@cereusdb/minimal 0.1.2 → 0.2.0

package/README.md

@@ -24,16 +24,25 @@ Included function families:
 
 Not included in this package:
 
+- Browser object stores for ranged remote Parquet reads and S3/GCS/Azure/HTTP providers
 - `ST_Transform`
 - S2 geography kernels
 - Raster `RS_*` functions
 
+## Object storage support
+
+Browser object stores are not included in `@cereusdb/minimal`. Use `@cereusdb/standard`, `@cereusdb/global`, or `@cereusdb/full` when you need `registerObjectStores()` and `registerParquetTable()`.
+
 ## JS / TS API
 
 Exports:
 
 - `CereusDB`
 - `CereusDBOptions`
+- `ObjectStoreRegistryConfig`
+- `ObjectStoreConfig`
+- `ObjectStoreProvider`
+- `RegisterParquetTableOptions`
 - `RasterFormat`
 - `QueryResult`
 
@@ -41,6 +50,7 @@ Main types:
 
 ```ts
 type RasterFormat = 'geotiff' | 'tiff';
+type ObjectStoreProvider = 'http' | 's3' | 'gcs' | 'azure';
 
 interface CereusDBOptions {
   wasmUrl?: string;
@@ -51,6 +61,24 @@ interface CereusDBOptions {
     | BufferSource
     | WebAssembly.Module
     | Promise<Response>;
+  objectStores?: ObjectStoreRegistryConfig;
+}
+
+interface ObjectStoreRegistryConfig {
+  maxConcurrency?: number;
+  stores: ObjectStoreConfig[];
+}
+
+interface ObjectStoreConfig {
+  name?: string;
+  provider: ObjectStoreProvider;
+  url: string;
+  options?: Record<string, string | number | boolean>;
+}
+
+interface RegisterParquetTableOptions {
+  fileExtension?: string;
+  targetPartitions?: number;
 }
 ```
 
@@ -62,6 +90,12 @@ class CereusDB {
   sql(query: string): Promise<Uint8Array>;
   sqlJSON(query: string): Promise<Record<string, unknown>[]>;
   registerRemoteParquet(name: string, url: string): Promise<void>;
+  registerObjectStores(config: ObjectStoreRegistryConfig): void;
+  registerParquetTable(
+    name: string,
+    url: string,
+    options?: RegisterParquetTableOptions,
+  ): Promise<void>;
   registerFile(name: string, file: File): Promise<void>;
   registerGeoJSON(name: string, geojson: string | object): void;
   registerRaster(name: string, data: BufferSource, format: RasterFormat): void;
@@ -77,6 +111,7 @@ API notes:
 - `sql()` returns Arrow IPC bytes as `Uint8Array`.
 - `sqlJSON()` returns parsed JSON rows.
 - `registerFile()` supports `.parquet`, `.geoparquet`, `.geojson`, and `.json` in this package.
+- Browser object-store methods are part of the shared wrapper, but object-store registration requires `@cereusdb/standard`, `@cereusdb/global`, or `@cereusdb/full`.
 - `registerRaster()` and `registerGeoTIFF()` are part of the shared wrapper, but raster registration requires `@cereusdb/full`.
 
 ## Example

package/dist/index.d.ts

@@ -13,8 +13,33 @@ export interface CereusDBOptions {
     wasmUrl?: string;
     /** Preloaded WASM bytes/module for Node or custom loaders. */
     wasmSource?: RequestInfo | URL | Response | BufferSource | WebAssembly.Module | Promise<Response>;
+    /** Browser-backed object stores to register at startup. */
+    objectStores?: ObjectStoreRegistryConfig;
 }
 export type RasterFormat = 'geotiff' | 'tiff';
+export type ObjectStoreProvider = 'http' | 's3' | 'gcs' | 'azure';
+export interface ObjectStoreRegistryConfig {
+    /** Maximum concurrent browser fetches used by object_store. */
+    maxConcurrency?: number;
+    /** Stores registered by URL prefix. */
+    stores: ObjectStoreConfig[];
+}
+export interface ObjectStoreConfig {
+    /** Optional diagnostic name. */
+    name?: string;
+    /** Backing provider. */
+    provider: ObjectStoreProvider;
+    /** URL prefix, for example https://host, s3://bucket, gs://bucket. */
+    url: string;
+    /** Upstream object_store option keys and primitive values. */
+    options?: Record<string, string | number | boolean>;
+}
+export interface RegisterParquetTableOptions {
+    /** File extension used during listing discovery. Defaults to .parquet. */
+    fileExtension?: string;
+    /** Optional DataFusion target partition count. */
+    targetPartitions?: number;
+}
 export declare class CereusDB {
     private inner;
     private constructor();
@@ -36,6 +61,14 @@ export declare class CereusDB {
      * The server must support CORS.
      */
     registerRemoteParquet(name: string, url: string): Promise<void>;
+    /**
+     * Register browser-backed object stores for ranged and listing reads.
+     */
+    registerObjectStores(config: ObjectStoreRegistryConfig): void;
+    /**
+     * Register a remote Parquet object or prefix through DataFusion's listing table path.
+     */
+    registerParquetTable(name: string, url: string, options?: RegisterParquetTableOptions): Promise<void>;
     /**
      * Register a local file (from File API / drag-and-drop) as a table.
      * Currently supports Parquet, GeoJSON, and GeoTIFF rasters.
@@ -61,4 +94,5 @@ export declare class CereusDB {
     tables(): string[];
     /** Version string. */
     version(): string;
+    private objectStoreApi;
 }

