@gscdump/engine 0.39.0 → 0.40.2

package/README.md

@@ -37,21 +37,20 @@ Optional peers (install only what your runtime needs):
 | `@gscdump/engine/scope` | Multi-tenant scope predicates. |
 | `@gscdump/engine/arrow` | Apache Arrow utilities for engine result conversion. |
 | `@gscdump/engine/node` | Node-only DuckDB handle. |
-| `@gscdump/engine/node-harness` | Node test harness for engine integration tests. |
 | `@gscdump/engine/filesystem` | Node-only `DataSource` + `ManifestStore` adapters. |
-| `@gscdump/engine/http` | Read-only HTTP `DataSource` (signed URLs, Range requests). |
 | `@gscdump/engine/hyparquet` | Pure-JS `ParquetCodec`. |
 | `@gscdump/engine/r2` | Cloudflare R2 `DataSource` (structurally typed against `R2Bucket`). |
 | `@gscdump/engine/r2-manifest` | R2-backed `ManifestStore` for hosted deployments. |
-| `@gscdump/engine/inspection-sqlite-node` | Node SQLite adapter for URL-inspection cache. |
-| `@gscdump/engine/inspection-sqlite-browser` | Browser (wa-sqlite) adapter for URL-inspection cache. |
+| `@gscdump/engine/iceberg` | Edge-safe GSC Iceberg schema/catalog/append sink. |
+| `@gscdump/engine/sink-node` | Node-only Iceberg overwrite/delete recovery writer. |
+| `@gscdump/engine/source` | Query-source contracts and factories. |
 
 ## Stability
 
 | Surface | Stability |
 |---|---|
 | `createStorageEngine` and storage contracts (`StorageEngine`, `Row`, `WriteCtx`, ...) | Public |
-| Adapters (`/node`, `/filesystem`, `/http`, `/hyparquet`, `/r2`) | Public |
+| Adapters (`/node`, `/filesystem`, `/hyparquet`, `/r2`, `/r2-manifest`) | Public |
 | Planner (`resolveToSQL`, `enumeratePartitions`) | Public |
 | Schema (`SCHEMAS`, `allTables`, ...) | Public |
 | Internals reached through `@gscdump/engine/<file>` paths not listed above | Private — may break between minors |
@@ -60,7 +59,6 @@ Optional peers (install only what your runtime needs):
 
 - [`gscdump`](../gscdump) — REST client + query builder (edge-safe peer dep).
 - [`@gscdump/analysis`](../analysis) — analyzers; consumes `StorageEngine` via `createEngine` factories.
-- [`@gscdump/engine-duckdb-node`](../engine-duckdb-node) — Node DuckDB analyzer adapter.
 - [`@gscdump/engine-duckdb-wasm`](../engine-duckdb-wasm) — DuckDB-WASM browser adapter.
 - [`@gscdump/engine-sqlite`](../engine-sqlite) — SQLite / D1 adapter.
 - [`@gscdump/cli`](../cli) — CLI wrapping engine + analysis.

package/dist/_chunks/errors.d.mts

@@ -103,8 +103,6 @@ declare const engineErrors: {
   readonly lockAcquireTimeout: (scope: string, timeoutMs: number) => EngineError;
 };
 declare function isEngineError(value: unknown): value is EngineError;
-/** The human-readable rendering of an `EngineError`, for logs and string sinks. */
-declare function formatEngineError(error: EngineError): string;
 /**
  * Re-raises an `EngineError` value as a generic `Error`, stashing the union under
  * `.engineError` for stack-walking. Used by the throwing wrappers over the
@@ -114,4 +112,4 @@ declare function formatEngineError(error: EngineError): string;
  * mapper so the class identity callers match on is preserved.
  */
 declare function engineErrorToException(error: EngineError): Error;
-export { EngineError, EngineErrorKind, engineErrorToException, engineErrors, formatEngineError, isEngineError };
+export { EngineError, EngineErrorKind, engineErrorToException, engineErrors, isEngineError };

package/dist/_chunks/index.d.mts

@@ -105,14 +105,5 @@ declare function createEngineQuerySource(options: EngineQuerySourceOptions): Ana
 declare function runAnalyzerWithEngine(deps: {
   engine: StorageEngine;
 }, ctx: TenantCtx, params: AnalysisParams, registry: AnalyzerRegistry): Promise<AnalysisResult>;
-interface TypedQuery<TRow> {
-  state: BuilderState;
-  readonly __row?: TRow;
-}
-declare function typedQuery<TRow>(state: BuilderState): TypedQuery<TRow>;
-declare function queryRows<TRow = QueryRow>(source: AnalysisQuerySource, query: BuilderState | TypedQuery<TRow>): Promise<TRow[]>;
-declare function queryComparisonRows<TRow = QueryRow>(source: AnalysisQuerySource, current: BuilderState | TypedQuery<TRow>, previous: BuilderState | TypedQuery<TRow>): Promise<{
-  current: TRow[];
-  previous: TRow[];
-}>;
-export { AttachedTableMissingError, AttachedTableRunner, AttachedTableSourceOptions, CreateSqlQuerySourceOptions, ENGINE_QUERY_CAPABILITIES, EngineQuerySourceOptions, TypedQuery, createAttachedTableSource, createEngineQuerySource, createSqlQuerySource, queryComparisonRows, queryRows, rewriteForTableSource, runAnalyzerWithEngine, typedQuery };
+declare function queryRows<TRow = QueryRow>(source: AnalysisQuerySource, state: BuilderState): Promise<TRow[]>;
+export { AttachedTableMissingError, AttachedTableRunner, AttachedTableSourceOptions, CreateSqlQuerySourceOptions, ENGINE_QUERY_CAPABILITIES, EngineQuerySourceOptions, createAttachedTableSource, createEngineQuerySource, createSqlQuerySource, queryRows, rewriteForTableSource, runAnalyzerWithEngine };

