@laitszkin/apollo-toolkit 4.1.4 → 5.0.0

package/packages/tools/docs-to-voice/dist/index.js

@@ -3,112 +3,19 @@ import fs from 'node:fs';
 import path from 'node:path';
 import https from 'node:https';
 import http from 'node:http';
+import { SystemError, UserInputError, createToolRunner } from '../../../tool-utils/dist/index.js';
 const DEFAULT_API_ENDPOINT = 'https://dashscope-intl.aliyuncs.com/api/v1/services/aigc/multimodal-generation/generation';
 const DEFAULT_API_MODEL = 'qwen3-tts';
 const DEFAULT_API_VOICE = 'Cherry';
-function parseArgs(args) {
-    const parsed = {
-        inputText: null,
-        inputFile: null,
-        projectDir: '.',
-        projectName: null,
-        outputName: null,
-        mode: 'say',
-        voice: null,
-        rate: null,
-        speechRate: null,
-        apiEndpoint: DEFAULT_API_ENDPOINT,
-        apiModel: DEFAULT_API_MODEL,
-        apiVoice: DEFAULT_API_VOICE,
-        apiKey: null,
-        maxChars: null,
-        noAutoProsody: false,
-        force: false,
-        help: false,
-    };
-    for (let i = 0; i < args.length; i++) {
-        const arg = args[i];
-        if (arg === '--help' || arg === '-h') {
-            parsed.help = true;
-            continue;
-        }
-        if (arg.startsWith('--')) {
-            const eqIndex = arg.indexOf('=');
-            let key;
-            let value;
-            if (eqIndex !== -1) {
-                key = arg.slice(2, eqIndex);
-                value = arg.slice(eqIndex + 1);
-            }
-            else {
-                key = arg.slice(2);
-                value = args[++i] || '';
-            }
-            switch (key) {
-                case 'text':
-                    parsed.inputText = value;
-                    break;
-                case 'input':
-                case 'input-file':
-                    parsed.inputFile = value;
-                    break;
-                case 'project-dir':
-                    parsed.projectDir = value;
-                    break;
-                case 'project-name':
-                    parsed.projectName = value;
-                    break;
-                case 'output-name':
-                    parsed.outputName = value;
-                    break;
-                case 'engine':
-                case 'mode':
-                    parsed.mode = value.toLowerCase();
-                    break;
-                case 'voice':
-                    parsed.voice = value;
-                    break;
-                case 'rate':
-                    parsed.rate = value;
-                    break;
-                case 'speech-rate':
-                    parsed.speechRate = value;
-                    break;
-                case 'api-endpoint':
-                    parsed.apiEndpoint = value;
-                    break;
-                case 'api-model':
-                    parsed.apiModel = value;
-                    break;
-                case 'api-voice':
-                    parsed.apiVoice = value;
-                    break;
-                case 'api-key':
-                    parsed.apiKey = value;
-                    break;
-                case 'max-chars':
-                    parsed.maxChars = value;
-                    break;
-                case 'no-auto-prosody':
-                    parsed.noAutoProsody = true;
-                    break;
-                case 'force':
-                    parsed.force = true;
-                    break;
-            }
-        }
-    }
-    return parsed;
-}
-function readInputText(opts) {
-    if (opts.inputFile) {
-        const inputPath = path.resolve(opts.inputFile);
+function readInputText(inputFile, inputText) {
+    if (inputFile) {
+        const inputPath = path.resolve(inputFile);
         if (!fs.existsSync(inputPath)) {
-            throw new Error(`Input file not found: ${inputPath}`);
+            throw new UserInputError(`Input file not found: ${inputPath}`);
         }
         return fs.readFileSync(inputPath, 'utf-8');
     }
-    return opts.inputText || '';
+    return inputText || '';
 }
 function splitSentences(rawText) {
     const endings = new Set(['。', '！', '？', '!', '?', '；', ';']);
@@ -299,7 +206,7 @@ function applySpeechRateToAudio(outputPath, speechRate) {
     catch (err) {
         if (fs.existsSync(tmpPath))
             fs.unlinkSync(tmpPath);
-        throw new Error(`ffmpeg failed while applying --speech-rate: ${err instanceof Error ? err.message : 'unknown error'}`);
