Languages and Alphabets
The Dasher concept works with almost any language. Dasher operates in Unicode and supports over sixty alphabets out of the box.
Supported Language Families
European Languages
- English
- French
- German
- Spanish
- Italian
- Portuguese
- Dutch
- Swedish
- Danish
- Norwegian
- Polish
- Czech
- Hungarian
- Finnish
- Greek
- Russian
Asian Languages
- Japanese (Hiragana)
- Chinese (partial support)
- Thai
- Mongolian
Dasher can generate complicated multi-part characters for languages like Thai by combining Unicode components.
Middle-Eastern Languages
- Arabic
- Hebrew
- Persian (Farsi)
Indian Languages
- Hindi
- Bengali
- Tamil
- Telugu
- Kannada
- Malayalam
- Gujarati
- Punjabi
- Marathi
- Oriya
- Assamese
- Urdu
- Kashmiri
- Sinhala
African Languages
- Swahili
- Afrikaans
- Zulu
- Xhosa
- Sepedi
- Sesotho
- Setswana
- Setswana
- Tsonga
- Tswana
- Venda
- Ndebele
- Swati
How to Switch Languages
- Open Dasher
- Select Options → Alphabet
- Choose your desired language from the list
All available alphabet files are included with Dasher. New alphabets can be added by
creating an appropriate alphabet.xml file.
Training Texts
Dasher learns from example text to provide accurate predictions. Training texts are provided for about fifty languages including European, Asian, African, and Semitic languages.
Why Training Texts Matter
Training texts teach Dasher the probability distribution of letters and words in your language. A well-trained Dasher will:
- Predict common words and phrases more accurately
- Offer better suggestions based on context
- Adapt to your personal writing style over time
Personalizing Dasher
We encourage you to personalize Dasher by supplying a training file written in your own style. This is especially useful for:
- Specialized vocabulary (technical, medical, legal)
- Regional dialects and slang
- Personal writing patterns
Special Language Implementations
Japanese Dasher
Japanese is implemented using Hiragana (60 or 83 character alphabets). Dasher can convert Japanese Shift-JIS to UTF8 and ensures Hiragana files are correct for the chosen alphabet.
Chinese Dasher
Partial support is available in version 3, with complete solutions planned for future versions. Chinese implementation uses Pinyin-based selection with character prediction.
Combining Characters
Dasher supports languages with combining characters (like Thai) by treating each component as a separate selectable element, then combining them appropriately in the output.
Creating Custom Alphabets
If your language is not already supported, you can create a custom alphabet by:
1. Create an alphabet file
Define your alphabet in an `alphabet.xml` file, specifying all characters and their order.
2. Provide training text
Create a text file (300K or more) of natural writing in your language to help Dasher learn character probabilities.
3. Install the files
Place the alphabet and training files in the appropriate Dasher directory and restart Dasher.
Full Guide: Creating Custom Alphabets →
Resources
Creating Alphabets Guide
Full guide to creating custom alphabets and training texts for Dasher.
Unicode Reference
Unicode Consortium - Official Unicode charts and standards for all languages.
Unicode Fonts
Alan Wood's Unicode Fonts - Information about Unicode fonts for various languages. Dasher requires Unicode fonts that support your language.
Existing Alphabets
Dasher on GitHub - View and download existing alphabet files from the repository.
Training Text Sources
Project Gutenberg - Free public domain books in many languages, useful for creating training texts.
Need Help?
If you need help with a specific language or want to contribute an alphabet file, please contact us on GitHub Discussions.