Data Explorer: Support opening gzipped csv files with DuckDB by decom…

…pressing in memory in NodeJS (#5901)

Addresses #5332. This works around duckdb-wasm's limitation that it
cannot read directly from gzipped CSV/TSV files by unzipping in memory
and registering as a virtual file. This is okay for small files but for
very large gzipped files will feel a bit slow and use a bunch of memory
(we will have to tackle that problem via #5889).

### Release Notes

#### New Features

- Support opening gzipped CSV/TSV files in the data explorer with .gz
extension from the file explorer pane (#5332).

#### Bug Fixes

- N/A

e2e: @:data-explorer

extensions/positron-duckdb/src/extension.ts

@@ -53,6 +53,7 @@ import {
 import * as duckdb from '@duckdb/duckdb-wasm';
 import * as path from 'path';
 import * as fs from 'fs';
+import * as zlib from 'zlib';
 import Worker from 'web-worker';
 import { Table, Vector } from 'apache-arrow';
 import { pathToFileURL } from 'url';
@@ -417,6 +418,7 @@ class ColumnProfileEvaluator {
 		const numRows = Number(stats.get('num_rows'));
 		const nullCount = Number(stats.get(`null_count_${field}`));
 
+
 		return {
 			values,
 			counts,
@@ -1243,6 +1245,12 @@ export class DataExplorerRpcHandler {
 	 * @param catalogName The table name to use in the DuckDB catalog.
 	 */
 	async createTableFromFilePath(filePath: string, catalogName: string) {
+		let fileExt = path.extname(filePath);
+		const isGzipped = fileExt === '.gz';
+
+		if (isGzipped) {
+			fileExt = path.extname(filePath.slice(0, -3));
+		}
 
 		const getCsvImportQuery = (_filePath: string, options: Array<String>) => {
 			return `CREATE OR REPLACE TABLE ${catalogName} AS
@@ -1273,15 +1281,22 @@ export class DataExplorerRpcHandler {
 			}
 		};
 
-		const fileExt = path.extname(filePath);
-
 		// Read the entire contents and register it as a temp file
 		// to avoid file handle caching in duckdb-wasm
-		const fileContents = fs.readFileSync(filePath, { encoding: null });
-		const virtualPath = path.basename(filePath);
+		let fileContents = fs.readFileSync(filePath, { encoding: null });
+		if (isGzipped) {
+			fileContents = zlib.gunzipSync(fileContents);
+		}
+
+		// For gzipped files, use the base name without the .gz extension
+		const virtualPath = isGzipped ?
+			path.basename(filePath, '.gz') :
+			path.basename(filePath);
+
 		await this.db.db.registerFileBuffer(virtualPath, fileContents);
 		try {
-			if (fileExt === '.parquet' || fileExt === '.parq') {
+			const baseExt = path.extname(virtualPath);
+			if (baseExt === '.parquet' || baseExt === '.parq') {
 				// Always create a view for Parquet files
 				const query = `CREATE OR REPLACE TABLE ${catalogName} AS
 				SELECT * FROM parquet_scan('${virtualPath}');`;

src/vs/workbench/contrib/positronDataExplorerEditor/browser/positronDataExplorerEditor.contribution.ts

@@ -17,6 +17,7 @@ import { PositronDataExplorerEditor } from './positronDataExplorerEditor.js';
 import { PositronDataExplorerEditorInput } from './positronDataExplorerEditorInput.js';
 import { registerPositronDataExplorerActions } from './positronDataExplorerActions.js';
 import { extname } from '../../../../base/common/resources.js';
+import { posix } from '../../../../base/common/path.js';
 import { IPositronDataExplorerService } from '../../../services/positronDataExplorer/browser/interfaces/positronDataExplorerService.js';
 import { PositronDataExplorerUri } from '../../../services/positronDataExplorer/common/positronDataExplorerUri.js';
 
@@ -76,15 +77,20 @@ class PositronDataExplorerContribution extends Disposable {
 			}
 		));
 
-		const DUCKDB_SUPPORTED_EXTENSIONS = ['parquet', 'parq', 'csv', 'tsv'];
+		const DUCKDB_SUPPORTED_EXTENSIONS = ['parquet', 'parq', 'csv', 'tsv', 'gz'];
 
 		this._register(editorResolverService.registerEditor(
 			`*.{${DUCKDB_SUPPORTED_EXTENSIONS.join(',')}}`,
 			editorInfo,
 			{
 				singlePerResource: true,
 				canSupportResource: resource => {
-					return DUCKDB_SUPPORTED_EXTENSIONS.includes(extname(resource).substring(1));
+					let fileExt = extname(resource).substring(1);
+					if (fileExt === 'gz') {
+						// Strip the .gz and get the actual extension
+						fileExt = posix.extname(resource.path.slice(0, -3)).substring(1);
+					}
+					return DUCKDB_SUPPORTED_EXTENSIONS.includes(fileExt);
 				}
 			},
 			{