+        throw new SystemError(`ffmpeg failed while applying --speech-rate: ${err instanceof Error ? err.message : 'unknown error'}`, undefined, { cause: err });
     }
 }
 function splitTextForTts(text, maxChars) {
@@ -344,7 +251,7 @@ function splitTextForTts(text, maxChars) {
 }
 function concatAudioFiles(partPaths, outputPath) {
     if (partPaths.length === 0) {
-        throw new Error('No chunk audio generated for concatenation.');
+        throw new SystemError('No chunk audio generated for concatenation.');
     }
     if (partPaths.length === 1) {
         fs.copyFileSync(partPaths[0], outputPath);
@@ -359,7 +266,7 @@ function concatAudioFiles(partPaths, outputPath) {
         execSync(`ffmpeg -hide_banner -loglevel error -y -f concat -safe 0 -i "${listFile}" -c:a copy "${outputPath}"`, { stdio: 'ignore', timeout: 120000 });
     }
     catch (err) {
-        throw new Error(`ffmpeg concat failed: ${err instanceof Error ? err.message : 'unknown error'}`);
+        throw new SystemError(`ffmpeg concat failed: ${err instanceof Error ? err.message : 'unknown error'}`, undefined, { cause: err });
     }
     finally {
         try {
@@ -415,104 +322,104 @@ function requestAlibabaCloudTTS(endpoint, apiKey, model, voice, text) {
                     const audioData = audio.data || '';
                     const audioFormat = audio.format || audio.mime_type || '';
                     if (!audioUrl && !audioData) {
-                        reject(new Error('API response does not contain output.audio.url or output.audio.data'));
+                        reject(new SystemError('API response does not contain output.audio.url or output.audio.data'));
                         return;
                     }
                     resolve({ audioUrl, audioData, audioFormat });
                 }
                 catch {
-                    reject(new Error('API response is not valid JSON.'));
+                    reject(new SystemError('API response is not valid JSON.'));
                 }
             });
             res.on('error', reject);
         });
         req.on('error', reject);
-        req.on('timeout', () => { req.destroy(); reject(new Error('API request timed out')); });
+        req.on('timeout', () => { req.destroy(); reject(new SystemError('API request timed out')); });
         req.write(payload);
         req.end();
     });
 }
-export async function docsToVoiceHandler(args, context) {
-    const stdout = context.stdout || process.stdout;
-    const stderr = context.stderr || process.stderr;
-    try {
-        const opts = parseArgs(args);
-        if (opts.help) {
-            stdout.write(`Usage: apltk docs-to-voice [options]
-
-Convert text into audio and sentence timelines.
-
-Options:
-  --input, --input-file <path>  Path to input text file
-  --text <string>               Raw text input
-  --project-dir <path>          Root project directory (default: .)
-  --project-name <name>         Folder name under DIR/audio/
-  --output-name <name>          Output filename
-  --engine, --mode <mode>       TTS mode: say (default) | api
-  --voice <name>                macOS say voice
-  --rate <wpm>                  macOS say rate
-  --speech-rate <factor>        Speech rate multiplier (e.g. 1.2)
-  --api-endpoint <url>          Alibaba Cloud TTS endpoint
-  --api-model <name>            Alibaba Cloud model (default: qwen3-tts)
-  --api-voice <name>            Alibaba Cloud voice (default: Cherry)
-  --api-key <key>               Alibaba Cloud API key
-  --max-chars <n>               Max chars per TTS chunk (0 disables)
-  --no-auto-prosody             Disable punctuation pause enhancement
-  --force                       Overwrite existing files
-`);
-            return 0;
-        }
-        if (opts.mode !== 'say' && opts.mode !== 'api') {
-            stderr.write('Error: --mode must be one of: say, api\n');
-            return 1;
+const schema = {
+    options: {
+        'text': { type: 'string' },
+        'input': { type: 'string' },
+        'input-file': { type: 'string' },
+        'project-dir': { type: 'string', default: '.' },
+        'project-name': { type: 'string' },
+        'output-name': { type: 'string' },
+        'engine': { type: 'string' },
+        'mode': { type: 'string', default: 'say' },
+        'voice': { type: 'string' },
+        'rate': { type: 'string' },
+        'speech-rate': { type: 'string' },
+        'api-endpoint': { type: 'string', default: DEFAULT_API_ENDPOINT },
+        'api-model': { type: 'string', default: DEFAULT_API_MODEL },
+        'api-voice': { type: 'string', default: DEFAULT_API_VOICE },
