lojbantrans: Proof-of-concept Lojban translator

UPDATE: We now use spaCy combined with word vectors, resulting in faster performance and more reliable translations.

---

lojbantrans is a Python-based proof-of-concept command-line tool for translating English text into Lojban and Lojban text into English using a glossword dictionary. It can be used standalone or as a starting point.
It produces readable, Lojban-inspired output that’s great for learners and conlang explorers.

Documentation

• --text: The Lojban text you want to translate
• --verbose: The verbose mode describing each word translation in detail
• --file: The text file you want to import
• --save: The text file you want to save to
• --reverse: The Lojban-English translator mode

---

🚀 How to Use

1. Install dependencies

pip install -r requirements.txt

And also this for downloading the spaCy model.

spacy download en_core_web_lg

2. Run the tool

python app.py --text "Your English text here"

Use --reverse to translate from Lojban to English:

python app.py --text "coi" --reverse

Or translate from a file:

python app.py --file input.txt --save output.txt

---

💡 Example

python app.py --text "Cats are small mammals"

Output:

Lojban Translation:
u'i lo mlatu cmalu mabru cu zasti

---

📂 Project Files

• app.py – Main CLI script
• valsi_glosswords.json – Glossword-to-Lojban dictionary (required)

---

⚙️ Features

• ✅ Sentence splitting
• ✅ Lemmatization (handles plurals like "cats")
• ✅ Synonym fallback
• ✅ Number-to-Lojban digit conversion
• ✅ Verbose mode for debugging translations
• ✅ File input/output support
• ✅ Reverse mode (Lojban → English)
• ✅ Multi-word gloss detection

---

🧠 Notes

• Common English glue words like "is", "the", and "a" are ignored automatically.
• Translations follow a basic lo ... cu ... Lojban bridi structure.
• Outputs are Lojban-ish (aka Lojbish) and use valid roots, but aren't guaranteed to be grammatically perfect.
• Great for Lojban learners, conlangers, AI language tinkerers, and fans of logical languages.

---

🌱 Future Features

• Grammar-aware parsing with tools like camxes or jbofihe
• Attitudinal customization (--attitude) for expressive tone