@databricks/zerobus-ingest-sdk 0.0.1 → 0.1.0

package/build.rs

@@ -0,0 +1,5 @@
+extern crate napi_build;
+
+fn main() {
+    napi_build::setup();
+}

package/index.d.ts

@@ -0,0 +1,387 @@
+/* tslint:disable */
+/* eslint-disable */
+
+/* auto-generated by NAPI-RS */
+
+/**
+ * Record serialization format.
+ *
+ * Specifies how records should be encoded when ingested into the stream.
+ */
+export const enum RecordType {
+  /** JSON encoding - records are JSON-encoded strings */
+  Json = 0,
+  /** Protocol Buffers encoding - records are binary protobuf messages */
+  Proto = 1
+}
+/**
+ * Configuration options for the Zerobus stream.
+ *
+ * These options control stream behavior including recovery, timeouts, and inflight limits.
+ */
+export interface StreamConfigurationOptions {
+  /**
+   * Maximum number of unacknowledged requests that can be in flight.
+   * Default: 10,000
+   */
+  maxInflightRequests?: number
+  /**
+   * Enable automatic stream recovery on transient failures.
+   * Default: true
+   */
+  recovery?: boolean
+  /**
+   * Timeout for recovery operations in milliseconds.
+   * Default: 15,000 (15 seconds)
+   */
+  recoveryTimeoutMs?: number
+  /**
+   * Delay between recovery retry attempts in milliseconds.
+   * Default: 2,000 (2 seconds)
+   */
+  recoveryBackoffMs?: number
+  /**
+   * Maximum number of recovery attempts before giving up.
+   * Default: 4
+   */
+  recoveryRetries?: number
+  /**
+   * Timeout for flush operations in milliseconds.
+   * Default: 300,000 (5 minutes)
+   */
+  flushTimeoutMs?: number
+  /**
+   * Timeout waiting for server acknowledgments in milliseconds.
+   * Default: 60,000 (1 minute)
+   */
+  serverLackOfAckTimeoutMs?: number
+  /**
+   * Record serialization format.
+   * Use RecordType.Json for JSON encoding or RecordType.Proto for Protocol Buffers.
+   * Default: RecordType.Proto (Protocol Buffers)
+   */
+  recordType?: number
+}
+/**
+ * Properties of the target Delta table for ingestion.
+ *
+ * Specifies which Unity Catalog table to write to and optionally the schema descriptor
+ * for Protocol Buffers encoding.
+ */
+export interface TableProperties {
+  /** Full table name in Unity Catalog (e.g., "catalog.schema.table") */
+  tableName: string
+  /**
+   * Optional Protocol Buffer descriptor as a base64-encoded string.
+   * If not provided, JSON encoding will be used.
+   */
+  descriptorProto?: string
+}
+/**
+ * JavaScript headers provider callback wrapper.
+ *
+ * Allows TypeScript code to provide custom authentication headers
+ * by implementing a getHeaders() function.
+ */
+export interface JsHeadersProvider {
+  /** JavaScript function: () => Promise<Array<[string, string]>> */
+  getHeadersCallback: (...args: any[]) => any
+}
+/**
+ * Custom error type for Zerobus operations.
+ *
+ * This error type includes information about whether the error is retryable,
+ * which helps determine if automatic recovery can resolve the issue.
+ */
+export declare class ZerobusError {
+  /** Returns true if this error can be automatically retried by the SDK. */
+  get isRetryable(): boolean
+  /** Get the error message. */
+  get message(): string
+}
+/**
+ * A stream for ingesting data into a Databricks Delta table.
+ *
+ * The stream manages a bidirectional gRPC connection, handles acknowledgments,
+ * and provides automatic recovery on transient failures.
+ *
+ * # Example
+ *
+ * ```typescript
+ * const stream = await sdk.createStream(tableProps, clientId, clientSecret, options);
+ * const ackPromise = await stream.ingestRecord(Buffer.from([1, 2, 3]));
+ * const offset = await ackPromise;
+ * await stream.close();
+ * ```
+ */
+export declare class ZerobusStream {
+  /**
+   * Ingests a single record into the stream.
