Skip to content

Latest commit

 

History

History
87 lines (64 loc) · 3.05 KB

README.md

File metadata and controls

87 lines (64 loc) · 3.05 KB

fuzzig

Fuzzy finder algorithms in Zig based on the Smith-Waterman algorithm, inspired by fzf.

Example

const std = @import("std");
const fuzzig = @import("fuzzig");

pub fn main() !void {
    var gpa = std.heap.GeneralPurposeAllocator(.{}){};
    defer _ = gpa.deinit();
    const allocator = gpa.allocator();

    // The default implementation needs to know the maximum haystack and needle
    // sizes so that it can allocate all memory contiguously. Smaller haystack
    // or needle strings will use subsets of the allocated memory.
    var searcher = try fuzzig.Ascii.init(
        allocator,
        128, // haystack max size
        32, // needle max size
        .{ .case_sensitive = false },
    );
    defer searcher.deinit();

    const haystack = "Hello World";
    const needle = "world";

    const score = searcher.score(haystack, needle);
    std.debug.print("Score: {d}\n", .{score.?});

    // Get a traceback of the character positions that were matched
    const match = searcher.scoreMatches(haystack, needle);
    std.debug.print(
        "Score with traceback: {d} {any}\n",
        .{ match.score.?, match.matches },
    );
}

Output

Score: 104
Score with traceback: 104 { 6, 7, 8, 9, 10 }

The traceback shows the indices of the haystack that were matched, useful for generating visual feedback.

See the AsciiOptions struct for a list of available options.

Design

The module defines an Algorithm generic type, which accepts the element type of the array to be fuzzy searched, the score type and values, and an algorithm implementation. The implementation must define an eqlFunc, a scoreFunc and a bonusFunc used to test for equality between tokens, for determining the score of two matching tokens, and for determining any in-places bonuses respectively.

  • Algorithms only have score and scoreMatches as public functions.
  • If not matches are detected, the score will be null.

This package currently implements ASCII and Unicode (compile with -Dunicode) fuzzy finders. The fuzzy finder implementation should be flexible enough to extend to other data types or character encodings in downstream packages by providing your own wrapper around AlgorithmType. See the ASCII and Unicode implementations for examples.

Usage

The library was written with Zig 0.12.0-dev.3541+05b185811, but there is likely a lot of flexibility with versioning.

To use in a Zig project, add it to your build.zig.zon

    // ...
    .dependencies = .{
        .fuzzig = .{
            .url = "https://github.com/fjebaker/fuzzig/archive/main.tar.gz",
            .hash = "" // get with `zig fetch`
        },
    },
    // ...

Then add the module to your build step in build.zig:

    // ...
    const fuzzig = b.dependency("fuzzig", .{}).module("fuzzig");

    my_exe_or_lib.root_module.addImport("fuzzig", fuzzig);
    // ...