#!/bin/bash
# MiniMax TTS → 钉钉语音消息
# 依赖: xxd, ffmpeg

set -e

# 配置
MINIMAX_API_KEY="sk-cp-WDKUJN0CM7byxgNX8nzr5E8JOe0c3jZP_YVBt200sbYt9bEqsRAeY4O7VldTSg0RBYhNvneKLIttYHDy3YM6m04XWAz4JRW0ABlFHSKXKpuPgZPU02k0MfY"
DINGTALK_APP_KEY="dingklemniq8uqk5qbgx"
DINGTALK_APP_SECRET="_8EHgyhvHRHRMx6fZbh9LNpQoxyYl3At0b-fXXlQiahwupbt9oY5P6Grj8IM9Dx8"
USER_ID="121922510028034588"

# 临时文件
TMP_DIR="/tmp/dingtalk_voice_$$"
mkdir -p "$TMP_DIR"

cleanup() {
    rm -rf "$TMP_DIR"
}
trap cleanup EXIT

# Step 1: 获取钉钉 Access Token
echo "[1/6] 获取钉钉 Access Token..."
TOKEN_RESPONSE=$(curl -s -X POST 'https://api.dingtalk.com/v1.0/oauth2/accessToken' \
    -H 'Content-Type: application/json' \
    -d "{\"appKey\":\"$DINGTALK_APP_KEY\",\"appSecret\":\"$DINGTALK_APP_SECRET\"}")

ACCESS_TOKEN=$(echo "$TOKEN_RESPONSE" | python3 -c "import sys,json; print(json.load(sys.stdin).get('accessToken',''))")
if [ -z "$ACCESS_TOKEN" ]; then
    echo "❌ 获取 Access Token 失败: $TOKEN_RESPONSE"
    exit 1
fi
echo "✓ 获取成功"

# Step 2: 调用 MiniMax TTS
TEXT="$1"
if [ -z "$TEXT" ]; then
    echo "用法: $0 <要转换的文本>"
    exit 1
fi

echo "[2/6] 生成 TTS 音频..."
TTS_RESPONSE=$(curl -s -X POST "https://api.minimaxi.com/v1/t2a_v2" \
    -H "Authorization: Bearer $MINIMAX_API_KEY" \
    -H "Content-Type: application/json" \
    -d "{
        \"model\": \"speech-2.8-hd\",
        \"text\": \"$TEXT\",
        \"stream\": false,
        \"voice_setting\": {
            \"voice_id\": \"female-tianmei\",
            \"speed\": 1,
            \"vol\": 1,
            \"pitch\": 0,
            \"emotion\": \"happy\"
        },
        \"audio_setting\": {
            \"sample_rate\": 32000,
            \"bitrate\": 128000,
            \"format\": \"mp3\",
            \"channel\": 1
        }
    }")

# 用 Python 解析响应（处理特殊字符）
TTS_RESULT=$(python3 << PYEOF
import json
import sys
data = json.loads("""$TTS_RESPONSE""")
hex_audio = data.get('data', {}).get('audio', '')
audio_len = data.get('extra_info', {}).get('audio_length', 0)
print(f"{hex_audio}\n{audio_len}")
PYEOF
)

HEX_AUDIO=$(echo "$TTS_RESULT" | head -1)
AUDIO_LENGTH=$(echo "$TTS_RESULT" | tail -1)

if [ -z "$HEX_AUDIO" ]; then
    echo "❌ TTS 调用失败: $TTS_RESPONSE"
    exit 1
fi
echo "✓ TTS 生成成功，时长: ${AUDIO_LENGTH}ms"

# Step 3: 转换格式 hex → MP3
echo "[3/6] 转换格式..."
echo "$HEX_AUDIO" | xxd -r -p > "$TMP_DIR/voice.mp3"

# 获取时长（秒）
DURATION=$(ffprobe -v error -show_entries format=duration -of default=noprint_wrappers=1:nokey=1 "$TMP_DIR/voice.mp3" 2>/dev/null | awk '{printf "%.0f", $1}')
if [ -z "$DURATION" ] || [ "$DURATION" = "0" ]; then
    DURATION=$(echo "scale=0; $AUDIO_LENGTH / 1000" | bc)
fi
echo "✓ MP3 生成完成，时长: ${DURATION}秒"

# Step 4: 上传 MP3 到钉钉
echo "[4/6] 上传音频到钉钉..."
UPLOAD_RESPONSE=$(curl -s -X POST "https://oapi.dingtalk.com/media/upload?access_token=$ACCESS_TOKEN&type=voice" \
    -F "media=@$TMP_DIR/voice.mp3;type=audio/mpeg;filename=voice.mp3")

MEDIA_ID=$(echo "$UPLOAD_RESPONSE" | python3 -c "import sys,json; print(json.load(sys.stdin).get('media_id',''))")
if [ -z "$MEDIA_ID" ]; then
    echo "❌ 上传失败: $UPLOAD_RESPONSE"
    exit 1
fi
echo "✓ 上传成功，mediaId: $MEDIA_ID"

# Step 5: 发送语音消息
echo "[5/6] 发送语音消息..."
SEND_RESPONSE=$(curl -s -X POST "https://api.dingtalk.com/v1.0/robot/oToMessages/batchSend" \
    -H "Content-Type: application/json" \
    -H "x-acs-dingtalk-access-token: $ACCESS_TOKEN" \
    -d "{
        \"robotCode\": \"$DINGTALK_APP_KEY\",
        \"userIds\": [\"$USER_ID\"],
        \"msgKey\": \"sampleAudio\",
        \"msgParam\": \"{\\\"mediaId\\\":\\\"$MEDIA_ID\\\",\\\"duration\\\":\\\"$DURATION\\\"}\"
    }")

PROCESS_KEY=$(echo "$SEND_RESPONSE" | python3 -c "import sys,json; print(json.load(sys.stdin).get('processQueryKey',''))")
if [ -n "$PROCESS_KEY" ]; then
    echo "✓ 语音消息发送成功！"
    echo "[6/6] 完成 ✓"
else
    echo "❌ 发送失败: $SEND_RESPONSE"
    exit 1
fi