package/dist/_chunks/index2.d.mts

@@ -30,9 +30,6 @@ declare function defaultStartDate(): string;
 declare function periodOf(params: AnalysisParams): AnalysisPeriod;
 declare function comparisonOf(params: AnalysisParams): ComparisonPeriod;
 declare function resolveWindow(opts: ResolveWindowOptions): ResolvedWindow;
-/** Convert a ResolvedWindow into the AnalysisPeriod / ComparisonPeriod shape. */
-declare function windowToPeriod(w: ResolvedWindow): AnalysisPeriod;
-declare function windowToComparisonPeriod(w: ResolvedWindow): ComparisonPeriod | undefined;
 interface PadTimeseriesOptions<T> {
   /** ISO date (YYYY-MM-DD), inclusive lower bound. */
   startDate: string;
@@ -54,4 +51,4 @@ type DateRowShape = Record<string, unknown> & {
  * once. Existing dates keep all their rows (grouped timeseries safe).
  */
 declare function padTimeseries<T extends DateRowShape = DateRowShape>(rows: readonly T[], options: PadTimeseriesOptions<T>): T[];
-export { AnalysisPeriod, ComparisonMode, ComparisonPeriod, PadTimeseriesOptions, ResolveWindowOptions, ResolvedWindow, WindowPreset, comparisonOf, defaultEndDate, defaultStartDate, padTimeseries, periodOf, resolveWindow, windowToComparisonPeriod, windowToPeriod };
+export { AnalysisPeriod, ComparisonMode, ComparisonPeriod, PadTimeseriesOptions, ResolveWindowOptions, ResolvedWindow, WindowPreset, comparisonOf, defaultEndDate, defaultStartDate, padTimeseries, periodOf, resolveWindow };

package/dist/_chunks/pg-adapter.d.mts

@@ -14,12 +14,6 @@ declare const pgResolverAdapter: ResolverAdapter<PgTableKey>;
  * accidental adapter caching that would lock in a stale `{{FILES}}` set.
  */
 interface ResolverAdapterOptions {
-  /**
-   * Deprecated compatibility flag. Fact tables no longer carry
-   * `query_canonical`; canonical reads derive from `query_dim`, or from a
-   * canonical rollup relation when `queryCanonicalSource: 'column'` is set.
-   */
-  canonicalFallback?: boolean;
   /**
    * `queryDim` reads canonical from a joined query dimension. `column` is only
    * for derived canonical rollup relations whose primary relation already

package/dist/_chunks/resolver.mjs

@@ -455,7 +455,6 @@ function createParquetResolverAdapter(options = {}) {
 	return createResolverAdapter({
 		...PG_BASE_CONFIG,
 		tableLabel: "parquet-resolver-adapter",
-		canonicalFallback: options.canonicalFallback ?? false,
 		queryCanonicalSource: options.queryCanonicalSource ?? "queryDim",
 		tableRef: (tk) => sql.raw(`read_parquet({{FILES}}, union_by_name = true) AS "${tk}"`),
 		queryDimTableRef: () => sql.raw("read_parquet({{QUERY_DIM}}, union_by_name = true) AS \"query_dim\"")
@@ -468,7 +467,6 @@ function createIcebergResolverAdapter(options = {}) {
 		includeSiteId: true,
 		includeSearchType: true,
 		tableLabel: "iceberg-resolver-adapter",
-		canonicalFallback: options.canonicalFallback ?? false,
 		queryCanonicalSource: options.queryCanonicalSource ?? "queryDim",
 		tableRef: (tk) => sql.raw(`"${tk}"`),
 		queryDimTableRef: () => sql.raw("\"query_dim\"")
@@ -481,7 +479,6 @@ function createR2SqlResolverAdapter(options = {}) {
 		includeSiteId: true,
 		includeSearchType: true,
 		tableLabel: "r2-sql-resolver-adapter",
-		canonicalFallback: options.canonicalFallback ?? false,
 		queryCanonicalSource: options.queryCanonicalSource ?? "queryDim",
 		capabilities: {
 			regex: false,

package/dist/_chunks/schema.d.mts

@@ -2115,7 +2115,6 @@ declare const TABLE_METADATA: Record<TableName, {
 }>;
 declare const SCHEMAS: Record<TableName, TableSchema>;
 declare function currentSchemaVersion(table: TableName): number;
-declare function schemaFor(table: TableName): TableSchema;
 /**
  * DATE column names for a table. The single schema-derived source every read
  * path uses to build the legacy-VARCHAR date canonicalization (see
@@ -2146,4 +2145,4 @@ declare function naturalKeyColumns(table: TableName): readonly string[];
  */
 declare function dedupeByNaturalKey(table: TableName, rows: readonly Row[]): Row[];
 declare function dimensionToColumn(dim: string, _table: TableName): string;
-export { type ColumnDef$1 as ColumnDef, type ColumnType, DrizzleSchema, SCHEMAS, TABLE_METADATA, type TableSchema$1 as TableSchema, allTables, countries, currentSchemaVersion, dateColumnsFor, dates, dedupeByNaturalKey, dimensionToColumn, drizzleSchema, hourly_pages, inferTable, naturalKeyColumns, page_queries, pages, queries, schemaFor, search_appearance, search_appearance_page_queries, search_appearance_pages, search_appearance_queries };
+export { type ColumnDef$1 as ColumnDef, type ColumnType, DrizzleSchema, SCHEMAS, TABLE_METADATA, type TableSchema$1 as TableSchema, allTables, countries, currentSchemaVersion, dateColumnsFor, dates, dedupeByNaturalKey, dimensionToColumn, drizzleSchema, hourly_pages, inferTable, naturalKeyColumns, page_queries, pages, queries, search_appearance, search_appearance_page_queries, search_appearance_pages, search_appearance_queries };

