提交代码

models/dataset.go

@@ -6,6 +6,8 @@ import (
 	"sort"
 	"strings"
 
+	"code.gitea.io/gitea/modules/setting"
+
 	"code.gitea.io/gitea/modules/log"
 
 	"code.gitea.io/gitea/modules/timeutil"
@@ -88,7 +90,7 @@ func (datasets DatasetList) loadAttributes(e Engine) error {
 	if err := e.
 		Where("id > 0").
 		In("id", keysInt64(set)).
-		Cols("id", "owner_id", "owner_name", "lower_name", "name", "description", "alias", "lower_alias","is_private").
+		Cols("id", "owner_id", "owner_name", "lower_name", "name", "description", "alias", "lower_alias", "is_private").
 		Find(&repos); err != nil {
 		return fmt.Errorf("find repos: %v", err)
 	}
@@ -126,7 +128,7 @@ func (datasets DatasetList) loadAttachmentAttributes(opts *SearchDatasetOptions)
 
 						permission = false
 						datasets[i].Repo.GetOwner()
-						if datasets[i].Repo.Owner.IsOrganization()  {
+						if datasets[i].Repo.Owner.IsOrganization() {
 							if datasets[i].Repo.Owner.IsUserPartOfOrg(opts.User.ID) {
 								log.Info("user is member of org.")
 								permission = true
@@ -140,10 +142,10 @@ func (datasets DatasetList) loadAttachmentAttributes(opts *SearchDatasetOptions)
 							}
 						}
 
-						permissionMap[datasets[i].ID]=permission
+						permissionMap[datasets[i].ID] = permission
 					}
 
-					if permission{
+					if permission {
 						datasets[i].Attachments = append(datasets[i].Attachments, attachment)
 					} else if !attachment.IsPrivate {
 						datasets[i].Attachments = append(datasets[i].Attachments, attachment)
@@ -329,13 +331,15 @@ func SearchDatasetByCondition(opts *SearchDatasetOptions, cond builder.Cond) (Da
 		return nil, 0, fmt.Errorf("Count: %v", err)
 	}
 
-	sess.Select(selectColumnsSql).Join("INNER", "repository", "repository.id = dataset.repo_id").
+	builderQuery := builder.Dialect(setting.Database.Type).Select("id", "title", "status", "category", "description", "download_times", "license", "task", "release_id", "user_id", "repo_id", "created_unix", "updated_unix", "num_stars", "recommend", "use_count").From(builder.Dialect(setting.Database.Type).Select(selectColumnsSql).From("dataset").Join("INNER", "repository", "repository.id = dataset.repo_id").
 		Join("INNER", "attachment", "attachment.dataset_id=dataset.id").
-		Where(cond).OrderBy(opts.SearchOrderBy.String())
+		Where(cond), "d").OrderBy(opts.SearchOrderBy.String())
+
 	if opts.PageSize > 0 {
-		sess.Limit(opts.PageSize, (opts.Page-1)*opts.PageSize)
+		builderQuery.Limit(opts.PageSize, (opts.Page-1)*opts.PageSize)
 	}
-	if err = sess.Find(&datasets); err != nil {
+
+	if err = sess.SQL(builderQuery).Find(&datasets); err != nil {
 		return nil, 0, fmt.Errorf("Dataset: %v", err)
 	}
 

models/dataset_reference.go

@@ -1,42 +1,87 @@
 package models
 
-import "code.gitea.io/gitea/modules/timeutil"
+import (
+	"strconv"
+	"strings"
+
+	"code.gitea.io/gitea/modules/timeutil"
+)
 
 type DatasetReference struct {
 	ID          int64              `xorm:"pk autoincr"`
-	RepoID      int64              `xorm:"INDEX unique(repo_dataset)"`
-	DatasetID   int64              `xorm:"INDEX  unique(repo_dataset)"`
+	RepoID      int64              `xorm:"INDEX unique"`
+	DatasetID   string             `xorm:"TEXT"`
 	CreatedUnix timeutil.TimeStamp `xorm:"INDEX created"`
 }
 
