Metadata-Version: 2.1
Name: sonatoki
Version: 0.1.1
Summary: ilo li moku e toki li pana e sona ni: ni li toki ala toki pona?
Author-Email: "jan Kekan San (@gregdan3)" <gregory.danielson3@gmail.com>
License: AGPL-3.0-or-later
Requires-Python: >=3.8
Requires-Dist: unidecode>=1.3.6
Requires-Dist: regex>=2023.12.25
Requires-Dist: typing-extensions>=4.11.0
Requires-Dist: nltk>=3.8.1; extra == "nltk"
Provides-Extra: nltk
Description-Content-Type: text/markdown

# sona toki

## What is **sona toki**?

This library, "Language Knowledge," helps you identify whether a message is in Toki Pona. No grammar checking, yet, which means this more checks whether a given message has enough Toki Pona words.

I wrote it with a variety of scraps and lessons learned from a prior project, [ilo pi toki pona taso, "toki-pona-only tool"](https://github.com/gregdan3/ilo-pi-toki-pona-taso). That tool will be rewritten to use this library shortly.

If you've ever worked on a similar project, you know the question "is this message in [language]" is not a consistent one- the environment, time, preferences of the speaker, and much more, can all alter whether a given message is "in toki pona," and this applies to essentially any language.

This project "solves" that complex problem by offering a highly configurable and incredibly lazy parser

## Quick Start

Install with your preferred Python package manager. Example:

```sh
pdm init  # if your pyproject.toml doesn't exist yet
pdm add sonatoki
```

Then get started with a script along these lines:

```py
from sonatoki.Filters import (
    Numerics,
    Syllabic,
    NimiLinku,
    Alphabetic,
    ProperName,
    Punctuations,
)
from sonatoki.Scorers import SoftScaling
from sonatoki.Cleaners import ConsecutiveDuplicates
from sonatoki.Tokenizers import word_tokenize_tok
from sonatoki.Preprocessors import URLs, DiscordEmotes

def main():
    ilo = Ilo(
        preprocessors=[URLs, DiscordEmotes],
        ignoring_filters=[Numerics, Punctuations],
        scoring_filters=[NimiLinku, Syllabic, ProperName, Alphabetic],
        cleaners=[ConsecutiveDuplicates],
        scorer=SoftScaling,
        tokenizer=word_tokenize_tok,
    )
    ilo.is_toki_pona("imagine how is touch the sky")  # False
    ilo.is_toki_pona("o pilin insa e ni: sina pilin e sewi")  # True
    ilo.is_toki_pona("I Think I Can Evade Detection")  # False

if __name__ == "__main__":
    main()
```

`Ilo` is highly configurable by design, so I recommend exploring the `Preprocessors`, `Filters`, and `Scorers` modules. The `Cleaners` module only contains one cleaner, which I recommend using. The `Tokenizers` module contains several other word tokenizers, but their performance will be worse than the dedicated Toki Pona tokenizer `word_tokenize_tok`.

## Development

1. Install [pdm](https://github.com/pdm-project/pdm)
1. `pdm install --dev`
1. Open any file you like!

## FAQ

### Why isn't this README/library written in Toki Pona?

The intent is to show our methodology to the Unicode Consortium, particularly to the Script Encoding Working Group (previously the Script Ad Hoc Group). As far as we're aware, zero members of the committee know Toki Pona, which unfortunately means we fall back on English.

After our proposal has been examined and a result given by the committee, I will translate this file and library into Toki Pona, with a note left behind for those who do not understand it.

### What's the deal with the tokenizers?

The Toki Pona tokenizer `word_tokenize_tok` is very specific in always separating writing characters from punctuation, and leaving contiguous punctuation as contiguous- this is a level of precision that NLTK's English tokenizer does not want for several reasons, such as that English words can have "punctuation" characters in them.

Toki Pona doesn't have any mid-word symbols when rendered in the Latin alphabet, so a more aggressive tokenizer is highly desirable.

The other tokenizers are provided as a comparison case more than anything. I do not recommend their use.

### Aren't there a lot of false positives?

Yes. It's up to you to use this tool responsibly on input you've done your best to clean, and better, use stronger filters before weaker ones. For now though, here's a list of relevant false positives:

- `ProperName` will errantly match text in languages without a capital/lowercase distinction, artificially inflating the scores.
- `Alphabetic` will match a _lot_ of undesirable text- it essentially allows 14 letters of the English alphabet.

### Don't some of the cleaners/filters conflict?

Yes. Some do so

- `ConsecutiveDuplicates` may errantly change a word's validity. For example, "manna" is phonotactically invalid in Toki Pona, but would become "mana" which is valid.
- `ConsecutiveDuplicates` will not work correctly with syllabaries (alphabets, but representing a pair of consonant and vowel).

You'll notice a _lot_ of these are troubles regarding the application of latin alphabet filters to non-latin text. Working on it!
