@soulcraft/brainy 0.54.4 → 0.54.5

package/dist/coreTypes.d.ts

@@ -378,6 +378,12 @@ export interface StorageAdapter {
     deleteVerb(id: string): Promise<void>;
     saveMetadata(id: string, metadata: any): Promise<void>;
     getMetadata(id: string): Promise<any | null>;
+    /**
+     * Get multiple metadata objects in batches (prevents socket exhaustion)
+     * @param ids Array of IDs to get metadata for
+     * @returns Promise that resolves to a Map of id -> metadata
+     */
+    getMetadataBatch?(ids: string[]): Promise<Map<string, any>>;
     /**
      * Save verb metadata to storage
      * @param id The ID of the verb

package/dist/storage/adapters/s3CompatibleStorage.d.ts

@@ -106,6 +106,11 @@ export declare class S3CompatibleStorage extends BaseStorage {
      * Initialize the storage adapter
      */
     init(): Promise<void>;
+    /**
+     * Auto-cleanup legacy /index folder during initialization
+     * This removes old index data that has been migrated to _system
+     */
+    private cleanupLegacyIndexFolder;
     /**
      * Initialize write buffers for high-volume scenarios
      */
@@ -332,6 +337,15 @@ export declare class S3CompatibleStorage extends BaseStorage {
      * Save noun metadata to storage
      */
     saveNounMetadata(id: string, metadata: any): Promise<void>;
+    /**
+     * Get multiple metadata objects in batches (CRITICAL: Prevents socket exhaustion)
+     * This is the solution to the metadata reading socket exhaustion during initialization
+     */
+    getMetadataBatch(ids: string[]): Promise<Map<string, any>>;
+    /**
+     * Get multiple verb metadata objects in batches (prevents socket exhaustion)
+     */
+    getVerbMetadataBatch(ids: string[]): Promise<Map<string, any>>;
     /**
      * Get noun metadata from storage
      */
@@ -371,8 +385,9 @@ export declare class S3CompatibleStorage extends BaseStorage {
      */
     private getCurrentStatisticsKey;
     /**
-     * Get the legacy statistics key (for backward compatibility)
+     * Get the legacy statistics key (DEPRECATED - /index folder is auto-cleaned)
      * @returns The legacy statistics key
+     * @deprecated Legacy /index folder is automatically cleaned on initialization
      */
     private getLegacyStatisticsKey;
     /**

