Skip to main content

docling-av-transcriber

独立的 Docling 兼容音/视频转结构化文档库,复用 Docling 文档模型,同时把底层语音识别切换为阿里云百炼 ASR(Paraformer、SenseVoice 等)。

特性

  • 支持 WAV/MP3/FLAC 等音频,以及 MP4/AVI/MOV 等视频(先抽取音轨)
  • AliyunBailianAsrClient 统一封装百炼 ASR 接口,默认使用 paraformer-v1
  • 输出 DoclingDocument,可无缝接入 Docling 生态
  • 通过环境变量 ALIYUN_BAILIAN_API_KEY 管理密钥
  • 结构化目录,易于扩展或替换模型

快速开始

pip install -e .  # 或普通 pip install
export ALIYUN_BAILIAN_API_KEY=sk-xxxx
python examples/basic_usage.py sample.mp4 --language zh

调试模式

为了更好地诊断问题,我们提供了增强日志的调试脚本:

python examples/debug_usage.py sample.mp4 --language zh --debug

调试日志将输出到控制台和 debug_transcription.log 文件中,包含详细的处理步骤和错误信息。

模块

  • docling_av_transcriber.api:顶层 API (transcribe_file/transcribe_bytes)
  • docling_av_transcriber.models:ASR 客户端抽象与百炼实现
  • docling_av_transcriber.media:输入验证与音轨抽取
  • docling_av_transcriber.pipelines:轻量 Pipeline 与 DoclingDocument 构建

获取抽取音轨与 Docling 文档

在 RAG / 检索场景中,通常需要同时拿到结构化文本与抽取后的 WAV 音轨以便上传 OSS 或 Supabase。
自 v0.x 起可以直接调用新的 *_with_artifacts API:

from docling_av_transcriber import transcribe_file_with_artifacts

result = transcribe_file_with_artifacts("sample.mp4", language="zh")
doc = result.document          # DoclingDocument,可直接入库或向量化
wav_path = result.audio_path   # 16kHz/单声道 WAV 临时文件

# 将 wav_path 上传到自定义存储后即可手动清理
# wav_path.unlink(missing_ok=True)

若是内存字节流,可调用 transcribe_bytes_with_artifacts(data, filename="input.mp3"),内部会先写入临时文件并统一转码到 16kHz/单声道 WAV。
需要注意 audio_path 来自 tempfile.mkstemp,生命周期由调用方管理:将文件上传或复制后请记得删除,以避免 /tmp 目录堆积。

测试

pip install -e .[dev]
pytest -q

发布

  1. python -m build
  2. twine check dist/*
  3. twine upload dist/*

更多细节见 CHANGELOG.md 与示例脚本。

Metadata

Release files for docling-av-transcriber 0.1.2

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Built distribution (wheel)

Table of built distributions (wheels) for docling-av-transcriber 0.1.2
File Interpreter ABI Platform
docling_av_transcriber-0.1.2-py3-none-any.whl Python 3 none any Details

Release files / docling_av_transcriber-0.1.2-py3-none-any.whl

Download URL docling_av_transcriber-0.1.2-py3-none-any.whl
Size 23.4 kB
Tags Python 3
SHA-256 checksum
How to use checksums
2c9a6d09284052a393d9be1b018995ffd449afbfd09fcd10210b75a0ea5ea60d
BLAKE2b-256 checksum
How to use checksums
f09093c21bb627ada0638087fb9a1eb93b884d4ac07f256a65d549032ba7afbd
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/6.2.0 CPython/3.12.8

Security for docling-av-transcriber 0.1.2

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

No provenance found

None of the files in this release were uploaded with provenance metadata.

A note on package security

PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.

To learn more about protecting yourself from malicious actors, visit our package security documentation.

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page