@byted-las/contextlake-openclaw 1.0.4 → 1.0.6

package/dist/src/commands/cli.d.ts

@@ -1,6 +1,6 @@
 import { ContextLakeConfig } from '../utils/config';
 export declare function getCliCommands(pluginConfig: ContextLakeConfig, logger: any): {
-    connectAction: (datasource_name: string, options: any) => Promise<void>;
+    connectAction: (datasource_name: string, url: string, options: any) => Promise<void>;
     ingestAction: (datasource_name: string) => Promise<void>;
     searchAction: (query: any, options: any) => Promise<void>;
     listAction: (options: any) => Promise<void>;

package/dist/src/commands/cli.js

@@ -34,28 +34,22 @@ function parseMetadata(metadata) {
 }
 function getCliCommands(pluginConfig, logger) {
     return {
-        connectAction: async (datasource_name, options) => {
-            logger.info(`[${new Date().toISOString()}] [ContextLake] CLI connect started`, { datasource_name, options });
+        connectAction: async (datasource_name, url, options) => {
+            logger.info(`[${new Date().toISOString()}] [ContextLake] CLI connect started`, { datasource_name, url, options });
             try {
                 const params = {
                     datasource_name,
-                    vendor: options.vendor,
+                    url,
                     endpoint: options.endpoint,
                     access_key: options.ak,
                     secret_key: options.sk,
                     region: options.region,
-                    bucket: options.bucket,
-                    prefix: options.prefix,
                     sample_rows: parseInt(options.sampleRows),
                 };
                 // eslint-disable-next-line no-console
                 console.log(`[contextlake connect] Connecting to datasource "${datasource_name}"...`);
                 // eslint-disable-next-line no-console
-                console.log(`  vendor: ${params.vendor}`);
-                // eslint-disable-next-line no-console
-                console.log(`  bucket: ${params.bucket}`);
-                // eslint-disable-next-line no-console
-                console.log(`  prefix: ${params.prefix}`);
+                console.log(`  url: ${params.url}`);
                 const result = await (0, profiler_1.connectDataSource)(params);
                 // eslint-disable-next-line no-console
                 console.log(JSON.stringify(result, null, 2));

package/dist/src/commands/index.js

@@ -26,6 +26,10 @@ function registerAll(ctx, logger) {
         logger.info(`[${new Date().toISOString()}] [ContextLake] Tool registered: ${tools.readS3ObjectTool.name}`);
         ctx.registerTool(tools.writeLanceCatalogTool);
         logger.info(`[${new Date().toISOString()}] [ContextLake] Tool registered: ${tools.writeLanceCatalogTool.name}`);
+        ctx.registerTool(tools.readLanceCatalogTool);
+        logger.info(`[${new Date().toISOString()}] [ContextLake] Tool registered: ${tools.readLanceCatalogTool.name}`);
+        ctx.registerTool(tools.generatePresignedUrlTool);
+        logger.info(`[${new Date().toISOString()}] [ContextLake] Tool registered: ${tools.generatePresignedUrlTool.name}`);
         ctx.registerTool(tools.listDatasourceTool);
         logger.info(`[${new Date().toISOString()}] [ContextLake] Tool registered: ${tools.listDatasourceTool.name}`);
         for (const lasTool of tools.lasTools) {
@@ -46,15 +50,12 @@ function registerAll(ctx, logger) {
                 .description('Manage ContextLake knowledge base');
             const commands = (0, cli_1.getCliCommands)(pluginConfig, logger);
             // connect -- data source profiling (las-data-profiler)
-            contextlake.command('connect <datasource_name>')
+            contextlake.command('connect <datasource_name> <url>')
                 .description('Connect to a data source and profile its structure, schemas, and media metadata into LanceDB')
-                .requiredOption('--vendor <vendor>', 'Data source type: volcengine | alibaba | tencent | aws | local')
                 .option('--endpoint <url>', 'S3 Endpoint URL (not needed for local)')
                 .option('--ak <credential_id>', 'Credential ID for the data source')
                 .option('--sk <credential_value>', 'Credential value for the data source')
                 .option('--region <region>', 'Region identifier (e.g. cn-beijing)')
-                .requiredOption('--bucket <bucket>', 'Bucket name (or local root directory for local vendor)')
-                .requiredOption('--prefix <prefix>', 'Path prefix to limit scan scope')
                 .option('--sample-rows <number>', 'Number of rows to sample per structured file', '100')
                 .action(commands.connectAction);
             // Ingest

