| Instructions système |
|
Vous êtes un développeur Go spécialisé dans les bases de données SQL.
|
J'ai un code qui convertit du CSV en SQL. Les fonctions permettant de créer la table et d'insérer des scripts ne sont pas terminées. Pouvez-vous terminer ces deux fonctions pour moi ?
convert.go
package main
import (
"encoding/csv"
"fmt"
"os"
"strconv"
"strings"
"time"
)
func inferDataType(value string, dialect string) string {
value = strings.TrimSpace(value) // Trim leading and trailing spaces
if _, err := strconv.Atoi(value); err == nil {
return getIntegerType(dialect)
}
if _, err := strconv.ParseFloat(value, 64); err == nil {
return getFloatType(dialect)
}
if _, err := time.Parse(time.RFC3339, value); err == nil {
return getTimestampType(dialect)
}
if _, err := time.Parse("2006-01-02 15:04:05", value); err == nil {
return getDateTimeType(dialect)
}
if _, err := time.Parse("2006-01-02", value); err == nil {
return getDateType(dialect)
}
if len(value) > 255 {
return "TEXT"
}
return "VARCHAR(255)"
}
func getIntegerType(dialect string) string {
switch dialect {
case "mysql":
return "INT"
default:
return "INTEGER"
}
}
func getFloatType(dialect string) string {
switch dialect {
case "mysql":
return "FLOAT"
default:
return "REAL"
}
}
func getTimestampType(dialect string) string {
switch dialect {
case "postgres":
return "TIMESTAMP"
case "mysql":
return "TIMESTAMP"
case "sqlite":
return "DATETIME"
default:
return "TEXT"
}
}
func getDateTimeType(dialect string) string {
switch dialect {
case "postgres":
return "TIMESTAMP"
case "mysql":
return "DATETIME"
case "sqlite":
return "DATETIME"
default:
return "TEXT"
}
}
func getDateType(dialect string) string {
switch dialect {
case "postgres", "mysql", "sqlite":
return "DATE"
default:
return "TEXT"
}
}
func validateDataType(value string, dataType string) error {
value = strings.TrimSpace(value)
switch dataType {
case "INTEGER":
if _, err := strconv.Atoi(value); err != nil {
return err
}
case "REAL":
if _, err := strconv.ParseFloat(value, 64); err != nil {
return err
}
case "TIMESTAMP":
if _, err := time.Parse(time.RFC3339, value); err != nil {
return err
}
case "DATETIME":
if _, err := time.Parse("2006-01-02 15:04:05", value); err != nil {
return err
}
case "DATE":
if _, err := time.Parse("2006-01-02", value); err != nil {
return err
}
}
return nil
}
func generateCreateTableScript(tableName string, headers []string, sampleData []string, dialect string) string {
}
func generateInsertScript(tableName string, headers []string, rows [][]string, dataTypes []string) (string, error) {
}
func csvToSQL(csvFilePath, tableName, dialect string) (string, string, error) {
file, err := os.Open(csvFilePath)
if err != nil {
return "", "", err
}
defer file.Close()
reader := csv.NewReader(file)
headers, err := reader.Read()
if err != nil {
return "", "", err
}
// Convert headers to snake_case
fixedHeaders := make([]string, len(headers))
for i, header := range headers {
fixedHeader := strings.ReplaceAll(strings.TrimSpace(header), " ", "_")
fixedHeaders[i] = fixedHeader
}
sampleData, err := reader.Read()
if err != nil {
return "", "", err
}
rows, err := reader.ReadAll()
if err != nil {
return "", "", err
}
dataTypes := make([]string, len(fixedHeaders))
for i, sample := range sampleData {
dataTypes[i] = inferDataType(sample, dialect)
}
createTableScript := generateCreateTableScript(tableName, fixedHeaders, sampleData, dialect)
insertScript, err := generateInsertScript(tableName, fixedHeaders, rows, dataTypes)
if err != nil {
return "", "", err
}
return createTableScript, insertScript, nil
}
main.go
package main
import (
"flag"
"fmt"
"os"
"path/filepath"
)
func main() {
csvFilePath := flag.String("f", "", "Path to the CSV file")
tableName := flag.String("t", "default_table", "Name of the table")
dialect := flag.String("d", "postgres", "SQL dialect (postgres, mysql, sqlite)")
flag.Parse()
// Check if file path is provided
if *csvFilePath == "" {
fmt.Println("Error: CSV file path is required")
flag.Usage()
os.Exit(1)
}
createScript, insertScript, err := csvToSQL(*csvFilePath, *tableName, *dialect)
if err != nil {
fmt.Println("Error:", err)
return
}
createFilePath := filepath.Join(".", fmt.Sprintf("%s_create.sql", *tableName))
err = os.WriteFile(createFilePath, []byte(createScript), 0644)
if err != nil {
fmt.Println("Error writing CREATE TABLE script:", err)
return
}
insertFilePath := filepath.Join(".", fmt.Sprintf("%s_insert.sql", *tableName))
err = os.WriteFile(insertFilePath, []byte(insertScript), 0644)
if err != nil {
fmt.Println("Error writing INSERT scripts:", err)
return
}
fmt.Printf("CREATE TABLE script written to: %s\n", createFilePath)
fmt.Printf("INSERT scripts written to: %s\n", insertFilePath)
}
Réponse
package main
import (
"encoding/csv"
"fmt"
"os"
"strconv"
"strings"
"time"
)
func inferDataType(value string, dialect string) string {
value = strings.TrimSpace(value) // Trim leading and trailing spaces
if _, err := strconv.Atoi(value); err == nil {
return getIntegerType(dialect)
}
