Audiobook Canvas is an Android application that converts written text into a narrated audiobook. The app breaks a text file into manageable pieces, identifies the speaker for each line, generates speech audio, and stitches everything into a final audio file.

• Text chunking and sanitizing – The app reads a text file, splits it into small blocks, and cleans up OCR artifacts or extra whitespace. It can handle languages where the dialogue start and end markers are the same.
• Character recognition – For each text block, the app sends a request to an OpenAI completion endpoint to perform Named Entity Recognition. This determines which character speaks each line of dialogue.
• Text to speech – Lines can be synthesized either by Android's built‑in TextToSpeech or by OpenAI's speech synthesis endpoint. Voice samples are stored so they can be reused across projects.
• Audio mixing – Individual WAV files are combined into a single audio book using the android_audio_mixer library. Background music can be added in parallel with narration.
• Room database – Projects, text blocks, and generated audio files are persisted locally using Room. Repositories and view models provide a clean separation between data and UI layers.
• Navigation architecture – A single navigation graph describes the user flow from creating a project to listening to the final mix.

The app relies on several external services and libraries:

• OpenAI APIs – Used for character recognition and optional text‑to‑speech synthesis.
• OkHttp and Gson – Handle HTTP requests and JSON parsing.
• android_audio_mixer – Mixes multiple WAV tracks into a final audio file.
• Android TextToSpeech – Provides offline speech synthesis.

To enable OpenAI calls, users must provide their API key in the app settings.

app/                 # Android application module
├── src/main/java/   # Application code (repositories, view models, UI)
├── src/main/res/    # Layouts, navigation graph, strings
arch/                # Design documents and SQL model
build.gradle         # Project build configuration

This project uses Gradle. Run ./gradlew assembleDebug to build the debug APK. Running the full unit test suite requires network access for Gradle to download dependencies.

1. Clone the repository and open it in Android Studio.
2. Set your OpenAI API key in the app's preferences screen.
3. Create a new project inside the app and select a text file to process.
4. Review and generate character lines, then synthesize speech and mix the final audiobook.