+   *
+   * This method accepts either:
+   * - A Protocol Buffer encoded record as a Buffer (Vec<u8>)
+   * - A JSON string
+   *
+   * This method BLOCKS until the record is sent to the SDK's internal landing zone,
+   * then returns a Promise for the server acknowledgment. This allows you to send
+   * many records immediately without waiting for acknowledgments:
+   *
+   * ```typescript
+   * let lastAckPromise;
+   * for (let i = 0; i < 1000; i++) {
+   *     // This call blocks until record is sent (in SDK)
+   *     lastAckPromise = stream.ingestRecord(record);
+   * }
+   * // All 1000 records are now in the SDK's internal queue
+   * // Wait for the last acknowledgment
+   * await lastAckPromise;
+   * // Flush to ensure all records are acknowledged
+   * await stream.flush();
+   * ```
+   *
+   * # Arguments
+   *
+   * * `payload` - The record data. Accepts:
+   *   - Buffer (low-level proto bytes)
+   *   - string (low-level JSON string)
+   *   - Protobuf message object with .encode() method (high-level, auto-serializes)
+   *   - Plain JavaScript object (high-level, auto-stringifies to JSON)
+   *
+   * # Returns
+   *
+   * A Promise that resolves to the offset ID when the server acknowledges the record.
+   */
+  ingestRecord(payload: unknown): Promise<bigint>
+  /**
+   * Ingests multiple records as a single atomic batch.
+   *
+   * This method accepts an array of records (Protocol Buffer buffers or JSON strings)
+   * and ingests them as a batch. The batch receives a single acknowledgment from
+   * the server with all-or-nothing semantics.
+   *
+   * Similar to ingestRecord(), this BLOCKS until the batch is sent to the SDK's
+   * internal landing zone, then returns a Promise for the server acknowledgment.
+   *
+   * # Arguments
+   *
+   * * `records` - Array of record data (Buffer for protobuf, string for JSON)
+   *
+   * # Returns
+   *
+   * Promise resolving to:
+   * - `bigint`: offset ID for non-empty batches
+   * - `null`: for empty batches
+   *
+   * # Example
+   *
+   * ```typescript
+   * const buffers = records.map(r => Buffer.from(encode(r)));
+   * const offsetId = await stream.ingestRecords(buffers);
+   *
+   * if (offsetId !== null) {
+   *   console.log('Batch acknowledged at offset:', offsetId);
+   * }
+   * ```
+   */
+  ingestRecords(records: Array<unknown>): Promise<bigint | null>
+  /**
+   * Flushes all pending records and waits for acknowledgments.
+   *
+   * This method ensures all previously ingested records have been sent to the server
+   * and acknowledged. It's useful for checkpointing or ensuring data durability.
+   *
+   * # Errors
+   *
+   * - Timeout errors if flush takes longer than configured timeout
+   * - Network errors if the connection fails during flush
+   */
+  flush(): Promise<void>
+  /**
+   * Closes the stream gracefully.
+   *
+   * This method flushes all pending records, waits for acknowledgments, and then
+   * closes the underlying gRPC connection. Always call this method when done with
+   * the stream to ensure data integrity.
+   *
+   * # Errors
+   *
+   * - Returns an error if some records could not be acknowledged
+   * - Network errors during the close operation
+   */
+  close(): Promise<void>
+  /**
+   * Gets the list of unacknowledged records.
+   *
+   * This method should only be called after a stream failure to retrieve records
+   * that were sent but not acknowledged by the server. These records can be
+   * re-ingested into a new stream.
+   *
+   * # Returns
+   *
+   * An array of Buffers containing the unacknowledged record payloads.
+   */
+  getUnackedRecords(): Promise<Array<Buffer>>
+  /**
+   * Gets unacknowledged records grouped by their original batches.
+   *
+   * This preserves the batch structure from ingestion:
+   * - Each ingestRecord() call → 1-element batch
+   * - Each ingestRecords() call → N-element batch
+   *
+   * Should only be called after stream failure. All records returned as Buffers
+   * (JSON strings are converted to UTF-8 bytes).
