

@ -0,0 +1,112 @@
// Copyright GoFrame gf Author(https://goframe.org). All Rights Reserved.
//
// This Source Code Form is subject to the terms of the MIT License.
// If a copy of the MIT was not distributed with this file,
// You can obtain one at https://github.com/gogf/gf.
package gendao
import (
"testing"
"github.com/gogf/gf/v2/database/gdb"
"github.com/gogf/gf/v2/test/gtest"
)
func Test_SQLite_CreateTable_Basic(t *testing.T) {
gtest.C(t, func(t *gtest.T) {
parser := &SQLiteParser{}
sql := `
CREATE TABLE users (
id INTEGER PRIMARY KEY AUTOINCREMENT,
name TEXT NOT NULL,
email TEXT,
age INTEGER DEFAULT 0,
score REAL DEFAULT 0.0,
is_active BOOLEAN NOT NULL DEFAULT 1
);
`
tables := make(map[string]map[string]*gdb.TableField)
err := processSQL(parser, sql, tables)
t.AssertNil(err)
fields := tables["users"]
t.Assert(len(fields), 6)
t.Assert(fields["id"].Key, "PRI")
t.Assert(fields["id"].Extra, "auto_increment")
t.Assert(fields["id"].Null, false)
t.Assert(fields["name"].Null, false)
t.Assert(fields["email"].Null, true)
t.Assert(fields["age"].Default, "0")
})
}
func Test_SQLite_AlterTable_AddColumn(t *testing.T) {
gtest.C(t, func(t *gtest.T) {
parser := &SQLiteParser{}
sql := `
CREATE TABLE users (
id INTEGER PRIMARY KEY AUTOINCREMENT,
name TEXT NOT NULL
);
ALTER TABLE users ADD COLUMN email TEXT;
ALTER TABLE users ADD COLUMN phone TEXT DEFAULT '';
`
tables := make(map[string]map[string]*gdb.TableField)
err := processSQL(parser, sql, tables)
t.AssertNil(err)
fields := tables["users"]
t.Assert(len(fields), 4)
t.Assert(fields["email"].Name, "email")
t.Assert(fields["phone"].Name, "phone")
})
}
func Test_SQLite_AlterTable_DropColumn(t *testing.T) {
gtest.C(t, func(t *gtest.T) {
parser := &SQLiteParser{}
sql := `
CREATE TABLE users (
id INTEGER PRIMARY KEY AUTOINCREMENT,
name TEXT NOT NULL,
old_col TEXT,
email TEXT
);
ALTER TABLE users DROP COLUMN old_col;
`
tables := make(map[string]map[string]*gdb.TableField)
err := processSQL(parser, sql, tables)
t.AssertNil(err)
fields := tables["users"]
t.Assert(len(fields), 3)
_, ok := fields["old_col"]
t.Assert(ok, false)
t.Assert(fields["name"].Name, "name")
t.Assert(fields["email"].Name, "email")
})
}
func Test_SQLite_RenameColumn(t *testing.T) {
gtest.C(t, func(t *gtest.T) {
parser := &SQLiteParser{}
sql := `
CREATE TABLE users (
id INTEGER PRIMARY KEY,
old_name TEXT NOT NULL
);
ALTER TABLE users RENAME COLUMN old_name TO new_name;
`
tables := make(map[string]map[string]*gdb.TableField)
err := processSQL(parser, sql, tables)
t.AssertNil(err)
fields := tables["users"]
_, ok := fields["old_name"]
t.Assert(ok, false)
t.Assert(fields["new_name"].Name, "new_name")
})
}

