Quick Guide

Contents

Quick Guide#

This project was conducted by Yu-Lin Chen at the Institute of Information Science, Academia Sinica (2025).
Use the sidebar to switch chapters, or click the links below for key sections.
The goal is to build a scalable, handoff-friendly, controllable, and verifiable Qt .ts translation workflow.


Quick Start#

  1. Go to Online GPT (or use Local LLM).

  2. Configure the translation model:

    • Online: enter your API Key, confirm Base URL and Model

    • Local: configure the model and generate a Notebook for local execution

  3. Upload the .ts file to be translated.

  4. (Optional) Upload a Glossary (CSV / ODS, columns: en, zh or English Term, Chinese Term).

  5. (Optional) Upload a previous .ts file: when <source> is identical, existing translations will be reused to reduce cost.

  6. Set Batch mode and the maximum number of entries, and run a small test first.

  7. After confirming format and results, increase the batch size and download the output qgis_zh-Hant.ts.

Tip

You may upload multiple glossary files; they will be merged automatically.
It is recommended to test with a small sample (e.g., 100–300 entries) before running the full translation.