 func GetDatasetIdsByRepoID(repoID int64) []int64 {
 	var datasets []int64
+	var datasetIds []string
 	_ = x.Table("dataset_reference").Where("repo_id=?", repoID).
-		Cols("dataset_reference.dataset_id").Find(&datasets)
+		Cols("dataset_reference.dataset_id").Find(&datasetIds)
+	if len(datasetIds) > 0 {
+		for _, datasetIdStr := range strings.Split(datasetIds[0], ",") {
+			datasetId, err := strconv.ParseInt(datasetIdStr, 10, 64)
+			if err != nil {
+				continue
+			}
+			datasets = append(datasets, datasetId)
+		}
+	}
+
 	return datasets
 }
 
-func DeleteDatasetIdsByRepoID(repoID int64, datasetIds []int64) error {
-	if len(datasetIds) == 0 {
-		return nil
+func HasReferenceDataset(repoID int64) bool {
+
+	var datasetIds []string
+	_ = x.Table("dataset_reference").Where("repo_id=?", repoID).
+		Cols("dataset_reference.dataset_id").Find(&datasetIds)
+	return len(datasetIds) > 0
+}
+
+func getReferenceDatasetStr(repoID int64) string {
+
+	var datasetIds []string
+	_ = x.Table("dataset_reference").Where("repo_id=?", repoID).
+		Cols("dataset_reference.dataset_id").Find(&datasetIds)
+	if len(datasetIds) > 0 {
+		return datasetIds[0]
 	}
-	_, err := x.In("dataset_id", datasetIds).And("repo_id", repoID).Delete(new(DatasetReference))
+	return ""
+}
+
+func deleteDatasetIdsByRepoID(repoID int64) error {
+
+	_, err := x.Where("repo_id", repoID).Delete(new(DatasetReference))
 	return err
 }
 
 func NewDatasetIdsByRepoID(repoID int64, datasetIds []int64) error {
-	var datasetReference []DatasetReference
+	if len(datasetIds) == 0 { //关联数据集数组为空，不需要处理
+		return nil
+	}
+	var datasetsStrArray []string
 	for _, datasetId := range datasetIds {
-		datasetReference = append(datasetReference, DatasetReference{
-			DatasetID: datasetId,
-			RepoID:    repoID,
-		})
+		datasetsStrArray = append(datasetsStrArray, strconv.FormatInt(datasetId, 10))
 	}
-	_, err := x.Insert(datasetReference)
-	return err
-}
 
-func DeleteReferenceDatasetByDatasetID(datasetId int64) error {
-	_, err := x.Delete(&DatasetReference{DatasetID: datasetId})
+	newDatasetStr := strings.Join(datasetsStrArray, ",")
+	oldDatasetStr := getReferenceDatasetStr(repoID)
+	if newDatasetStr == oldDatasetStr { //关联数据集无变化，不需要处理
+		return nil
+	}
+	if oldDatasetStr != "" { //已经存在关联数据集
+		err := deleteDatasetIdsByRepoID(repoID)
+		if err != nil {
+			return err
+		}
+	}
+	datasetReference := DatasetReference{
+		DatasetID: newDatasetStr,
+		RepoID:    repoID,
+	}
+
+	_, err := x.Insert(datasetReference)
 	return err
 }

models/models.go

@@ -145,7 +145,7 @@ func init() {
 		new(OrgStatistic),
 		new(SearchRecord),
 		new(AiModelConvert),
-		new(CloudbrainTemp),
+		new(DatasetReference),
 	)
 
 	tablesStatistic = append(tablesStatistic,

modules/dataset/dataset.go

@@ -27,7 +27,7 @@ func IsShowDataSetOfCurrentRepo(repoID int64) bool {
 	if dataset != nil {
 		return true
 	}
-	if len(models.GetDatasetIdsByRepoID(repoID)) > 0 {
+	if models.HasReferenceDataset(repoID) {
 		return false
 	}
 	return true

routers/repo/dataset.go

@@ -274,7 +274,16 @@ func ReferenceDatasetDelete(ctx *context.Context) {
 	repoID := ctx.Repo.Repository.ID
 	datasetId, _ := strconv.ParseInt(ctx.Params(":id"), 10, 64)
 