View File

@ -0,0 +1,302 @@
// Copyright GoFrame gf Author(https://goframe.org). All Rights Reserved.
//
// This Source Code Form is subject to the terms of the MIT License.
// If a copy of the MIT was not distributed with this file,
// You can obtain one at https://github.com/gogf/gf.
package gendao
import (
"testing"
"github.com/gogf/gf/v2/database/gdb"
"github.com/gogf/gf/v2/test/gtest"
)
// ===========================
// Common parser utilities tests
// ===========================
func Test_splitSQLStatements(t *testing.T) {
gtest.C(t, func(t *gtest.T) {
stmts := splitSQLStatements("CREATE TABLE t1 (id INT); ALTER TABLE t1 ADD COLUMN name VARCHAR(100);")
t.Assert(len(stmts), 2)
t.AssertIN("CREATE TABLE t1 (id INT)", stmts)
})
}
func Test_splitSQLStatements_WithComments(t *testing.T) {
gtest.C(t, func(t *gtest.T) {
sql := `
-- This is a comment
CREATE TABLE t1 (id INT);
/* Block comment */
ALTER TABLE t1 ADD COLUMN name VARCHAR(100);
`
stmts := splitSQLStatements(sql)
t.Assert(len(stmts), 2)
})
}
func Test_splitSQLStatements_WithQuotedSemicolon(t *testing.T) {
gtest.C(t, func(t *gtest.T) {
sql := `CREATE TABLE t1 (id INT, name VARCHAR(100) DEFAULT 'a;b');`
stmts := splitSQLStatements(sql)
t.Assert(len(stmts), 1)
})
}
func Test_classifyStatement(t *testing.T) {
gtest.C(t, func(t *gtest.T) {
t.Assert(classifyStatement("CREATE TABLE users (id INT)"), SQLStatementCreateTable)
t.Assert(classifyStatement("CREATE TEMPORARY TABLE tmp (id INT)"), SQLStatementCreateTable)
t.Assert(classifyStatement("ALTER TABLE users ADD COLUMN email VARCHAR(100)"), SQLStatementAlterTable)
t.Assert(classifyStatement("ALTER TABLE users RENAME TO customers"), SQLStatementRenameTable)
t.Assert(classifyStatement("DROP TABLE IF EXISTS users"), SQLStatementDropTable)
t.Assert(classifyStatement("RENAME TABLE old_name TO new_name"), SQLStatementRenameTable)
t.Assert(classifyStatement("COMMENT ON COLUMN users.name IS 'User name'"), SQLStatementComment)
t.Assert(classifyStatement("SELECT * FROM users"), SQLStatementUnknown)
t.Assert(classifyStatement("INSERT INTO users VALUES (1)"), SQLStatementUnknown)
})
}
func Test_unquoteIdentifier(t *testing.T) {
gtest.C(t, func(t *gtest.T) {
t.Assert(unquoteIdentifier("`users`"), "users")
t.Assert(unquoteIdentifier(`"users"`), "users")
t.Assert(unquoteIdentifier("[users]"), "users")
t.Assert(unquoteIdentifier("users"), "users")
})
}
func Test_extractTableName(t *testing.T) {
gtest.C(t, func(t *gtest.T) {
t.Assert(extractTableName("CREATE TABLE users"), "users")
t.Assert(extractTableName("CREATE TABLE IF NOT EXISTS users"), "users")
t.Assert(extractTableName("CREATE TABLE `users`"), "users")
t.Assert(extractTableName("CREATE TABLE mydb.users"), "users")
t.Assert(extractTableName("CREATE TEMPORARY TABLE temp_users"), "temp_users")
})
}
func Test_applyDropTable(t *testing.T) {
gtest.C(t, func(t *gtest.T) {
