Selextrans

It is an easy-to-use app which can translate selected text without copying manually.

Python

This app is developed with Python 3.10.

Platform Support

• Linux (Debian/Ubuntu). If you use other Linux distributions, I suggest you to checkout the Dependencies carefully. The latter part of the document will only explain the use of apt for dependency configuration.
• Windows

Quick Start

Here is a quick start for Linux users who only use pip and apt to configure env. Run the following script, then you can jump to How to Use.

git clone https://github.com/Old-Farmer/Selextrans.git
cd Selextrans
./setup.sh # Set up before first running the app
./run.py # run the app

Dependencies

Linux

• xclip
• tesseract ocr engine (no data file need to be downloaded)

sudo apt install xclip tesseract-ocr

Windows

• tesseract ocr engine (no data file need to be downloaded). Check the website Home · UB-Mannheim/tesseract Wiki · GitHub and install Tesseract. Just Install it in the default configure, and choose a path to install it. Do not forget to set "tesseract_cmd" in data/settings.json which will be explained in Settings with details.

Requirements for Python packages

• requests
• pyperclip
• pynput
• langid
• googletrans>=4.0.0rc1
• httpcore
• tkinter
• pytesseract
• tqdm
• openai
• tiktoken
• pyqt5
• pyautogui
• translators

# For linux
sudo apt install python3-tk # For tkinter
pip install -r requirements.txt --upgrade

# For Windows
pip install -r requirements.txt --upgrade

How to Use

Text translate

Select text, then translation results will be shown.

Or just type sth. into the input textbox, press Enter, then translation results will be shown.

Screenshot translate

Press F4, select a screen area, then translation results will be shown. Auto source language is not supported here.

To select an area, you need to press the left mouse button, drag the mouse to select the desired area, and finally release the button to complete the selection. You can press the right mouse button or Esc to cancel a selection.

Refreshing

F5 for refreshing

Command

F1 for Toggling the command entry.

Input help into the command entry for more information.

Run and Setup

Directly run the run.py.

Attention: if you first run the app, run setup.py first to do some setups, which will take some time.

Settings

All Settings is in data/settings.json. Please restart or refresh the app if you change the file.

{
    "https_proxy": "http://127.0.0.1:7890/",
    "request_url_for_baidu_api": "https://fanyi-api.baidu.com/api/trans/vip/translate",
    "private_key_for_baidu_api": "",
    "appid_for_baidu_api": "",
    "openai_api_key": "",
    "engine": "translators",
    "translators_engine_service": "baidu",
    "mode": "dark",
    "text_translate_trigger_mode": "select",
    "text_translate_shortcut_key": "<f2>",
    "screenshot_translate_shortcut_key": "<f4>",
    "copy_key": "<ctrl>+c",
    "tesseract_cmd": "tesseract"
}

"engine" can be set as "google", "baidu_api", "translators" or "openai_api".

"google" engine (default engine) uses googletrans lib, so using it is totally free and no configuration is required. Please ensure that you can access the Google Translate website.

"baidu_api" engine uses 百度通用文本翻译API (Baidu General Text Translation API), and requires "private_key_for_baidu_api"and "appid_for_baidu_api" to be set ( The corresponding website is 百度翻译开放平台 ).

"openai_api" engine requires "openai_api_key" to be set. (The corresponding website is https://platform.openai.com/account/api-keys) "openai_api" engine is not as fast as other engines, so the app shows translation results token by token.

"translators" engine is based on translators lib (https://github.com/UlionTse/translators). "translators_engine_service" should be set for choosing a specific service. Now selextrans supports "baidu" and "youdao". If you want more, add specifc "languages_for_translators_xxx" kv.(please refer chapter Customize languages).

"mode" can be set as "dark" or "light".

"text_translate_trigger_mode" can be set as "select" or "key". If it is set as "key", you should press the key specified by "text_translate_shortcut_key" to trigger translation.

"xx_shortcut_key" specify shortcut key combinations for some functions. Key combinations are sequences of key identifiers separated by "+". Key identifiers are either single characters representing a keyboard key, such as "a", or special key names identified by names enclosed by brackets, such as "<ctrl>". You can set your own shortcut key combinations for convenience.

"copy_key" is used for text translate to copy selected text to this app. The default setting "<ctrl>+c" can work for most softwares, except some softwares, e.g. command line programs. If you really want to use this app with specific softwares, set "copy_key" as you want.

"tesseract_cmd" defines the command to start tesseract, which can be the executable file path or just the command "tesseract". For Windows, if you do not set the environment variable for tesseract, set "tesseract_cmd" as the executable file tesseract.exe path. Use / or \\ as separators.

Attention: Googletrans lib and Translators lib needs your network proxy setting if you have set it in your computer. This software can automatically detect you proxy setting. But if the automatic detection can't work for you computer, you can manually set "https_proxy" to specify the proxy. Leave it empty if you never set network proxy in your computer or you want automatic detection to work.

Language Support

• Chinese (simplified) 简体中文
• Chinese (traditional) 繁體中文
• English
• French Français
• Russian Русскийязык
• Japanese 日本語
• Korean 한국어

Customize languages

If you want to add other languages, Please follow the following steps:

1. Find the data/configs.json file.
2. Focus on "languages", "languages_for_google", "languages_for_baidu_api", "languages_for_tessertact". If you never use baidu api, you can ignore "languages_for_baidu_api". The last three kv pairs are perfectly matching with the first one. For example, "Chinese (Simplified)" in "languages" <=> "zh-cn" in "languages_for_google".
3. What you should do is to add sth and to keep the match. Add a language name in "languages", and then add the corresponding language code referenced by Googletrans: Free and Unlimited Google translate API for Python — Googletrans 3.0.0 documentation in "languages_for_google" at the same index. Likewise, if you want to use baidu api, add the corresponding language code referenced by 百度翻译开放平台 in "languages_for_baidu_api" at the same index. Last, add the corresponding language code in "languages_for_tessertact" at the same index, and download the corresponding data file to data/tessdata, referenced by Traineddata Files for Version 4.00 + | tessdoc.

Bugs list

• Note that translators engine may have some bugs, which are out of my ablility to fix.