#nom #grapheme #cluster #unicode #string #spans #unicode-characters

nom-grapheme-clusters

Adapter that allows nom to account for unicode grapheme clusters

9 releases (4 breaking)

0.5.2 Sep 29, 2023
0.5.1 Sep 24, 2023
0.5.0 May 21, 2023
0.4.1 May 20, 2023
0.1.0 Dec 22, 2022

#5 in #grapheme

Custom license

105KB
3K SLoC

nom-grapheme-clusters

A library that adds support to strings for nom such that:

  1. An "element" of such string is a grapheme cluster/segment, not a byte nor an UTF-8 character. A grapheme cluster is a sequence of one or more characters such that they are displayed as what a human would feel like a single character should be. For instance, "m̤̊" is a single grapheme cluster, but not a single character although it looks like a single written character.

  2. Grapheme clusters location at the input string (i.e. the "source code") are tracked by their absolute position, line and column. All of the three data are counted in terms of grapheme clusters as well. For instance, in the string "m̤̊ef", the grapheme cluster "e" is at position 1, line 0 and column 1 (all of them begin at 0), because "m̤̊" counts as a single thing.

Main-branch Docs

https://brunoczim.github.io/nom-grapheme-clusters/nom_grapheme_clusters/

Example

use nom_grapheme_clusters::{Source, Span, SpanContent, tag_table};
use nom::{IResult, bytes::complete::tag, combinator::map};

#[derive(Debug, Clone)]
struct Tags {
    smth: Span,
    atn: Span,
}

#[derive(Debug, Clone, PartialEq, Eq)]
struct ParsedAtn {
    span: Span,
}

#[derive(Debug, Clone, PartialEq, Eq)]
struct ParsedSmth {
    span: Span,
}

fn parse_smth<'a>(
    tags: &'a Tags
) -> impl FnMut(Span) -> IResult<Span, ParsedAtn> + 'a {
    map(tag(&tags.smth), |span| ParsedAtn { span })
}

fn parse_atn<'a>(
    tags: &'a Tags
) -> impl FnMut(Span) -> IResult<Span, ParsedAtn> + 'a {
    map(tag(&tags.atn), |span| ParsedAtn { span })
}

fn main() {
    let tags = tag_table! {
        Tags {
            smth: "smth",
            atn: "atn̩̊",
        }
    };
    let source = Source::new("file.txt", "atn̩̊smtha");
   
    let span0 = source.full_span();
    let (span1, parsed) = parse_atn(&tags)(span0).unwrap();
    assert_eq!(parsed.span.as_str(), "atn̩̊");
    assert_eq!(parsed.span.start().position(), 0);
    assert_eq!(parsed.span.start().line(), 0);
    assert_eq!(parsed.span.start().column(), 0);
    assert_eq!(parsed.span.len(), 3);
    assert_eq!(parsed.span.end().position(), 3);
    assert_eq!(parsed.span.end().line(), 0);
    assert_eq!(parsed.span.end().column(), 3);
   
    let (span2, parsed) = parse_smth(&tags)(span1).unwrap();
    assert_eq!(parsed.span.as_str(), "smth");
    assert_eq!(parsed.span.start().position(), 3);
    assert_eq!(parsed.span.start().line(), 0);
    assert_eq!(parsed.span.start().column(), 3);
    assert_eq!(parsed.span.len(), 4);
    assert_eq!(parsed.span.end().position(), 7);
    assert_eq!(parsed.span.end().line(), 0);
    assert_eq!(parsed.span.end().column(), 7);
   
    let result = parse_atn(&tags)(span2);
    assert!(result.is_err());
    println!("{}", result.unwrap_err());
}

Dependencies

~315–570KB