+   *
+   * # Returns
+   *
+   * Array of batches, where each batch is an array of Buffers
+   *
+   * # Example
+   *
+   * ```typescript
+   * try {
+   *   await stream.ingestRecords(batch1);
+   *   await stream.ingestRecords(batch2);
+   * } catch (error) {
+   *   const unackedBatches = await stream.getUnackedBatches();
+   *
+   *   // Re-ingest with new stream
+   *   for (const batch of unackedBatches) {
+   *     await newStream.ingestRecords(batch);
+   *   }
+   * }
+   * ```
+   */
+  getUnackedBatches(): Promise<Array<Array<Buffer>>>
+}
+/**
+ * The main SDK for interacting with the Databricks Zerobus service.
+ *
+ * This is the entry point for creating ingestion streams to Delta tables.
+ *
+ * # Example
+ *
+ * ```typescript
+ * const sdk = new ZerobusSdk(
+ *   "https://workspace-id.zerobus.region.cloud.databricks.com",
+ *   "https://workspace.cloud.databricks.com"
+ * );
+ *
+ * const stream = await sdk.createStream(
+ *   { tableName: "catalog.schema.table" },
+ *   "client-id",
+ *   "client-secret"
+ * );
+ * ```
+ */
+export declare class ZerobusSdk {
+  /**
+   * Creates a new Zerobus SDK instance.
+   *
+   * # Arguments
+   *
+   * * `zerobus_endpoint` - The Zerobus API endpoint URL
+   *   (e.g., "https://workspace-id.zerobus.region.cloud.databricks.com")
+   * * `unity_catalog_url` - The Unity Catalog endpoint URL
+   *   (e.g., "https://workspace.cloud.databricks.com")
+   *
+   * # Errors
+   *
+   * - Invalid endpoint URLs
+   * - Failed to extract workspace ID from the endpoint
+   */
+  constructor(zerobusEndpoint: string, unityCatalogUrl: string)
+  /**
+   * Creates a new ingestion stream to a Delta table.
+   *
+   * This method establishes a bidirectional gRPC connection to the Zerobus service
+   * and prepares it for data ingestion. By default, it uses OAuth 2.0 Client Credentials
+   * authentication. For custom authentication (e.g., Personal Access Tokens), provide
+   * a custom headers_provider.
+   *
+   * # Arguments
+   *
+   * * `table_properties` - Properties of the target table including name and optional schema
+   * * `client_id` - OAuth 2.0 client ID (ignored if headers_provider is provided)
+   * * `client_secret` - OAuth 2.0 client secret (ignored if headers_provider is provided)
+   * * `options` - Optional stream configuration (uses defaults if not provided)
+   * * `headers_provider` - Optional custom headers provider for authentication.
+   *   If not provided, uses OAuth with client_id and client_secret.
+   *
+   * # Returns
+   *
+   * A Promise that resolves to a ZerobusStream ready for data ingestion.
+   *
+   * # Errors
+   *
+   * - Authentication failures (invalid credentials)
+   * - Invalid table name or insufficient permissions
+   * - Network connectivity issues
+   * - Schema validation errors
+   *
+   * # Examples
+   *
+   * OAuth authentication (default):
+   * ```typescript
+   * const stream = await sdk.createStream(
+   *   { tableName: "catalog.schema.table" },
+   *   "client-id",
+   *   "client-secret"
+   * );
+   * ```
+   *
+   * Custom authentication with headers provider:
+   * ```typescript
+   * const headersProvider = {
+   *   getHeadersCallback: async () => [
+   *     ["authorization", `Bearer ${myToken}`],
+   *     ["x-databricks-zerobus-table-name", tableName]
+   *   ]
+   * };
+   * const stream = await sdk.createStream(
+   *   { tableName: "catalog.schema.table" },
+   *   "", // ignored
+   *   "", // ignored
+   *   undefined,
+   *   headersProvider
+   * );
+   * ```
+   */
+  createStream(tableProperties: TableProperties, clientId: string, clientSecret: string, options?: StreamConfigurationOptions | undefined | null, headersProvider?: JsHeadersProvider | undefined | null): Promise<ZerobusStream>
+  /**
+   * Recreates a stream with the same configuration and re-ingests unacknowledged batches.
