Add Mongolian language model compatible with Spacy NLP toolkit as alpha support.

spaCy is a free, open-source library for advanced Natural Language Processing framework in Python, which helps the developers to build real-world production-ready applications. The spaCy framework does not support the Mongolian language yet, not even alpha support.

In spaCy, languages marked as "alpha support" usually only include tokenization rules and various other rules and language data.

The Mongolian language has two alphabets, Mongolian Cyrillic and Traditional Mongolian scripts. In Mongolia, Mongolian Cyrillic is the official alphabet right now. This project includes the Mongolian Cyrillic transcript.

Added Mongolian language into spaCy framework as
- code: mn - name: Mongolian

• stop_words.py
  List of most common words of the Mongolian language that are often useful to filter out, for example, "ба" or "нь." Matching tokens will return True for is_stop.

• tokenizer_exceptions.py
  Some Mongolian language's special-case rules for the tokenizer, for example, abbreviations with punctuation, like "г.м," abbreviations with upper cases, like "УИХ."

• punctuation.py
  Regular expressions for splitting tokens. It includes rules for infixes.

• lex_attrs.py
  Custom functions for setting lexical attributes on tokens, e.g., like_num, which includes numeric words in the Mongolian language, like "мянга" or "зуу."

• examples.py
  Some sample words, sentences in the Mongolian language for testing inside the spaCy.

• __init.py
  Create classes for Mongolian by linking the above language pieces of information.

• test_tokenizer.py
  Test cases for tokenization for text in the Mongolian language.

• test.py

```
import spacy
from spacy.lang.mn import Mongolian
# importing the Mongolian language model

nlp = Mongolian()
# defining Mongolian NLP

print(nlp.lang) 
# Output:
    # mn
print(nlp) 
# Output:
    # <spacy.lang.mn.Mongolian object at 0x0000018A469C5130>

# Some Mongolian cyrillic text:
my_txt = ("Монгол хэл нь үндсэн хоёр бичиг бүхий байна."+
            "Кирил үсгийг 1940 оноос л үзэж эхлэв.")

my_doc = nlp(my_txt)

# Printing some token attributes:
print('Token\t\tStop\tPunct\tDigit\tLikeNum\tTitle')
for token in my_doc:
    print(token.text,'\t\t',token.is_stop,'\t',token.is_punct,'\t',token.is_digit,'\t',token.like_num,'\t', token.is_title)
# Output:
    # Token           Stop    Punct   Digit   LikeNum Title
    # Монгол           False   False   False   False   True
    # хэл              False   False   False   False   False
    # нь               True    False   False   False   False
    # үндсэн           False   False   False   False   False
    # хоёр             False   False   False   True    False
    # бичиг            False   False   False   False   False
    # бүхий            False   False   False   False   False
    # байна            True    False   False   False   False
    # .                False   True    False   False   False
    # Кирил            False   False   False   False   True
    # үсгийг           False   False   False   False   False
    # 1940             False   False   True    True    False
    # оноос            False   False   False   False   False
    # л                True    False   False   False   False
    # үзэж             False   False   False   False   False
    # эхлэв            False   False   False   False   False
    # .                False   True    False   False   False

# Remove stop words & punctuations:
cleaned = [token for token in my_doc if not token.is_stop and not token.is_punct]

for token in cleaned:
    print(token.text)
# Output:
    # Монгол
    # хэл
    # үндсэн
    # хоёр
    # бичиг
    # бүхий
    # Кирил
    # үсгийг
    # 1940
    # оноос
    # үзэж
    # эхлэв
    ```

No, not yet. Contributing to the spaCy community needs a lot of effort, and it is going in progress.

Because this package has not been officially included in spaCy yet, you need to clone the whole package and build it locally to use in your applications.

You can download the source from below link and build it locally, and then you can use it in your applications. https://github.com/bayarra/spaCy

• Required tools and packages:
  • C++ Build Tools
    For Windows:
    • Download and install Visual Studio Build Tools
    • Install Desktop development with C++
  • python
  • cython. pip install Cython
  • and more...
• Build the spaCy
  • Run the command python setup.py build_ext --inplace

https://github.com/bayarra/spacy.lang.mn/blob/main/Project%20Presentation.pptx

Official site: https://spacy.io/
spaCy 101: https://spacy.io/usage/spacy-101
Trained Models & Pipelines: https://spacy.io/models
Models & Languages: https://spacy.io/usage/models
spaCy source on the github: https://github.com/explosion/spaCy
Adding models for new languages: explosion/spaCy#3056
Contributing: https://github.com/explosion/spaCy/blob/master/CONTRIBUTING.md
spaCy tests: https://github.com/explosion/spaCy/blob/master/spacy/tests/README.md