louieh / dict-tiny

A tiny command-line translator was implemented in Plumbum. Youdao.com + Google Translation + Gemini + OpenAI

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Dict-tiny

PyPI version GitHub Workflow Status License: MIT Downloads

A command-line tool that integrates Youdao Dict, Google Translate, DeepL Translator, Gemini and OpenAI.

Just for fun :)

Features

  • Translate Chinese and English words using youdao.com.
  • Google Translate API.
  • Gemini.
  • OpenAI.
  • Use the above features in interactive mode.

interactive_mode

Installing

Install with pip. (Python >= 3.9)

$ pip install dict-tiny

Upgrading

$ pip install --upgrade dict-tiny

Options

$ dict-tiny

Usage:
    dict-tiny [SWITCHES] words...

Gemini:
    --gemini                        Use Gemini API
    --gemini-key VALUE:str          Indicate gemini api key
    --gemini-model VALUE:str        Select gemini model; the default is gemini-
                                    pro
    --img-path VALUE:ExistingFile   The path of image

GoogleTranslate:
    --detect-language               Detect the language of the given text
    -g, --google                    Use Google Translate

LLM:
    --dialog-turns VALUE:[1..20]    Number of conversations turns; the default
                                    is 10
    --max-output-tokens VALUE:int   The maximum number of tokens to include in a
                                    candidate.
    --temperature VALUE:float       Controls the randomness of the output

Meta-switches:
    -h, --help                      Prints this help message and quits
    --help-all                      Prints help messages of all sub-commands and
                                    quits
    -v, --version                   Prints the program's version and quits

OpenAI:
    --api-version VALUE:str         Azure openai api version
    --azure-base-url VALUE:str      Azure openai base url
    --azure-endpoint VALUE:str      Azure endpoint
    --openai                        Use OpenAI API
    --openai-key VALUE:str          Indicate openai api key
    --openai-model VALUE:str        Select openai model; the default is
                                    gpt-3.5-turbo

Switches:
    -c, --clipboard                 Use the contents of the clipboard.
    --default-translator VALUE:str  Set default translator
    -i, --interactive               Interactive mode
    --source-language VALUE:str     What language you want to translate
    --target-language VALUE:str     What language you want to translate into

YoudaoDict:
    -m, --more                      Get more details
    -y, --youdao                    Use Youdao Dictionary, currently only
                                    supports English or Chinese words

Details and examples

Youdao Dict

Add -y / --youdao to use Youdao Dict:

$ dict-tiny -y book

>>> YoudaoDict <<<
book
======
英[bʊk]美[bʊk]
n. 书,书籍;本子,簿册;(长篇作品的)篇,卷,部;装订成册之物;赌局,打赌;账册,账簿
v. 预订,预约;(警方)将……记录在案;(裁判)记名警告
【名】 (Book)(英)布克,(瑞典)博克,(朝)北(人名)
$ dict-tiny -y 书

>>> YoudaoDict <<<
书
===
[shū]
n.book;letter;script
vt.write

Use -m/--more to get more detail translation for the word:

$ dict-tiny -y 曾经 -m

>>> YoudaoDict <<<
曾经
====
[céng jīng]
adv.once;ever

more detail:
======== 副词 ========
1:
  (表示有过某些行为或情况) once:
  He once lived in Shanghai.
  他曾经在上海住过。
  She has taken part in a major battle for oil.
  她曾经参加过石油大会战。
$ dict-tiny -y dictionary -m

>>> YoudaoDict <<<
dictionary
============
英[ˈdɪkʃən(ə)ri]美[ˈdɪkʃəneri]
n. 字典,词典;专业词典,术语大全;电子词典;双语词典

more detail (collins):
======== N-COUNT 可数名词 ========
 · 词典

A dictionary is a book in which the words and phrases of a language are listed alphabetically, together with their meanings or their translations in another language.


 例: ...a Spanish-English dictionary.
     …一本西班牙语—英语词典。

Note:

  • Some words have a lot of translation that may occupy the entire screen.
  • Only supports English or Chinese words currently.

Google Translate

Add -g / --google to use Google Translate:

$ dict-tiny -g book

>>> GoogleTranslate <<<
book
======
output: 书
detected language: en

Add --target-language to specify the language to translate results into:

$ dict-tiny -g operation system --target-language ja

>>> GoogleTranslate <<<
operation system
==================
output: オペレーションシステム
detected language: en

Add --source-language to specify the language of the text to be translated, but in most cases you don't need to do this because the google translate api automatically detects the source language type.

And if you give a wrong source language, the translation result may not be what you expect.

So, of course, you can add --detect-language to detect the language type:

$ dict-tiny -g --detect-language español

>>> GoogleTranslate <<<
español
=========
confidence: 0.49805447459220886
input: español
language: es
name: Spanish

Note:

  • Make sure Google services are available in your network environment.

  • The source and target languages for Google Translate are identified using the iso-639-1 codes.

