@mastra/mongodb 1.12.1 → 1.13.0-alpha.0

package/dist/index.js

@@ -1,6 +1,6 @@
 import { v4 } from '@lukeed/uuid';
 import { MastraError, ErrorCategory, ErrorDomain } from '@mastra/core/error';
-import { createVectorErrorId, AgentsStorage, TABLE_AGENTS, TABLE_AGENT_VERSIONS, createStorageErrorId, normalizePerPage, calculatePagination, BackgroundTasksStorage, TABLE_BACKGROUND_TASKS, BlobStore, TABLE_SKILL_BLOBS, DatasetsStorage, TABLE_DATASETS, TABLE_DATASET_ITEMS, TABLE_DATASET_VERSIONS, ensureDate, safelyParseJSON, TABLE_EXPERIMENTS, TABLE_EXPERIMENT_RESULTS, ExperimentsStorage, MCPClientsStorage, TABLE_MCP_CLIENTS, TABLE_MCP_CLIENT_VERSIONS, MCPServersStorage, TABLE_MCP_SERVERS, TABLE_MCP_SERVER_VERSIONS, MemoryStorage, TABLE_THREADS, TABLE_MESSAGES, TABLE_RESOURCES, NotificationsStorage, TABLE_NOTIFICATIONS, ObservabilityStorage, TABLE_SPANS, listTracesArgsSchema, toTraceSpans, PromptBlocksStorage, TABLE_PROMPT_BLOCKS, TABLE_PROMPT_BLOCK_VERSIONS, SchedulesStorage, TABLE_SCHEDULES, TABLE_SCHEDULE_TRIGGERS, ScorerDefinitionsStorage, TABLE_SCORER_DEFINITIONS, TABLE_SCORER_DEFINITION_VERSIONS, ScoresStorage, TABLE_SCORERS, SkillsStorage, TABLE_SKILLS, TABLE_SKILL_VERSIONS, WorkflowsStorage, TABLE_WORKFLOW_SNAPSHOT, WorkspacesStorage, TABLE_WORKSPACES, TABLE_WORKSPACE_VERSIONS, MastraCompositeStore, TraceStatus, parseDuration, normalizeScheduleTarget, transformScoreRow as transformScoreRow$1 } from '@mastra/core/storage';
+import { createVectorErrorId, AgentsStorage, TABLE_AGENTS, TABLE_AGENT_VERSIONS, createStorageErrorId, normalizePerPage, calculatePagination, BackgroundTasksStorage, TABLE_BACKGROUND_TASKS, BlobStore, TABLE_SKILL_BLOBS, DatasetsStorage, TABLE_DATASETS, TABLE_DATASET_ITEMS, TABLE_DATASET_VERSIONS, ensureDate, safelyParseJSON, hasErrorCode, TABLE_EXPERIMENTS, TABLE_EXPERIMENT_RESULTS, ExperimentsStorage, MCPClientsStorage, TABLE_MCP_CLIENTS, TABLE_MCP_CLIENT_VERSIONS, MCPServersStorage, TABLE_MCP_SERVERS, TABLE_MCP_SERVER_VERSIONS, MemoryStorage, TABLE_THREADS, TABLE_MESSAGES, TABLE_RESOURCES, NotificationsStorage, TABLE_NOTIFICATIONS, ObservabilityStorage, TABLE_SPANS, listTracesArgsSchema, toTraceSpans, PromptBlocksStorage, TABLE_PROMPT_BLOCKS, TABLE_PROMPT_BLOCK_VERSIONS, SchedulesStorage, TABLE_SCHEDULES, TABLE_SCHEDULE_TRIGGERS, ScorerDefinitionsStorage, TABLE_SCORER_DEFINITIONS, TABLE_SCORER_DEFINITION_VERSIONS, ScoresStorage, TABLE_SCORERS, SkillsStorage, TABLE_SKILLS, TABLE_SKILL_VERSIONS, WorkflowsStorage, TABLE_WORKFLOW_SNAPSHOT, WorkspacesStorage, TABLE_WORKSPACES, TABLE_WORKSPACE_VERSIONS, MastraCompositeStore, TraceStatus, parseDuration, normalizeScheduleTarget, transformScoreRow as transformScoreRow$1 } from '@mastra/core/storage';
 import { MastraVector, validateUpsertInput, validateVectorValues } from '@mastra/core/vector';
 import { MongoClient, MongoBulkWriteError } from 'mongodb';
 import { BaseFilterTranslator } from '@mastra/core/vector/filter';
@@ -13,7 +13,7 @@ import { skillSnapshotFieldValuesEqual } from '@mastra/core/storage/domains/skil
 
 // package.json
 var package_default = {
-  version: "1.12.1"};
+  version: "1.13.0-alpha.0"};
 var MongoDBFilterTranslator = class extends BaseFilterTranslator {
   getSupportedOperators() {
     return {
@@ -2216,6 +2216,11 @@ var MongoDBDatasetsStorage = class extends DatasetsStorage {
         keys: { datasetId: 1, datasetVersion: 1 },
         options: { name: "idx_dataset_items_datasetid_version" }
       },
+      {
+        collection: TABLE_DATASET_ITEMS,
+        keys: { datasetId: 1, externalId: 1, datasetVersion: 1 },
+        options: { name: "idx_dataset_items_datasetid_externalid_version", sparse: true }
+      },
       {
         collection: TABLE_DATASET_ITEMS,
         keys: { datasetId: 1, validTo: 1, isDeleted: 1, createdAt: -1, id: 1 },
@@ -2286,6 +2291,7 @@ var MongoDBDatasetsStorage = class extends DatasetsStorage {
       id: row.id,
       datasetId: row.datasetId,
       datasetVersion: row.datasetVersion,
+      externalId: row.externalId ?? null,
       organizationId: row.organizationId ?? null,
       projectId: row.projectId ?? null,
       input: typeof row.input === "string" ? safelyParseJSON(row.input) : row.input,