tables := map[string]map[string]*gdb.TableField{
"users": {},
"logs": {},
}
applyDropTable("DROP TABLE IF EXISTS users", tables)
t.Assert(len(tables), 1)
_, ok := tables["users"]
t.Assert(ok, false)
})
}
func Test_applyRenameTable_MySQL(t *testing.T) {
gtest.C(t, func(t *gtest.T) {
tables := map[string]map[string]*gdb.TableField{
"old_name": {"id": {Index: 0, Name: "id", Type: "int"}},
}
applyRenameTable("RENAME TABLE old_name TO new_name", tables)
t.Assert(len(tables), 1)
_, ok := tables["new_name"]
t.Assert(ok, true)
})
}
func Test_applyRenameTable_PgSQL(t *testing.T) {
gtest.C(t, func(t *gtest.T) {
tables := map[string]map[string]*gdb.TableField{
"old_name": {"id": {Index: 0, Name: "id", Type: "int"}},
}
applyRenameTable("ALTER TABLE old_name RENAME TO new_name", tables)
t.Assert(len(tables), 1)
_, ok := tables["new_name"]
t.Assert(ok, true)
})
}
// ===========================
// Abnormal/edge-case parsing tests
// ===========================
func Test_processSQL_OnlyDMLStatements(t *testing.T) {
gtest.C(t, func(t *gtest.T) {
parser := &MySQLParser{}
sql := `
INSERT INTO users (id, name) VALUES (1, 'Alice');
INSERT INTO users (id, name) VALUES (2, 'Bob');
DELETE FROM users WHERE id = 1;
UPDATE users SET name = 'Charlie' WHERE id = 2;
SELECT * FROM users;
`
tables := make(map[string]map[string]*gdb.TableField)
err := processSQL(parser, sql, tables)
t.AssertNil(err)
t.Assert(len(tables), 0)
})
}
func Test_processSQL_EmptySQL(t *testing.T) {
gtest.C(t, func(t *gtest.T) {
parser := &MySQLParser{}
tables := make(map[string]map[string]*gdb.TableField)
// Empty string
err := processSQL(parser, "", tables)
t.AssertNil(err)
t.Assert(len(tables), 0)
// Only whitespace and newlines
err = processSQL(parser, " \n\n \t ", tables)
t.AssertNil(err)
t.Assert(len(tables), 0)
})
}
func Test_processSQL_OnlyComments(t *testing.T) {
gtest.C(t, func(t *gtest.T) {
parser := &MySQLParser{}
sql := `
-- This is a line comment
/* This is a block comment */
-- Another comment
`
tables := make(map[string]map[string]*gdb.TableField)
err := processSQL(parser, sql, tables)
t.AssertNil(err)
t.Assert(len(tables), 0)
})
}
func Test_processSQL_AlterNonExistentTable(t *testing.T) {
gtest.C(t, func(t *gtest.T) {
parser := &MySQLParser{}
sql := `
ALTER TABLE non_existent ADD COLUMN email VARCHAR(200);
ALTER TABLE non_existent DROP COLUMN name;
ALTER TABLE non_existent MODIFY COLUMN name VARCHAR(200);
`
tables := make(map[string]map[string]*gdb.TableField)
err := processSQL(parser, sql, tables)
t.AssertNil(err)
t.Assert(len(tables), 0)
})
}
func Test_processSQL_DropNonExistentTable(t *testing.T) {
gtest.C(t, func(t *gtest.T) {
parser := &MySQLParser{}
sql := `DROP TABLE IF EXISTS non_existent;`
tables := make(map[string]map[string]*gdb.TableField)
err := processSQL(parser, sql, tables)
t.AssertNil(err)
t.Assert(len(tables), 0)
