pdf2zh

English | [简体中文](README_zh-CN.md)

PDFMathTranslate

PDF scientific paper translation and bilingual comparison.

• 📊 Preserve formulas, charts, table of contents, and annotations (preview).
• 🌐 Support multiple languages, and diverse translation services.
• 🤖 Provides commandline tool, interactive user interface, and Docker

Feel free to provide feedback in GitHub Issues, Telegram Group or QQ Group.

Updates

• [Nov. 26 2024] CLI now supports online file(s) (by @reycn)
  
• [Nov. 24 2024] ONNX support to reduce dependency sizes (by @Wybxc)
  
• [Nov. 23 2024] 🌟 Public Service online! (by @Byaidu)
  
• [Nov. 23 2024] Firewall for preventing web bots (by @Byaidu)
  
• [Nov. 22 2024] GUI now supports Italian, and has been improved (by @Byaidu, @reycn)
  
• [Nov. 22 2024] You can now share your deployed service to others (by @Zxis233)
  
• [Nov. 22 2024] Now supports Tencent Translation (by @hellofinch)
  
• [Nov. 21 2024] GUI now supports downloading dual-document (by @reycn)
  
• [Nov. 20 2024] 🌟 Demo online! (by @reycn)
  

Preview

Public Service 🌟

Free Service (https://pdf2zh.com/)

You can try our public service online without installation.

Hugging Face Demo

You can try our demo on HuggingFace without installation. Note that the computing resources of the demo are limited, so please avoid abusing them.

Installation and Usage

We provide three methods for using this project: Commandline, GUI, and Docker.

Method I. Commandline

1. Python installed (3.8 <= version <= 3.12)

2. Install our package:

     pip install pdf2zh
   

3. Execute translation, files generated in current working directory:

     pdf2zh document.pdf
   

Method II. GUI

1. Python installed (3.8 <= version <= 3.12)

2. Install our package:

     pip install pdf2zh
   

3. Start using in browser:

     pdf2zh -i
   

4. If your browswer has not been started automatically, goto

   http://localhost:7860/
   

   

See documentation for GUI for more details.

Method III. Docker

1. Pull and run:

   docker pull byaidu/pdf2zh
   docker run -d -p 7860:7860 byaidu/pdf2zh
   

2. Open in browser:

   http://localhost:7860/
   

For docker deployment on cloud service:

Advanced Options

Execute the translation command in the command line to generate the translated document example-zh.pdf and the bilingual document example-dual.pdf in the current working directory. Use Google as the default translation service.

In the following table, we list all advanced options for reference:

Option	Function	Example
files	Local files	pdf2zh ~/local.pdf
links	Online files	pdf2zh http://arxiv.org/paper.pdf
-i	Enter GUI	pdf2zh -i
-p	Partial document translation	pdf2zh example.pdf -p 1
-li	Source language	pdf2zh example.pdf -li en
-lo	Target language	pdf2zh example.pdf -lo zh
-s	Translation service	pdf2zh example.pdf -s deepl
-t	Multi-threads	pdf2zh example.pdf -t 1
-o	Output dir	pdf2zh example.pdf -o output
-f, -c	Exceptions	pdf2zh example.pdf -f "(MS.*)"

Some services require setting environmental variables.

Full / partial document translation

• Entire document

  pdf2zh example.pdf
  

• Part of the document

  pdf2zh example.pdf -p 1-3,5
  

Specify source and target languages

See Google Languages Codes, DeepL Languages Codes

pdf2zh example.pdf -li en -lo ja

Translate with Different Services

• DeepL

See DeepL

Set ENVs to construct an endpoint like: {DEEPL_SERVER_URL}/translate

• DEEPL_SERVER_URL (Optional), e.g., export DEEPL_SERVER_URL=https://api.deepl.com

• DEEPL_AUTH_KEY, e.g., export DEEPL_AUTH_KEY=xxx

  pdf2zh example.pdf -s deepl
  

• DeepLX

See DeepLX

Set ENVs to construct an endpoint like: {DEEPL_SERVER_URL}/translate

• DEEPLX_SERVER_URL (Optional), e.g., export DEEPLX_SERVER_URL=https://api.deeplx.org

• DEEPLX_AUTH_KEY, e.g., export DEEPLX_AUTH_KEY=xxx

  pdf2zh example.pdf -s deeplx
  

• Ollama

See Ollama

Set ENVs to construct an endpoint like: {OLLAMA_HOST}/api/chat

• OLLAMA_HOST (Optional), e.g., export OLLAMA_HOST=https://localhost:11434

  pdf2zh example.pdf -s ollama:gemma2
  

• LLM with OpenAI compatible schemas (OpenAI / SiliconCloud / Zhipu)

See SiliconCloud, Zhipu

Set ENVs to construct an endpoint like: {OPENAI_BASE_URL}/chat/completions

• OPENAI_BASE_URL (Optional), e.g., export OPENAI_BASE_URL=https://api.openai.com/v1

• OPENAI_API_KEY, e.g., export OPENAI_API_KEY=xxx

  pdf2zh example.pdf -s openai:gpt-4o
  

• Azure

See Azure Text Translation

Following ENVs are required:

• AZURE_APIKEY, e.g., export AZURE_APIKEY=xxx
• AZURE_ENDPOINT, e.g, export AZURE_ENDPOINT=https://api.translator.azure.cn/

• AZURE_REGION, e.g., export AZURE_REGION=chinaeast2

  pdf2zh example.pdf -s azure
  

• Tencent Machine Translation

See Tencent Machine Translation

Following ENVs are required:

• TENCENT_SECRET_ID, e.g., export TENCENT_SECRET_ID=AKIDxxx

• TENCENT_SECRET_KEY, e.g, export TENCENT_SECRET_KEY=xxx

  pdf2zh example.pdf -s tencent
  

Translate wih exceptions

Use regex to specify formula fonts and characters that need to be preserved:

pdf2zh example.pdf -f "(CM[^RT].*|MS.*|.*Ital)" -c "(\(|\||\)|\+|=|\d|[\u0080-\ufaff])"

Specify threads

Use -t to specify how many threads to use in translation:

pdf2zh example.pdf -t 1

TODO

• [ ] Parse layout with PaddleX, PaperMage, SAM2

• [ ] Fix page rotation, table of contents, format of list

• [ ] Fix pixel formula in old paper

• [ ] Support multiple language with Noto Font, Ubuntu Font

• [ ] Retry except KeyboardInterrupt

Acknowledgements

• Document merging: PyMuPDF

• Document parsing: Pdfminer.six

• Document extraction: MinerU

• Multi-threaded translation: MathTranslate

• Layout parsing: DocLayout-YOLO

• Document standard: PDF Explained, PDF Cheat Sheets

Contributors

Star History