package/dist/_chunks/schema.mjs

@@ -219,9 +219,6 @@ const SCHEMAS = Object.fromEntries(METRIC_TABLES.map((t) => [t, tableSchemaFrom(
 function currentSchemaVersion(table) {
 	return SCHEMAS[table].version;
 }
-function schemaFor(table) {
-	return SCHEMAS[table];
-}
 function dateColumnsFor(table) {
 	return SCHEMAS[table].columns.filter((c) => c.type === "DATE").map((c) => c.name);
 }
@@ -259,4 +256,4 @@ function dimensionToColumn(dim, _table) {
 	if (dim === "queryCanonical") return "query_canonical";
 	return dim;
 }
-export { SCHEMAS, TABLE_METADATA, allTables, countries, currentSchemaVersion, dateColumnsFor, dates, dedupeByNaturalKey, dimensionToColumn, drizzleSchema, hourly_pages, inferTable, naturalKeyColumns, page_queries, pages, queries, schemaFor, search_appearance, search_appearance_page_queries, search_appearance_pages, search_appearance_queries };
+export { SCHEMAS, TABLE_METADATA, allTables, countries, currentSchemaVersion, dateColumnsFor, dates, dedupeByNaturalKey, dimensionToColumn, drizzleSchema, hourly_pages, inferTable, naturalKeyColumns, page_queries, pages, queries, search_appearance, search_appearance_page_queries, search_appearance_pages, search_appearance_queries };

package/dist/_chunks/source.mjs

@@ -137,21 +137,7 @@ async function runAnalyzerWithEngine(deps, ctx, params, registry) {
 		searchType: params.searchType ?? "web"
 	}), params, registry);
 }
-function typedQuery(state) {
-	return { state };
-}
-function isTypedQuery(value) {
-	return "state" in value;
-}
-async function queryRows(source, query) {
-	const state = isTypedQuery(query) ? query.state : query;
+async function queryRows(source, state) {
 	return await source.queryRows(state);
 }
-async function queryComparisonRows(source, current, previous) {
-	const [currentRows, previousRows] = await Promise.all([queryRows(source, current), queryRows(source, previous)]);
-	return {
-		current: currentRows,
-		previous: previousRows
-	};
-}
-export { AttachedTableMissingError, ENGINE_QUERY_CAPABILITIES, createAttachedTableSource, createEngineQuerySource, createSqlQuerySource, queryComparisonRows, queryRows, rewriteForTableSource, runAnalyzerWithEngine, typedQuery };
+export { AttachedTableMissingError, ENGINE_QUERY_CAPABILITIES, createAttachedTableSource, createEngineQuerySource, createSqlQuerySource, queryRows, rewriteForTableSource, runAnalyzerWithEngine };

package/dist/errors.d.mts

@@ -1,2 +1,2 @@
-import { EngineError, EngineErrorKind, engineErrorToException, engineErrors, formatEngineError, isEngineError } from "./_chunks/errors.mjs";
-export { EngineError, EngineErrorKind, engineErrorToException, engineErrors, formatEngineError, isEngineError };
+import { EngineError, EngineErrorKind, engineErrorToException, engineErrors, isEngineError } from "./_chunks/errors.mjs";
+export { EngineError, EngineErrorKind, engineErrorToException, engineErrors, isEngineError };

package/dist/errors.mjs

@@ -190,13 +190,10 @@ const ENGINE_ERROR_KINDS = /* @__PURE__ */ new Set([
 function isEngineError(value) {
 	return typeof value === "object" && value !== null && ENGINE_ERROR_KINDS.has(value.kind) && typeof value.message === "string";
 }
-function formatEngineError(error) {
-	return error.message;
-}
 function engineErrorToException(error) {
 	const exception = new Error(error.message);
 	if ("cause" in error && error.cause !== void 0) exception.cause = error.cause;
 	exception.engineError = error;
 	return exception;
 }
-export { engineErrorToException, engineErrors, formatEngineError, isEngineError };
+export { engineErrorToException, engineErrors, isEngineError };

