Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/dev99problems/alfred-translayta
Alfred Translation workflow for en-uk, uk-en translations
https://github.com/dev99problems/alfred-translayta
alfred alfred-translayta alfred-workflows alfy favorites language translate translation ukrainian
Last synced: 4 months ago
JSON representation
Alfred Translation workflow for en-uk, uk-en translations
- Host: GitHub
- URL: https://github.com/dev99problems/alfred-translayta
- Owner: dev99problems
- License: mit
- Created: 2019-09-25T13:08:18.000Z (over 5 years ago)
- Default Branch: main
- Last Pushed: 2023-10-18T21:31:16.000Z (over 1 year ago)
- Last Synced: 2024-10-12T08:32:04.342Z (4 months ago)
- Topics: alfred, alfred-translayta, alfred-workflows, alfy, favorites, language, translate, translation, ukrainian
- Language: JavaScript
- Homepage:
- Size: 2.68 MB
- Stars: 5
- Watchers: 1
- Forks: 0
- Open Issues: 4
-
Metadata Files:
- Readme: README.md
- License: license
Awesome Lists containing this project
README
# alfred-translayta
![version] ![MIT][license] ![last-commit]
[version]: https://img.shields.io/npm/v/alfred-translayta
[license]: https://img.shields.io/badge/license-MIT-blue.svg
[last-commit]: https://img.shields.io/github/last-commit/dev99problems/alfred-translayta> **Translate** with Google Translate in Alfred (*en-uk* / *uk-en*). **Save** to favorites. Practice **learning** the words.
Alfred Translayta is an ***Alfred workflow***, which lets you easily translate πΊπ¦ π π¬π§ , save words to favorites and have convenient way to review and memorize them later.
## Features
### Translation
* ***Auto-detection*** of the translation direction
* Correction of *misspelled words*
* Foremost app ***selected text translation***
* ***Pasting*** translation into the foremost app
* ***Copying*** translation into the clipboard
* ***Local caching*** of the last translation response### Favorites
* *Add/Edit/Delete* favorites
* *Learning* mode
* *Search* through normal/learning mode
* *Import/Export* favorites from/to local file## Prerequisites
* [Node.js 14](https://nodejs.org) or higher
* [Alfred (3,4,5)](https://www.alfredapp.com/)
* [Alfred Powerpack License](https://www.alfredapp.com/powerpack/)## Install
```
$ npm install --global alfred-translayta
```## Usage
Open Alfred and type `t` **or** press `Cmd+Shift+K` to activate workflow
Translate/Auto-correction
* start typing phrase in English or Ukrainian
* hit `Enter` on list item with `auto-correction` suggestion
* enter ` ` (2 spaces) to get latest search results out of cache
* select any word of phrase and hit `Cmd-Shift-K` to get translation

Copy/Paste
* navigate to any translation and hit `Enter` to copy to `clipboard`
* navigate to any translation and hit `Shift-Enter` to paste it into the `foremost` application

Favorites
* navigate to the item named `/add` and hit `Enter` to add word to `Favs`
* to get access to your `Favs` list enter `.` in workflow, aka `t .`
* to edit/remove/copy any word from the list hit `Enter` on any item in your `Favs` list

Learning mode
* to activate learning mode for your `Favs`, enter `..`, aka `t ..`
* in both `Favs` and `Learning` modes you can search by typing after `..`,
e.g. `t ..brick` or `t .. brick` will filter the list* in `Learning` mode you can navigate to any item and press and hold `Option` key to make it's translation visible

## HotKeys
`Cmd+Shift+K` β to open workflow main window
`Cmd+Shift+J` β to open favorites list (shortcut for `Cmd+Shift+K` + and entered `.`)
## Testimonials
**Visual**
All icons are taken from [flaticon.com](https://www.flaticon.com/), so many kudos and thanks to [the authors](/icons/testimonials.md)