-	err := models.DeleteDatasetIdsByRepoID(repoID, []int64{datasetId})
+	oldDatasetIds := models.GetDatasetIdsByRepoID(repoID)
+
+	var newDatasetIds []int64
+
+	for _, tempDatasetId := range oldDatasetIds {
+		if datasetId != tempDatasetId {
+			newDatasetIds = append(newDatasetIds, datasetId)
+		}
+	}
+	err := models.NewDatasetIdsByRepoID(repoID, newDatasetIds)
 	if err != nil {
 		ctx.JSON(http.StatusOK, models.BaseErrorMessage("dataset.cancel_reference_dataset_fail"))
 	}
@@ -284,29 +293,7 @@ func ReferenceDatasetDelete(ctx *context.Context) {
 
 func ReferenceDatasetPost(ctx *context.Context, form auth.ReferenceDatasetForm) {
 	repoID := ctx.Repo.Repository.ID
-	datasetsID := models.GetDatasetIdsByRepoID(repoID)
-
-	var newDataset []int64
-	var deletedDataset []int64
-
-	for _, id := range datasetsID {
-		if !contains(form.DatasetID, id) {
-			deletedDataset = append(deletedDataset, id)
-		}
-	}
-
-	for _, id := range form.DatasetID {
-		if !contains(datasetsID, id) {
-			newDataset = append(newDataset, id)
-		}
-	}
-
-	err := models.DeleteDatasetIdsByRepoID(repoID, deletedDataset)
-
-	if err != nil {
-		ctx.JSON(http.StatusOK, models.BaseErrorMessage(ctx.Tr("dataset.reference_dataset_fail")))
-	}
-	err = models.NewDatasetIdsByRepoID(repoID, newDataset)
+	err := models.NewDatasetIdsByRepoID(repoID, form.DatasetID)
 	if err != nil {
 		ctx.JSON(http.StatusOK, models.BaseErrorMessage("dataset.reference_dataset_fail"))
 	}
@@ -315,15 +302,6 @@ func ReferenceDatasetPost(ctx *context.Context, form auth.ReferenceDatasetForm)
 
 }
 
-func contains(s []int64, e int64) bool {
-	for _, a := range s {
-		if a == e {
-			return true
-		}
-	}
-	return false
-}
-
 func EditDatasetPost(ctx *context.Context, form auth.EditDatasetForm) {
 	ctx.Data["PageIsDataset"] = true
 
@@ -467,13 +445,23 @@ func MyDatasets(ctx *context.Context) {
 }
 
 func datasetMultiple(ctx *context.Context, opts *models.SearchDatasetOptions) {
-	datasets, count, err := datasetMultipleGet(ctx, opts)
+	page := ctx.QueryInt("page")
+	keyword := strings.Trim(ctx.Query("q"), " ")
+	opts.Keyword = keyword
+	if opts.SearchOrderBy.String() == "" {
+		opts.SearchOrderBy = models.SearchOrderByRecentUpdated
+	}
 
-	datasetMultipleResult(ctx, err, datasets, count)
+	opts.RecommendOnly = ctx.QueryBool("recommend")
+	opts.ListOptions = models.ListOptions{
+		Page:     page,
+		PageSize: setting.UI.DatasetPagingNum,
+	}
+	opts.JustNeedZipFile = true
+	opts.User = ctx.User
 
-}
+	datasets, count, err := models.SearchDataset(opts)
 
-func datasetMultipleResult(ctx *context.Context, err error, datasets models.DatasetList, count int64) {
 	if err != nil {
 		ctx.ServerError("datasets", err)
 		return
@@ -496,52 +484,31 @@ func datasetMultipleResult(ctx *context.Context, err error, datasets models.Data
 	})
 }
 