@@ -2317,7 +2323,8 @@ var MongoDBDatasetsStorage = class extends DatasetsStorage {
   // --- Dataset CRUD ---
   async createDataset(input) {
     try {
-      const id = randomUUID();
+      const id = input.id ?? randomUUID();
+      if (input.id !== void 0) this.validateCallerDefinedDatasetId(input.id);
       const now = /* @__PURE__ */ new Date();
       const collection = await this.getCollection(TABLE_DATASETS);
       const record = {
@@ -2342,6 +2349,10 @@ var MongoDBDatasetsStorage = class extends DatasetsStorage {
       await collection.insertOne(record);
       return this.transformDatasetRow(record);
     } catch (error) {
+      if (input.id !== void 0 && hasErrorCode(error, /* @__PURE__ */ new Set([11e3]))) {
+        const existing = await this.getDatasetById({ id: input.id });
+        if (existing) return this.resolveExistingDataset(existing, { ...input, id: input.id });
+      }
       if (error instanceof MastraError) throw error;
       throw new MastraError(
         {
@@ -2655,6 +2666,7 @@ var MongoDBDatasetsStorage = class extends DatasetsStorage {
             id: args.id,
             datasetId: args.datasetId,
             datasetVersion: newVersion,
+            externalId: existing.externalId ?? null,
             organizationId: parentOrganizationId,
             projectId: parentProjectId,
             validTo: null,
@@ -2751,6 +2763,7 @@ var MongoDBDatasetsStorage = class extends DatasetsStorage {
             id,
             datasetId,
             datasetVersion: newVersion,
+            externalId: existing.externalId ?? null,
             organizationId: parentOrganizationId,
             projectId: parentProjectId,
             validTo: null,
@@ -2792,37 +2805,27 @@ var MongoDBDatasetsStorage = class extends DatasetsStorage {
   // --- Batch operations ---
   async _doBatchInsertItems(input) {
     try {
-      const dataset = await this.getDatasetById({ id: input.datasetId });
-      if (!dataset) {
+      if (input.items.length === 0) return [];
+      const externalIds = [...new Set(input.items.flatMap((item) => item.externalId ? [item.externalId] : []))];
+      if (externalIds.length > 0 && !await this.#connector.supportsTransactions()) {
         throw new MastraError({
-          id: createStorageErrorId("MONGODB", "BULK_ADD_ITEMS", "DATASET_NOT_FOUND"),
+          id: "DATASET_ITEM_IDENTITY_REQUIRES_TRANSACTIONS",
           domain: ErrorDomain.STORAGE,
           category: ErrorCategory.USER,
           details: { datasetId: input.datasetId }
         });
       }
-      if (!input.items || input.items.length === 0) {
-        return [];
-      }
-      const now = /* @__PURE__ */ new Date();
-      const versionId = randomUUID();
-      const itemsWithIds = input.items.map((itemInput) => ({
-        generatedId: randomUUID(),
-        itemInput
-      }));
       const datasetsCollection = await this.getCollection(TABLE_DATASETS);
       const itemsCollection = await this.getCollection(TABLE_DATASET_ITEMS);
       const versionsCollection = await this.getCollection(TABLE_DATASET_VERSIONS);
-      let newVersion = 0;
-      let organizationId = null;
-      let projectId = null;
+      let resolvedItems = [];
       await this.#connector.withTransaction(async (session) => {
-        const result = await datasetsCollection.findOneAndUpdate(
+        const dataset = await datasetsCollection.findOneAndUpdate(
           { id: input.datasetId },
-          { $inc: { version: 1 } },
+          { $inc: { identityRevision: 1 } },
           { session, returnDocument: "after" }
         );
-        if (!result) {
+        if (!dataset) {
           throw new MastraError({
             id: createStorageErrorId("MONGODB", "BULK_ADD_ITEMS", "DATASET_NOT_FOUND"),
             domain: ErrorDomain.STORAGE,
@@ -2830,54 +2833,52 @@ var MongoDBDatasetsStorage = class extends DatasetsStorage {
             details: { datasetId: input.datasetId }
           });
         }
-        newVersion = result.version;
-        organizationId = result.organizationId ?? null;
-        projectId = result.projectId ?? null;
