ThaiNLP.NET 0.1.3

thainlp.net

Thai NLP in .NET

Features

Word Tokenization

• newmm - Dictionary-based maximal matching word segmentation constrained by Thai Character Cluster (TCC) boundaries
• API similar to PyThaiNLP for easy migration from Python

Subword Tokenization

• TCC (Thai Character Cluster) tokenization for breaking text into character clusters

Number to Thai Word Conversion

• NumToThaiWord - Convert numbers to Thai text representation
• BahtText - Convert numbers to Thai currency format (Baht and Satang)

Installation

From NuGet (Recommended)

dotnet add package ThaiNLP.NET

Or via Package Manager:

Install-Package ThaiNLP.NET

From Source

Build the project:

dotnet build

Usage

Word Tokenization (newmm)

Basic usage:

using Thainlp;

// Simple tokenization
var tokens = WordTokenizer.Tokenize("ประเทศไทยมีอากาศดี");
// Output: ["ประเทศ", "ไทย", "มี", "อากาศ", "ดี"]

// With more options
var tokens = WordTokenizer.WordTokenize(
    text: "โอเคบ่พวกเรารักภาษาบ้านเกิด",
    engine: "newmm",
    keepWhitespace: true
);
// Output: ["โอเค", "บ่", "พวกเรา", "รัก", "ภาษา", "บ้านเกิด"]

Custom Dictionary

using Thainlp;
using System.Collections.Generic;

// Create custom dictionary
var customWords = new List<string> { "ชินโซ", "อาเบะ" };
var customDict = new Trie(customWords);

// Use with tokenizer
var tokens = WordTokenizer.WordTokenize(
    "ชินโซ อาเบะ เกิด 21 กันยายน",
    customDict: customDict
);

TCC (Thai Character Cluster) Tokenization

using Thainlp;

// Tokenize into character clusters
var clusters = TCC.Segment("ประเทศไทย");
// Output: ["ป", "ระ", "เท", "ศ", "ไ", "ท", "ย"]

// Get cluster positions
var positions = TCC.GetPositions("ประเทศไทย");

Legacy Subword API

using Thainlp;

// Original TCC implementation
var clusters = Subword.tcc("ประเทศไทย");
var positions = Subword.tcc_pos("ประเทศไทย");

Number to Thai Word Conversion

using Thainlp;

// Convert number to Thai words
string text = NumToWord.NumToThaiWord(112);
// Output: หนึ่งร้อยสิบสอง

string negative = NumToWord.NumToThaiWord(-273);
// Output: ลบสองร้อยเจ็ดสิบสาม

// Convert to Thai Baht currency format
string baht = NumToWord.BahtText(5611116.50);
// Output: ห้าล้านหกแสนหนึ่งหมื่นหนึ่งพันหนึ่งร้อยสิบหกบาทห้าสิบสตางค์

string simple = NumToWord.BahtText(116);
// Output: หนึ่งร้อยสิบหกบาทถ้วน

API Compatibility with PyThaiNLP

This library provides an API similar to PyThaiNLP:

Testing

Run the test suite:

dotnet test

Creating a Release

The project is configured to automatically create GitHub releases and publish to NuGet when a version tag is pushed.

Prerequisites

1. Create a NuGet API key at nuget.org
2. Add the API key as a secret in your GitHub repository settings:
   • Go to Settings → Secrets and variables → Actions
   • Add a new repository secret named NUGET_API_KEY
   • Paste your NuGet API key as the value

Release Process

1. Update the version in thainlp/Thainlp.csproj:

   <Version>0.1.0</Version>
   

2. Commit your changes:

   git commit -am "Bump version to 0.1.0"
   git push
   

3. Create and push a version tag:

   git tag v0.1.0
   git push origin v0.1.0
   

The GitHub Actions workflow will automatically:

• Build the project
• Run tests
• Create the NuGet package
• Create a GitHub release with the package attached
• Publish to NuGet

Continuous Integration

Every push to any branch triggers the CI workflow which:

• Builds the project
• Runs tests
• Creates the NuGet package as an artifact (not published)

License

See LICENSE file for details.