package/dist/index.js

@@ -29,7 +29,11 @@ export class CereusDB {
             await init({ module_or_path: source });
         }
         const inner = WasmCereusDB.create();
-        return new CereusDB(inner);
+        const db = new CereusDB(inner);
+        if (options?.objectStores !== undefined) {
+            db.registerObjectStores(options.objectStores);
+        }
+        return db;
     }
     /**
      * Execute a SQL query and return results as Arrow IPC bytes.
@@ -51,6 +55,18 @@ export class CereusDB {
     async registerRemoteParquet(name, url) {
         await this.inner.register_remote_parquet(name, url);
     }
+    /**
+     * Register browser-backed object stores for ranged and listing reads.
+     */
+    registerObjectStores(config) {
+        this.objectStoreApi().register_object_stores(config);
+    }
+    /**
+     * Register a remote Parquet object or prefix through DataFusion's listing table path.
+     */
+    async registerParquetTable(name, url, options = {}) {
+        await this.objectStoreApi().register_parquet_table(name, url, options);
+    }
     /**
      * Register a local file (from File API / drag-and-drop) as a table.
      * Currently supports Parquet, GeoJSON, and GeoTIFF rasters.
@@ -105,4 +121,12 @@ export class CereusDB {
     version() {
         return this.inner.version();
     }
+    objectStoreApi() {
+        const api = this.inner;
+        if (typeof api.register_object_stores !== 'function' ||
+            typeof api.register_parquet_table !== 'function') {
+            throw new Error('Browser object stores are not available in this CereusDB build');
+        }
+        return api;
+    }
 }

package/dist/wasm/cereusdb.d.ts

@@ -27,11 +27,19 @@ export class CereusDB {
      * Requires the full GDAL-enabled build.
      */
     register_geotiff_buffer(table_name: string, data: Uint8Array): void;
+    /**
+     * Register browser-backed object stores for ranged/listing reads.
+     */
+    register_object_stores(config: any): void;
     /**
      * Register a Uint8Array containing Parquet data as a named table.
      * Use for files obtained via the browser File API.
      */
     register_parquet_buffer(table_name: string, data: Uint8Array): Promise<void>;
+    /**
+     * Register a remote Parquet object or prefix as a DataFusion listing table.
+     */
+    register_parquet_table(table_name: string, table_url: string, options: any): Promise<void>;
     /**
      * Register a raster buffer as a single-column raster table.
      * Requires the full GDAL-enabled build.
@@ -81,7 +89,9 @@ export interface InitOutput {
     readonly cereusdb_drop_table: (a: number, b: number, c: number) => [number, number];
     readonly cereusdb_register_geojson: (a: number, b: number, c: number, d: number, e: number) => [number, number];
     readonly cereusdb_register_geotiff_buffer: (a: number, b: number, c: number, d: number, e: number) => [number, number];
+    readonly cereusdb_register_object_stores: (a: number, b: any) => [number, number];
     readonly cereusdb_register_parquet_buffer: (a: number, b: number, c: number, d: number, e: number) => any;
+    readonly cereusdb_register_parquet_table: (a: number, b: number, c: number, d: number, e: number, f: any) => any;
     readonly cereusdb_register_raster_buffer: (a: number, b: number, c: number, d: number, e: number, f: number, g: number) => [number, number];
     readonly cereusdb_register_remote_parquet: (a: number, b: number, c: number, d: number, e: number) => any;
     readonly cereusdb_sql: (a: number, b: number, c: number) => any;
@@ -97,7 +107,7 @@ export interface InitOutput {
     readonly emscripten_builtin_free: (a: number) => void;
     readonly emscripten_builtin_malloc: (a: number) => number;
     readonly __cpp_exception: WebAssembly.Tag;
-    readonly wasm_bindgen__closure__destroy__hcd35ec5c66a281ba: (a: number, b: number) => void;
+    readonly wasm_bindgen__closure__destroy__h8db467765ce434ac: (a: number, b: number) => void;
     readonly wasm_bindgen__convert__closures_____invoke__h7573a8a64f443504: (a: number, b: number, c: any) => [number, number];
     readonly wasm_bindgen__convert__closures_____invoke__h3bcc27440de746c0: (a: number, b: number, c: any, d: any) => void;
     readonly __wbindgen_malloc: (a: number, b: number) => number;