-        const docs = itemsWithIds.map(({ generatedId, itemInput }) => ({
-          id: generatedId,
-          datasetId: input.datasetId,
-          datasetVersion: newVersion,
-          organizationId,
-          projectId,
-          validTo: null,
-          isDeleted: false,
-          input: itemInput.input,
-          groundTruth: itemInput.groundTruth ?? null,
-          expectedTrajectory: itemInput.expectedTrajectory ?? null,
-          toolMocks: itemInput.toolMocks ?? null,
-          requestContext: itemInput.requestContext ?? null,
-          metadata: itemInput.metadata ?? null,
-          source: itemInput.source ?? null,
-          createdAt: now,
-          updatedAt: now
-        }));
-        await itemsCollection.insertMany(docs, { session });
-        await versionsCollection.insertOne(
-          {
-            id: versionId,
-            datasetId: input.datasetId,
-            version: newVersion,
-            createdAt: now
-          },
-          { session }
+        const historyRows = externalIds.length === 0 ? [] : await itemsCollection.find({ datasetId: input.datasetId, externalId: { $in: externalIds } }, { session }).sort({ datasetVersion: 1 }).toArray();
+        const plan = this.planDatasetItemBatch(
+          input.items,
+          historyRows.map((row) => this.transformItemRowFull(row)),
+          randomUUID
+        );
+        const resolved = new Map(
+          [...plan.existingCurrentItems].map(([id, row]) => [id, this.datasetItemFromRow(row)])
         );
+        if (plan.inserts.length > 0) {
+          const now = /* @__PURE__ */ new Date();
+          const newVersion = Number(dataset.version) + 1;
+          await datasetsCollection.updateOne(
+            { id: input.datasetId },
+            { $set: { version: newVersion, updatedAt: now } },
+            { session }
+          );
+          for (const insert of plan.inserts) {
+            const item = {
+              id: insert.id,
+              datasetId: input.datasetId,
+              datasetVersion: newVersion,
+              externalId: insert.item.externalId ?? null,
+              organizationId: dataset.organizationId ?? null,
+              projectId: dataset.projectId ?? null,
+              input: insert.item.input,
+              groundTruth: insert.item.groundTruth,
+              expectedTrajectory: insert.item.expectedTrajectory,
+              toolMocks: insert.item.toolMocks,
+              requestContext: insert.item.requestContext,
+              metadata: insert.item.metadata,
+              source: insert.item.source,
+              createdAt: now,
+              updatedAt: now
+            };
+            await itemsCollection.insertOne({ ...item, validTo: null, isDeleted: false }, { session });
+            resolved.set(item.id, item);
+          }
+          await versionsCollection.insertOne(
+            { id: randomUUID(), datasetId: input.datasetId, version: newVersion, createdAt: now },
+            { session }
+          );
+        }
+        resolvedItems = plan.resolvedIds.map((id) => resolved.get(id));
       });
-      return itemsWithIds.map(({ generatedId, itemInput }) => ({
-        id: generatedId,
-        datasetId: input.datasetId,
-        datasetVersion: newVersion,
-        organizationId,
-        projectId,
-        input: itemInput.input,
-        groundTruth: itemInput.groundTruth,
-        expectedTrajectory: itemInput.expectedTrajectory,
-        toolMocks: itemInput.toolMocks,
-        requestContext: itemInput.requestContext,
-        metadata: itemInput.metadata,
-        source: itemInput.source,
-        createdAt: now,
-        updatedAt: now
-      }));
+      return resolvedItems;
     } catch (error) {
       if (error instanceof MastraError) throw error;
       throw new MastraError(
@@ -2936,6 +2937,7 @@ var MongoDBDatasetsStorage = class extends DatasetsStorage {
           id: item.id,
           datasetId: input.datasetId,
           datasetVersion: newVersion,
+          externalId: item.externalId ?? null,
           organizationId: parentOrganizationId,
           projectId: parentProjectId,
           validTo: null,