Rename IcebergMetadata to TableMetadata

Author: platypii

src/iceberg.js

@@ -13,13 +13,13 @@ export { avroData } from './avro.data.js'
  * Reads data from the Iceberg table with optional row-level delete processing.
  * Row indices are zero-based and rowEnd is exclusive.
  *
- * @import {IcebergMetadata, Schema} from '../src/types.js'
+ * @import {TableMetadata, Schema} from '../src/types.js'
  * @param {object} options
  * @param {string} options.tableUrl - Base S3 URL of the table.
  * @param {number} [options.rowStart] - The starting global row index to fetch (inclusive).
  * @param {number} [options.rowEnd] - The ending global row index to fetch (exclusive).
  * @param {string} [options.metadataFileName] - Name of the Iceberg metadata file.
- * @param {IcebergMetadata} [options.metadata] - Pre-fetched Iceberg metadata.
+ * @param {TableMetadata} [options.metadata] - Pre-fetched Iceberg metadata.
  * @param {RequestInit} [options.requestInit] - Optional fetch request initialization.
  * @returns {Promise<Array<Record<string, any>>>} Array of data records.
  */

src/iceberg.manifest.js

@@ -3,9 +3,9 @@ import { fetchAvroRecords } from './iceberg.fetch.js'
 /**
  * Returns manifest entries for the current snapshot.
  *
- * @import {IcebergMetadata, Manifest, ManifestEntry} from '../src/types.js'
+ * @import {TableMetadata, Manifest, ManifestEntry} from '../src/types.js'
  * @typedef {{ url: string, entries: ManifestEntry[] }[]} ManifestList
- * @param {IcebergMetadata} metadata
+ * @param {TableMetadata} metadata
  * @param {RequestInit} [requestInit]
  * @returns {Promise<ManifestList>}
  */

src/iceberg.metadata.js

@@ -82,12 +82,12 @@ export function icebergListVersions({ tableUrl, requestInit }) {
  * Fetches the iceberg table metadata.
  * If metadataFileName is not provided, uses icebergLatestVersion to get the version hint.
  *
- * @import {IcebergMetadata} from '../src/types.js'
+ * @import {TableMetadata} from '../src/types.js'
  * @param {object} options
  * @param {string} options.tableUrl - Base URL of the table (e.g. "s3://my-bucket/path/to/table")
  * @param {string} [options.metadataFileName] - Name of the metadata JSON file
  * @param {RequestInit} [options.requestInit] - Optional fetch request initialization
- * @returns {Promise<IcebergMetadata>} The table metadata as a JSON object
+ * @returns {Promise<TableMetadata>} The table metadata as a JSON object
  */
 export async function icebergMetadata({ tableUrl, metadataFileName, requestInit }) {
   if (!metadataFileName) {

src/types.d.ts

@@ -1,5 +1,5 @@
 
-export interface IcebergMetadata {
+export interface TableMetadata {
   'format-version': number
   'table-uuid': string
   location: string
@@ -11,31 +11,34 @@ export interface IcebergMetadata {
   'default-spec-id': number
   'partition-specs': PartitionSpec[]
   'last-partition-id': number
-  'default-sort-order-id': number
-  'sort-orders': SortOrder[]
-  properties: object
+  properties: Record<string, string>
   'current-snapshot-id': number
-  refs: object
   snapshots: Snapshot[]
-  statistics: TableStatistics[]
   'snapshot-log': SnapshotLog[]
   'metadata-log': MetadataLog[]
+  'sort-orders': SortOrder[]
+  'default-sort-order-id': number
+  refs: object
+  statistics: TableStatistics[]
+  'partition-statistics'?: PartitionStatistics[]
+  'next-row-id'?: bigint
 }
 
 export interface Schema {
-  type: string
+  type: 'struct'
   'schema-id': number
+  'identifier-field-ids'?: number[]
   fields: Field[]
 }
 
 interface Field {
   id: number
   name: string
   required: boolean
-  type: string
+  type: string // TODO
   doc?: string
-  'initial-default': string
-  'write-default': string
+  'initial-default': any
+  'write-default': any
 }
 
 export interface PartitionSpec {
@@ -56,20 +59,34 @@ export type PartitionTransform =
   'year' |
   'month' |
   'day' |
-  'hour'
+  'hour' |
+  'void'
+interface PartitionStatistics {
+  'snapshot-id': bigint
+  'statistics-path': string
+  'file-size-in-bytes': bigint
+}
 
 interface SortOrder {
   'order-id': number
-  'fields': unknown[]
+  'fields': SortField[]
+}
+interface SortField {
+  transform: string
+  'source-id'?: number
+  'source-ids'?: number[] // V3
+  'direction': 'asc' | 'desc'
+  'null-order': 'nulls-first' | 'nulls-last'
 }
 
 export interface Snapshot {
-  'sequence-number': number
   'snapshot-id': number
+  'parent-snapshot-id'?: number
+  'sequence-number': number
   'timestamp-ms': number
   summary: {
     operation: string
-    'spark.app.id': string
+    // 'spark.app.id'?: string
     'added-data-files': string
     'added-records': string
     'added-files-size': string
@@ -111,13 +128,15 @@ export interface Manifest {
   sequence_number?: bigint
   min_sequence_number?: bigint
   added_snapshot_id: bigint
-  added_data_files_count: number
-  existing_data_files_count: number
-  deleted_data_files_count: number
+  added_files_count: number
+  existing_files_count: number
+  deleted_files_count: number
   added_rows_count: bigint
   existing_rows_count: bigint
   deleted_rows_count: bigint
   partitions?: FieldSummary[]
+  // key_metadata?: unknown
+  first_row_id?: bigint
 }
 
 export interface ManifestEntry {
@@ -130,18 +149,25 @@ export interface ManifestEntry {
 
 interface FieldSummary {
   'contains-null': boolean
+  'contains-nan'?: boolean
+  'lower-bound'?: unknown
+  'upper-bound'?: unknown
 }
 
 export interface DataFile {
   content: 0 | 1 | 2 // 0=data, 1=position_delete, 2=equality_delete
   file_path: string
   file_format: string
-  partition: Record<number, PartitionSpec> // indexed by field id
+  partition: Record<number, unknown> // indexed by field id
   record_count: bigint
   file_size_in_bytes: bigint
   split_offsets: bigint[]
   equality_ids?: number[]
-  sort_order_id: number
+  sort_order_id?: number
+  first_row_id?: bigint
+  referenced_data_file?: string
+  content_offset?: bigint
+  content_size_in_bytes?: bigint
 }
 
 export interface FilePositionDelete {