@elevasis/core 0.7.0 → 0.8.0

package/src/execution/engine/tools/integration/server/adapters/apify/fetch/get-dataset-items/index.ts

@@ -1,92 +1,92 @@
-import { ToolingError } from '../../../../../../types'
-import { withRetry, createHttpError, DEFAULT_RETRY_POLICY } from '../../../../../../../../../platform/resilience'
-import type { ExecutionContext } from '../../../../../../../base/types'
-
-interface ApifyCredentials {
-  apiToken: string
-}
-
-export interface GetDatasetItemsParams {
-  datasetId: string
-  maxItems?: number
-  offset?: number
-}
-
-export interface GetDatasetItemsResult {
-  items: unknown[]
-  totalCount: number
-}
-
-const APIFY_API_BASE = 'https://api.apify.com/v2'
-
-/**
- * Fetch items from an Apify dataset
- *
- * Retrieves items from a dataset by ID. Useful for:
- * - Recovering data after a runActor timeout (actor keeps running on Apify)
- * - Fetching results from a previously completed run
- * - Paginating through large datasets with offset/maxItems
- *
- * @param credentials - Apify API token
- * @param params - Dataset ID and optional pagination
- * @param context - Execution context for logging
- * @returns Dataset items and total count
- */
-export async function getDatasetItems(
-  credentials: ApifyCredentials,
-  params: GetDatasetItemsParams,
-  context?: ExecutionContext
-): Promise<GetDatasetItemsResult> {
-  const { apiToken } = credentials
-  const { datasetId, maxItems, offset } = params
-
-  if (!datasetId) {
-    throw new ToolingError('validation_error', 'Missing required parameter: datasetId', {
-      integration: 'apify',
-      method: 'getDatasetItems'
-    })
-  }
-
-  let datasetUrl = `${APIFY_API_BASE}/datasets/${datasetId}/items?format=json`
-  if (maxItems) {
-    datasetUrl += `&limit=${maxItems}`
-  }
-  if (offset) {
-    datasetUrl += `&offset=${offset}`
-  }
-
-  if (context?.logger) {
-    context.logger.info(
-      `[ApifyAdapter] Fetching dataset items: organizationId=${context.organizationId} executionId=${context.executionId} datasetId=${datasetId} maxItems=${maxItems} offset=${offset}`
-    )
-  }
-
-  const datasetResponse = await withRetry(async () => {
-    const response = await fetch(datasetUrl, {
-      headers: { Authorization: `Bearer ${apiToken}` }
-    })
-
-    if (!response.ok) {
-      throw await createHttpError(response, {
-        integration: 'apify',
-        method: 'getDatasetItems',
-        datasetId
-      })
-    }
-
-    return response.json()
-  }, DEFAULT_RETRY_POLICY)
-
-  const items = Array.isArray(datasetResponse) ? datasetResponse : []
-
-  if (context?.logger) {
-    context.logger.info(
-      `[ApifyAdapter] Dataset items fetched: organizationId=${context.organizationId} executionId=${context.executionId} datasetId=${datasetId} itemCount=${items.length}`
-    )
-  }
-
-  return {
-    items,
-    totalCount: items.length
-  }
-}
+import { ToolingError } from '../../../../../../types'
+import { withRetry, createHttpError, DEFAULT_RETRY_POLICY } from '../../../../../../../../../platform/resilience'
+import type { ExecutionContext } from '../../../../../../../base/types'
+
+interface ApifyCredentials {
+  apiToken: string
+}
+
+export interface GetDatasetItemsParams {
+  datasetId: string
+  maxItems?: number
+  offset?: number
+}
+
+export interface GetDatasetItemsResult {
+  items: unknown[]
+  totalCount: number
+}
+
+const APIFY_API_BASE = 'https://api.apify.com/v2'