package/dist/index.d.mts

@@ -1,7 +1,7 @@
 import { CodecCtx, CompactionThresholds, CompactionTier, DataSource, EngineOptions, FileSetRef, GcCtx, Grain, ListLiveFilter, LockScope, ManifestEntry, ManifestPurgeResult, ManifestStore, ParquetCodec, PurgeFilter, PurgeResult, PurgeUrlsResult, QueryCtx, QueryExecuteOptions, QueryExecuteResult, QueryExecutor, QueryProfiler, QueryResult, QuerySpan, Row, RunSQLOptions, SearchType, StorageEngine, SyncState, SyncStateDetail, SyncStateFilter, SyncStateKind, SyncStateScope, TableName, TenantCtx, Watermark, WatermarkFilter, WatermarkScope, WriteCtx, WriteResult, enumeratePartitions } from "./_chunks/storage.mjs";
 import { DuckDBFactory, DuckDBHandle, canonicalEmptyParquetSchema, createDuckDBCodec, createDuckDBExecutor } from "./_chunks/duckdb.mjs";
 import { ColumnDef, ColumnType, DrizzleSchema, SCHEMAS, TABLE_METADATA, TableSchema, allTables, countries, currentSchemaVersion, dates, dimensionToColumn, drizzleSchema, hourly_pages, inferTable, page_queries, pages, queries } from "./_chunks/schema.mjs";
-import { EngineError, EngineErrorKind, engineErrorToException, engineErrors, formatEngineError, isEngineError } from "./_chunks/errors.mjs";
+import { EngineError, EngineErrorKind, engineErrorToException, engineErrors, isEngineError } from "./_chunks/errors.mjs";
 import { InspectionVerdict, SchedulePolicy, ScheduleState, fixedPolicy, inspectionPolicy, sitemapPolicy } from "./schedule.mjs";
 import { IcebergTableName, Sink, SinkCapabilities, SinkCloseResult, SinkOptions, SinkSlice, SinkWriteResult } from "./_chunks/sink.mjs";
 import { GscApiRow, IngestOptions, RowAccumulator, RowAccumulatorOptions, assembleDatesRow, createRowAccumulator, toPath, toSumPosition, transformGscRow } from "./ingest.mjs";
@@ -109,7 +109,6 @@ interface CreateIngestAccumulatorOptions extends RowAccumulatorOptions {
   ctx: IngestAccumulatorCtx;
   hooks: IngestAccumulatorHooks;
 }
-declare function createNoopIngestAccumulator(): IngestAccumulator;
 declare function createIngestAccumulator(opts: CreateIngestAccumulatorOptions): IngestAccumulator;
 declare function dayPartition(date: string): string;
 /**
@@ -209,4 +208,4 @@ declare const MIN_SYNC_IMPRESSIONS = 1;
 declare const MIN_COUNTRY_IMPRESSIONS = 10;
 declare const MAX_SITEMAP_URLS_PER_SITE = 50000;
 declare const MAX_TRACKED_URLS_PER_SITE = 200000;
-export { type CodecCtx, type ColumnDef, type ColumnType, type CompactionThresholds, type CompactionTier, type CreateIngestAccumulatorOptions, DEFAULT_SEARCH_TYPE, type DataSource, type DateWeight, type DrizzleSchema, type DuckDBFactory, type DuckDBHandle, ENGINE_QUERY_CAPABILITIES, EngineError, EngineErrorKind, type EngineOptions, FILES_PLACEHOLDER, type FileSetRef, type FinalizeOptions, type FinalizeResult, type GcCtx, type Grain, type GscApiRow, type InMemorySink, type IngestAccumulator, type IngestAccumulatorCtx, type IngestAccumulatorEngine, type IngestAccumulatorHooks, type IngestOptions, type InspectionVerdict, type ListLiveFilter, type LockScope, MAX_DAY_BYTES, MAX_GSC_PAGES_R2, MAX_SITEMAP_URLS_PER_SITE, MAX_TRACKED_URLS_PER_SITE, MIN_COUNTRY_IMPRESSIONS, MIN_SYNC_IMPRESSIONS, type ManifestEntry, type ManifestPurgeResult, type ManifestStore, type ParquetCodec, type PurgeFilter, type PurgeResult, type PurgeUrlsResult, type QueryCtx, type QueryExecuteOptions, type QueryExecuteResult, type QueryExecutor, type QueryProfiler, type QueryResult, type QuerySpan, ROW_LIMIT_R2, type ResolvedQuery, type Row, type RowAccumulator, type RowAccumulatorOptions, type RunSQLOptions, SCHEMAS, type SchedulePolicy, type ScheduleState, type SearchType, type Sink, type SinkCapabilities, type SinkCloseResult, type SinkOptions, type SinkSlice, type SinkWriteResult, type StorageEngine, type StoredRow, type SyncState, type SyncStateDetail, type SyncStateFilter, type SyncStateKind, type SyncStateScope, type SyncTableName, TABLES_BY_SEARCH_TYPE, TABLE_METADATA, TABLE_TIERS, TIER_PRIORITY, type