    $ dict-tiny -g book --target-language zh --source-language en
    
    >>> GoogleTranslate <<<
    book
    ======
    output: 书
    source language: en

    You can also enter the ISO language name:

    $ dict-tiny -g book --target-language German --source-language English
    
    >>> GoogleTranslate <<<
    book
    ======
    output: Buch
    source language: English
  • Set the environment variable $DICT_TINY_TARGET_LAN so that you do not have to manually specify the target language each time. if you specify the target language parameter on the command line with --target-language, the setting in the environment variable will be overwritten.

  • The default target-language for Goole Translate are English .

Gemini

Add --gemini to use Gemini.

Use --gemini-model to specify the model used by Gemini. You can also set the environment variable $DICT_TINY_GEMINI_MODEL. The default model is gemini-pro.

Use --gemini-key to specify the Gemini api key, or set the environment variable $DICT_TINY_GEMINI_API_KEY.

Note:

  • If you use gemini-pro-vision model, you need to use --img-path to specify the path of image.

  • Use --max-output-tokens to specify the maximum output token.

  • Use --temperature to specify the temperature.

OpenAI

Add --openai to use OpenAI.

Use --openai-model to specify the model used by OpenAI. You can also set the environment variable $DICT_TINY_OPENAI_MODEL. The default model is gpt-3.5-turbo.

Use --openai-key to specify the OpenAI api key, or set the environment variable $DICT_TINY_OPENAI_API_KEY.

$ dict-tiny --openai who are you?

>>> OpenAI-gpt-3.5-turbo <<<
I am an AI language model created by OpenAI. I can assist you with various tasks, answer questions, and
engage in conversations on a wide range of topics. How may I assist you today?

Note:

  • Currently only text generation models are supported.

  • Use --max-output-tokens to specify the maximum output token.

  • Use --temperature to specify the temperature.

Interactive mode

Each of the above functions can be entered into its interactive mode by adding the -i . Ctrl + d to exit interactive mode.

In interactive mode you can:

  • Use above features in a continuous interactive manner.
  • Press Tab for word auto-completion (using Youdao's auto-completion function, currently only supports Chinese, English, French, Korean, Japanese)
  • For Gemini and OpenAI, the dialog context is maintained in interactive mode, use --dialog-turns to specify the number of dialog turns to maintain, the default is 10.
  • All settings cannot be changed after entering interactive mode, such as target-language, model, img-path, temperature, etc., unless you exit to change the settings and re-enter.

Other

Default behavior

  • Youdao Dict is the default translator, which means Youdao Dict will be used when no translator is specified.

    $ dict-tiny 机器学习
    
    >>> YoudaoDict <<<
    机器学习
    ======
    machine learning

    You can use the environment variable $DICT_TINY_DEFAULT_TRANS to set the default translator. There are four options described above: YoudaoDict, GoogleTranslate, Gemini, OpenAI. Case insensitive.

  • For Youdao Dict and Google translate, if the target language is not specified, Chinese and English will be used as the target language for each other.

  • In non-interactive mode, multiple translators can be specified at the same time, for example, YoudaoDict and GoogleTranslate can be used at the same time.

    $ dict-tiny formulation -y -g
    
    >>> YoudaoDict <<<
    formulation
    =============
    英[ˌfɔːmjuˈleɪʃ(ə)n]美[ˌfɔːrmjuˈleɪʃ(ə)n]
    n. (政策、计划等的)制定,构想;(想法的)阐述方式,表达方法;(药品或化妆品的)配方,配方产品
    >>> GoogleTranslate <<<
    formulation
    =============
    output: 公式
    detected language: en

Use clipboard content

Use -c/--clipboard to use the contents of the clipboard:

$ dict-tiny -c -y

>>> YoudaoDict <<<
encounter
===========
英[ɪnˈkaʊntə(r)]美[ɪnˈkaʊntər]
v. 遭遇;偶遇,邂逅
n. 偶遇,邂逅;经历,体验;冲突;比赛,交锋

Note:

  • -c/--clipboard has low priority. If you add -c and a word at the same time, the switch -c will be ignored.

Environment variables

name default description
DICT_TINY_TARGET_LAN en Specify the target language to be used in Google translate.
DICT_TINY_DEFAULT_TRANS youdaodict Specify the default translator.
youdaodict googletranslate gemini openai
DICT_TINY_GEMINI_MODEL gemini-pro Specify the model to be used in the Gemini.
DICT_TINY_GEMINI_API_KEY Specify Gemini API key.
DICT_TINY_OPENAI_MODEL gpt-3.5-turbo Specify the model to be used in the OpenAI.
DICT_TINY_OPENAI_API_KEY Specify OpenAI API key.

License

MIT

About

A tiny command-line translator was implemented in Plumbum. Youdao.com + Google Translation + Gemini + OpenAI

License:MIT License


Languages

Language:Python 100.0%