package/dist/src/commands/slashcmd.js

@@ -133,18 +133,13 @@ function getSlashCommands(pluginConfig, logger) {
             const args = rawArgs.split(' ').filter((arg) => arg.trim() !== '');
             logger.info(`[${new Date().toISOString()}] [ContextLake] Slash command profiler started`, { args });
             try {
-                if (args.length < 4) {
-                    return { text: `**Error:** Missing arguments. Usage: /contextlake-profiler <datasource_name> <vendor> <bucket> <prefix> [endpoint] [ak] [sk] [region]` };
-                }
-                const [datasource_name, vendor, bucket, prefix, endpoint, access_key, secret_key, region] = args;
-                if (!['volcengine', 'alibaba', 'tencent', 'aws', 'local'].includes(vendor)) {
-                    return { text: `**Error:** Invalid vendor. Must be one of: volcengine, alibaba, tencent, aws, local` };
+                if (args.length < 2) {
+                    return { text: `**Error:** Missing arguments. Usage: /contextlake-profiler <datasource_name> <url> [endpoint] [ak] [sk] [region]` };
                 }
+                const [datasource_name, url, endpoint, access_key, secret_key, region] = args;
                 const params = {
                     datasource_name,
-                    vendor: vendor,
-                    bucket,
-                    prefix,
+                    url,
                     endpoint,
                     access_key,
                     secret_key,

package/dist/src/commands/tools.d.ts

@@ -9,6 +9,8 @@ export declare function getAgentTools(pluginConfig: ContextLakeConfig, logger: a
     listDatasourceTool: AnyAgentTool;
     listS3ObjectsTool: AnyAgentTool;
     readS3ObjectTool: AnyAgentTool;
+    generatePresignedUrlTool: AnyAgentTool;
     writeLanceCatalogTool: AnyAgentTool;
+    readLanceCatalogTool: AnyAgentTool;
     lasTools: AnyAgentTool[];
 };

