can't scan 1 (float32) into float32

[nasa1024]

what should i fix this error?

• model

type ScriptCategory struct {
	bun.BaseModel `bun:"script_category,alias:sc" json:"-"` 

	Id         int64   `bun:"id,pk" json:"id"`  
	Name       string  `bun:"name" json:"name"` 
	Sort       float32 `bun:"sort" json:"sort"`
	BaseFields `json:"-"`
}

• sql

CREATE TABLE `script_category` (
  `id` bigint(20) NOT NULL,
  `name` varchar(255) NOT NULL DEFAULT '',
  `sort` float NOT NULL DEFAULT '0' ,
  `created_at` int(11) unsigned NOT NULL DEFAULT '0' ,
  `updated_at` int(11) unsigned NOT NULL DEFAULT '0' ,
  PRIMARY KEY (`id`)
) ENGINE=InnoDB DEFAULT CHARSET=utf8mb4 COLLATE=utf8mb4_0900_ai_ci ;

• function

func QueryCategory(ctx context.Context, db *bun.DB) (list []model.ScriptCategory, err error) {
	err = db.NewSelect().Model(&model.ScriptCategory{}).
		ColumnExpr("sc.id, sc.name, sc.sort").
		Order("sort asc").
		Scan(ctx, &list)
	if err != nil {
		err = fmt.Errorf("NewSelect ScriptCategory: %v", err)
		return
	}
	return
}

• error log

[bun]  10:34:57.703   SELECT               19.081ms  SELECT sc.id, sc.name, sc.sort FROM `script_category` AS `sc` ORDER BY `sort` asc 	  *fmt.wrapError: sql: Scan error on column index 2, name "sort": bun: can't scan 1 (float32) into float32 

[CL-Jeremy]

Second this. In addition to this, for me at least, it doesn't work either if I try to scan float32 into float64. I had to effectively leave the field out completely. I'm using Microsoft SQL Server for that matter (I'll have to check which exact version with our IT department). I'd say this much that this is not limited to MySQL.

[CL-Jeremy]

Found out that the root problem in my case actually lies in https://github.com/denisenkom/go-mssqldb.

According to the spec, the scanner function in database/sql should always return the value in double precision, regardless of the precision of the original field. It might be somewhat counter-intuitive (since the logical value of the bits representation does get interpreted differently), but type casts from float32 to float64 then back to float32 retains the precision, so there's nothing wrong with the go spec.

In line 355, this conversion is missing, whereas in line 396 and other cases, the functions comply to the go spec. Adding the needed type cast in line 355 solves my issue.

I suppose that it probably makes more sense to look at https://github.com/go-sql-driver/mysql for your issue.

[github-actions]

This issue has been automatically marked as stale because it has not had activity in the last 30 days. If there is no update within the next 7 days, this issue will be closed.