Go Say

You write it, we say it.

Learning goals

• Use of "os/exec" to both run and manage an external Text to Speech program and to play the output from the host's audio card
  • Use of os.Pipe
• Use of https://github.com/spf13/viper and https://github.com/spf13/cobra
• Queue structure for TTS
• Cancel context for os.CommandCtx operations
• Packaging a golang binary into a .deb archive for amd64 and arm64

Resources

• https://www.dolthub.com/blog/2022-11-28-go-os-exec-patterns/
• https://github.com/spf13/cobra/blob/v1.7.0/user_guide.md#user-guide
• Piper
• Aplay

CLI API design

• gosay "will say this text": Uses Piper TTS and aplay to read the argument aloud
  • arg -m || -model: use provided model by name or path. Assumed model config resides next to model
• echo "Will also say this text" | gosay: Same as above but reading from stdin
• gosay list models: Prints a list of available models sourced from piper project. Should print an * for models already present in model directory
• gosay list models --installed: Prints a list of models present in the user's GOSAY_MODELS directory. Defaults to ${HOME}/.config/gosay/model gosay download model <model name>: Downloads the piper model and .config file from piper to GOSAY_MODELS directroy gosay set-default <model name>: Sets model as default model and persists setting to ${HOME}/.config/gosay/config.json gosay server -port 6543: Starts an HTTP server on port (default 6543) gosay check: Checks that all dependencies and configuration are setup. If an error occurs a helpful message is printed to the console for the user.