Reggy

Friendly regular expressions for text analytics. Typical regex features are removed/adjusted to make natural language queries easier. Able to incrementally match streaming text.

API Usage

// use the high-level Pattern API for simple use cases
let mut p = Pattern::new("dogs?").unwrap();
assert_eq!(p.findall("cat dog dogs cats"), vec![(4, 7), (8, 12)])

// transpile to normal (https://docs.rs/regex/) syntax
let ast = Ast::parse(r"do(gg)*|(!CAT|CAR)").unwrap();
assert_eq!(r"(?i:do(?:gg)*|(?-i:CAT|CAR))", ast.to_regex());

// perform an incremental search with several patterns at once
let money = Ast::parse(r"$(\d?\d?\d,)*\d?\d?\d.\d\d").unwrap();
let people = Ast::parse(r"(!(John|Jane) Doe)").unwrap();

let mut search = Search::new(&[money, people]);

// call step() to begin searching a stream
let jane_match = Match { pos: (0, 8), id: 1 };
assert_eq!(search.step("Jane Doe paid John"), vec![jane_match]);

// call step() again to continue with the same search state
// note "John Doe" matches across the step boundary
let john_match = Match { pos: (14, 22), id: 1 };
let money_match_1 = Match { pos: (23, 33), id: 0 };
assert_eq!(search.step(" Doe $45,700.66 instead of $499.00"), vec![john_match, money_match_1]);

// call finish() to retrieve any pending matches once the stream is done
let money_match_2 = Match { pos: (45, 52), id: 0 };
assert_eq!(search.finish(), vec![money_match_2] );

Pattern Language

Reggy is case-insensitive by default. Spaces match any amount of whitespace (i.e. \s+). All the reserved characters mentioned below (\, (, ), ?, |, *, +, and !) may be escaped with a backslash for a literal match. Patterns are surrounded by implicit unicode word boundaries (i.e. \b).

Examples

Make a letter optional with ?

dogs? matches dog and dogs

Create two or more options with |

dog|cat matches dog and cat

Perform operations on groups of characters with (...)

the qualit(y|ies) required matches the quality required and the qualities required

the only( one)? around matches the only around and the only one around

Create a case-sensitive group with (!...)

United States of America|(!USA) matches USA, not usa

Match digits with \d

\d.\d\d matches 3.14

Match zero-or-more characters with *, or one-or-more characters with +

$(\d?\d?\d,)*\d?\d?\d.\d\d matches $20.66 and $4,670,055.32