})
}
func Test_processSQL_MixedDDLAndDML(t *testing.T) {
gtest.C(t, func(t *gtest.T) {
parser := &MySQLParser{}
sql := `
INSERT INTO logs (msg) VALUES ('starting migration');
CREATE TABLE users (
id INT NOT NULL AUTO_INCREMENT,
name VARCHAR(100) NOT NULL,
PRIMARY KEY (id)
);
INSERT INTO users (name) VALUES ('Alice');
ALTER TABLE users ADD COLUMN email VARCHAR(200);
UPDATE users SET email = 'alice@example.com' WHERE id = 1;
DELETE FROM logs WHERE msg = 'starting migration';
`
tables := make(map[string]map[string]*gdb.TableField)
err := processSQL(parser, sql, tables)
t.AssertNil(err)
// Only DDL statements should be processed; DML should be skipped.
t.Assert(len(tables), 1)
fields := tables["users"]
t.Assert(len(fields), 3)
t.Assert(fields["id"].Key, "PRI")
t.Assert(fields["email"].Name, "email")
})
}
func Test_processSQL_CommentOnNonExistentTable(t *testing.T) {
gtest.C(t, func(t *gtest.T) {
parser := &PgSQLParser{}
sql := `COMMENT ON COLUMN non_existent.col1 IS 'some comment';`
tables := make(map[string]map[string]*gdb.TableField)
err := processSQL(parser, sql, tables)
t.AssertNil(err)
t.Assert(len(tables), 0)
})
}
func Test_processSQL_RenameNonExistentTable(t *testing.T) {
gtest.C(t, func(t *gtest.T) {
parser := &MySQLParser{}
sql := `RENAME TABLE non_existent TO new_name;`
tables := make(map[string]map[string]*gdb.TableField)
err := processSQL(parser, sql, tables)
t.AssertNil(err)
t.Assert(len(tables), 0)
})
}
func Test_processSQL_DropColumnFromNonExistentTable(t *testing.T) {
gtest.C(t, func(t *gtest.T) {
parser := &MySQLParser{}
sql := `
CREATE TABLE users (id INT, name VARCHAR(100), PRIMARY KEY (id));
ALTER TABLE orders DROP COLUMN status;
`
tables := make(map[string]map[string]*gdb.TableField)
err := processSQL(parser, sql, tables)
t.AssertNil(err)
// users table should still exist, orders ALTER should be silently ignored.
t.Assert(len(tables), 1)
t.Assert(len(tables["users"]), 2)
})
}
// ===========================
// CheckLocalTypeForFieldType Tests
// ===========================
func Test_CheckLocalTypeForFieldType(t *testing.T) {
gtest.C(t, func(t *gtest.T) {
tests := []struct {
fieldType string
expected string
}{
{"int(10)", "int"},
{"int(10) unsigned", "uint"},
{"bigint(20)", "int64"},
{"bigint(20) unsigned", "uint64"},
{"tinyint(1)", "int"},
{"varchar(100)", "string"},
{"text", "string"},
{"datetime", "datetime"},
{"timestamp", "datetime"},
{"timestamptz", "datetime"},
{"date", "date"},
{"time", "time"},
{"json", "json"},
{"jsonb", "jsonb"},
{"float", "float64"},
{"double", "float64"},
{"decimal(10,2)", "string"},
{"bool", "bool"},
{"boolean", "bool"},
{"blob", "[]byte"},
{"binary(16)", "[]byte"},
{"bit(1)", "bool"},
}
for _, tt := range tests {
localType, err := gdb.CheckLocalTypeForFieldType(tt.fieldType)
t.AssertNil(err)
t.Assert(string(localType), tt.expected)
}
})
}

