glance-cli 0.8.6 → 0.8.8

package/CHANGELOG.md

@@ -5,6 +5,43 @@ All notable changes to glance-cli will be documented in this file.
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
 and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
+## [0.8.8] - 2026-01-02
+
+### Fixed
+- **Voice System Improvements**: Fixed voice listing to show only working voices with proper language mapping
+- **Text Cleaner Optimization**: Made text cleaning less aggressive to prevent legitimate content removal
+- **Cache System**: Resolved overly aggressive cleaning of cache-related terms and content
+- **--prefer-quality Flag**: Fixed issue where flag wasn't affecting automatic model selection
+- **Voice Validation**: Improved voice name/ID validation and error messages
+
+### Improved
+- **Voice Listing**: Clean, organized display by language (🇺🇸 English, 🇫🇷 French, 🇪🇸 Spanish, 🇭🇹 Haitian Creole)
+- **Error Messages**: More specific and helpful voice synthesis error messages
+- **Text Processing**: Smarter cleaning that preserves paragraph structure and legitimate content
+- **Model Selection**: --prefer-quality now correctly prioritizes OpenAI/Gemini over Ollama
+
+### Technical Details
+- Enhanced voice mapping system to show only available voices in user's ElevenLabs account
+- Refined binary artifact detection to avoid false positives from ANSI color codes
+- Improved cache system stability with better error handling
+- More conservative text cleaning patterns to preserve legitimate content
+
+## [0.8.7] - 2026-01-02
+
+### Fixed
+- **Smart Text Cleaning**: Resolved aggressive text cleaning that was destroying paragraph structure
+- **Paragraph Formatting**: Restored beautiful paragraph breaks and section formatting from v0.8.3
+- **Binary Artifact Detection**: Fixed false positives from ANSI color codes in formatted output
+- **AI Response Processing**: Preserves natural paragraph structure in AI-generated summaries
+- **Terminal Output**: Smart sanitization that only applies nuclear cleaning when actual artifacts detected
+- Perfect balance: Maintains v0.8.6 security features while restoring v0.8.3 formatting beauty
+
+### Technical Details
+- Modified `sanitizeOutputForTerminal()` to ignore ANSI escape sequences in artifact detection
+- Updated summarizer functions to preserve paragraph structure in clean AI responses
+- Implemented conditional cleaning: nuclear mode only for actual binary artifacts
+- Smart cleaning preserves newlines, sections, and natural content flow
+
 ## [0.8.6] - 2026-01-02
 
 ### Fixed

package/README.md

@@ -118,6 +118,15 @@ glance https://article.com --audio-output summary.mp3
 
 # Read aloud immediately
 glance https://blog.com --read
+
+# Use specific voice
+glance https://news.com --voice nova --read
+
+# French content with French voice
+glance https://lemonde.fr --voice antoine -l fr --read
+
+# Spanish content with Spanish voice
+glance https://elpais.com --voice isabella -l es --read
 ```
 
 ### Advanced Features
@@ -216,6 +225,36 @@ glance https://api-docs.com --ask "How do I authenticate?" --format plain
 
 ---
 
+## 🎤 Voice & Language Support
+
+Glance supports **multi-language voice synthesis** with ElevenLabs integration:
+
+```bash
+# List available voices by language
+glance --list-voices
+
+# English voices (default)
+glance https://news.com --voice nova --read       # Energetic
+glance https://docs.com --voice onyx --read       # Deep, authoritative
+
+# French voices  
+glance https://lemonde.fr --voice antoine -l fr --read
+
+# Spanish voices
+glance https://elpais.com --voice isabella -l es --read
+
+# Haitian Creole support
+glance https://example.ht --voice nova -l ht --read
+```
+
+**Voice Quality Levels:**
+- **Free**: System TTS (macOS Say, Windows SAPI, Linux espeak)
+- **Premium**: ElevenLabs voices with natural pronunciation (requires API key)
+
+Set `ELEVENLABS_API_KEY` environment variable for premium voices.
+
+---
+
 ## 📚 All Options
 
 ```bash
@@ -243,8 +282,9 @@ glance <url> --ask "question" # Custom Q&A
 # Voice/Audio
 --read, --speak               # Read aloud (text-to-speech)
 --audio-output <file>         # Save as MP3
---voice <name>                # Choose voice (alloy, echo, nova, etc.)
---list-voices                 # Show available voices
+--voice <name>                # Choose voice (nova, onyx, antoine, etc.)
+--list-voices                 # Show available voices by language
+-l, --language <code>         # Language for voice selection (en, fr, es, ht)
 
 # Advanced
 --screenshot <file>           # Capture screenshot