TableName, type TableSchema, type TableTier, type TenantCtx, type TieredTableName, WEIGHT_PRIORITY, type Watermark, type WatermarkFilter, type WatermarkScope, type WriteCtx, type WriteResult, allTables, assembleDatesRow, bindLiterals, canonicalEmptyParquetSchema, coerceRow, coerceRows, collectSpans, countries, createDuckDBCodec, createDuckDBExecutor, createIcebergResolverAdapter, createInMemorySink, createIngestAccumulator, createNoopIngestAccumulator, createParquetResolverAdapter, createQueryProfiler, createR2SqlResolverAdapter, createRowAccumulator, createSqlQuerySource, createStorageEngine, currentSchemaVersion, dates, dayPartition, dimensionToColumn, drizzleSchema, engineErrorToException, engineErrors, enumeratePartitions, fixedPolicy, formatEngineError, formatLiteral, getDateWeight, getTableTier, getTablesForTier, hourPartition, hourly_pages, inferLegacyTier, inferSearchType, inferTable, inspectionPolicy, isEngineError, objectKey, page_queries, pages, parseEnabledSearchTypes, pgResolverAdapter, queries, rebuildDailyFromHourly, resolveParquetSQL, sitemapPolicy, substituteNamedFiles, toPath, toSumPosition, transformGscRow, validateEnabledSearchTypes, validateEnabledSearchTypesResult };
+export { type CodecCtx, type ColumnDef, type ColumnType, type CompactionThresholds, type CompactionTier, type CreateIngestAccumulatorOptions, DEFAULT_SEARCH_TYPE, type DataSource, type DateWeight, type DrizzleSchema, type DuckDBFactory, type DuckDBHandle, ENGINE_QUERY_CAPABILITIES, EngineError, EngineErrorKind, type EngineOptions, FILES_PLACEHOLDER, type FileSetRef, type FinalizeOptions, type FinalizeResult, type GcCtx, type Grain, type GscApiRow, type InMemorySink, type IngestAccumulator, type IngestAccumulatorCtx, type IngestAccumulatorEngine, type IngestAccumulatorHooks, type IngestOptions, type InspectionVerdict, type ListLiveFilter, type LockScope, MAX_DAY_BYTES, MAX_GSC_PAGES_R2, MAX_SITEMAP_URLS_PER_SITE, MAX_TRACKED_URLS_PER_SITE, MIN_COUNTRY_IMPRESSIONS, MIN_SYNC_IMPRESSIONS, type ManifestEntry, type ManifestPurgeResult, type ManifestStore, type ParquetCodec, type PurgeFilter, type PurgeResult, type PurgeUrlsResult, type QueryCtx, type QueryExecuteOptions, type QueryExecuteResult, type QueryExecutor, type QueryProfiler, type QueryResult, type QuerySpan, ROW_LIMIT_R2, type ResolvedQuery, type Row, type RowAccumulator, type RowAccumulatorOptions, type RunSQLOptions, SCHEMAS, type SchedulePolicy, type ScheduleState, type SearchType, type Sink, type SinkCapabilities, type SinkCloseResult, type SinkOptions, type SinkSlice, type SinkWriteResult, type StorageEngine, type StoredRow, type SyncState, type SyncStateDetail, type SyncStateFilter, type SyncStateKind, type SyncStateScope, type SyncTableName, TABLES_BY_SEARCH_TYPE, TABLE_METADATA, TABLE_TIERS, TIER_PRIORITY, type TableName, type TableSchema, type TableTier, type TenantCtx, type TieredTableName, WEIGHT_PRIORITY, type Watermark, type WatermarkFilter, type WatermarkScope, type WriteCtx, type WriteResult, allTables, assembleDatesRow, bindLiterals, canonicalEmptyParquetSchema, coerceRow, coerceRows, collectSpans, countries, createDuckDBCodec, createDuckDBExecutor, createIcebergResolverAdapter, createInMemorySink, createIngestAccumulator, createParquetResolverAdapter, createQueryProfiler, createR2SqlResolverAdapter, createRowAccumulator, createSqlQuerySource, createStorageEngine, currentSchemaVersion, dates, dayPartition, dimensionToColumn, drizzleSchema, engineErrorToException, engineErrors, enumeratePartitions, fixedPolicy, formatLiteral, getDateWeight, getTableTier, getTablesForTier, hourPartition, hourly_pages, inferLegacyTier, inferSearchType, inferTable, inspectionPolicy, isEngineError, objectKey, page_queries, pages, parseEnabledSearchTypes, pgResolverAdapter, queries, rebuildDailyFromHourly, resolveParquetSQL, sitemapPolicy, substituteNamedFiles, toPath, toSumPosition, transformGscRow, validateEnabledSearchTypes, validateEnabledSearchTypesResult };