-func datasetMultipleGet(ctx *context.Context, opts *models.SearchDatasetOptions) (models.DatasetList, int64, error) {
-	page := ctx.QueryInt("page")
-	keyword := strings.Trim(ctx.Query("q"), " ")
-	orderBy := models.SearchOrderByRecentUpdated
-	opts.Keyword = keyword
-	opts.SearchOrderBy = orderBy
-	opts.RecommendOnly = ctx.QueryBool("recommend")
-	opts.ListOptions = models.ListOptions{
-		Page:     page,
-		PageSize: setting.UI.DatasetPagingNum,
-	}
-	opts.JustNeedZipFile = true
-	opts.User = ctx.User
-
-	datasets, count, err := models.SearchDataset(opts)
-	return datasets, count, err
-}
-
 func CurrentRepoDatasetMultiple(ctx *context.Context) {
-
+	datasetIds := models.GetDatasetIdsByRepoID(ctx.Repo.Repository.ID)
+	searchOrderBy := getSearchOrderByInValues(datasetIds)
 	opts := &models.SearchDatasetOptions{
 		RepoID:         ctx.Repo.Repository.ID,
 		NeedAttachment: true,
 		CloudBrainType: ctx.QueryInt("type"),
-		DatasetIDs:     models.GetDatasetIdsByRepoID(ctx.Repo.Repository.ID),
+		DatasetIDs:     datasetIds,
+		SearchOrderBy:  searchOrderBy,
 	}
-	datasetList, count, err := datasetMultipleGet(ctx, opts)
 
-	if len(datasetList) > 0 {
-		var convertDatasetList models.DatasetList
+	datasetMultiple(ctx, opts)
 
-		for _, dataset := range datasetList {
-			if dataset.RepoID == ctx.Repo.Repository.ID && len(convertDatasetList) == 0 {
-				convertDatasetList = append(convertDatasetList, dataset)
-			}
-		}
-		for _, dataset := range datasetList {
-			if dataset.RepoID != ctx.Repo.Repository.ID {
-				convertDatasetList = append(convertDatasetList, dataset)
-			}
-		}
-		datasetMultipleResult(ctx, err, convertDatasetList, count)
-	} else {
-		datasetMultipleResult(ctx, err, datasetList, count)
-	}
+}
 
+func getSearchOrderByInValues(datasetIds []int64) models.SearchOrderBy {
+	if len(datasetIds) == 0 {
+		return ""
+	}
+	searchOrderBy := "CASE id "
+	for i, id := range datasetIds {
+		searchOrderBy += fmt.Sprintf(" WHEN %d THEN %d", id, i+1)
+	}
+	searchOrderBy += " ELSE 0 END"
+	return models.SearchOrderBy(searchOrderBy)
 }
 
 func MyDatasetsMultiple(ctx *context.Context) {
@@ -589,17 +556,18 @@ func MyFavoriteDatasetMultiple(ctx *context.Context) {
 }
 func ReferenceDataset(ctx *context.Context) {
 	MustEnableDataset(ctx)
+	datasetIds := models.GetDatasetIdsByRepoID(ctx.Repo.Repository.ID)
 	opts := &models.SearchDatasetOptions{
-		DatasetIDs:     models.GetDatasetIdsByRepoID(ctx.Repo.Repository.ID),
+		DatasetIDs:     datasetIds,
 		NeedAttachment: false,
 		CloudBrainType: models.TypeCloudBrainAll,
 		ListOptions: models.ListOptions{
 			Page:     1,
 			PageSize: setting.RepoMaxReferenceDatasetNum,
 		},
-		SearchOrderBy: models.SearchOrderByRecentUpdated,
+		SearchOrderBy: getSearchOrderByInValues(datasetIds),
 	}
-	datasets, count, err := models.SearchDataset(opts)
+	datasets, _, err := models.SearchDataset(opts)
 
 	if err != nil {
 		ctx.ServerError("SearchDatasets", err)
@@ -608,7 +576,7 @@ func ReferenceDataset(ctx *context.Context) {
 
 	ctx.Data["Datasets"] = repository.ConvertToDatasetWithStar(ctx, datasets)
 	ctx.Data["PageIsDataset"] = true
-	ctx.Data["Total"] = count
+	ctx.Data["MaxReferenceDatasetNum"] = setting.RepoMaxReferenceDatasetNum
 	ctx.HTML(200, tplReference)
 
 }