+        'api-key': { type: 'string' },
+        'max-chars': { type: 'string' },
+        'no-auto-prosody': { type: 'boolean', default: false },
+        'force': { type: 'boolean', default: false },
+    },
+    allowPositionals: true,
+    usage: 'apltk docs-to-voice [options]',
+    description: 'Convert text into audio and sentence timelines.',
+    handler: async (values, _positionals, context) => {
+        const stdout = context.stdout ?? process.stdout;
+        const stderr = context.stderr ?? process.stderr;
+        // Resolve args (previously handled by parseCliArgs)
+        const inputText = values['text'] ?? null;
+        const inputFile = values['input'] || values['input-file'] || null;
+        const projectDir = values['project-dir'] || '.';
+        const projectName = values['project-name'] ?? null;
+        const outputName = values['output-name'] ?? null;
+        const mode = (values['mode'] || values['engine'] || 'say').toLowerCase();
+        const voice = values['voice'] ?? null;
+        const rate = values['rate'] ?? null;
+        const speechRate = values['speech-rate'] ?? null;
+        const apiEndpoint = values['api-endpoint'] || DEFAULT_API_ENDPOINT;
+        const apiModel = values['api-model'] || DEFAULT_API_MODEL;
+        const apiVoice = values['api-voice'] || DEFAULT_API_VOICE;
+        const apiKey = values['api-key'] ?? null;
+        const maxChars = values['max-chars'] ?? null;
+        const noAutoProsody = !!values['no-auto-prosody'];
+        const force = !!values['force'];
+        if (mode !== 'say' && mode !== 'api') {
+            throw new UserInputError('--mode must be one of: say, api');
         }
-        const sourceText = readInputText(opts);
+        const sourceText = readInputText(inputFile, inputText);
         if (!sourceText.trim()) {
-            stderr.write('Error: No text content found for conversion.\n');
-            return 1;
+            throw new UserInputError('No text content found for conversion.');
         }
         // Resolve output directory
-        const projectDir = path.resolve(opts.projectDir);
-        const projectName = opts.projectName || path.basename(projectDir);
-        if (!projectName) {
-            stderr.write('Error: Unable to determine project name.\n');
-            return 1;
+        const resolvedProjectDir = path.resolve(projectDir);
+        const resolvedProjectName = projectName || path.basename(resolvedProjectDir);
+        if (!resolvedProjectName) {
+            throw new UserInputError('Unable to determine project name.');
         }
-        const outputDir = path.join(projectDir, 'audio', projectName);
+        const outputDir = path.join(resolvedProjectDir, 'audio', resolvedProjectName);
         fs.mkdirSync(outputDir, { recursive: true });
         const timestamp = new Date().toISOString().replace(/[:.]/g, '-').slice(0, 19);
-        const outputName = opts.outputName || `voice-${timestamp}`;
-        const hasExtension = outputName.includes('.');
-        if (opts.mode === 'say') {
-            // macOS say mode
-            const textChunks = splitTextForTts(sourceText, opts.maxChars ? parseInt(opts.maxChars, 10) || null : null);
-            if (textChunks.length === 0) {
-                stderr.write('Error: No text content found for conversion.\n');
-                return 1;
-            }
-            // Check if `say` is available
-            try {
-                execSync('which say', { stdio: 'ignore' });
-            }
-            catch {
-                stderr.write("Error: macOS 'say' command not found.\n");
-                return 1;
-            }
-            const finalOutputName = hasExtension ? outputName : `${outputName}.aiff`;
+        const outName = outputName || `voice-${timestamp}`;
+        const hasExtension = outName.includes('.');
+        if (mode === 'say') {
+            const finalOutputName = hasExtension ? outName : `${outName}.aiff`;
             const outputPath = path.join(outputDir, finalOutputName);
-            if (fs.existsSync(outputPath) && !opts.force) {
-                stderr.write(`Error: Output already exists: ${outputPath}. Use --force to overwrite.\n`);
-                return 1;
+            if (fs.existsSync(outputPath) && !force) {
+                throw new UserInputError(`Output already exists: ${outputPath}. Use --force to overwrite.`);
             }
             // Build prosody-enhanced text
-            const chunks = opts.noAutoProsody ? textChunks : textChunks.map(buildAutoProsodyText);