package/dist/src/commands/tools.js

@@ -253,16 +253,10 @@ Example User Queries:
                 type: 'object',
                 properties: {
                     datasource_name: { type: 'string', description: 'Name of the data source' },
-                    vendor: { type: 'string', enum: ['volcengine', 'alibaba', 'tencent', 'aws', 'local'], description: 'Data source type' },
-                    endpoint: { type: 'string', description: 'S3 Endpoint URL (not needed for local)' },
-                    access_key: { type: 'string', description: 'Credential ID for the data source' },
-                    secret_key: { type: 'string', description: 'Credential value for the data source' },
-                    region: { type: 'string', description: 'Region identifier (e.g. cn-beijing)' },
-                    bucket: { type: 'string', description: 'Bucket name (or local root directory for local vendor)' },
-                    prefix: { type: 'string', description: 'Path prefix to limit scan scope' },
+                    url: { type: 'string', description: 'Data source URL (e.g. tos://bucket/prefix, oss://..., s3://..., file:///path)' },
                     sample_rows: { type: 'integer', description: 'Number of rows to sample per structured file' }
                 },
-                required: ['datasource_name', 'vendor', 'bucket', 'prefix'],
+                required: ['datasource_name', 'url'],
                 additionalProperties: false
             },
             async execute(toolCallId, params) {
@@ -295,17 +289,15 @@ Example User Queries:
             parameters: {
                 type: 'object',
                 properties: {
-                    vendor: { type: 'string', enum: ['volcengine', 'alibaba', 'tencent', 'aws', 'local'] },
-                    bucket: { type: 'string' },
+                    url: { type: 'string', description: 'Data source URL (e.g. tos://bucket/prefix, oss://..., file:///path)' },
+                    vendor: { type: 'string', enum: ['volcengine', 'alibaba', 'tencent', 'aws', 'local'], description: 'Required if url is not provided' },
+                    bucket: { type: 'string', description: 'Required if url is not provided' },
                     prefix: { type: 'string' },
                     endpoint: { type: 'string' },
-                    access_key: { type: 'string' },
-                    secret_key: { type: 'string' },
-                    region: { type: 'string' },
                     maxKeys: { type: 'integer' },
                     continuationToken: { type: 'string' }
                 },
-                required: ['vendor', 'bucket'],
+                required: [],
                 additionalProperties: false
             },
             async execute(toolCallId, params) {
@@ -326,22 +318,40 @@ Example User Queries:
             parameters: {
                 type: 'object',
                 properties: {
-                    vendor: { type: 'string', enum: ['volcengine', 'alibaba', 'tencent', 'aws', 'local'] },
-                    bucket: { type: 'string' },
-                    key: { type: 'string' },
+                    url: { type: 'string', description: 'Full URL to the object (e.g. tos://bucket/path/to/key.txt)' },
+                    vendor: { type: 'string', enum: ['volcengine', 'alibaba', 'tencent', 'aws', 'local'], description: 'Required if url is not provided' },
+                    bucket: { type: 'string', description: 'Required if url is not provided' },
+                    key: { type: 'string', description: 'Required if url is not provided' },
                     endpoint: { type: 'string' },
-                    access_key: { type: 'string' },
-                    secret_key: { type: 'string' },
-                    region: { type: 'string' },
                     maxBytes: { type: 'integer' }
                 },
-                required: ['vendor', 'bucket', 'key'],
+                required: [],
                 additionalProperties: false
             },
             async execute(toolCallId, params) {
                 let actualParams = params.params || params;
                 try {
-                    const buf = await (0, s3_tools_1.readS3Object)(actualParams, actualParams.key, actualParams.maxBytes);
+                    // Extract key from url if provided
+                    let key = actualParams.key;
+                    if (actualParams.url && !key) {
+                        try {
+                            if (actualParams.url.startsWith('file://')) {
+                                // Key is not strictly needed for file://, bucket contains the path in parseS3Url
+                                key = '';
+                            }
+                            else {
+                                const parsedUrl = new URL(actualParams.url);
+                                key = parsedUrl.pathname.replace(/^\//, '');
+                            }
+                        }
+                        catch (e) {
+                            // let it fail in readS3Object
+                        }
+                    }
+                    if (!key && !actualParams.url?.startsWith('file://')) {
+                        throw new Error('key is required or must be part of the url');
+                    }
+                    const buf = await (0, s3_tools_1.readS3Object)(actualParams, key, actualParams.maxBytes);
                     // Return as base64 string
                     return { content: [{ type: "text", text: buf.toString('base64') }], details: { length: buf.length } };
                 }
@@ -350,6 +360,52 @@ Example User Queries:
                 }
             }
         },
+        generatePresignedUrlTool: {
+            name: 'generate-presigned-url',
+            label: 'Generate Presigned URL',
+            description: 'Generate a presigned HTTP URL for an S3/TOS object, allowing temporary public access',
+            parameters: {
+                type: 'object',
+                properties: {
+                    url: { type: 'string', description: 'Full URL to the object (e.g. tos://bucket/path/to/key.txt)' },
+                    vendor: { type: 'string', enum: ['volcengine', 'alibaba', 'tencent', 'aws', 'local'], description: 'Required if url is not provided' },
+                    bucket: { type: 'string', description: 'Required if url is not provided' },
+                    key: { type: 'string', description: 'Required if url is not provided' },
+                    endpoint: { type: 'string' },
+                    expiresIn: { type: 'integer', description: 'Expiration time in seconds (default 3600)' }
+                },
+                required: [],
+                additionalProperties: false
+            },
+            async execute(toolCallId, params) {
+                let actualParams = params.params || params;
+                try {
+                    let key = actualParams.key;
+                    if (actualParams.url && !key) {
+                        try {
+                            if (actualParams.url.startsWith('file://')) {
+                                key = '';
+                            }
+                            else {
+                                const parsedUrl = new URL(actualParams.url);
+                                key = parsedUrl.pathname.replace(/^\//, '');
+                            }
+                        }
+                        catch (e) {
+                            // let it fail in getPresignedUrl
+                        }
+                    }
+                    if (!key && !actualParams.url?.startsWith('file://')) {
+                        throw new Error('key is required or must be part of the url');
+                    }
+                    const url = await (0, s3_tools_1.getPresignedUrl)(actualParams, key, actualParams.expiresIn);
+                    return { content: [{ type: "text", text: url }], details: { url } };
+                }
+                catch (e) {
+                    return { content: [{ type: "text", text: String(e.message) }], details: { error: e.message } };
+                }
+            }
+        },
         writeLanceCatalogTool: {
             name: 'write-lance-catalog',
             label: 'Write LanceDB Catalog',
@@ -374,6 +430,32 @@ Example User Queries:
                     return { content: [{ type: "text", text: String(e.message) }], details: { error: e.message } };
                 }
             }
+        },
+        readLanceCatalogTool: {
+            name: 'read-lance-catalog',
+            label: 'Read LanceDB Catalog',
+            description: 'Read records from a local LanceDB table for validation or ingestion processes',
+            parameters: {
+                type: 'object',
+                properties: {
+                    db_path: { type: 'string', description: 'Path to the local LanceDB database' },
+                    table_name: { type: 'string', description: 'Name of the table to read' },
+                    limit: { type: 'integer', description: 'Maximum number of records to return' },
+                    filter: { type: 'string', description: 'SQL-like filter string (e.g., "category = \'structured\'")' }
+                },
+                required: ['db_path', 'table_name'],
+                additionalProperties: false
+            },
+            async execute(toolCallId, params) {
+                let actualParams = params.params || params;
+                try {
+                    const results = await (0, lance_tools_1.readLanceCatalog)(actualParams);
+                    return { content: [{ type: "text", text: JSON.stringify(results) }], details: { count: results.length, data: results } };
+                }
+                catch (e) {
+                    return { content: [{ type: "text", text: String(e.message) }], details: { error: e.message } };
+                }
+            }
         }
     };
 }

