openpecha / pyewts Goto Github PK

View Code? Open in Web Editor NEW

19.0 3.0 3.0 83 KB

Tibetan Unicode to Wylie converter. (EWTS-Extended Wylie Transliteration Scheme)

License: Apache License 2.0

Python 100.00%

tibetan wylie

pyewts's Introduction

Python Tibetan Unicode to Wylie (EWTS) Converter

Description • Installation • Examples • Changes • License • Maintenance • Owner

Description

The goal of this code is to provide a library to convert back and forth between Tibetan Unicode and EWTS. The code is adapted from Java ewts-converter.

It also provides a conversion from the ACIP Transliteration to EWTS.

Installation

pip install pyewts

Examples

Convert Wylie to Unicode

import pyewts

converter = pyewts.pyewts()
print(converter.toUnicode("ba b+ba [a] ba\\u0f0b"))
# བ་བྦ་a་བ་

Convert Unicode to Wylie

print(converter.toWylie("༼༽"))
# ()

Catch Wylie warnings

>>> orig = """dangs
... zhwa
... dwang
... rma
... tshe
... phywa
... dge
... rgya
... dwags
... (rtse mgron)"""
>>> 
>>> print(orig)
dangs
zhwa
dwang
rma
tshe
phywa
dge
rgya
dwags
(rtse mgron)
>>> warns = []
>>> res = converter.toUnicode(orig, warns)
>>> print(res)
དངས
ཞྭ
དྭང
རྨ
ཚེ
ཕྱྭ
དགེ
རྒྱ
དྭགས
༼རྩེ་མགྲོན༽
>>> print(warns)
['line 1: "dangs": Syllable should probably be "dngas".']

See demo.py

Changes

See CHANGELOG.md.

License

Maintenance

Build the source dist:

rm -rf dist/
python3 setup.py clean sdist

and upload on twine (version >= 1.11.0) with:

twine upload dist/*

Owner

@eroux

pyewts's People

Contributors

Stargazers

Watchers

Forkers

riggy2013 yongruiq khawa-angx

pyewts's Issues

Python tools for Tibetan latin to unicode

Hi Elie. Nice to see you're using Python, as I remember you were mainly using lower level languages.

I am glad to see more work being done in this field. Just so you know I also have a Python tool for solving similar problems: https://github.com/ironhouzi/pytib

I hope we could perhaps join our efforts, so that we could benefit from one another.

pytib is far from feature complete and has a different approach than pyewts. I am also curious to see how compatible pytib handles EWTS, as pytib supports dynamic configuration of Latin definitions. pytib has just been a pet project I started when I first learned Python, but as a professional Python programmer, I gave it a big rewrite last year. Still, I think it could use a whole lot more polish, so learning from other skilled developers solving similar problems is very inspiring.

The main difference between the two is definitely the algorithm. I've chosen a more analytical approach, and you're using lookup tables. I think there are pros and cons to both approaches. While there's a benefit of getting rudimentary spell checking from using the analytical approach, the performance is not spectacular. I would assume lookup tables gives good performance, which is the current challenge I'm trying to tackle through concurrent processing. While I find the translation function is pretty OK, the implementation code that utilizes the parse() function for parsing documents has been implemented rather quickly. It uses line based handling instead of character based handling, which seems like a better approach for managing correct Tibetan punctuation and will also need to be figured out before I can do any work on concurrent Latin-Tibetan parsing.

Looking forward to learning from you.

below Tibetan syllables can't be converted.

I used pyewts to convert some text and here are the Tibetan syllables with errors.

dangs
zhwa
dwang
rma
tshe
phywa
dge
rgya
dwags

I don't have much time to look into it by myself. Will someone pls fix them?

Interestingly, it can convert left parenthesis but not the right one. For example,

(rtse mgron)

will be converted to

༼རྩེ་མགྲོན)

unresolved reference

my Python IDE tells me of two unresolved references for TransConverter on lines 804 and 806.
The library seems to be working nonetheless, so it is a low priority issue.