v1.5.0

bilibili/Index-anisorav1.5.0Mar 12, 2026by amirivojdan

AI Summary

A major feature release introducing a Web UI, dependency parser, TextRank keyword extraction, and enhanced spacing normalization.

Key Highlights

Web UI for Easy Experimentation via `shekar serve`.
Dependency Parser for syntactic analysis of Persian sentences.
TextRank Keyword Extraction method added.
Enhanced Spacing Normalizer for plurals and superlatives.

New Features

Web UI for Easy Experimentation
Dependency Parser
TextRank Keyword Extraction
Enhanced Spacing Normalizer

Full Release Notes

# What's New in v1.5.0

This release introduces new features, enhancements, and bug fixes to Shekar's Persian text processing toolkit.

## Highlights

- **Web UI for Easy Experimentation** – launch an interactive interface with a single command: `shekar serve -p 8080`. Explore and test Shekar's features without writing any code.
- **Dependency Parser** – added a new dependency parser module for syntactic analysis of Persian sentences.
- **TextRank Keyword Extraction** – TextRank is now available as a keyword extraction method alongside the existing options.
- **Enhanced Spacing Normalizer** – now correctly normalizes attached plurals and superlatives (e.g., `آدم‌ها` → `آدمها`, `نازک‌ترین` → `نازکترین`).

![Shekar WebUI Demo](https://github.com/amirivojdan/shekar/blob/d5d34f6fdfce6075224d70818eb508268fa6ed49/assets/webui-demo.gif)

## Bug Fixes

- Fixed a bug in the stemmer.
- Fixed a bug in RAKE keyword extraction.
- Fixed a bug in `PunctuationSpacingNormalizer`.