sqlite_skim - SQLite Fuzzy Matching Extension

sqlite_skim is a Rust-based SQLite loadable extension that provides a fuzzy matching scoring function, skim_score, powered by the skim-rs library. This allows you to integrate powerful fuzzy string matching into your SQL queries, enabling features like search ranking and intelligent sorting of results.

Features

• skim_score(query_string, choice_string): Returns an integer score representing how well choice_string matches query_string using the skim-rs fuzzy matching algorithm.

Requirements

• Rustup (for managing Rust toolchains)
• Rust nightly toolchain
• SQLite

Building

To build the sqlite_skim extension, follow these steps:

1. Clone the repository:

   git clone https://github.com/tzachar/sqlite_skim.git
   cd sqlite_skim/sqlite_skim
   

   (Replace https://github.com/tzachar/sqlite_skim.git with the actual repository URL once available)

2. Set the Rust toolchain to nightly: This project requires the nightly Rust toolchain due to dependencies within skim-rs.

   rustup override set nightly-2026-02-11
   

3. Build in release mode:

   cargo build --release
   

   This will produce the loadable SQLite extension library. The path to the library will typically be:

   • Linux: target/release/libsqlite_skim.so
   • macOS: target/release/libsqlite_skim.dylib
   • Windows: target/release/sqlite_skim.dll

Usage

Once the extension library is built, you can load it into your SQLite environment and start using the skim_score function.

1. Load the extension in SQLite:

   Open the SQLite CLI or integrate into your application's SQLite connection:

   SELECT load_extension('/path/to/your/sqlite_skim/target/release/libsqlite_skim.so');
   

   Replace /path/to/your/sqlite_skim/target/release/libsqlite_skim.so with the actual absolute path to the compiled library file.

2. Using the skim_score function:

   The skim_score function takes two string arguments: the query_string and the choice_string.

   SELECT skim_score('query', 'choice string');
   

   Examples:

   -- Basic fuzzy matching
   SELECT skim_score('apple', 'I like apples');
   SELECT skim_score('data', 'database management');
   SELECT skim_score('db', 'database');
   
   -- No match will return 0
   SELECT skim_score('xyz', 'abc');
   

3. Sorting results with skim_score:

   You can use skim_score in your ORDER BY clause to sort results based on their relevance to a query.

   Imagine a table products with a name column:

   CREATE TABLE products (id INTEGER PRIMARY KEY, name TEXT);
   INSERT INTO products (name) VALUES
       ('Apple MacBook Pro'),
       ('Banana Smoothie Maker'),
       ('Red Delicious Apples'),
       ('Pineapple Slicer');
   
   SELECT
       name,
       skim_score('apple', name) AS score
   FROM
       products
   ORDER BY
       score DESC;
   

   This query would return products related to 'apple', with the most relevant (highest score) first.