Search Engine Awesome

• Search Benchmark
• quickwit-oss/tantivy
  • MIT, Rust
  • 等同于基于 Rust 的 Lucene
• 基于 Tantivy
  • toshi-search/Toshi
    • MIT, Rust
  • mosuka/bayard
    • MIT, Rust
  • quickwit-oss/quickwit
    • AGPLv3, Rust
    • Like Elasticsearch, but highly reliable & cost-efficient for log management.
  • lnx-search/lnx
    • MIT, Rust
• valeriansaliou/sonic
  • MPL-2.0, Rust
  • schema-less search backend
  • 非常小巧，对于一个词组默认索引 1000 条记录
• groonga/groonga
  • LGPL-2.1, C
  • 有 MySQL 和 PostgreSQL 插件 - 非常易用
• manticoresoftware/manticoresearch
  • GPL-2.0, C++
  • Database for search
  • MySQL 协议和语法
  • 支持 行存、列存、文档存储
  • forked from Sphinx 2.3.2 in 2017
  • vs MeiliSearch https://www.reddit.com/r/selfhosted/comments/w89tgh/comment/ihq798e/
• typesense/typesense
  • GPL-3.0, C++
  • 不支持 CJK
    • Support for writing systems without spaces between words typesense/typesense#228
• pisa-engine/pisa
  • Performant Indexes and Search for Academia
• Apache Lucene Core
• olivernn/lunr.js
• Apache nutch
• sphinxsearch
  • v3 后不开源
• blevesearch/bleve
  • Apache-2.0, Go
• mosuka/blast
  • 基于 bleve
• os-fulltext-search-solutions
• gajus/liqe
  • Lucene-like parser and search engine
• blugelabs/bluge
  • Apache-2.0, Go
  • indexing library for Go
• Bluge Based
  • prabhatsharma/zinc
    • Apache-2.0, Go+Vue
    • alternative to elasticsearch
    • 目标是日志分析，不是 ES 兼容 https://github.com/prabhatsharma/zinc/issues/52#issuecomment-1000064449
  • mosuka/phalanx
    • Apache-2.0, Go
    • cloud-native distributed search engine
• MeiliSearch
  • MIT, Rust
  • 索引要能放入 RAM - 小数据量场景
  • BucketSort
• wibyweb/wiby
  • GPLv2
• wilsonzlin/edgesearch
  • MIT, Rust
  • WASM code at Cloudflare Worker
• jameslittle230/stork
  • Apache-2.0, Rust
• tinysearch/tinysearch
  • Apache-2.0,MIT, Rust
• Web Search
  • https://searchhut.org/
    • 提交域名索引 https://searchhut.org/about
    • https://sr.ht/~sircmpwn/searchhut/
      • GPLv3, Go
    • https://news.ycombinator.com/item?id=32104609
    • Crawler SearchHut Bot 0.0 (GNU AGPL 3.0); https://sr.ht/~sircmpwn/searchhut; <[email protected]>
  • StractOrg/stract
    • AGPLv3, Rust, Svelte
    • axum web framework, rocksdb
• yacy/yacy_search_server
  • GPLv2+, Java
  • Distributed Peer-to-Peer Web Search Engine and Intranet Search Appliance
• mwmbl/mwmbl
  • AGPLv3, Python
• searx/searx
  • AGPLv3, Python
  • Privacy-respecting, hackable metasearch engine
  • https://searx.org/
  • https://searx.space/
• List of search engines
• metasearch engine
  • 聚合其他搜索引擎
• beir-cellar/beir
  • Heterogeneous Benchmark for Information Retrieval
• Web
  • pagefind
  • nearform/lyra
    • @nearform/lyra
    • in-memory, typo-tolerant, full-text search engine
  • fuzzyjs
  • lucaong/minisearch

Lcoal Search

• xapian/xapian
• The Lemur Project
• A Local Search Engine
• Building Monocle, a universal personal search engine for life
• recoll
  • desktop full-text search tool
  • HN

中文分词

• ik
• jieba
  • fxsjy/jieba
    • Pythone
• hightman/scws
  • PHP
  • Simple Chinese Word Segmentation
• go-ego/gse
  • 字典 https://github.com/go-ego/gse/tree/master/data/dict/zh

Library

• Golang
  • FST - finite state transducer
    • blevesearch/vellum
      • couchbase/vellum
• BurntSushi/fst
  • Rust
• Extract
  • ndaidong/article-parser
    • https://demos.pwshub.com/article-parser

Tech

• TF-IDF
• FST - finite state transducer - 有限状态传感器
  • term -> id
  • 实现推荐 - O(length(querry))
  • 共享尾部
  • 内存少于 tire
• DFSA - Deterministric acyclic finite state acceptor - 确定无环有限状态接收机
  • tire-tree
    • 适合英文词典 - 字符集少，唯一前缀少
• Double Array Trie
  • 适合做中文词典，内存占用小
  • https://linux.thai.net/~thep/datrie/datrie.html
• Ternary Search Tree
• Ragel
  • adrian-thurston/ragel
  • https://www.colm.net/open-source/ragel/

Story

• Index 1,600,000,000 Keys with Automata and Rust

Misc

• RaRe-Technologies/gensim
  • LGPL-2.1, Python
  • Topic Modelling for Humans
• oborchers/Fast_Sentence_Embeddings
• mayabot/mynlp