pidgin

build non-recursive grammars

Pidgin is a Rust library for generating non-recursive grammars which allow one to convert strings to parse trees.

Pidgin's grammars are implemented underlyingly as regular expressions with named matching groups. Rust's regular expression engine is very fast but it has a limitation which makes parsing hierarchically structured patterns with repeating components difficult: you cannot use a particular match name more than once in a pattern. This makes a simple grammar such as

foo -> bar baz | baz bar
bar -> '1'
baz -> '2'

problematic. The foo rule uses both bar and baz twice.

Pidgin allows you to work around this restriction by managing group renaming.

Another disadvantage of regular expressions is that the more expressive they are the more difficult they are to read. Pidgin allows you to construct expressive regular expressions without obscuring your intention.

Example

#![recursion_limit="256"]
#[macro_use]
extern crate pidgin;

fn experiment() -> Result<(), Box<Error>> {
    let date = grammar!{
    
        // comments are legal in grammars
    
        (?ibB)   // default flags for all rules -- case insensitive and enforce leading and trailing word boundaries
    
        // the master rule; it has multiple alternates
        date -> <weekday> (",") <month> <monthday> (",") <year>
        date -> <month> <monthday> | <weekday> | <monthday> <month> <year>
        date -> <month> <monthday> (",") <year>
        date -> <numeric_date>
    
        // sub-rules
    
        numeric_date -> <year> ("/") <numeric_month> ("/") <numeric_day>
        numeric_date -> <year> ("-") <numeric_month> ("-") <numeric_day>
        numeric_date -> <numeric_month> ("/") <numeric_day> ("/") <year>
        numeric_date -> <numeric_month> ("-") <numeric_day> ("-") <year>
        numeric_date -> <numeric_day> ("/") <numeric_month> ("/") <year>
        numeric_date -> <numeric_day> ("-") <numeric_month> ("-") <year>
    
        year    => r(r"\b[12][0-9]{3}|[0-9]{2}\b")
        weekday => [
                "Sunday Monday Tuesday Wednesday Thursday Friday Saturday"
                    .split(" ")
                    .into_iter()
                    .flat_map(|s| vec![s, &s[0..2], &s[0..3]])
                    .collect::<Vec<_>>()
            ]
        weekday     => (?-i) [["M", "T", "W", "R", "F", "S", "U"]]
        monthday    => [(1..=31).into_iter().collect::<Vec<_>>()]
        numeric_day => [
                (1..=31)
                    .into_iter()
                    .flat_map(|i| vec![i.to_string(), format!("{:02}", i)])
                    .collect::<Vec<_>>()
            ]
        month => [
                vec![
                    "January",
                    "February",
                    "March",
                    "April",
                    "May",
                    "June",
                    "July",
                    "August",
                    "September",
                    "October",
                    "November",
                    "December",
                ].into_iter().flat_map(|s| vec![s, &s[0..3]]).collect::<Vec<_>>()
            ]
        numeric_month => [
                (1..=31)
                    .into_iter()
                    .flat_map(|i| vec![i.to_string(), format!("{:02}", i)])
                    .collect::<Vec<_>>()
            ]
    };

    // compile this into a matcher
    // (uncompiled grammar's can be used as elements of other grammars)
    let matcher = date.matcher()?;
    
    // we let whitespace vary
    assert!(matcher.is_match(" June   6,    1969 "));

    // we made it case-insensitive
    assert!(matcher.is_match("june 6, 1969"));

    // but we want to respect word boundaries
    assert!(!matcher.is_match("jejune 6, 1969"));

    // we can inspect the parse tree
    let m = matcher.parse("2018/10/6").unwrap();
    assert!(m.name("numeric_date").is_some());
    assert_eq!(m.name("year").unwrap().as_str(), "2018");

    let m = matcher.parse("Friday").unwrap();
    assert!(!m.name("numeric_date").is_some());
    assert!(m.name("weekday").is_some());

    // still more crazy things we allow
    assert!(matcher.is_match("F"));
    assert!(matcher.is_match("friday"));
    assert!(matcher.is_match("Fri"));

    // but we said single-letter days had to be capitalized
    assert!(!matcher.is_match("f"));

    Ok(())
}

Limitations

Recursion

Because you must define sub-grammars before you use them, and because Rust's regular expression engine and formalism cannot represent it in any case, you cannot produce a recursive grammar in Pidgin. Something like this is impossible:

XP -> XP conj XP

Named Captures in Regex Elements

Pidgin allows you to provide regular expression elements to rules. Such an element is a Rust expression which may be converted to a String via to_string bracketed in the grammar! macro body by r( and ), so, for example r(r"\d+"). Regex elements allow you to insert regex expressions into the grammar that the grammar formalism doesn't provide itself, like anchors and character classes.

Rust's regular expression engine does not allow the reuse of names in named capturing groups, however. (?<foo>bar) baz (?<foo>qux) won't compile into a usable regex because the name "foo" is reused. The following grammar produces just this situation:

grammar!{
    TOP -> <foo> ("bar") <foo>
    foo => r("(?P<baz>foo)")
};

The solution is to avoid named capturing groups in grammar regex elements. The grammar itself will handle the named capturing, so you should never need to do this.

Benchmarks

The principle motivation behind Pidgin is simply to produce something like an abstract syntax tree one can use to better understand matched text, but the regular expressions its [vec] elements generate generally match as well as or better than naively constructed regular expressions represented as alternations of expressions provided the expressions are bounded. "Bounded" here means the expression must match from the first character of the string to the last. An unbounded expression simply must match somewhere in the string.

// good
let rx = regex::Regex::new("\A(?:cat|camel|canteloupe)\z").unwrap();

// better!
let rx = grammar!{ TOP => r("\A") [["cat", "camel", "canteloupe"]] r("\z") }.rx().unwrap();

In the unbounded case, the naive regular expression performs better.

There is a simple benchmark suite included in the benches/ directory demonstrating this. The suite compares matching speed for "naive" regular expressions of the form foo|bar|baz to those generated by Pidgin, which involve no backtracking and, in this case, would be something like foo|ba[rz]. The benchmark suite compares non-matching to matching and bounded to unbounded patterns. The mean match/non-match times for these 8 cases were

The better time is bolded for each pair.

The full report can be generated by running

cargo bench

Full API

The complete Pidgin API is available at https://docs.rs/pidgin/.

Acknowledgments

In writing this I received considerable advice from TurkeyMcMac.