package/dist/index.mjs

@@ -3,7 +3,7 @@ import { DEFAULT_SEARCH_TYPE, dayPartition, hourPartition, inferLegacyTier, infe
 import { SCHEMAS, TABLE_METADATA, allTables, countries, currentSchemaVersion, dates, dimensionToColumn, drizzleSchema, hourly_pages, inferTable, page_queries, pages, queries } from "./_chunks/schema.mjs";
 import { enumeratePartitions } from "./_chunks/compaction.mjs";
 import { FILES_PLACEHOLDER, resolveParquetSQL, substituteNamedFiles } from "./_chunks/parquet-plan.mjs";
-import { engineErrorToException, engineErrors, formatEngineError, isEngineError } from "./errors.mjs";
+import { engineErrorToException, engineErrors, isEngineError } from "./errors.mjs";
 import { bindLiterals, formatLiteral } from "./sql-bind.mjs";
 import { MAX_DAY_BYTES, canonicalEmptyParquetSchema, createDuckDBCodec, createDuckDBExecutor, createStorageEngine } from "./_chunks/engine.mjs";
 import { assembleDatesRow, createRowAccumulator, toPath, toSumPosition, transformGscRow } from "./ingest.mjs";
@@ -13,12 +13,6 @@ import { rebuildDailyFromHourly } from "./rollups.mjs";
 import { fixedPolicy, inspectionPolicy, sitemapPolicy } from "./schedule.mjs";
 import { ENGINE_QUERY_CAPABILITIES, createSqlQuerySource } from "./_chunks/source.mjs";
 import { err, ok, unwrapResult } from "gscdump/result";
-const NOOP_RESULT = {
-	flushed: 0,
-	recovered: 0,
-	failed: 0,
-	rowsWritten: 0
-};
 function scopeOf(ctx, table, date) {
 	return {
 		userId: String(ctx.userId),
@@ -28,16 +22,6 @@ function scopeOf(ctx, table, date) {
 		...ctx.searchType !== void 0 ? { searchType: ctx.searchType } : {}
 	};
 }
-function createNoopIngestAccumulator() {
-	return {
-		push() {
-			return false;
-		},
-		async finalize() {
-			return NOOP_RESULT;
-		}
-	};
-}
 function createIngestAccumulator(opts) {
 	const { engine, ctx, hooks, ...accOpts } = opts;
 	const acc = createRowAccumulator(accOpts);
@@ -331,4 +315,4 @@ const MIN_SYNC_IMPRESSIONS = 1;
 const MIN_COUNTRY_IMPRESSIONS = 10;
 const MAX_SITEMAP_URLS_PER_SITE = 5e4;
 const MAX_TRACKED_URLS_PER_SITE = 2e5;
-export { DEFAULT_SEARCH_TYPE, ENGINE_QUERY_CAPABILITIES, FILES_PLACEHOLDER, MAX_DAY_BYTES, MAX_GSC_PAGES_R2, MAX_SITEMAP_URLS_PER_SITE, MAX_TRACKED_URLS_PER_SITE, MIN_COUNTRY_IMPRESSIONS, MIN_SYNC_IMPRESSIONS, ROW_LIMIT_R2, SCHEMAS, TABLES_BY_SEARCH_TYPE, TABLE_METADATA, TABLE_TIERS, TIER_PRIORITY, WEIGHT_PRIORITY, allTables, assembleDatesRow, bindLiterals, canonicalEmptyParquetSchema, coerceRow, coerceRows, collectSpans, countries, createDuckDBCodec, createDuckDBExecutor, createIcebergResolverAdapter, createInMemorySink, createIngestAccumulator, createNoopIngestAccumulator, createParquetResolverAdapter, createQueryProfiler, createR2SqlResolverAdapter, createRowAccumulator, createSqlQuerySource, createStorageEngine, currentSchemaVersion, dates, dayPartition, dimensionToColumn, drizzleSchema, engineErrorToException, engineErrors, enumeratePartitions, fixedPolicy, formatEngineError, formatLiteral, getDateWeight, getTableTier, getTablesForTier, hourPartition, hourly_pages, inferLegacyTier, inferSearchType, inferTable, inspectionPolicy, isEngineError, objectKey, page_queries, pages, parseEnabledSearchTypes, pgResolverAdapter, queries, rebuildDailyFromHourly, resolveParquetSQL, sitemapPolicy, substituteNamedFiles, toPath, toSumPosition, transformGscRow, validateEnabledSearchTypes, validateEnabledSearchTypesResult };
+export { DEFAULT_SEARCH_TYPE, ENGINE_QUERY_CAPABILITIES, FILES_PLACEHOLDER, MAX_DAY_BYTES, MAX_GSC_PAGES_R2, MAX_SITEMAP_URLS_PER_SITE, MAX_TRACKED_URLS_PER_SITE, MIN_COUNTRY_IMPRESSIONS, MIN_SYNC_IMPRESSIONS, ROW_LIMIT_R2, SCHEMAS, TABLES_BY_SEARCH_TYPE, TABLE_METADATA, TABLE_TIERS, TIER_PRIORITY, WEIGHT_PRIORITY, allTables, assembleDatesRow, bindLiterals, canonicalEmptyParquetSchema, coerceRow, coerceRows, collectSpans, countries, createDuckDBCodec, createDuckDBExecutor, createIcebergResolverAdapter, createInMemorySink, createIngestAccumulator, createParquetResolverAdapter, createQueryProfiler, createR2SqlResolverAdapter, createRowAccumulator, createSqlQuerySource, createStorageEngine, currentSchemaVersion, dates, dayPartition, dimensionToColumn, drizzleSchema, engineErrorToException, engineErrors, enumeratePartitions, fixedPolicy, formatLiteral, getDateWeight, getTableTier, getTablesForTier, hourPartition, hourly_pages, inferLegacyTier, inferSearchType, inferTable, inspectionPolicy, isEngineError, objectKey, page_queries, pages, parseEnabledSearchTypes, pgResolverAdapter, queries, rebuildDailyFromHourly, resolveParquetSQL, sitemapPolicy, substituteNamedFiles, toPath, toSumPosition, transformGscRow, validateEnabledSearchTypes, validateEnabledSearchTypesResult };

package/dist/ingest.d.mts

@@ -16,8 +16,6 @@ interface GscApiRow {
   position: number;
 }
 interface IngestOptions {
-  /** @deprecated Canonical query data is built through query_dim, not fact rows. */
-  normalizeQuery?: (query: string) => string | null | undefined;
   /** Date for one-day `searchAppearance` total queries, whose keys omit date. */
   date?: string;
   /** Search appearance filter used for contextual second-step rows. */

package/dist/period/index.d.mts

@@ -1,2 +1,2 @@
-import { AnalysisPeriod, ComparisonMode, ComparisonPeriod, PadTimeseriesOptions, ResolveWindowOptions, ResolvedWindow, WindowPreset, comparisonOf, defaultEndDate, defaultStartDate, padTimeseries, periodOf, resolveWindow, windowToComparisonPeriod, windowToPeriod } from "../_chunks/index2.mjs";
-export { AnalysisPeriod, ComparisonMode, ComparisonPeriod, PadTimeseriesOptions, ResolveWindowOptions, ResolvedWindow, WindowPreset, comparisonOf, defaultEndDate, defaultStartDate, padTimeseries, periodOf, resolveWindow, windowToComparisonPeriod, windowToPeriod };
+import { AnalysisPeriod, ComparisonMode, ComparisonPeriod, PadTimeseriesOptions, ResolveWindowOptions, ResolvedWindow, WindowPreset, comparisonOf, defaultEndDate, defaultStartDate, padTimeseries, periodOf, resolveWindow } from "../_chunks/index2.mjs";
+export { AnalysisPeriod, ComparisonMode, ComparisonPeriod, PadTimeseriesOptions, ResolveWindowOptions, ResolvedWindow, WindowPreset, comparisonOf, defaultEndDate, defaultStartDate, padTimeseries, periodOf, resolveWindow };