package/dist/src/lib/actions/lance-tools.d.ts

@@ -3,4 +3,11 @@ export interface LanceWriteParams {
     table_name: string;
     records: any[];
 }
+export interface LanceReadParams {
+    db_path: string;
+    table_name: string;
+    limit?: number;
+    filter?: string;
+}
 export declare function writeLanceCatalog(params: LanceWriteParams): Promise<void>;
+export declare function readLanceCatalog(params: LanceReadParams): Promise<any[]>;

package/dist/src/lib/actions/lance-tools.js

@@ -34,7 +34,9 @@ var __importStar = (this && this.__importStar) || (function () {
 })();
 Object.defineProperty(exports, "__esModule", { value: true });
 exports.writeLanceCatalog = writeLanceCatalog;
+exports.readLanceCatalog = readLanceCatalog;
 const lancedb = __importStar(require("@lancedb/lancedb"));
+const fs = __importStar(require("fs"));
 async function writeLanceCatalog(params) {
     if (!params.records || params.records.length === 0) {
         return;
@@ -49,3 +51,23 @@ async function writeLanceCatalog(params) {
         await db.createTable(params.table_name, params.records);
     }
 }
+async function readLanceCatalog(params) {
+    if (!fs.existsSync(params.db_path)) {
+        throw new Error(`Database not found at ${params.db_path}`);
+    }
+    const db = await lancedb.connect(params.db_path);
+    const tableNames = await db.tableNames();
+    if (!tableNames.includes(params.table_name)) {
+        throw new Error(`Table '${params.table_name}' not found in database`);
+    }
+    const table = await db.openTable(params.table_name);
+    let query = table.query();
+    if (params.filter) {
+        query = query.where(params.filter);
+    }
+    if (params.limit && params.limit > 0) {
+        query = query.limit(params.limit);
+    }
+    const results = await query.toArray();
+    return results;
+}