+
+/**
+ * Fetch items from an Apify dataset
+ *
+ * Retrieves items from a dataset by ID. Useful for:
+ * - Recovering data after a runActor timeout (actor keeps running on Apify)
+ * - Fetching results from a previously completed run
+ * - Paginating through large datasets with offset/maxItems
+ *
+ * @param credentials - Apify API token
+ * @param params - Dataset ID and optional pagination
+ * @param context - Execution context for logging
+ * @returns Dataset items and total count
+ */
+export async function getDatasetItems(
+  credentials: ApifyCredentials,
+  params: GetDatasetItemsParams,
+  context?: ExecutionContext
+): Promise<GetDatasetItemsResult> {
+  const { apiToken } = credentials
+  const { datasetId, maxItems, offset } = params
+
+  if (!datasetId) {
+    throw new ToolingError('validation_error', 'Missing required parameter: datasetId', {
+      integration: 'apify',
+      method: 'getDatasetItems'
+    })
+  }
+
+  let datasetUrl = `${APIFY_API_BASE}/datasets/${datasetId}/items?format=json`
+  if (maxItems) {
+    datasetUrl += `&limit=${maxItems}`
+  }
+  if (offset) {
+    datasetUrl += `&offset=${offset}`
+  }
+
+  if (context?.logger) {
+    context.logger.info(
+      `[ApifyAdapter] Fetching dataset items: organizationId=${context.organizationId} executionId=${context.executionId} datasetId=${datasetId} maxItems=${maxItems} offset=${offset}`
+    )
+  }
+
+  const datasetResponse = await withRetry(async () => {
+    const response = await fetch(datasetUrl, {
+      headers: { Authorization: `Bearer ${apiToken}` }
+    })
+
+    if (!response.ok) {
+      throw await createHttpError(response, {
+        integration: 'apify',
+        method: 'getDatasetItems',
+        datasetId
+      })
+    }
+
+    return response.json()
+  }, DEFAULT_RETRY_POLICY)
+
+  const items = Array.isArray(datasetResponse) ? datasetResponse : []
+
+  if (context?.logger) {
+    context.logger.info(
+      `[ApifyAdapter] Dataset items fetched: organizationId=${context.organizationId} executionId=${context.executionId} datasetId=${datasetId} itemCount=${items.length}`
+    )
+  }
+
+  return {
+    items,
+    totalCount: items.length
+  }
+}

package/src/execution/engine/tools/integration/server/adapters/apify/fetch/run-actor/index.ts

@@ -1,218 +1,218 @@
-import { ToolingError } from '../../../../../../types'
-import { withRetry, createHttpError, DEFAULT_RETRY_POLICY } from '../../../../../../../../../platform/resilience'
-import type { ExecutionContext } from '../../../../../../../base/types'
-
-interface ApifyCredentials {
-  apiToken: string
-}
-
-export interface RunActorParams {
-  actorId: string
-  input?: Record<string, unknown>
-  timeoutSecs?: number
-  pollIntervalSecs?: number
-  maxItems?: number
-}
-
-export interface RunActorResult {
-  status: 'SUCCEEDED' | 'FAILED' | 'ABORTED' | 'TIMED_OUT'
-  runId: string
-  datasetId: string
-  items: unknown[]
-  totalCount: number
-  executionTimeMs: number
-}
-
-const APIFY_API_BASE = 'https://api.apify.com/v2'
-
-/**
- * Run an Apify actor, wait for completion, and fetch results
- *
- * Combined operation that:
- * 1. Starts the actor run with provided input
- * 2. Polls for completion (SUCCEEDED, FAILED, ABORTED, TIMED-OUT)
- * 3. Fetches dataset items when complete
- *
- * TIMEOUT BEHAVIOR:
- * If actor exceeds timeoutSecs, returns status='TIMED_OUT' with empty items array.
