jieba-php

Why we included this project

PHP developers who need Chinese text segmentation for search or text analysis have few mature options, and jieba-php brings the well-known Jieba algorithm to that ecosystem. It offers three modes: precise segmentation for analysis, a full mode that surfaces every possible word, and a search-engine mode that re-splits long words for better recall when building inverted indexes. Out-of-vocabulary terms are handled by an HMM with Viterbi decoding, and you can supply custom dictionaries to fix domain-specific words the built-in lexicon misses. It also supports traditional Chinese and other CJK text, and bundles TF-IDF keyword extraction and part-of-speech tagging, so it works as a small toolkit rather than a single splitter. When you want fast, cheap segmentation without calling an LLM, this is a solid Composer-installable option.