package/dist/wasm/cereusdb.js

@@ -110,6 +110,22 @@ export class CereusDB {
             throw takeFromExternrefTable0(ret[0]);
         }
     }
+    /**
+     * Register browser-backed object stores for ranged/listing reads.
+     * @param {any} config
+     */
+    register_object_stores(config) {
+        let ret;
+        __wbg_termination_guard();
+        try {
+            ret = wasm.cereusdb_register_object_stores(this.__wbg_ptr, config);;
+        } catch(e) {
+            __wbg_handle_catch(e);
+        }
+        if (ret[1]) {
+            throw takeFromExternrefTable0(ret[0]);
+        }
+    }
     /**
      * Register a Uint8Array containing Parquet data as a named table.
      * Use for files obtained via the browser File API.
@@ -131,6 +147,27 @@ export class CereusDB {
         }
         return ret;
     }
+    /**
+     * Register a remote Parquet object or prefix as a DataFusion listing table.
+     * @param {string} table_name
+     * @param {string} table_url
+     * @param {any} options
+     * @returns {Promise<void>}
+     */
+    register_parquet_table(table_name, table_url, options) {
+        const ptr0 = passStringToWasm0(table_name, wasm.__wbindgen_malloc, wasm.__wbindgen_realloc);
+        const len0 = WASM_VECTOR_LEN;
+        const ptr1 = passStringToWasm0(table_url, wasm.__wbindgen_malloc, wasm.__wbindgen_realloc);
+        const len1 = WASM_VECTOR_LEN;
+        let ret;
+        __wbg_termination_guard();
+        try {
+            ret = wasm.cereusdb_register_parquet_table(this.__wbg_ptr, ptr0, len0, ptr1, len1, options);;
+        } catch(e) {
+            __wbg_handle_catch(e);
+        }
+        return ret;
+    }
     /**
      * Register a raster buffer as a single-column raster table.
      * Requires the full GDAL-enabled build.
@@ -491,8 +528,8 @@ function __wbg_get_imports() {
             return ret;
         },
         __wbindgen_cast_0000000000000001: function(arg0, arg1) {
-            // Cast intrinsic for `Closure(Closure { dtor_idx: 4230, function: Function { arguments: [Externref], shim_idx: 19295, ret: Result(Unit), inner_ret: Some(Result(Unit)) }, mutable: true }) -> Externref`.
-            const ret = makeMutClosure(arg0, arg1, wasm.wasm_bindgen__closure__destroy__hcd35ec5c66a281ba, wasm_bindgen__convert__closures_____invoke__h7573a8a64f443504);
+            // Cast intrinsic for `Closure(Closure { dtor_idx: 4368, function: Function { arguments: [Externref], shim_idx: 20024, ret: Result(Unit), inner_ret: Some(Result(Unit)) }, mutable: true }) -> Externref`.
+            const ret = makeMutClosure(arg0, arg1, wasm.wasm_bindgen__closure__destroy__h8db467765ce434ac, wasm_bindgen__convert__closures_____invoke__h7573a8a64f443504);
             return ret;
         },
         __wbindgen_cast_0000000000000002: function(arg0, arg1) {

package/dist/wasm/cereusdb_bg.wasm.d.ts

@@ -14,7 +14,9 @@ export const cereusdb_create: () => [number, number, number];
 export const cereusdb_drop_table: (a: number, b: number, c: number) => [number, number];
 export const cereusdb_register_geojson: (a: number, b: number, c: number, d: number, e: number) => [number, number];
 export const cereusdb_register_geotiff_buffer: (a: number, b: number, c: number, d: number, e: number) => [number, number];
+export const cereusdb_register_object_stores: (a: number, b: any) => [number, number];
 export const cereusdb_register_parquet_buffer: (a: number, b: number, c: number, d: number, e: number) => any;
+export const cereusdb_register_parquet_table: (a: number, b: number, c: number, d: number, e: number, f: any) => any;
 export const cereusdb_register_raster_buffer: (a: number, b: number, c: number, d: number, e: number, f: number, g: number) => [number, number];
 export const cereusdb_register_remote_parquet: (a: number, b: number, c: number, d: number, e: number) => any;
 export const cereusdb_sql: (a: number, b: number, c: number) => any;
@@ -30,7 +32,7 @@ export const abort: () => void;
 export const emscripten_builtin_free: (a: number) => void;
 export const emscripten_builtin_malloc: (a: number) => number;
 export const __cpp_exception: WebAssembly.Tag;
-export const wasm_bindgen__closure__destroy__hcd35ec5c66a281ba: (a: number, b: number) => void;
+export const wasm_bindgen__closure__destroy__h8db467765ce434ac: (a: number, b: number) => void;
 export const wasm_bindgen__convert__closures_____invoke__h7573a8a64f443504: (a: number, b: number, c: any) => [number, number];
 export const wasm_bindgen__convert__closures_____invoke__h3bcc27440de746c0: (a: number, b: number, c: any, d: any) => void;
 export const __wbindgen_malloc: (a: number, b: number) => number;

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@cereusdb/minimal",
-  "version": "0.1.2",
+  "version": "0.2.0",
   "description": "Minimal CereusDB browser package with GEOS-backed spatial SQL",
   "type": "module",
   "main": "./dist/index.js",