ToolGood.Words is a sensitive word detection and filtering component that uses string matching algorithms to identify and filter prohibited content in text.
The tool addresses the problem of efficiently detecting sensitive words, illegal content, and profanity in large volumes of text. It employs Aho-Corasick and trie-tree based algorithms to achieve high-speed filtering. The core implementations include multiple search classes optimized for different use cases: StringSearch variants return matched strings, WordsSearch variants return detailed results including position and sequence information, and IllegalWordsSearch provides specialized filtering with configurable skip-word handling and blacklist support. The tool also supports wildcard matching with partial regex syntax including dots, question marks, brackets, and pipe operators.
Developers should choose this tool when building content moderation systems that require fast, accurate detection across large datasets. It suits projects handling user-generated content, chat filtering, or document scanning where performance matters. The tool includes utility functions beyond core filtering: simplified-traditional Chinese conversion, full-width to half-width character normalization, pinyin conversion and matching, and fuzzy search capabilities. These supplementary features make it useful for applications serving Chinese-language audiences. The README does not compare this tool to alternatives, so no comparative guidance can be offered.
Development activity shows consistent maintenance with multiple specialized implementations addressing different performance requirements. The project maintains separate optimized variants including pointer-based versions for performance-critical scenarios. The codebase includes serialization support through Save and Load methods to accelerate initialization in production environments. The tool provides both basic and extended search classes, allowing developers to select implementations matching their specific needs rather than forcing a one-size-fits-all approach.