package/dist/period/index.mjs

@@ -96,25 +96,6 @@ function resolveWindow(opts) {
 	}
 	return result;
 }
-function windowToPeriod(w) {
-	return {
-		startDate: w.start,
-		endDate: w.end
-	};
-}
-function windowToComparisonPeriod(w) {
-	if (!w.comparison) return void 0;
-	return {
-		current: {
-			startDate: w.start,
-			endDate: w.end
-		},
-		previous: {
-			startDate: w.comparison.start,
-			endDate: w.comparison.end
-		}
-	};
-}
 const DEFAULT_FILL = {
 	clicks: 0,
 	impressions: 0,
@@ -147,4 +128,4 @@ function padTimeseries(rows, options) {
 	}
 	return result;
 }
-export { comparisonOf, defaultEndDate, defaultStartDate, padTimeseries, periodOf, resolveWindow, windowToComparisonPeriod, windowToPeriod };
+export { comparisonOf, defaultEndDate, defaultStartDate, padTimeseries, periodOf, resolveWindow };

package/dist/resolver/index.d.mts

@@ -48,8 +48,6 @@ interface SqlFragmentsConfig<TableKey extends string> {
   queryDimTableRef?: () => SQL;
   /** When true, also joins query_dim on site_id (SQLite/D1 dimension table). */
   queryDimSiteScoped?: boolean;
-  /** @deprecated Canonical reads now derive from query_dim or canonical rollups. */
-  canonicalFallback?: boolean;
 }
 interface SqlFragments<TableKey extends string> {
   METRIC_NAMES: Metric[];
@@ -175,11 +173,6 @@ interface ResolveExtraFn {
 interface RunOptimizedQueryOptions {
   /** Overlay tried per extra before the live SQL; absent → today's live path. */
   resolveExtra?: ResolveExtraFn;
-  /**
-   * Deprecated compatibility flag. Canonical reads now derive from query_dim,
-   * not a nullable fact-table `query_canonical` column.
-   */
-  canonicalFallback?: boolean;
   /**
    * Versioned query dimension backing canonical reads over fact rows. Required
    * whenever a query groups or filters by `queryCanonical`; a canonical rollup
@@ -218,11 +211,6 @@ interface RunOptimizedQueryOptions {
     coversThrough?: string;
     queryDim?: CanonicalQueryDimSource;
   };
-  /**
-   * @deprecated Canonical-source misses are recorded on `source.fallbacks`;
-   * raw daily fallback is controlled by `primarySourceFallback: 'raw'`.
-   */
-  canonicalSourceFallback?: 'raw';
   /** Optional version gates for the query dimension backing `canonicalSource`. */
   canonicalRequirements?: {
     normalizerVersion?: number;

package/dist/schema.d.mts

@@ -1,2 +1,2 @@
-import { ColumnDef, ColumnType, DrizzleSchema, SCHEMAS, TABLE_METADATA, TableSchema, allTables, countries, currentSchemaVersion, dateColumnsFor, dates, dedupeByNaturalKey, dimensionToColumn, drizzleSchema, hourly_pages, inferTable, naturalKeyColumns, page_queries, pages, queries, schemaFor, search_appearance, search_appearance_page_queries, search_appearance_pages, search_appearance_queries } from "./_chunks/schema.mjs";
-export { type ColumnDef, type ColumnType, type DrizzleSchema, SCHEMAS, TABLE_METADATA, type TableSchema, allTables, countries, currentSchemaVersion, dateColumnsFor, dates, dedupeByNaturalKey, dimensionToColumn, drizzleSchema, hourly_pages, inferTable, naturalKeyColumns, page_queries, pages, queries, schemaFor, search_appearance, search_appearance_page_queries, search_appearance_pages, search_appearance_queries };
+import { ColumnDef, ColumnType, DrizzleSchema, SCHEMAS, TABLE_METADATA, TableSchema, allTables, countries, currentSchemaVersion, dateColumnsFor, dates, dedupeByNaturalKey, dimensionToColumn, drizzleSchema, hourly_pages, inferTable, naturalKeyColumns, page_queries, pages, queries, search_appearance, search_appearance_page_queries, search_appearance_pages, search_appearance_queries } from "./_chunks/schema.mjs";
+export { type ColumnDef, type ColumnType, type DrizzleSchema, SCHEMAS, TABLE_METADATA, type TableSchema, allTables, countries, currentSchemaVersion, dateColumnsFor, dates, dedupeByNaturalKey, dimensionToColumn, drizzleSchema, hourly_pages, inferTable, naturalKeyColumns, page_queries, pages, queries, search_appearance, search_appearance_page_queries, search_appearance_pages, search_appearance_queries };

package/dist/schema.mjs

@@ -1,2 +1,2 @@
-import { SCHEMAS, TABLE_METADATA, allTables, countries, currentSchemaVersion, dateColumnsFor, dates, dedupeByNaturalKey, dimensionToColumn, drizzleSchema, hourly_pages, inferTable, naturalKeyColumns, page_queries, pages, queries, schemaFor, search_appearance, search_appearance_page_queries, search_appearance_pages, search_appearance_queries } from "./_chunks/schema.mjs";
-export { SCHEMAS, TABLE_METADATA, allTables, countries, currentSchemaVersion, dateColumnsFor, dates, dedupeByNaturalKey, dimensionToColumn, drizzleSchema, hourly_pages, inferTable, naturalKeyColumns, page_queries, pages, queries, schemaFor, search_appearance, search_appearance_page_queries, search_appearance_pages, search_appearance_queries };
+import { SCHEMAS, TABLE_METADATA, allTables, countries, currentSchemaVersion, dateColumnsFor, dates, dedupeByNaturalKey, dimensionToColumn, drizzleSchema, hourly_pages, inferTable, naturalKeyColumns, page_queries, pages, queries, search_appearance, search_appearance_page_queries, search_appearance_pages, search_appearance_queries } from "./_chunks/schema.mjs";
+export { SCHEMAS, TABLE_METADATA, allTables, countries, currentSchemaVersion, dateColumnsFor, dates, dedupeByNaturalKey, dimensionToColumn, drizzleSchema, hourly_pages, inferTable, naturalKeyColumns, page_queries, pages, queries, search_appearance, search_appearance_page_queries, search_appearance_pages, search_appearance_queries };

