Open Source Go Search Software

fzf is a general-purpose command-line fuzzy finder. It's an interactive Unix filter for command-line that can be used with any list; files, command history, processes, hostnames, bookmarks, git commits, etc. fzf will launch an interactive finder, read the list from STDIN, and write the selected item to STDOUT. Without STDIN pipe, fzf will use find command to fetch the list of files excluding hidden ones. (You can override the default command with FZF_DEFAULT_COMMAND). fzf by default starts in fullscreen mode, but you can make it start below the cursor with the height option. Unless otherwise specified, fzf starts in "extended-search mode" where you can type in multiple search terms delimited by spaces. Fuzzy completion for files and directories can be triggered if the word before the cursor ends with the trigger sequence, which is by default **. Fuzzy completion for PIDs is provided for the kill command. In this case, there is no trigger sequence.

PanSou is a high-performance cloud-storage resource search API built around concurrent Telegram and plugin-based discovery. It searches multiple configured sources in parallel and aggregates links into a normalized result format. Supported link categories include Baidu, Aliyun, Quark, Tianyi, UC, 115, PikPak, Xunlei, 123 Cloud, magnet links, ED2K, and others. Results are ranked using source priority, freshness, and keyword signals. An asynchronous plugin system allows new search providers to be added without changing the core service. Memory and disk caching improve repeated-query performance, while optional JWT authentication can protect the API. PanSou is written in Go and can be deployed directly or through Docker and Docker Compose.

Efficient indexing and searching (1M Weibo 500M data is indexed in 28 seconds, search response time is 1.65 milliseconds, and search QPS is 19K). Support Chinese word segmentation (concurrent word segmentation using the sego word segmentation package, speed 27MB/sec). Support to calculate the proximity distance of keywords in the text (token proximity). When a request to add a document to the index comes in, the main coroutine will send the text to be segmented to a word segmentation coroutine through a channel, and the coroutine will segment the text and send it to a word segmentation through another channel. Indexer coroutines. The indexer coroutine builds an inverted index from the search keyword to the document, and the inverted index table is stored in memory for quick recall. The main coroutine receives the user's request, segments the request phrase in the main coroutine, and then sends it to the indexer through the channel.