package/dist/src/lib/actions/las-tools.js

@@ -2,8 +2,35 @@
 Object.defineProperty(exports, "__esModule", { value: true });
 exports.getLasTools = getLasTools;
 const las_api_1 = require("./las-api");
+const s3_tools_1 = require("./s3-tools");
 function getLasTools(pluginConfig, logger) {
     const apiClient = new las_api_1.LasApiClient(pluginConfig, logger);
+    const processUrl = async (url) => {
+        if (!url)
+            return url;
+        if (url.startsWith('tos://')) {
+            // LAS operators prefer native tos:// paths when supported, leave as is
+            return url;
+        }
+        else if (url.startsWith('oss://') || url.startsWith('s3://') || url.startsWith('cos://') || url.startsWith('file://')) {
+            // Need presigned HTTP url for other vendors
+            logger.info(`[LasTools] Presigning URL for vendor: ${url}`);
+            try {
+                // If it's a file:// we also presign it to file:// which might not be supported by remote LAS, 
+                // but local files typically need to be uploaded to TOS first. We'll leave file:// to fail or be handled elsewhere.
+                if (url.startsWith('file://'))
+                    return url;
+                const urlParts = new URL(url);
+                const key = urlParts.pathname.replace(/^\//, '');
+                return await (0, s3_tools_1.getPresignedUrl)({ url }, key, 3600);
+            }
+            catch (e) {
+                logger.warn(`[LasTools] Failed to presign URL: ${url}`, { error: e.message });
+                return url; // fallback to original
+            }
+        }
+        return url;
+    };
     const callApi = async (method, args) => {
         try {
             // @ts-ignore
@@ -33,6 +60,9 @@ Parameters in data:
                 required: ['data']
             },
             async execute(toolCallId, params) {
+                if (params.data?.image) {
+                    params.data.image = await processUrl(params.data.image);
+                }
                 return await callApi('process', ['las_image_resample', params.data]);
             }
         },
@@ -53,6 +83,9 @@ Parameters in data:
                 required: ['data']
             },
             async execute(toolCallId, params) {
+                if (params.data?.input_path) {
+                    params.data.input_path = await processUrl(params.data.input_path);
+                }
                 return await callApi('process', ['las_audio_extract_and_split', params.data]);
             }
         },
@@ -71,6 +104,9 @@ Parameters in data:
                 required: ['data']
             },
             async execute(toolCallId, params) {
+                if (params.data?.input_path) {
+                    params.data.input_path = await processUrl(params.data.input_path);
+                }
                 return await callApi('process', ['las_audio_convert', params.data]);
             }
         },
@@ -89,6 +125,9 @@ Parameters in data:
                 required: ['data']
             },
             async execute(toolCallId, params) {
+                if (params.data?.audio?.url) {
+                    params.data.audio.url = await processUrl(params.data.audio.url);
+                }
                 return await callApi('submitAndPoll', ['las_asr_pro', params.data]);
             }
         },
@@ -105,6 +144,9 @@ Parameters in data:
                 required: ['data']
             },
             async execute(toolCallId, params) {
+                if (params.data?.audio?.url) {
+                    params.data.audio.url = await processUrl(params.data.audio.url);
+                }
                 return await callApi('submitAndPoll', ['las_seed_2_0', params.data]);
             }
         },
@@ -128,6 +170,13 @@ Parameters:
                 required: ['model', 'input']
             },
             async execute(toolCallId, params) {
+                if (params.input && Array.isArray(params.input)) {
+                    for (const item of params.input) {
+                        if (item.type === 'image_url' && item.image_url?.url) {
+                            item.image_url.url = await processUrl(item.image_url.url);
+                        }
+                    }
+                }
                 return await callApi('multimodalEmbedding', [
                     params.model,
                     params.input,
@@ -151,6 +200,9 @@ Parameters in data:
                 required: ['data']
             },
             async execute(toolCallId, params) {
+                if (params.data?.video_url) {
+                    params.data.video_url = await processUrl(params.data.video_url);
+                }
                 return await callApi('submitAndPoll', ['las_long_video_understand', params.data]);
             }
         },
@@ -169,6 +221,9 @@ Parameters in data:
                 required: ['data']
             },
             async execute(toolCallId, params) {
+                if (params.data?.url) {
+                    params.data.url = await processUrl(params.data.url);
+                }
                 return await callApi('submitAndPoll', ['las_pdf_parse_doubao', params.data]);
             }
         },
@@ -187,6 +242,9 @@ Parameters in data:
                 required: ['data']
             },
             async execute(toolCallId, params) {
+                if (params.data?.video_url) {
+                    params.data.video_url = await processUrl(params.data.video_url);
+                }
                 return await callApi('submitAndPoll', ['las_video_resize', params.data]);
             }
         }