View File

@ -20,14 +20,20 @@ import (
"github.com/gogf/gf/v2/text/gstr"
)
// generateStructDefinitionInput holds parameters for generating a Go struct definition
// from database table fields.
type generateStructDefinitionInput struct {
CGenDaoInternalInput
TableName string // Table name.
StructName string // Struct name.
FieldMap map[string]*gdb.TableField // Table field map.
IsDo bool // Is generating DTO struct.
TableName string // Original database table name.
StructName string // Go struct name (CamelCase of table name).
FieldMap map[string]*gdb.TableField // Map of column name to field metadata.
IsDo bool // Whether generating a DO struct (uses g.Meta orm tag).
}
// generateStructDefinition generates a complete Go struct definition string from table fields.
// It returns the struct source code and a list of additional import paths needed
// by custom type mappings. The fields are rendered in a table-aligned format
// using tablewriter for consistent code formatting.
func generateStructDefinition(ctx context.Context, in generateStructDefinitionInput) (string, []string) {
var appendImports []string
buffer := bytes.NewBuffer(nil)
@ -59,6 +65,10 @@ func generateStructDefinition(ctx context.Context, in generateStructDefinitionIn
return buffer.String(), appendImports
}
// getTypeMappingInfo looks up a database field type in the type mapping configuration.
// It handles exact matches first, then tries to extract the base type name from
// parameterized types like "varchar(255)" or "numeric(10,2) unsigned".
// Returns the mapped Go type name and its import path (if any).
func getTypeMappingInfo(
ctx context.Context, fieldType string, inTypeMapping map[DBFieldTypeName]CustomAttributeType,
) (typeNameStr, importStr string) {
@ -105,9 +115,17 @@ func generateStructFieldDefinition(
}
if localTypeNameStr == "" {
localTypeName, err = in.DB.CheckLocalTypeForField(ctx, field.Type, nil)
if err != nil {
panic(err)
if in.DB != nil {
localTypeName, err = in.DB.CheckLocalTypeForField(ctx, field.Type, nil)
if err != nil {
panic(err)
}
} else {
// SQL file mode: use standalone type checking without database connection.
localTypeName, err = gdb.CheckLocalTypeForFieldType(field.Type)
if err != nil {
panic(err)
}
}
localTypeNameStr = string(localTypeName)
switch localTypeName {
@ -181,11 +199,12 @@ func generateStructFieldDefinition(
return attrLines, appendImport
}
// FieldNameCase defines the naming convention for converting field names to Go identifiers.
type FieldNameCase string
const (
FieldNameCaseCamel FieldNameCase = "CaseCamel"
FieldNameCaseCamelLower FieldNameCase = "CaseCamelLower"
FieldNameCaseCamel FieldNameCase = "CaseCamel" // PascalCase: "user_name" -> "UserName"
FieldNameCaseCamelLower FieldNameCase = "CaseCamelLower" // camelCase: "user_name" -> "userName"
)
// formatFieldName formats and returns a new field name that is used for golang codes generating.