+            const textChunks = splitTextForTts(sourceText, maxChars ? parseInt(maxChars, 10) || null : null);
+            if (textChunks.length === 0) {
+                throw new UserInputError('No text content found for conversion.');
+            }
+            const chunks = noAutoProsody ? textChunks : textChunks.map(buildAutoProsodyText);
             if (chunks.length === 1) {
                 // Single say command
                 const sayArgs = ['-o', outputPath];
-                if (opts.voice)
-                    sayArgs.push('-v', opts.voice);
-                if (opts.rate)
-                    sayArgs.push('-r', opts.rate);
+                if (voice)
+                    sayArgs.push('-v', voice);
+                if (rate)
+                    sayArgs.push('-r', rate);
                 const tmpFile = path.join(fs.mkdtempSync('docs-to-voice-'), 'input.txt');
                 fs.mkdirSync(path.dirname(tmpFile), { recursive: true });
                 fs.writeFileSync(tmpFile, chunks[0], 'utf-8');
@@ -525,7 +432,7 @@ Options:
                 }
                 catch (err) {
                     const msg = err instanceof Error ? err.message : 'unknown error';
-                    throw new Error(`say mode failed: ${msg}`);
+                    throw new SystemError(`say mode failed: ${msg}`);
                 }
                 finally {
                     try {
@@ -543,15 +450,15 @@ Options:
                 try {
                     for (let i = 0; i < chunks.length; i++) {
                         const partPath = path.join(tempDir, `part-${String(i + 1).padStart(4, '0')}${partExt}`);
-                        const sayArgs = ['-o', partPath];
-                        if (opts.voice)
-                            sayArgs.push('-v', opts.voice);
-                        if (opts.rate)
-                            sayArgs.push('-r', opts.rate);
-                        const tmpFile = path.join(tempDir, `chunk-${i}.txt`);
-                        fs.writeFileSync(tmpFile, chunks[i], 'utf-8');
-                        sayArgs.push('-f', tmpFile);
-                        execSync(`say ${sayArgs.map((a) => (a.includes(' ') ? `"${a}"` : a)).join(' ')}`, { stdio: 'ignore', timeout: 300000 });
+                        const sArgs = ['-o', partPath];
+                        if (voice)
+                            sArgs.push('-v', voice);
+                        if (rate)
+                            sArgs.push('-r', rate);
+                        const tFile = path.join(tempDir, `chunk-${i}.txt`);
+                        fs.writeFileSync(tFile, chunks[i], 'utf-8');
+                        sArgs.push('-f', tFile);
+                        execSync(`say ${sArgs.map((a) => (a.includes(' ') ? `"${a}"` : a)).join(' ')}`, { stdio: 'ignore', timeout: 300000 });
                         partPaths.push(partPath);
                     }
                     concatAudioFiles(partPaths, outputPath);
@@ -564,10 +471,10 @@ Options:
                 }
             }
             // Apply speech rate if requested
-            if (opts.speechRate) {
-                const rate = parseFloat(opts.speechRate);
-                if (rate > 0)
-                    applySpeechRateToAudio(outputPath, rate);
+            if (speechRate) {
+                const rateVal = parseFloat(speechRate);
+                if (rateVal > 0)
+                    applySpeechRateToAudio(outputPath, rateVal);
             }
             // Write timeline files
             writeTimelineFiles(sourceText, outputPath, null);
@@ -575,23 +482,20 @@ Options:
         }
         else {
             // API mode
-            const apiKey = opts.apiKey;
             if (!apiKey) {
-                stderr.write('Error: --api-key is required for api mode.\n');
-                return 1;
+                throw new UserInputError('--api-key is required for api mode.');
             }
             const sentences = splitSentences(sourceText);
             if (sentences.length === 0) {
-                stderr.write('Error: No text content found for conversion.\n');
-                return 1;
+                throw new UserInputError('No text content found for conversion.');
             }
-            const maxChars = opts.maxChars ? parseInt(opts.maxChars, 10) || null : null;
+            const maxCharsNum = maxChars ? parseInt(maxChars, 10) || null : null;
             const requestItems = [];
             for (let si = 0; si < sentences.length; si++) {
                 const sentence = sentences[si];