- * Workflow developers should handle this by either:
- * - Increasing timeoutSecs for long-running actors
- * - Checking run status manually via Apify console
- * - Retrying with exponential backoff
- *
- * @param credentials - Apify API token
- * @param params - Actor ID, input, timeout, poll interval
- * @param context - Execution context for logging
- * @returns Run status, dataset items, and metadata
- */
-export async function runActor(
-  credentials: ApifyCredentials,
-  params: RunActorParams,
-  context?: ExecutionContext
-): Promise<RunActorResult> {
-  const { apiToken } = credentials
-  const { actorId, input = {}, timeoutSecs = 300, pollIntervalSecs = 10, maxItems } = params
-
-  if (!actorId) {
-    throw new ToolingError('validation_error', 'Missing required parameter: actorId', {
-      integration: 'apify',
-      method: 'runActor'
-    })
-  }
-
-  const startTime = Date.now()
-  const timeoutMs = timeoutSecs * 1000
-
-  // Step 1: Start the actor run
-  // URL-encode the actorId to handle slashes (e.g., "apify/hello-world" -> "apify%2Fhello-world")
-  const encodedActorId = encodeURIComponent(actorId)
-  // Pass timeout as query param to set the actor's server-side execution timeout on Apify
-  const runUrl = `${APIFY_API_BASE}/acts/${encodedActorId}/runs?timeout=${timeoutSecs}`
-
-  const runResponse = await withRetry(async () => {
-    const response = await fetch(runUrl, {
-      method: 'POST',
-      headers: {
-        Authorization: `Bearer ${apiToken}`,
-        'Content-Type': 'application/json'
-      },
-      body: JSON.stringify(input)
-    })
-
-    if (!response.ok) {
-      throw await createHttpError(response, {
-        integration: 'apify',
-        method: 'runActor',
-        actorId
-      })
-    }
-
-    return response.json()
-  }, DEFAULT_RETRY_POLICY)
-
-  const runId = runResponse.data?.id
-  const defaultDatasetId = runResponse.data?.defaultDatasetId
-
-  if (!runId) {
-    throw new ToolingError('api_error', 'No run ID returned from Apify', {
-      integration: 'apify',
-      method: 'runActor',
-      actorId
-    })
-  }
-
-  if (context?.logger) {
-    context.logger.info(
-      `[ApifyAdapter] Actor run started: organizationId=${context.organizationId} executionId=${context.executionId} actorId=${actorId} runId=${runId}`
-    )
-  }
-
-  // Step 2: Poll for completion with retry logic
-  const statusUrl = `${APIFY_API_BASE}/actor-runs/${runId}`
-  let status = 'RUNNING'
-  let datasetId = defaultDatasetId
-
-  while (status === 'RUNNING' || status === 'READY') {
-    // Check timeout
-    if (Date.now() - startTime > timeoutMs) {
-      return {
-        status: 'TIMED_OUT',
-        runId,
-        datasetId: datasetId || '',
-        items: [],
-        totalCount: 0,
-        executionTimeMs: Date.now() - startTime
-      }
-    }
-
-    // Wait before polling
-    await new Promise((resolve) => setTimeout(resolve, pollIntervalSecs * 1000))
-
-    // Check status with retry for transient network errors
-    const statusData = await withRetry(async () => {
-      const response = await fetch(statusUrl, {
-        headers: { Authorization: `Bearer ${apiToken}` }
-      })
-
-      if (!response.ok) {
-        throw await createHttpError(response, {
-          integration: 'apify',
-          method: 'runActor',
-          runId
-        })
-      }
-
-      return response.json()
-    }, DEFAULT_RETRY_POLICY)
-
-    status = statusData.data?.status
-    datasetId = statusData.data?.defaultDatasetId || datasetId
-
-    if (context?.logger) {
-      context.logger.info(
-        `[ApifyAdapter] Polling status: organizationId=${context.organizationId} runId=${runId} status=${status} elapsedMs=${Date.now() - startTime}`
-      )
-    }
-  }
-
-  // Step 3: Handle final status
-  if (status !== 'SUCCEEDED') {
-    // Map Apify status to our result status
-    const resultStatus = status === 'ABORTED' ? 'ABORTED' : 'FAILED'
-    return {
-      status: resultStatus,