+   *
+   * This method is the recommended approach for recovering from stream failures. It:
+   * 1. Retrieves all unacknowledged batches from the failed stream
+   * 2. Creates a new stream with identical configuration
+   * 3. Re-ingests all unacknowledged batches in order
+   * 4. Returns the new stream ready for continued ingestion
+   *
+   * # Arguments
+   *
+   * * `stream` - The failed or closed stream to recreate
+   *
+   * # Returns
+   *
+   * A Promise that resolves to a new ZerobusStream with all unacknowledged batches re-ingested.
+   *
+   * # Errors
+   *
+   * - Failed to retrieve unacknowledged batches from the original stream
+   * - Authentication failures when creating the new stream
+   * - Network connectivity issues during re-ingestion
+   *
+   * # Examples
+   *
+   * ```typescript
+   * try {
+   *   await stream.ingestRecords(batch);
+   * } catch (error) {
+   *   await stream.close();
+   *   // Recreate stream with all unacked batches re-ingested
+   *   const newStream = await sdk.recreateStream(stream);
+   *   // Continue ingesting with newStream
+   * }
+   * ```
+   */
+  recreateStream(stream: ZerobusStream): Promise<ZerobusStream>
+}

package/index.js

@@ -0,0 +1,318 @@
+/* tslint:disable */
+/* eslint-disable */
+/* prettier-ignore */
+
+/* auto-generated by NAPI-RS */
+
+const { existsSync, readFileSync } = require('fs')
+const { join } = require('path')
+
+const { platform, arch } = process
+
+let nativeBinding = null
+let localFileExisted = false
+let loadError = null
+
+function isMusl() {
+  // For Node 10
+  if (!process.report || typeof process.report.getReport !== 'function') {
+    try {
+      const lddPath = require('child_process').execSync('which ldd').toString().trim()
+      return readFileSync(lddPath, 'utf8').includes('musl')
+    } catch (e) {
+      return true
+    }
+  } else {
+    const { glibcVersionRuntime } = process.report.getReport().header
+    return !glibcVersionRuntime
+  }
+}
+
+switch (platform) {
+  case 'android':
+    switch (arch) {
+      case 'arm64':
+        localFileExisted = existsSync(join(__dirname, 'zerobus-sdk-ts.android-arm64.node'))
+        try {
+          if (localFileExisted) {
+            nativeBinding = require('./zerobus-sdk-ts.android-arm64.node')
+          } else {
+            nativeBinding = require('@databricks/zerobus-ingest-sdk-android-arm64')
+          }
+        } catch (e) {
+          loadError = e
+        }
+        break
+      case 'arm':
+        localFileExisted = existsSync(join(__dirname, 'zerobus-sdk-ts.android-arm-eabi.node'))
+        try {
+          if (localFileExisted) {
+            nativeBinding = require('./zerobus-sdk-ts.android-arm-eabi.node')
+          } else {
+            nativeBinding = require('@databricks/zerobus-ingest-sdk-android-arm-eabi')
+          }
+        } catch (e) {
+          loadError = e
+        }
+        break
+      default:
+        throw new Error(`Unsupported architecture on Android ${arch}`)
+    }
+    break
+  case 'win32':
+    switch (arch) {
+      case 'x64':
+        localFileExisted = existsSync(
+          join(__dirname, 'zerobus-sdk-ts.win32-x64-msvc.node')
+        )
+        try {
+          if (localFileExisted) {
+            nativeBinding = require('./zerobus-sdk-ts.win32-x64-msvc.node')
+          } else {
+            nativeBinding = require('@databricks/zerobus-ingest-sdk-win32-x64-msvc')
+          }
+        } catch (e) {
+          loadError = e
+        }
+        break
+      case 'ia32':
+        localFileExisted = existsSync(
+          join(__dirname, 'zerobus-sdk-ts.win32-ia32-msvc.node')
+        )
+        try {
+          if (localFileExisted) {
+            nativeBinding = require('./zerobus-sdk-ts.win32-ia32-msvc.node')
+          } else {