package/dist/src/lib/actions/profiler.d.ts

@@ -1,12 +1,13 @@
 export interface ConnectParams {
     datasource_name: string;
-    vendor: 'volcengine' | 'alibaba' | 'tencent' | 'aws' | 'local';
+    url: string;
+    vendor?: 'volcengine' | 'alibaba' | 'tencent' | 'aws' | 'local';
     endpoint?: string;
     access_key?: string;
     secret_key?: string;
     region?: string;
-    bucket: string;
-    prefix: string;
+    bucket?: string;
+    prefix?: string;
     sample_rows?: number;
 }
 export interface ConnectResult {

package/dist/src/lib/actions/profiler.js

@@ -98,12 +98,41 @@ function classifyFile(ext) {
 async function connectDataSource(params, _ctx) {
     if (!params.datasource_name)
         throw new Error('datasource_name is required');
-    if (!params.vendor)
-        throw new Error('vendor is required');
-    if (!params.bucket)
-        throw new Error('bucket is required');
-    if (params.prefix === undefined || params.prefix === null)
-        throw new Error('prefix is required');
+    if (!params.url)
+        throw new Error('url is required (e.g. tos://bucket/prefix)');
+    // Parse URL: tos://bucket/prefix
+    try {
+        if (params.url.startsWith('file://') || params.url.startsWith('/')) {
+            params.vendor = 'local';
+            const localPath = params.url.startsWith('file://') ? params.url.slice(7) : params.url;
+            params.bucket = localPath;
+            params.prefix = '.';
+        }
+        else {
+            const parsedUrl = new URL(params.url);
+            const protocol = parsedUrl.protocol.replace(':', '');
+            if (['tos', 'oss', 'cos', 's3'].includes(protocol)) {
+                if (protocol === 'tos')
+                    params.vendor = 'volcengine';
+                else if (protocol === 'oss')
+                    params.vendor = 'alibaba';
+                else if (protocol === 'cos')
+                    params.vendor = 'tencent';
+                else if (protocol === 's3')
+                    params.vendor = 'aws';
+                params.bucket = parsedUrl.hostname;
+                params.prefix = parsedUrl.pathname.replace(/^\//, ''); // Remove leading slash
+            }
+            else {
+                throw new Error(`Unsupported protocol: ${protocol}`);
+            }
+        }
+    }
+    catch (e) {
+        if (!params.vendor || !params.bucket || params.prefix === undefined) {
+            throw new Error(`Invalid url format: ${e.message}`);
+        }
+    }
     if (params.vendor !== 'local') {
         if (!params.endpoint && params.vendor !== 'aws')
             throw new Error(`endpoint is required for vendor "${params.vendor}"`);
@@ -139,7 +168,14 @@ async function connectDataSource(params, _ctx) {
         const allRecords = [];
         const scan_ts = new Date().toISOString() + 'Z';
         while (isTruncated) {
-            const response = await (0, s3_tools_1.listS3Objects)(params, params.prefix, 1000, continuationToken);
+            const response = await (0, s3_tools_1.listS3Objects)({
+                vendor: params.vendor,
+                bucket: params.bucket,
+                endpoint: params.endpoint,
+                access_key: params.access_key,
+                secret_key: params.secret_key,
+                region: params.region
+            }, params.prefix || '', 1000, continuationToken);
             for (const obj of response.Contents) {
                 const key = obj.Key || '';
                 if (key.endsWith('/'))

