Regenerate Audio Example

Regenerate a single demo audio example for EnglishConnect lessons.

Syntax Parsing

Parse user requests into lesson, pattern, and example:

| User says | Lesson | Pattern | Example | |-----------|--------|---------|---------| | L15 > P1 > E2 | 15 | 1 | 2 | | lesson 15 pattern 1 example 2 | 15 | 1 | 2 | | regenerate the 3rd example of pattern 2 in lesson 15 | 15 | 2 | 3 |

Prerequisites

TTS server must be running. If not, run /start-tts first.

Command

cd /home/smolen/dev/EnglishConnect/src/tools/demo-generator && \
  source /home/smolen/dev/EnglishConnect/src/backend/.venv/bin/activate && \
  python regenerate_example.py --lesson LESSON --pattern PATTERN --example EXAMPLE

Replace LESSON, PATTERN, EXAMPLE with parsed values.

Output

Files are saved to content/audio/ec1/demos/lesson-{NN}/ with deterministic names:

• lesson{NN}-pattern{N}-ex{N}.wav - audio file
• lesson{NN}-pattern{N}-ex{N}.json - metadata

Example: lesson15-pattern1-ex3.wav

Regenerating overwrites the existing file (no unique hashes).

Optional: Change Voices

Add --teacher-voice VOICE or --student-voice VOICE:

• speaker_a - Aria (female)
• speaker_b - Emma (female, upbeat)
• speaker_c - Davis (male)
• speaker_d - Grace (female, default teacher)
• speaker_e - James (male)
• speaker_f - Lily (female)

List Voices

cd /home/smolen/dev/EnglishConnect/src/tools/demo-generator && \
  source /home/smolen/dev/EnglishConnect/src/backend/.venv/bin/activate && \
  python regenerate_example.py --list-voices