package/dist/storage/adapters/s3CompatibleStorage.js

@@ -8,7 +8,7 @@ import { StorageCompatibilityLayer } from '../backwardCompatibility.js';
 import { StorageOperationExecutors } from '../../utils/operationUtils.js';
 import { BrainyError } from '../../errors/brainyError.js';
 import { CacheManager } from '../cacheManager.js';
-import { createModuleLogger } from '../../utils/logger.js';
+import { createModuleLogger, prodLog } from '../../utils/logger.js';
 import { getGlobalSocketManager } from '../../utils/adaptiveSocketManager.js';
 import { getGlobalBackpressure } from '../../utils/adaptiveBackpressure.js';
 import { getWriteBuffer } from '../../utils/writeBuffer.js';
@@ -230,6 +230,8 @@ export class S3CompatibleStorage extends BaseStorage {
             this.initializeBuffers();
             // Initialize request coalescer
             this.initializeCoalescer();
+            // Auto-cleanup legacy /index folder on initialization
+            await this.cleanupLegacyIndexFolder();
             this.isInitialized = true;
             this.logger.info(`Initialized ${this.serviceType} storage with bucket ${this.bucketName}`);
         }
@@ -238,6 +240,57 @@ export class S3CompatibleStorage extends BaseStorage {
             throw new Error(`Failed to initialize ${this.serviceType} storage: ${error}`);
         }
     }
+    /**
+     * Auto-cleanup legacy /index folder during initialization
+     * This removes old index data that has been migrated to _system
+     */
+    async cleanupLegacyIndexFolder() {
+        try {
+            // Check if there are any objects in the legacy index folder
+            const { ListObjectsV2Command } = await import('@aws-sdk/client-s3');
+            const listResponse = await this.s3Client.send(new ListObjectsV2Command({
+                Bucket: this.bucketName,
+                Prefix: this.indexPrefix,
+                MaxKeys: 1 // Just check if anything exists
+            }));
+            // If there are objects in the legacy index folder, clean them up
+            if (listResponse.Contents && listResponse.Contents.length > 0) {
+                prodLog.info(`🧹 Cleaning up legacy /index folder during initialization...`);
+                // Use the existing deleteObjectsWithPrefix function logic
+                const { ListObjectsV2Command, DeleteObjectsCommand } = await import('@aws-sdk/client-s3');
+                let continuationToken = undefined;
+                let totalDeleted = 0;
+                do {
+                    const listResponseBatch = await this.s3Client.send(new ListObjectsV2Command({
+                        Bucket: this.bucketName,
+                        Prefix: this.indexPrefix,
+                        ContinuationToken: continuationToken
+                    }));
+                    if (listResponseBatch.Contents && listResponseBatch.Contents.length > 0) {
+                        const objectsToDelete = listResponseBatch.Contents.map((obj) => ({
+                            Key: obj.Key
+                        }));
+                        await this.s3Client.send(new DeleteObjectsCommand({
+                            Bucket: this.bucketName,
+                            Delete: {
+                                Objects: objectsToDelete
+                            }
+                        }));
+                        totalDeleted += objectsToDelete.length;
+                    }
+                    continuationToken = listResponseBatch.NextContinuationToken;
+                } while (continuationToken);
+                prodLog.info(`✅ Cleaned up ${totalDeleted} legacy index objects`);
+            }
+            else {
+                prodLog.debug('No legacy /index folder found - already clean');
+            }
+        }
+        catch (error) {
+            // Don't fail initialization if cleanup fails
+            prodLog.warn('Failed to cleanup legacy /index folder:', error);
+        }
+    }
     /**
      * Initialize write buffers for high-volume scenarios
      */
@@ -1417,6 +1470,75 @@ export class S3CompatibleStorage extends BaseStorage {
             throw new Error(`Failed to save noun metadata for ${id}: ${error}`);
         }
     }
+    /**
+     * Get multiple metadata objects in batches (CRITICAL: Prevents socket exhaustion)
+     * This is the solution to the metadata reading socket exhaustion during initialization
+     */
+    async getMetadataBatch(ids) {
+        await this.ensureInitialized();
+        const results = new Map();
+        const batchSize = Math.min(this.getBatchSize(), 10); // Smaller batches for metadata to prevent socket exhaustion
+        // Process in smaller batches to avoid socket exhaustion
+        for (let i = 0; i < ids.length; i += batchSize) {
+            const batch = ids.slice(i, i + batchSize);
+            // Process batch with concurrency control
+            const batchPromises = batch.map(async (id) => {
+                try {
+                    const metadata = await this.getMetadata(id);
+                    return { id, metadata };
+                }
+                catch (error) {
+                    // Don't fail entire batch if one metadata read fails
+                    this.logger.debug(`Failed to read metadata for ${id}:`, error);
+                    return { id, metadata: null };
+                }
+            });
+            const batchResults = await Promise.all(batchPromises);
+            // Add results to map
+            for (const { id, metadata } of batchResults) {
+                if (metadata !== null) {
+                    results.set(id, metadata);
+                }
+            }
+            // Yield to prevent socket exhaustion between batches
+            await new Promise(resolve => setImmediate(resolve));
+        }
+        return results;
+    }
+    /**
+     * Get multiple verb metadata objects in batches (prevents socket exhaustion)
+     */
+    async getVerbMetadataBatch(ids) {
+        await this.ensureInitialized();
+        const results = new Map();