-                if (maxChars && sentence.length > maxChars) {
-                    for (let i = 0; i < sentence.length; i += maxChars) {
-                        requestItems.push({ sentenceIndex: si, text: sentence.slice(i, i + maxChars) });
+                if (maxCharsNum && sentence.length > maxCharsNum) {
+                    for (let i = 0; i < sentence.length; i += maxCharsNum) {
+                        requestItems.push({ sentenceIndex: si, text: sentence.slice(i, i + maxCharsNum) });
                     }
                 }
                 else {
@@ -599,8 +503,7 @@ Options:
                 }
             }
             if (requestItems.length === 0) {
-                stderr.write('Error: No text content found for conversion.\n');
-                return 1;
+                throw new UserInputError('No text content found for conversion.');
             }
             const tempDir = fs.mkdtempSync('docs-to-voice-api-');
             const partPaths = [];
@@ -610,7 +513,7 @@ Options:
             try {
                 for (let i = 0; i < requestItems.length; i++) {
                     const item = requestItems[i];
-                    const apiResult = await requestAlibabaCloudTTS(opts.apiEndpoint, apiKey, opts.apiModel, opts.apiVoice, item.text);
+                    const apiResult = await requestAlibabaCloudTTS(apiEndpoint, apiKey, apiModel, apiVoice, item.text);
                     const currentExt = apiResult.audioFormat || 'wav';
                     if (!partExt)
                         partExt = currentExt;
@@ -622,10 +525,10 @@ Options:
                         fs.writeFileSync(partPath, Buffer.from(apiResult.audioData, 'base64'));
                     }
                     else {
-                        throw new Error('No audio data in API response.');
+                        throw new SystemError('No audio data in API response.');
                     }
                     if (!fs.existsSync(partPath) || fs.statSync(partPath).size === 0) {
-                        throw new Error(`Failed to generate audio chunk ${i + 1}.`);
+                        throw new SystemError(`Failed to generate audio chunk ${i + 1}.`);
                     }
                     partPaths.push(partPath);
                     const partDuration = readDurationSeconds(partPath);
@@ -637,12 +540,11 @@ Options:
                     }
                 }
                 const finalOutputName = hasExtension
-                    ? outputName
-                    : `${outputName}.${partExt || 'wav'}`;
+                    ? outName
+                    : `${outName}.${partExt || 'wav'}`;
                 const outputPath = path.join(outputDir, finalOutputName);
-                if (fs.existsSync(outputPath) && !opts.force) {
-                    stderr.write(`Error: Output already exists: ${outputPath}. Use --force to overwrite.\n`);
-                    return 1;
+                if (fs.existsSync(outputPath) && !force) {
+                    throw new UserInputError(`Output already exists: ${outputPath}. Use --force to overwrite.`);
                 }
                 concatAudioFiles(partPaths, outputPath);
                 // Build timeline durations
@@ -670,12 +572,12 @@ Options:
                     }
                 }
                 // Apply speech rate if requested
-                if (opts.speechRate) {
-                    const rate = parseFloat(opts.speechRate);
-                    if (rate > 0) {
-                        applySpeechRateToAudio(outputPath, rate);
+                if (speechRate) {
+                    const rateVal = parseFloat(speechRate);
+                    if (rateVal > 0) {
+                        applySpeechRateToAudio(outputPath, rateVal);
                         if (timelineDurations) {
-                            timelineDurations = timelineDurations.map((d) => d / rate);
+                            timelineDurations = timelineDurations.map((d) => d / rateVal);
                         }
                     }
                 }
@@ -690,16 +592,11 @@ Options:
             }
         }
         return 0;
-    }
-    catch (err) {
-        const msg = err instanceof Error ? err.message : 'Unknown error';
-        stderr.write(`Error: ${msg}\n`);
-        return 1;
-    }
-}
+    },
+};
 export const tool = {
     name: 'docs-to-voice',
     category: 'media',
     description: 'Convert text into audio and sentence timelines.',
-    handler: docsToVoiceHandler,
+    handler: createToolRunner(schema),
 };