AWS Polly TTS demo

polly.sh

#!/usr/bin/env bash
#
# AWS Polly TTS demo
# Usage: polly.sh [--voice ID] [--out FILE] [--no-play] "text" | echo text | polly.sh
#
set -euo pipefail
export AWS_PAGER=""  # no less(1) between you and the JSON
REGION="${AWS_REGION:-eu-west-1}"
VOICE=Vicki; OUT=out.mp3; PLAY=1
PLAY_CMD=""; for c in afplay paplay aplay ffplay; do command -v "$c" >/dev/null && PLAY_CMD="$c" && break; done
while [ $# -gt 0 ]; do
  case "$1" in
    --voice)  VOICE="$2"; shift 2 ;;
    --out)    OUT="$2"; shift 2 ;;
    --no-play) PLAY=0; shift ;;
    -h|--help) sed -n '2,3p' "$0"; exit 0 ;;
    *) TEXT="${TEXT:+$TEXT }$1"; shift ;;
  esac
done
TEXT="${TEXT:-$(cat)}"
[ -n "$TEXT" ] || { echo "no text given" >&2; exit 1; }
aws sts get-caller-identity >/dev/null 2>&1 || { echo "AWS session not active. Run: aws login" >&2; exit 1; }
mkdir -p "$(dirname "$OUT")"
aws --region "$REGION" polly synthesize-speech --voice-id "$VOICE" --engine neural \
  --output-format mp3 --text "$TEXT" "$OUT"
if [ "$PLAY" = 1 ] && [ -n "$PLAY_CMD" ]; then
  case "$PLAY_CMD" in
    ffplay) ffplay -nodisp -autoexit "$OUT" >/dev/null 2>&1 ;;
    *)      "$PLAY_CMD" "$OUT" ;;
  esac
fi
ls -lh "$OUT"