package/dist/src/lib/actions/s3-tools.d.ts

@@ -1,11 +1,13 @@
 export interface S3Params {
-    vendor: 'volcengine' | 'alibaba' | 'tencent' | 'aws' | 'local';
+    url?: string;
+    vendor?: 'volcengine' | 'alibaba' | 'tencent' | 'aws' | 'local';
     endpoint?: string;
     access_key?: string;
     secret_key?: string;
     region?: string;
-    bucket: string;
+    bucket?: string;
 }
+export declare function parseS3Url(params: S3Params): S3Params;
 export declare function listS3Objects(params: S3Params, prefix: string, maxKeys?: number, continuationToken?: string): Promise<{
     Contents: any[];
     IsTruncated: boolean;
@@ -16,3 +18,4 @@ export declare function listS3Objects(params: S3Params, prefix: string, maxKeys?
     NextContinuationToken: string | undefined;
 }>;
 export declare function readS3Object(params: S3Params, key: string, maxBytes?: number): Promise<Buffer>;
+export declare function getPresignedUrl(params: S3Params, key: string, expiresIn?: number): Promise<string>;

package/dist/src/lib/actions/s3-tools.js

@@ -33,12 +33,48 @@ var __importStar = (this && this.__importStar) || (function () {
     };
 })();
 Object.defineProperty(exports, "__esModule", { value: true });
+exports.parseS3Url = parseS3Url;
 exports.listS3Objects = listS3Objects;
 exports.readS3Object = readS3Object;
+exports.getPresignedUrl = getPresignedUrl;
 const client_s3_1 = require("@aws-sdk/client-s3");
+const s3_request_presigner_1 = require("@aws-sdk/s3-request-presigner");
 const fs = __importStar(require("fs"));
 const path = __importStar(require("path"));
+function parseS3Url(params) {
+    if (params.url) {
+        if (params.url.startsWith('file://') || params.url.startsWith('/')) {
+            params.vendor = 'local';
+            const localPath = params.url.startsWith('file://') ? params.url.slice(7) : params.url;
+            params.bucket = localPath;
+        }
+        else {
+            const parsedUrl = new URL(params.url);
+            const protocol = parsedUrl.protocol.replace(':', '');
+            if (['tos', 'oss', 'cos', 's3'].includes(protocol)) {
+                if (protocol === 'tos')
+                    params.vendor = 'volcengine';
+                else if (protocol === 'oss')
+                    params.vendor = 'alibaba';
+                else if (protocol === 'cos')
+                    params.vendor = 'tencent';
+                else if (protocol === 's3')
+                    params.vendor = 'aws';
+                params.bucket = parsedUrl.hostname;
+                // Prefix is usually parsed separately or passed explicitly for listing
+            }
+            else {
+                throw new Error(`Unsupported protocol: ${protocol}`);
+            }
+        }
+    }
+    if (!params.vendor || !params.bucket) {
+        throw new Error('Could not determine vendor or bucket. Please provide a valid url or vendor/bucket directly.');
+    }
+    return params;
+}
 function createS3Client(params) {
+    params = parseS3Url(params);
     if (params.vendor === 'local')
         return null;
     let endpoint = params.endpoint;
@@ -80,6 +116,7 @@ function createS3Client(params) {
     });
 }
 async function listS3Objects(params, prefix, maxKeys = 1000, continuationToken) {
+    params = parseS3Url(params);
     if (params.vendor === 'local') {
         const root = params.bucket;
         const prefixPath = prefix && prefix !== '.' ? path.join(root, prefix) : root;
@@ -130,6 +167,7 @@ async function listS3Objects(params, prefix, maxKeys = 1000, continuationToken)
     };
 }
 async function readS3Object(params, key, maxBytes) {
+    params = parseS3Url(params);
     if (params.vendor === 'local') {
         const fullPath = path.join(params.bucket, key);
         if (maxBytes) {
@@ -165,3 +203,19 @@ async function readS3Object(params, key, maxBytes) {
     }
     return Buffer.alloc(0);
 }
+async function getPresignedUrl(params, key, expiresIn = 3600) {
+    params = parseS3Url(params);
+    if (params.vendor === 'local') {
+        const fullPath = path.join(params.bucket, key);
+        return `file://${fullPath}`;
+    }
+    const client = createS3Client(params);
+    if (!client)
+        throw new Error('Failed to create S3 client');
+    const command = new client_s3_1.GetObjectCommand({
+        Bucket: params.bucket,
+        Key: key
+    });
+    const signedUrl = await (0, s3_request_presigner_1.getSignedUrl)(client, command, { expiresIn });
+    return signedUrl;
+}