package/dist/sink-node.d.mts

@@ -17,9 +17,8 @@ interface OverwriteWriterCatalogConfig {
   catalogToken?: string;
 }
 /**
- * The JSON job wire-format shared with `scripts/iceberg-writer.py` — identical
- * to the shape `LocalIcebergSink` already emits, so one PyIceberg script backs
- * both the local sink and this writer. `op` is always `overwrite` here.
+ * The JSON job wire-format shared with `scripts/iceberg-writer.py`.
+ * `op` is always `overwrite` here.
  */
 interface OverwriteJob {
   op: 'overwrite';

package/dist/source/index.d.mts

@@ -1,3 +1,3 @@
 import { AnalysisQuerySource, AnalysisSourceKind, ExecuteSqlOptions, FileSet, QueryRow, SourceCapabilities } from "../_chunks/registry.mjs";
-import { AttachedTableMissingError, AttachedTableRunner, AttachedTableSourceOptions, CreateSqlQuerySourceOptions, ENGINE_QUERY_CAPABILITIES, EngineQuerySourceOptions, TypedQuery, createAttachedTableSource, createEngineQuerySource, createSqlQuerySource, queryComparisonRows, queryRows, rewriteForTableSource, runAnalyzerWithEngine, typedQuery } from "../_chunks/index.mjs";
-export { type AnalysisQuerySource, type AnalysisSourceKind, AttachedTableMissingError, type AttachedTableRunner, type AttachedTableSourceOptions, type CreateSqlQuerySourceOptions, ENGINE_QUERY_CAPABILITIES, EngineQuerySourceOptions, type ExecuteSqlOptions, type FileSet, type QueryRow, type SourceCapabilities, TypedQuery, createAttachedTableSource, createEngineQuerySource, createSqlQuerySource, queryComparisonRows, queryRows, rewriteForTableSource, runAnalyzerWithEngine, typedQuery };
+import { AttachedTableMissingError, AttachedTableRunner, AttachedTableSourceOptions, CreateSqlQuerySourceOptions, ENGINE_QUERY_CAPABILITIES, EngineQuerySourceOptions, createAttachedTableSource, createEngineQuerySource, createSqlQuerySource, queryRows, rewriteForTableSource, runAnalyzerWithEngine } from "../_chunks/index.mjs";
+export { type AnalysisQuerySource, type AnalysisSourceKind, AttachedTableMissingError, type AttachedTableRunner, type AttachedTableSourceOptions, type CreateSqlQuerySourceOptions, ENGINE_QUERY_CAPABILITIES, EngineQuerySourceOptions, type ExecuteSqlOptions, type FileSet, type QueryRow, type SourceCapabilities, createAttachedTableSource, createEngineQuerySource, createSqlQuerySource, queryRows, rewriteForTableSource, runAnalyzerWithEngine };

package/dist/source/index.mjs

@@ -1,2 +1,2 @@
-import { AttachedTableMissingError, ENGINE_QUERY_CAPABILITIES, createAttachedTableSource, createEngineQuerySource, createSqlQuerySource, queryComparisonRows, queryRows, rewriteForTableSource, runAnalyzerWithEngine, typedQuery } from "../_chunks/source.mjs";
-export { AttachedTableMissingError, ENGINE_QUERY_CAPABILITIES, createAttachedTableSource, createEngineQuerySource, createSqlQuerySource, queryComparisonRows, queryRows, rewriteForTableSource, runAnalyzerWithEngine, typedQuery };
+import { AttachedTableMissingError, ENGINE_QUERY_CAPABILITIES, createAttachedTableSource, createEngineQuerySource, createSqlQuerySource, queryRows, rewriteForTableSource, runAnalyzerWithEngine } from "../_chunks/source.mjs";
+export { AttachedTableMissingError, ENGINE_QUERY_CAPABILITIES, createAttachedTableSource, createEngineQuerySource, createSqlQuerySource, queryRows, rewriteForTableSource, runAnalyzerWithEngine };

package/package.json

@@ -1,7 +1,7 @@
 {
   "name": "@gscdump/engine",
   "type": "module",
-  "version": "0.39.0",
+  "version": "0.40.2",
   "description": "Append-only Parquet/DuckDB storage engine + planner + adapters for the gscdump pipeline. Node + edge runtimes; opt-in heavy peers.",
   "author": {
     "name": "Harlan Wilton",
@@ -181,9 +181,9 @@
   "dependencies": {
     "drizzle-orm": "1.0.0-rc.3",
     "proper-lockfile": "^4.1.2",
-    "@gscdump/contracts": "0.39.0",
-    "@gscdump/lakehouse": "0.39.0",
-    "gscdump": "0.39.0"
+    "@gscdump/contracts": "0.40.2",
+    "@gscdump/lakehouse": "0.40.2",
+    "gscdump": "0.40.2"
   },
   "devDependencies": {
     "@duckdb/duckdb-wasm": "^1.32.0",