xiaoxia/xiaoxia-saas

fix: CosyVoice全量重写 - 适配DashScope百炼API #212

Merged

xiaoxia merged 2 commits from fix/cosyvoice-full-rewrite into develop

2026-07-10 22:52:22 +08:00

Author	SHA1	Message	Date
用户CI Test	1d2aaad52f	fix: PR #212 P0+P1 问题修复 CI/CD Pipeline / Validate Code Quality And Tests (pull_request) Failing after 13s Details CI/CD Pipeline / Frontend Lint (pull_request) Successful in 2m46s Details CI/CD Pipeline / Build & Push Staging (Watchtower auto-deploy) (pull_request) Has been skipped Details CI/CD Pipeline / Build Production Runtime Images (pull_request) Has been skipped Details CI/CD Pipeline / Staging E2E Tests (pull_request) Has been skipped Details CI/CD Pipeline / Staging API Integration Tests (pull_request) Has been skipped Details CI/CD Pipeline / Deploy Production (pull_request) Has been skipped Details CI/CD Pipeline / Production Browser E2E (pull_request) Has been skipped Details - P0: 修复 process_synthesis_failure docstring 语法错误（三引号位置错误） - P1: _resynthesize_and_complete 补上 speed/volume 参数传递（从metadata读取） - P1: _poll_segment_tasks 适配新同步接口，缺失分段重新同步合成而非轮询 - 移除 time 依赖（不再轮询） - 并发重新合成缺失 audio_url 的分段 - 已有 audio_url 的分段直接复用 - 新增 1 个分段复用测试用例	2026-07-10 22:40:20 +08:00
用户CI Test	506eae98f1	fix: CosyVoice全量重写 - 适配DashScope百炼API CI/CD Pipeline / Validate Code Quality And Tests (pull_request) Failing after 4s Details CI/CD Pipeline / Frontend Lint (pull_request) Successful in 4m18s Details CI/CD Pipeline / Build & Push Staging (Watchtower auto-deploy) (pull_request) Has been skipped Details CI/CD Pipeline / Build Production Runtime Images (pull_request) Has been skipped Details CI/CD Pipeline / Staging E2E Tests (pull_request) Has been skipped Details CI/CD Pipeline / Staging API Integration Tests (pull_request) Has been skipped Details CI/CD Pipeline / Deploy Production (pull_request) Has been skipped Details CI/CD Pipeline / Production Browser E2E (pull_request) Has been skipped Details - 配置更新：base_url改为api/v1，model改为cosyvoice-v3.5-plus，新增clone_model=voice-enrollment - 音色克隆：适配customization接口 + voice-enrollment模型 + create_voice action - 状态查询：适配同路径query_voice action（POST），返回DEPLOYING/OK/UNDEPLOYED - 语音合成：适配SpeechSynthesizer非流式同步接口，直接返回audio_url - 新增audio_url_signer钩子：对私有bucket音频URL预签名（slash_safe=True） - voice_clone workflow适配：voice_id替代task_id，status=OK标记ready - tts_job workflow适配：同步接口直接返回，无task_id时自动重新合成兜底 - API端+Worker端依赖注入OSS预签名函数 - 单元测试：36个service测试 + 18个workflow测试，全部通过	2026-07-10 22:17:07 +08:00