RSLP Checker

This application provides a user interface and API for the RSLP (Portuguese Language Suffix Remover) algorithm, which is used to remove suffixes from Portuguese words in applications such as search engines, text analysis, and natural language processing.

System Architecture

flowchart LR
    Client([User]) -->|HTTP Request| Nginx(Nginx Proxy)
    Nginx -->|/ requests| Frontend[SvelteKit Frontend]
    Nginx -->|/api requests| Backend[RSLP Stemmer API]
    Frontend -->|POST /api/stem| Nginx
    Backend -->|Stemming Process| RSLPRules[(RSLP Rules)]
    subgraph Docker Compose
        Nginx
        Frontend
        Backend
        RSLPRules
    end

RSLP Process Flow

This flowchart illustrates the steps involved in the RSLP algorithm for stemming Portuguese words, based on this article: RSLP Stemmer (Removedor de Sufixos da Lingua Portuguesa):

flowchart LR
    Start([Begin]) --> wordEndsInPlural{Word ends in "s"?}
    wordEndsInPlural -- Yes --> pluralReduction[Plural Reduction]
    wordEndsInPlural -- No --> wordEndsInFeminine{Word ends in "a"?}
    pluralReduction --> wordEndsInFeminine
    wordEndsInFeminine -- Yes --> feminineReduction[Feminine Reduction]
    wordEndsInFeminine -- No --> augmentativeDiminutive{Word ends in "ão" or "inho"?}
    feminineReduction --> augmentativeDiminutive
    augmentativeDiminutive --> adverbReduction[Adverb Reduction]
    adverbReduction --> nounSuffixReduction[Noun Reduction]
    nounSuffixReduction --> isSuffixRemoved{Suffix removed}
    isSuffixRemoved -- Yes --> removeAccents[Remove Accent]
    isSuffixRemoved -- No --> verbSuffixReduction[Verb Reduction]
    verbSuffixReduction --> isSuffixRemoved2{Suffix removed}
    isSuffixRemoved2 -- Yes --> removeAccents
    isSuffixRemoved2 -- No --> removeVowel[Remove Vowel]
    removeVowel --> removeAccents
    removeAccents --> final([End])

Features

• Frontend: Built with SvelteJS and TailwindCSS, providing a modern and responsive user interface.
• Backend: Developed in TypeScript using Bun, which handles the API requests via ElysiaJS and implements the RSLP algorithm.

How to Use

The application was built with Docker in mind. To run the application, you need to have Docker installed on your machine. After installing Docker, you can follow these steps:

docker compose up -d

This will start the application in the background. After that, you can access the application through your browser at http://localhost.

API Reference

User Interface

GET /

The main web interface where you can interactively test the RSLP algorithm by entering Portuguese words and seeing their stemmed results.

REST API

Stem a Word

POST /api/stem

Stems a Portuguese word by removing its suffixes according to the RSLP algorithm.

Request Body:

{
  "text": "string"
}

Parameters:

Response:

{
  "original": "string",
  "stemmed": "string"
}

Examples

Single Word Example

Request:

curl -X POST http://localhost/api/stem \
  -H "Content-Type: application/json" \
  -d '{"text": "caminhando"}'

Response:

{
  "original": "caminhando",
  "stemmed": "caminh"
}

Multiple Words Example

Request:

curl -X POST http://localhost/api/stem \
  -H "Content-Type: application/json" \
  -d '{"text": "brasileiros aprendendo português"}'

Response:

{
  "original": "brasileiros aprendendo português",
  "stemmed": "brasil aprend portug"
}

RSLP Algorithm Overview

The RSLP (Removedor de Sufixos da Língua Portuguesa) algorithm reduces Portuguese words to their stem form by removing suffixes through a series of rule-based steps:

1. Plural Reduction: Removes plural endings (e.g., "casas" → "casa")
2. Adverb Reduction: Removes adverb suffixes (e.g., "felizmente" → "feliz")
3. Feminine Reduction: Converts feminine to masculine forms (e.g., "brasileira" → "brasileir")
4. Augmentative/Diminutive Reduction: Removes augmentative/diminutive suffixes (e.g., "cãozinho" → "cã")
5. Noun Suffix Reduction: Removes nominal suffixes (e.g., "pensamento" → "pens")
6. Verb Suffix Reduction: Removes verbal suffixes (e.g., "caminhando" → "caminh")
7. Vowel Removal: Removes final vowels (e.g., "menino" → "menin")