-      runId,
-      datasetId: datasetId || '',
-      items: [],
-      totalCount: 0,
-      executionTimeMs: Date.now() - startTime
-    }
-  }
-
-  // Step 4: Fetch dataset items
-  if (!datasetId) {
-    throw new ToolingError('api_error', 'No dataset ID available', {
-      integration: 'apify',
-      method: 'runActor',
-      runId
-    })
-  }
-
-  let datasetUrl = `${APIFY_API_BASE}/datasets/${datasetId}/items?format=json`
-  if (maxItems) {
-    datasetUrl += `&limit=${maxItems}`
-  }
-
-  const datasetResponse = await withRetry(async () => {
-    const response = await fetch(datasetUrl, {
-      headers: { Authorization: `Bearer ${apiToken}` }
-    })
-
-    if (!response.ok) {
-      throw await createHttpError(response, {
-        integration: 'apify',
-        method: 'runActor',
-        datasetId
-      })
-    }
-
-    return response.json()
-  }, DEFAULT_RETRY_POLICY)
-
-  const items = Array.isArray(datasetResponse) ? datasetResponse : []
-
-  if (context?.logger) {
-    context.logger.info(
-      `[ApifyAdapter] Actor run completed: organizationId=${context.organizationId} executionId=${context.executionId} actorId=${actorId} runId=${runId} itemCount=${items.length} executionTimeMs=${Date.now() - startTime}`
-    )
-  }
-
-  return {
-    status: 'SUCCEEDED',
-    runId,
-    datasetId,
-    items,
-    totalCount: items.length,
-    executionTimeMs: Date.now() - startTime
-  }
-}
+import { ToolingError } from '../../../../../../types'
+import { withRetry, createHttpError, DEFAULT_RETRY_POLICY } from '../../../../../../../../../platform/resilience'
+import type { ExecutionContext } from '../../../../../../../base/types'
+
+interface ApifyCredentials {
+  apiToken: string
+}
+
+export interface RunActorParams {
+  actorId: string
+  input?: Record<string, unknown>
+  timeoutSecs?: number
+  pollIntervalSecs?: number
+  maxItems?: number
+}
+
+export interface RunActorResult {
+  status: 'SUCCEEDED' | 'FAILED' | 'ABORTED' | 'TIMED_OUT'
+  runId: string
+  datasetId: string
+  items: unknown[]
+  totalCount: number
+  executionTimeMs: number
+}
+
+const APIFY_API_BASE = 'https://api.apify.com/v2'
+
+/**
+ * Run an Apify actor, wait for completion, and fetch results
+ *
+ * Combined operation that:
+ * 1. Starts the actor run with provided input
+ * 2. Polls for completion (SUCCEEDED, FAILED, ABORTED, TIMED-OUT)
+ * 3. Fetches dataset items when complete
+ *
+ * TIMEOUT BEHAVIOR:
+ * If actor exceeds timeoutSecs, returns status='TIMED_OUT' with empty items array.
+ * Workflow developers should handle this by either:
+ * - Increasing timeoutSecs for long-running actors
+ * - Checking run status manually via Apify console
+ * - Retrying with exponential backoff
+ *
+ * @param credentials - Apify API token
+ * @param params - Actor ID, input, timeout, poll interval
+ * @param context - Execution context for logging
+ * @returns Run status, dataset items, and metadata
+ */
+export async function runActor(
+  credentials: ApifyCredentials,
+  params: RunActorParams,
+  context?: ExecutionContext
+): Promise<RunActorResult> {
+  const { apiToken } = credentials
+  const { actorId, input = {}, timeoutSecs = 300, pollIntervalSecs = 10, maxItems } = params
+
+  if (!actorId) {
+    throw new ToolingError('validation_error', 'Missing required parameter: actorId', {
+      integration: 'apify',
+      method: 'runActor'
+    })
+  }
+
+  const startTime = Date.now()
+  const timeoutMs = timeoutSecs * 1000
+
+  // Step 1: Start the actor run
+  // URL-encode the actorId to handle slashes (e.g., "apify/hello-world" -> "apify%2Fhello-world")
+  const encodedActorId = encodeURIComponent(actorId)
+  // Pass timeout as query param to set the actor's server-side execution timeout on Apify
+  const runUrl = `${APIFY_API_BASE}/acts/${encodedActorId}/runs?timeout=${timeoutSecs}`
+
+  const runResponse = await withRetry(async () => {
+    const response = await fetch(runUrl, {
+      method: 'POST',
