Remove charset detection and utf8 decoding, extend octal transform to sequences

Signed-off-by: Jens Reinecke <[email protected]>

Author: jreineckearm

src/MIParser.ts

@@ -9,7 +9,6 @@
  *********************************************************************/
 import { Readable } from 'stream';
 import { IGDBBackend } from './types/gdb';
-import * as utf8 from 'utf8';
 import { NamedLogger } from './namedLogger';
 
 interface Command {
@@ -144,12 +143,17 @@ export class MIParser {
 
         let cstring = '';
         let octal = '';
+        let octalSequence: number[] = [];
         mainloop: for (c = this.next(); c; c = this.next()) {
             if (octal) {
                 octal += c;
                 if (octal.length == 3) {
-                    cstring += String.fromCodePoint(parseInt(octal, 8));
+                    octalSequence.push(parseInt(octal, 8));
                     octal = '';
+                    if (this.peek() !== '\\') {
+                        cstring += String.fromCodePoint(...octalSequence);
+                        octalSequence = [];
+                    }
                 }
                 continue;
             }
@@ -190,19 +194,7 @@ export class MIParser {
             }
         }
 
-        // Return received string without decoding if turned off.
-        if (!this.decodeUtf8) {
-            return cstring;
-        }
-
-        try {
-            return utf8.decode(cstring);
-        } catch (err) {
-            this.logger.error(
-                `Failed to decode cstring '${cstring}'. ${JSON.stringify(err)}`
-            );
-            return cstring;
-        }
+        return cstring;
     }
 
     protected handleString() {

src/gdb/GDBBackend.ts

@@ -28,10 +28,6 @@ import { compareVersions } from '../util/compareVersions';
 import { isProcessActive } from '../util/processes';
 import { NamedLogger } from '../namedLogger';
 
-// Error as returned by GDB when calling 'set charset' without argument
-const CHARSETS_ERROR_REGEXP =
-    /Requires an argument. Valid arguments are (.+)\./;
-
 type WriteCallback = (error: Error | null | undefined) => void;
 
 export class GDBBackend extends events.EventEmitter implements IGDBBackend {
@@ -61,49 +57,6 @@ export class GDBBackend extends events.EventEmitter implements IGDBBackend {
         return this.varMgr;
     }
 
-    /**
-     * Detect list of supported charsets by provoking error from GDB
-     * which returns the list.
-     */
-    private async getSupportedCharsets(): Promise<string[]> {
-        let charsetString: string | undefined;
-        try {
-            // Provoke error by missing argument
-            await this.sendGDBSet('charset');
-        } catch (error) {
-            const errMessage = (error as Error)?.message;
-            if (errMessage) {
-                const matches = CHARSETS_ERROR_REGEXP.exec(errMessage);
-                charsetString = matches ? matches[1] : undefined;
-            }
-        }
-        if (!charsetString) {
-            return [];
-        }
-        const charsets = charsetString
-            .split(',')
-            .map((charset) => charset.trim());
-        return charsets;
-    }
-
-    /**
-     * Detect if to apply UTF-8 decoding based on detecting available
-     * charsets in GDB.
-     *
-     * Return 'false' if only CP1252 (ASCII) and 'auto' like for some
-     * GDB variants for embedded on Windows.
-     * Note: Strictly speaking Windows-only, but keep this method OS
-     * agnostic to avoid potential trouble with web use-case.
-     */
-    private async shouldDecodeUTF8(): Promise<boolean> {
-        const supportedCharsets = await this.getSupportedCharsets();
-        return !(
-            supportedCharsets.length === 2 &&
-            supportedCharsets.includes('CP1252') &&
-            supportedCharsets.includes('auto')
-        );
-    }
-
     public async spawn(
         requestArgs: LaunchRequestArguments | AttachRequestArguments
     ) {
@@ -139,9 +92,6 @@ export class GDBBackend extends events.EventEmitter implements IGDBBackend {
         this.asyncRequestedExplicitly = !!(
             requestArgs.gdbAsync || requestArgs.gdbNonStop
         );
-        if (!(await this.shouldDecodeUTF8())) {
-            this.parser.decodeUtf8 = false;
-        }
         await this.setNonStopMode(requestArgs.gdbNonStop);
         await this.setAsyncMode(requestArgs.gdbAsync);
     }