View File

@ -62,7 +62,7 @@ type generateTableSingleInput struct {
// generateTableSingle generates dao files for a single table.
func generateTableSingle(ctx context.Context, in generateTableSingleInput) {
// Generating table data preparing.
fieldMap, err := in.DB.TableFields(ctx, in.TableName)
fieldMap, err := getTableFields(ctx, in.CGenDaoInternalInput, in.TableName)
if err != nil {
mlog.Fatalf(`fetching tables fields failed for table "%s": %+v`, in.TableName, err)
}

View File

@ -74,6 +74,8 @@ CONFIGURATION SUPPORT
CGenDaoBriefTypeMapping = `custom local type mapping for generated struct attributes relevant to fields of table`
CGenDaoBriefFieldMapping = `custom local type mapping for generated struct attributes relevant to specific fields of table`
CGenDaoBriefShardingPattern = `sharding pattern for table name, e.g. "users_?" will be replace tables "users_001,users_002,..." to "users" dao`
CGenDaoBriefSqlDir = `directory path of SQL DDL files for generating dao/do/entity without database connection`
CGenDaoBriefSqlType = `SQL dialect type when using sqlDir, options: mysql|pgsql|mssql|oracle|sqlite, default is "mysql"`
CGenDaoBriefGroup = `
specifying the configuration group name of database for generated ORM instance,
it's not necessary and the default value is "default"
@ -95,21 +97,23 @@ generated json tag case for model struct, cases are as follows:
CGenDaoBriefTplDaoDoPathPath = `template file path for dao do file`
CGenDaoBriefTplDaoEntityPath = `template file path for dao entity file`
tplVarTableName = `TplTableName`
tplVarTableNameCamelCase = `TplTableNameCamelCase`
tplVarTableNameCamelLowerCase = `TplTableNameCamelLowerCase`
tplVarTableSharding = `TplTableSharding`
tplVarTableShardingPrefix = `TplTableShardingPrefix`
tplVarTableFields = `TplTableFields`
tplVarPackageImports = `TplPackageImports`
tplVarImportPrefix = `TplImportPrefix`
tplVarStructDefine = `TplStructDefine`
tplVarColumnDefine = `TplColumnDefine`
tplVarColumnNames = `TplColumnNames`
tplVarGroupName = `TplGroupName`
tplVarDatetimeStr = `TplDatetimeStr`
tplVarCreatedAtDatetimeStr = `TplCreatedAtDatetimeStr`
tplVarPackageName = `TplPackageName`
// Template variable names used by gview for rendering Go file templates.
// These are passed to tplView.Assigns() and referenced in template files.
tplVarTableName = `TplTableName` // Original database table name.
tplVarTableNameCamelCase = `TplTableNameCamelCase` // PascalCase table name (e.g., "UserDetail").
tplVarTableNameCamelLowerCase = `TplTableNameCamelLowerCase` // camelCase table name (e.g., "userDetail").
tplVarTableSharding = `TplTableSharding` // Boolean: whether this is a sharding table.
tplVarTableShardingPrefix = `TplTableShardingPrefix` // Sharding table name prefix (e.g., "user_").
tplVarTableFields = `TplTableFields` // Generated table field definitions.
tplVarPackageImports = `TplPackageImports` // Generated import block string.
tplVarImportPrefix = `TplImportPrefix` // Go import path prefix for internal dao package.
tplVarStructDefine = `TplStructDefine` // Generated struct definition string.
tplVarColumnDefine = `TplColumnDefine` // Column struct field definitions for dao internal.
tplVarColumnNames = `TplColumnNames` // Column name-to-string assignments for dao internal.
tplVarGroupName = `TplGroupName` // Database configuration group name.
tplVarDatetimeStr = `TplDatetimeStr` // Current datetime string for file headers.
tplVarCreatedAtDatetimeStr = `TplCreatedAtDatetimeStr` // "Created at <datetime>" string (empty if WithTime is false).
tplVarPackageName = `TplPackageName` // Go package name for the generated file.
)
func init() {
@ -145,6 +149,8 @@ func init() {
`CGenDaoBriefTypeMapping`: CGenDaoBriefTypeMapping,
`CGenDaoBriefFieldMapping`: CGenDaoBriefFieldMapping,
`CGenDaoBriefShardingPattern`: CGenDaoBriefShardingPattern,
`CGenDaoBriefSqlDir`: CGenDaoBriefSqlDir,
`CGenDaoBriefSqlType`: CGenDaoBriefSqlType,
`CGenDaoBriefGroup`: CGenDaoBriefGroup,
`CGenDaoBriefJsonCase`: CGenDaoBriefJsonCase,
`CGenDaoBriefTplDaoIndexPath`: CGenDaoBriefTplDaoIndexPath,

View File

@ -864,6 +864,24 @@ const (
fieldTypeTimestampz = "timestamptz"
fieldTypeJson = "json"
fieldTypeJsonb = "jsonb"
// PostgreSQL specific types.
fieldTypeInt2 = "int2"
fieldTypeInt4 = "int4"
fieldTypeInteger = "integer"
fieldTypeInt8 = "int8"
fieldTypeFloat4 = "float4"
fieldTypeFloat8 = "float8"
fieldTypeDoublePrecision = "double precision"
fieldTypeBoolean = "boolean"
// Oracle specific types.
fieldTypeNumber = "number"
// MSSQL specific types.
fieldTypeDatetime2 = "datetime2"
fieldTypeDatetimeOffset = "datetimeoffset"
fieldTypeSmalldatetime = "smalldatetime"
)
var (