+      headers: {
+        Authorization: `Bearer ${apiToken}`,
+        'Content-Type': 'application/json'
+      },
+      body: JSON.stringify(input)
+    })
+
+    if (!response.ok) {
+      throw await createHttpError(response, {
+        integration: 'apify',
+        method: 'runActor',
+        actorId
+      })
+    }
+
+    return response.json()
+  }, DEFAULT_RETRY_POLICY)
+
+  const runId = runResponse.data?.id
+  const defaultDatasetId = runResponse.data?.defaultDatasetId
+
+  if (!runId) {
+    throw new ToolingError('api_error', 'No run ID returned from Apify', {
+      integration: 'apify',
+      method: 'runActor',
+      actorId
+    })
+  }
+
+  if (context?.logger) {
+    context.logger.info(
+      `[ApifyAdapter] Actor run started: organizationId=${context.organizationId} executionId=${context.executionId} actorId=${actorId} runId=${runId}`
+    )
+  }
+
+  // Step 2: Poll for completion with retry logic
+  const statusUrl = `${APIFY_API_BASE}/actor-runs/${runId}`
+  let status = 'RUNNING'
+  let datasetId = defaultDatasetId
+
+  while (status === 'RUNNING' || status === 'READY') {
+    // Check timeout
+    if (Date.now() - startTime > timeoutMs) {
+      return {
+        status: 'TIMED_OUT',
+        runId,
+        datasetId: datasetId || '',
+        items: [],
+        totalCount: 0,
+        executionTimeMs: Date.now() - startTime
+      }
+    }
+
+    // Wait before polling
+    await new Promise((resolve) => setTimeout(resolve, pollIntervalSecs * 1000))
+
+    // Check status with retry for transient network errors
+    const statusData = await withRetry(async () => {
+      const response = await fetch(statusUrl, {
+        headers: { Authorization: `Bearer ${apiToken}` }
+      })
+
+      if (!response.ok) {
+        throw await createHttpError(response, {
+          integration: 'apify',
+          method: 'runActor',
+          runId
+        })
+      }
+
+      return response.json()
+    }, DEFAULT_RETRY_POLICY)
+
+    status = statusData.data?.status
+    datasetId = statusData.data?.defaultDatasetId || datasetId
+
+    if (context?.logger) {
+      context.logger.info(
+        `[ApifyAdapter] Polling status: organizationId=${context.organizationId} runId=${runId} status=${status} elapsedMs=${Date.now() - startTime}`
+      )
+    }
+  }
+
+  // Step 3: Handle final status
+  if (status !== 'SUCCEEDED') {
+    // Map Apify status to our result status
+    const resultStatus = status === 'ABORTED' ? 'ABORTED' : 'FAILED'
+    return {
+      status: resultStatus,
+      runId,
+      datasetId: datasetId || '',
+      items: [],
+      totalCount: 0,
+      executionTimeMs: Date.now() - startTime
+    }
+  }
+
+  // Step 4: Fetch dataset items
+  if (!datasetId) {
+    throw new ToolingError('api_error', 'No dataset ID available', {
+      integration: 'apify',
+      method: 'runActor',
+      runId
+    })
+  }
+
+  let datasetUrl = `${APIFY_API_BASE}/datasets/${datasetId}/items?format=json`
+  if (maxItems) {
+    datasetUrl += `&limit=${maxItems}`
+  }
+
+  const datasetResponse = await withRetry(async () => {
+    const response = await fetch(datasetUrl, {
+      headers: { Authorization: `Bearer ${apiToken}` }
+    })
+
+    if (!response.ok) {
+      throw await createHttpError(response, {
+        integration: 'apify',
+        method: 'runActor',
+        datasetId
+      })
+    }
+
+    return response.json()
+  }, DEFAULT_RETRY_POLICY)
+
+  const items = Array.isArray(datasetResponse) ? datasetResponse : []
+
+  if (context?.logger) {
+    context.logger.info(
+      `[ApifyAdapter] Actor run completed: organizationId=${context.organizationId} executionId=${context.executionId} actorId=${actorId} runId=${runId} itemCount=${items.length} executionTimeMs=${Date.now() - startTime}`
+    )
+  }
+
+  return {
+    status: 'SUCCEEDED',
+    runId,
+    datasetId,
+    items,
+    totalCount: items.length,
+    executionTimeMs: Date.now() - startTime
+  }
+}