+        const batchSize = Math.min(this.getBatchSize(), 10); // Smaller batches for metadata to prevent socket exhaustion
+        // Process in smaller batches to avoid socket exhaustion
+        for (let i = 0; i < ids.length; i += batchSize) {
+            const batch = ids.slice(i, i + batchSize);
+            // Process batch with concurrency control
+            const batchPromises = batch.map(async (id) => {
+                try {
+                    const metadata = await this.getVerbMetadata(id);
+                    return { id, metadata };
+                }
+                catch (error) {
+                    // Don't fail entire batch if one metadata read fails
+                    this.logger.debug(`Failed to read verb metadata for ${id}:`, error);
+                    return { id, metadata: null };
+                }
+            });
+            const batchResults = await Promise.all(batchPromises);
+            // Add results to map
+            for (const { id, metadata } of batchResults) {
+                if (metadata !== null) {
+                    results.set(id, metadata);
+                }
+            }
+            // Yield to prevent socket exhaustion between batches
+            await new Promise(resolve => setImmediate(resolve));
+        }
+        return results;
+    }
     /**
      * Get noun metadata from storage
      */
@@ -1474,9 +1596,9 @@ export class S3CompatibleStorage extends BaseStorage {
             try {
                 // Import the GetObjectCommand only when needed
                 const { GetObjectCommand } = await import('@aws-sdk/client-s3');
-                console.log(`Getting metadata for ${id} from bucket ${this.bucketName}`);
+                prodLog.debug(`Getting metadata for ${id} from bucket ${this.bucketName}`);
                 const key = `${this.metadataPrefix}${id}.json`;
-                console.log(`Looking for metadata at key: ${key}`);
+                prodLog.debug(`Looking for metadata at key: ${key}`);
                 // Try to get the metadata from the metadata directory
                 const response = await this.s3Client.send(new GetObjectCommand({
                     Bucket: this.bucketName,
@@ -1484,20 +1606,20 @@ export class S3CompatibleStorage extends BaseStorage {
                 }));
                 // Check if response is null or undefined (can happen in mock implementations)
                 if (!response || !response.Body) {
-                    console.log(`No metadata found for ${id}`);
+                    prodLog.debug(`No metadata found for ${id}`);
                     return null;
                 }
                 // Convert the response body to a string
                 const bodyContents = await response.Body.transformToString();
-                console.log(`Retrieved metadata body: ${bodyContents}`);
+                prodLog.debug(`Retrieved metadata body: ${bodyContents}`);
                 // Parse the JSON string
                 try {
                     const parsedMetadata = JSON.parse(bodyContents);
-                    console.log(`Successfully retrieved metadata for ${id}:`, parsedMetadata);
+                    prodLog.debug(`Successfully retrieved metadata for ${id}:`, parsedMetadata);
                     return parsedMetadata;
                 }
                 catch (parseError) {
-                    console.error(`Failed to parse metadata for ${id}:`, parseError);
+                    prodLog.error(`Failed to parse metadata for ${id}:`, parseError);
                     return null;
                 }
             }
@@ -1510,7 +1632,7 @@ export class S3CompatibleStorage extends BaseStorage {
                         (error.message.includes('NoSuchKey') ||
                             error.message.includes('not found') ||
                             error.message.includes('does not exist')))) {
-                    console.log(`Metadata not found for ${id}`);
+                    prodLog.debug(`Metadata not found for ${id}`);
                     return null;
                 }
                 // For other types of errors, convert to BrainyError for better classification
@@ -1564,7 +1686,7 @@ export class S3CompatibleStorage extends BaseStorage {
             this.statisticsModified = false;
         }
         catch (error) {
-            console.error('Failed to clear storage:', error);
+            prodLog.error('Failed to clear storage:', error);
             throw new Error(`Failed to clear storage: ${error}`);
         }
     }
@@ -1659,8 +1781,9 @@ export class S3CompatibleStorage extends BaseStorage {
         return this.getStatisticsKeyForDate(new Date());
     }
     /**
-     * Get the legacy statistics key (for backward compatibility)
+     * Get the legacy statistics key (DEPRECATED - /index folder is auto-cleaned)
      * @returns The legacy statistics key
+     * @deprecated Legacy /index folder is automatically cleaned on initialization
      */
     getLegacyStatisticsKey() {
         return `${this.indexPrefix}${STATISTICS_KEY}.json`;
@@ -1872,11 +1995,12 @@ export class S3CompatibleStorage extends BaseStorage {
             // Import the GetObjectCommand only when needed
             const { GetObjectCommand } = await import('@aws-sdk/client-s3');
             // Try statistics locations in order of preference (but with timeout)
+            // NOTE: Legacy /index folder is auto-cleaned on init, so only check _system
             const keys = [
                 this.getCurrentStatisticsKey(),
                 // Only try yesterday if it's within 2 hours of midnight to avoid unnecessary calls
-                ...(this.shouldTryYesterday() ? [this.getStatisticsKeyForDate(this.getYesterday())] : []),
-                this.getLegacyStatisticsKey()
+                ...(this.shouldTryYesterday() ? [this.getStatisticsKeyForDate(this.getYesterday())] : [])
+                // Legacy fallback removed - /index folder is auto-cleaned on initialization
             ];
             let statistics = null;
             // Try each key with a timeout to prevent hanging