View File

@ -226,6 +226,13 @@ Default:
// GetFormattedDBTypeNameForField retrieves and returns the formatted database type name
// eg. `int(10) unsigned` -> `int`, `varchar(100)` -> `varchar`, etc.
func (c *Core) GetFormattedDBTypeNameForField(fieldType string) (typeName, typePattern string) {
return FormatDBTypeName(fieldType)
}
// FormatDBTypeName retrieves and returns the formatted database type name and pattern
// from raw field type string without requiring a database connection.
// eg. `int(10) unsigned` -> (`int`, `10`), `varchar(100)` -> (`varchar`, `100`).
func FormatDBTypeName(fieldType string) (typeName, typePattern string) {
match, _ := gregex.MatchString(`(.+?)\((.+)\)`, fieldType)
if len(match) == 3 {
typeName = gstr.Trim(match[1])
@ -246,11 +253,17 @@ func (c *Core) GetFormattedDBTypeNameForField(fieldType string) (typeName, typeP
// The `fieldType` is retrieved from ColumnTypes of db driver, example:
// UNSIGNED INT
func (c *Core) CheckLocalTypeForField(ctx context.Context, fieldType string, _ any) (LocalType, error) {
return CheckLocalTypeForFieldType(fieldType)
}
// CheckLocalTypeForFieldType checks and returns corresponding local type for given db field type string
// without requiring a database connection.
func CheckLocalTypeForFieldType(fieldType string) (LocalType, error) {
var (
typeName string
typePattern string
)
typeName, typePattern = c.GetFormattedDBTypeNameForField(fieldType)
typeName, typePattern = FormatDBTypeName(fieldType)
switch typeName {
case
fieldTypeBinary,
@ -268,7 +281,10 @@ func (c *Core) CheckLocalTypeForField(ctx context.Context, fieldType string, _ a
fieldTypeSmallint,
fieldTypeMediumInt,
fieldTypeMediumint,
fieldTypeSerial:
fieldTypeSerial,
fieldTypeInt2,
fieldTypeInt4,
fieldTypeInteger:
if gstr.ContainsI(fieldType, "unsigned") {
return LocalTypeUint, nil
}
@ -277,7 +293,8 @@ func (c *Core) CheckLocalTypeForField(ctx context.Context, fieldType string, _ a
case
fieldTypeBigInt,
fieldTypeBigint,
fieldTypeBigserial:
fieldTypeBigserial,
fieldTypeInt8:
if gstr.ContainsI(fieldType, "unsigned") {
return LocalTypeUint64, nil
}
@ -298,11 +315,15 @@ func (c *Core) CheckLocalTypeForField(ctx context.Context, fieldType string, _ a
fieldTypeDecimal,
fieldTypeMoney,
fieldTypeNumeric,
fieldTypeSmallmoney:
fieldTypeSmallmoney,
fieldTypeNumber:
return LocalTypeString, nil
case
fieldTypeFloat,
fieldTypeDouble:
fieldTypeDouble,
fieldTypeFloat4,
fieldTypeFloat8,
fieldTypeDoublePrecision:
return LocalTypeFloat64, nil
case
@ -317,7 +338,8 @@ func (c *Core) CheckLocalTypeForField(ctx context.Context, fieldType string, _ a
return LocalTypeInt64Bytes, nil
case
fieldTypeBool:
fieldTypeBool,
fieldTypeBoolean:
return LocalTypeBool, nil
case
@ -331,7 +353,10 @@ func (c *Core) CheckLocalTypeForField(ctx context.Context, fieldType string, _ a
case
fieldTypeDatetime,
fieldTypeTimestamp,
fieldTypeTimestampz:
fieldTypeTimestampz,
fieldTypeDatetime2,
fieldTypeDatetimeOffset,
fieldTypeSmalldatetime:
return LocalTypeDatetime, nil
case
@ -345,7 +370,10 @@ func (c *Core) CheckLocalTypeForField(ctx context.Context, fieldType string, _ a
default:
// Auto-detect field type, using key match.
switch {
case strings.Contains(typeName, "text") || strings.Contains(typeName, "char") || strings.Contains(typeName, "character"):
case strings.Contains(typeName, "text") || strings.Contains(typeName, "char") ||
strings.Contains(typeName, "character") || strings.Contains(typeName, "clob") ||
strings.Contains(typeName, "ntext") || strings.Contains(typeName, "xml") ||
strings.Contains(typeName, "string"):
return LocalTypeString, nil
case strings.Contains(typeName, "float") || strings.Contains(typeName, "double") || strings.Contains(typeName, "numeric"):
@ -354,7 +382,9 @@ func (c *Core) CheckLocalTypeForField(ctx context.Context, fieldType string, _ a
case strings.Contains(typeName, "bool"):
return LocalTypeBool, nil
case strings.Contains(typeName, "binary") || strings.Contains(typeName, "blob"):
case strings.Contains(typeName, "binary") || strings.Contains(typeName, "blob") ||
strings.Contains(typeName, "bytea") || strings.Contains(typeName, "image") ||
strings.Contains(typeName, "raw"):
return LocalTypeBytes, nil
case strings.Contains(typeName, "int"):