+            nativeBinding = require('@databricks/zerobus-ingest-sdk-win32-ia32-msvc')
+          }
+        } catch (e) {
+          loadError = e
+        }
+        break
+      case 'arm64':
+        localFileExisted = existsSync(
+          join(__dirname, 'zerobus-sdk-ts.win32-arm64-msvc.node')
+        )
+        try {
+          if (localFileExisted) {
+            nativeBinding = require('./zerobus-sdk-ts.win32-arm64-msvc.node')
+          } else {
+            nativeBinding = require('@databricks/zerobus-ingest-sdk-win32-arm64-msvc')
+          }
+        } catch (e) {
+          loadError = e
+        }
+        break
+      default:
+        throw new Error(`Unsupported architecture on Windows: ${arch}`)
+    }
+    break
+  case 'darwin':
+    localFileExisted = existsSync(join(__dirname, 'zerobus-sdk-ts.darwin-universal.node'))
+    try {
+      if (localFileExisted) {
+        nativeBinding = require('./zerobus-sdk-ts.darwin-universal.node')
+      } else {
+        nativeBinding = require('@databricks/zerobus-ingest-sdk-darwin-universal')
+      }
+      break
+    } catch {}
+    switch (arch) {
+      case 'x64':
+        localFileExisted = existsSync(join(__dirname, 'zerobus-sdk-ts.darwin-x64.node'))
+        try {
+          if (localFileExisted) {
+            nativeBinding = require('./zerobus-sdk-ts.darwin-x64.node')
+          } else {
+            nativeBinding = require('@databricks/zerobus-ingest-sdk-darwin-x64')
+          }
+        } catch (e) {
+          loadError = e
+        }
+        break
+      case 'arm64':
+        localFileExisted = existsSync(
+          join(__dirname, 'zerobus-sdk-ts.darwin-arm64.node')
+        )
+        try {
+          if (localFileExisted) {
+            nativeBinding = require('./zerobus-sdk-ts.darwin-arm64.node')
+          } else {
+            nativeBinding = require('@databricks/zerobus-ingest-sdk-darwin-arm64')
+          }
+        } catch (e) {
+          loadError = e
+        }
+        break
+      default:
+        throw new Error(`Unsupported architecture on macOS: ${arch}`)
+    }
+    break
+  case 'freebsd':
+    if (arch !== 'x64') {
+      throw new Error(`Unsupported architecture on FreeBSD: ${arch}`)
+    }
+    localFileExisted = existsSync(join(__dirname, 'zerobus-sdk-ts.freebsd-x64.node'))
+    try {
+      if (localFileExisted) {
+        nativeBinding = require('./zerobus-sdk-ts.freebsd-x64.node')
+      } else {
+        nativeBinding = require('@databricks/zerobus-ingest-sdk-freebsd-x64')
+      }
+    } catch (e) {
+      loadError = e
+    }
+    break
+  case 'linux':
+    switch (arch) {
+      case 'x64':
+        if (isMusl()) {
+          localFileExisted = existsSync(
+            join(__dirname, 'zerobus-sdk-ts.linux-x64-musl.node')
+          )
+          try {
+            if (localFileExisted) {
+              nativeBinding = require('./zerobus-sdk-ts.linux-x64-musl.node')
+            } else {
+              nativeBinding = require('@databricks/zerobus-ingest-sdk-linux-x64-musl')
+            }
+          } catch (e) {
+            loadError = e
+          }
+        } else {
+          localFileExisted = existsSync(
+            join(__dirname, 'zerobus-sdk-ts.linux-x64-gnu.node')
+          )
+          try {
+            if (localFileExisted) {
+              nativeBinding = require('./zerobus-sdk-ts.linux-x64-gnu.node')
+            } else {
+              nativeBinding = require('@databricks/zerobus-ingest-sdk-linux-x64-gnu')
+            }
+          } catch (e) {
+            loadError = e
+          }
+        }
+        break
+      case 'arm64':
+        if (isMusl()) {
+          localFileExisted = existsSync(
+            join(__dirname, 'zerobus-sdk-ts.linux-arm64-musl.node')
+          )
+          try {
+            if (localFileExisted) {