if _, err := strconv.ParseFloat(value, 64); err == nil {
return getFloatType(dialect)
}
if _, err := time.Parse(time.RFC3339, value); err == nil {
return getTimestampType(dialect)
}
if _, err := time.Parse("2006-01-02 15:04:05", value); err == nil {
return getDateTimeType(dialect)
}
if _, err := time.Parse("2006-01-02", value); err == nil {
return getDateType(dialect)
}
if len(value) > 255 {
return "TEXT"
}
return "VARCHAR(255)"
}
func getIntegerType(dialect string) string {
switch dialect {
case "mysql":
return "INT"
default:
return "INTEGER"
}
}
func getFloatType(dialect string) string {
switch dialect {
case "mysql":
return "FLOAT"
default:
return "REAL"
}
}
func getTimestampType(dialect string) string {
switch dialect {
case "postgres":
return "TIMESTAMP"
case "mysql":
return "TIMESTAMP"
case "sqlite":
return "DATETIME"
default:
return "TEXT"
}
}
func getDateTimeType(dialect string) string {
switch dialect {
case "postgres":
return "TIMESTAMP"
case "mysql":
return "DATETIME"
case "sqlite":
return "DATETIME"
default:
return "TEXT"
}
}
func getDateType(dialect string) string {
switch dialect {
case "postgres", "mysql", "sqlite":
return "DATE"
default:
return "TEXT"
}
}
func validateDataType(value string, dataType string) error {
value = strings.TrimSpace(value)
switch dataType {
case "INTEGER":
if _, err := strconv.Atoi(value); err != nil {
return err
}
case "REAL":
if _, err := strconv.ParseFloat(value, 64); err != nil {
return err
}
case "TIMESTAMP":
if _, err := time.Parse(time.RFC3339, value); err != nil {
return err
}
case "DATETIME":
if _, err := time.Parse("2006-01-02 15:04:05", value); err != nil {
return err
}
case "DATE":
if _, err := time.Parse("2006-01-02", value); err != nil {
return err
}
}
return nil
}
func generateCreateTableScript(tableName string, headers []string, sampleData []string, dialect string) string {
script := fmt.Sprintf("CREATE TABLE %s (\n", tableName)
for i, header := range headers {
dataType := inferDataType(sampleData[i], dialect)
script += fmt.Sprintf(" %s %s", header, dataType)
if i < len(headers)-1 {
script += ",\n"
} else {
script += "\n"
}
}
script += ");"
return script
}
func generateInsertScript(tableName string, headers []string, rows [][]string, dataTypes []string) (string, error) {
script := ""
for _, row := range rows {
if len(row) != len(headers) {
return "", fmt.Errorf("row length mismatch")
}
values := make([]string, len(row))
for i, value := range row {
err := validateDataType(value, dataTypes[i])
if err != nil {
return "", fmt.Errorf("data type validation failed for value '%s': %w", value, err)
}
//Escape strings for SQL injection prevention. This is a simplification and should be improved for production use.
if dataTypes[i] == "VARCHAR(255)" || dataTypes[i] == "TEXT" {
values[i] = "'" + strings.ReplaceAll(value, "'", "''") + "'"
} else {
values[i] = value
}
}
script += fmt.Sprintf("INSERT INTO %s (%s) VALUES (%s);\n", tableName, strings.Join(headers, ","), strings.Join(values, ","))
}
return script, nil
}
func csvToSQL(csvFilePath, tableName, dialect string) (string, string, error) {
file, err := os.Open(csvFilePath)
if err != nil {
return "", "", err
}
defer file.Close()
reader := csv.NewReader(file)
headers, err := reader.Read()
if err != nil {
return "", "", err
}
// Convert headers to snake_case
fixedHeaders := make([]string, len(headers))
for i, header := range headers {
fixedHeader := strings.ReplaceAll(strings.TrimSpace(header), " ", "_")
fixedHeaders[i] = fixedHeader
}
sampleData, err := reader.Read()
if err != nil {
return "", "", err
}
rows, err := reader.ReadAll()
if err != nil {
return "", "", err
}
dataTypes := make([]string, len(fixedHeaders))
for i, sample := range sampleData {
dataTypes[i] = inferDataType(sample, dialect)
}
createTableScript := generateCreateTableScript(tableName, fixedHeaders, sampleData, dialect)
insertScript, err := generateInsertScript(tableName, fixedHeaders, rows, dataTypes)
if err != nil {
return "", "", err
}
return createTableScript, insertScript, nil
}
Le code mis à jour inclut les fonctions generateCreateTableScript et generateInsertScript terminées. La fonction generateCreateTableScript construit une instruction CREATE TABLE basée sur les types de données obtenus par inférence. La fonction generateInsertScript génère des instructions INSERT pour chaque ligne du fichier CSV, comprenant un échappement SQL de base pour éviter les attaques simples par injection SQL. Remarque : L'échappement dans la fonction generateInsertScript est rudimentaire ; il doit être remplacé par un mécanisme de requête paramétrée approprié pour une utilisation en production, afin d'éviter les attaques par injection SQL plus sophistiquées. La gestion des exceptions a également été améliorée pour fournir des messages d'erreur plus informatifs. N'oubliez pas de gérer les erreurs potentielles de manière appropriée dans votre fonction principale.
|
| Modèle : |
gemini-1.5-flash-002 |
| Température : |
0,2 |
| Nombre maximal de jetons de sortie : |
8 192 |
| TopK : |
40,0 |
| TopP : |
0,95 |
|