+              nativeBinding = require('./zerobus-sdk-ts.linux-arm64-musl.node')
+            } else {
+              nativeBinding = require('@databricks/zerobus-ingest-sdk-linux-arm64-musl')
+            }
+          } catch (e) {
+            loadError = e
+          }
+        } else {
+          localFileExisted = existsSync(
+            join(__dirname, 'zerobus-sdk-ts.linux-arm64-gnu.node')
+          )
+          try {
+            if (localFileExisted) {
+              nativeBinding = require('./zerobus-sdk-ts.linux-arm64-gnu.node')
+            } else {
+              nativeBinding = require('@databricks/zerobus-ingest-sdk-linux-arm64-gnu')
+            }
+          } catch (e) {
+            loadError = e
+          }
+        }
+        break
+      case 'arm':
+        if (isMusl()) {
+          localFileExisted = existsSync(
+            join(__dirname, 'zerobus-sdk-ts.linux-arm-musleabihf.node')
+          )
+          try {
+            if (localFileExisted) {
+              nativeBinding = require('./zerobus-sdk-ts.linux-arm-musleabihf.node')
+            } else {
+              nativeBinding = require('@databricks/zerobus-ingest-sdk-linux-arm-musleabihf')
+            }
+          } catch (e) {
+            loadError = e
+          }
+        } else {
+          localFileExisted = existsSync(
+            join(__dirname, 'zerobus-sdk-ts.linux-arm-gnueabihf.node')
+          )
+          try {
+            if (localFileExisted) {
+              nativeBinding = require('./zerobus-sdk-ts.linux-arm-gnueabihf.node')
+            } else {
+              nativeBinding = require('@databricks/zerobus-ingest-sdk-linux-arm-gnueabihf')
+            }
+          } catch (e) {
+            loadError = e
+          }
+        }
+        break
+      case 'riscv64':
+        if (isMusl()) {
+          localFileExisted = existsSync(
+            join(__dirname, 'zerobus-sdk-ts.linux-riscv64-musl.node')
+          )
+          try {
+            if (localFileExisted) {
+              nativeBinding = require('./zerobus-sdk-ts.linux-riscv64-musl.node')
+            } else {
+              nativeBinding = require('@databricks/zerobus-ingest-sdk-linux-riscv64-musl')
+            }
+          } catch (e) {
+            loadError = e
+          }
+        } else {
+          localFileExisted = existsSync(
+            join(__dirname, 'zerobus-sdk-ts.linux-riscv64-gnu.node')
+          )
+          try {
+            if (localFileExisted) {
+              nativeBinding = require('./zerobus-sdk-ts.linux-riscv64-gnu.node')
+            } else {
+              nativeBinding = require('@databricks/zerobus-ingest-sdk-linux-riscv64-gnu')
+            }
+          } catch (e) {
+            loadError = e
+          }
+        }
+        break
+      case 's390x':
+        localFileExisted = existsSync(
+          join(__dirname, 'zerobus-sdk-ts.linux-s390x-gnu.node')
+        )
+        try {
+          if (localFileExisted) {
+            nativeBinding = require('./zerobus-sdk-ts.linux-s390x-gnu.node')
+          } else {
+            nativeBinding = require('@databricks/zerobus-ingest-sdk-linux-s390x-gnu')
+          }
+        } catch (e) {
+          loadError = e
+        }
+        break
+      default:
+        throw new Error(`Unsupported architecture on Linux: ${arch}`)
+    }
+    break
+  default:
+    throw new Error(`Unsupported OS: ${platform}, architecture: ${arch}`)
+}
+
+if (!nativeBinding) {
+  if (loadError) {
+    throw loadError
+  }
+  throw new Error(`Failed to load native binding`)
+}
+
+const { RecordType, ZerobusError, ZerobusStream, ZerobusSdk } = nativeBinding
+
+module.exports.RecordType = RecordType
+module.exports.ZerobusError = ZerobusError
+module.exports.ZerobusStream = ZerobusStream
+module.exports.ZerobusSdk = ZerobusSdk