schala/schala-lang/src/parsing/combinator.rs

use nom::{
    Err,
    branch::alt,
    bytes::complete::{take_till, tag},
    character::complete::{alpha1, alphanumeric0, not_line_ending,none_of, char, one_of, space0, space1, multispace0, line_ending},
    combinator::{opt, peek, not, value, map, recognize},
    error::{context, VerboseError, ParseError},
    multi::{fold_many1, many1, many0, separated_list1, separated_list0},
    sequence::{pair, tuple, preceded},
    IResult, Parser,
};
 use std::rc::Rc;

type ParseResult<'a, O> = IResult<&'a str, O, VerboseError<&'a str>>;

use crate::ast::*;

fn rc_string(s: &str) -> Rc<String> {
    Rc::new(s.to_string())
}

fn tok<'a, O>(input_parser: impl Parser<&'a str, O, VerboseError<&'a str>>) -> impl FnMut(&'a str)
-> IResult<&'a str, O, VerboseError<&'a str>> { 

    context("tok", 
    map(tuple((ws0, input_parser)), |(_, output)|
    output))
}

fn kw<'a>(keyword_str: &'static str) -> impl FnMut(&'a str) -> ParseResult<()> {
    context("keyword",
    tok(value((), tag(keyword_str))))
}


// whitespace does consume at least one piece of whitespace - use ws0 for maybe none
fn whitespace(input: &str) -> ParseResult<()> {
    context("whitespace",
    alt((
    block_comment,
    line_comment,
    value((), space1),
    )))(input)
}

fn ws0(input: &str) -> ParseResult<()> {
    context("WS0",
    value((), many0(whitespace)))(input)
}

fn line_comment(input: &str) -> ParseResult<()> {
    value((),
    tuple((tag("//"), not_line_ending)),
    )(input)
}

fn block_comment(input: &str) -> ParseResult<()> {
    context("Block-comment",
    value((),
    tuple((
        tag("/*"), 
        many0(alt((
            value((), none_of("*/")),
            value((), none_of("/*")),
            block_comment,
        ))),
        tag("*/")
    ))))(input)
}

fn statement_delimiter(input: &str) -> ParseResult<()> {
    tok(alt((
        value((), line_ending),
        value((), char(';'))
        ))
    )(input)
}

fn block(input: &str) -> ParseResult<Block> {
    context("block",
    map(
    tuple((
        tok(char('{')),
        many0(statement_delimiter),
        separated_list0(statement_delimiter, statement),
        many0(statement_delimiter),
        tok(char('}')),
    )), |(_, _, items, _, _)| items.into()))(input)
}

fn statement(input: &str) -> ParseResult<Statement> {
    context("Parsing-statement",
    map(expression, |expr| Statement {
        id: Default::default(),
        location: Default::default(),
        kind: StatementKind::Expression(expr),
    }))(input)
}

fn expression(input: &str) -> ParseResult<Expression> {
    map(pair(expression_kind, opt(type_anno)), |(kind, maybe_anno)| {
        Expression::new(Default::default(), kind)
    })(input)
}

fn type_anno(input: &str) -> ParseResult<TypeIdentifier> {
    preceded(kw(":"), type_identifier)(input)
}

fn type_identifier(input: &str) -> ParseResult<TypeIdentifier> {
    /*
    alt((
        tuple((kw("("), separated_list0(kw(","), type_identifier), kw(")"))),
        type_singleton_name
    ))(input)
    */
    unimplemented!()
}

fn type_singleton_name(input: &str) -> ParseResult<TypeSingletonName> {
    unimplemented!()
}

pub fn expression_kind(input: &str) -> ParseResult<ExpressionKind> {
    context("expression-kind", primary_expr)(input)
}

fn primary_expr(input: &str) -> ParseResult<ExpressionKind> {

    context("primary-expr",
    alt((
        number_literal,
        bool_literal,
        identifier_expr,
    )))(input)

}

fn identifier_expr(input: &str) -> ParseResult<ExpressionKind> {
    context("identifier-expr", map(qualified_identifier, ExpressionKind::Value))(input)
}

fn qualified_identifier(input: &str) -> ParseResult<QualifiedName> {
    tok(
    map(
    separated_list1(tag("::"), map(identifier, rc_string)),
    |items| QualifiedName { id: Default::default(), components: items }
    ))(input)
}

fn identifier(input: &str) -> ParseResult<&str> {
    recognize(
    tuple((
        alt((tag("_"), alpha1)),
        alphanumeric0,
    )))(input)
}

fn bool_literal(input: &str) -> ParseResult<ExpressionKind> {
    context("bool-literal",
    alt((
        map(kw("true"), |_| ExpressionKind::BoolLiteral(true)),
        map(kw("false"), |_| ExpressionKind::BoolLiteral(false)),
    )))(input)
}

fn number_literal(input: &str) -> ParseResult<ExpressionKind> {
    map(alt((tok(hex_literal), tok(bin_literal), tok(dec_literal))), ExpressionKind::NatLiteral)(input)
}

fn dec_literal(input: &str) -> ParseResult<u64> {
    map(digits(digit_group_dec), |chars: Vec<char>| {
        let s: String = chars.into_iter().collect();
        s.parse().unwrap()
    })(input)
}

fn hex_literal(input: &str) -> ParseResult<u64> {
    map(preceded(alt((tag("0x"), tag("0X"))), digits(digit_group_hex)), |chars: Vec<char>| {
        let s: String = chars.into_iter().collect();
        parse_hex(&s).unwrap()
    })(input)
}

fn bin_literal(input: &str) -> ParseResult<u64> {
    map(preceded(alt((tag("0b"), tag("0B"))), digits(digit_group_bin)), |chars: Vec<char>| {
        let s: String = chars.into_iter().collect();
        parse_binary(&s).unwrap()
    })(input)
}

fn digits<'a, E: ParseError<&'a str>>(
    digit_type: impl Parser<&'a str, Vec<char>, E>,
) -> impl FnMut(&'a str) -> IResult<&'a str, Vec<char>, E> {
    map(separated_list1(many1(char('_')), digit_type), |items: Vec<Vec<char>>| {
        items.into_iter().flatten().collect()
    })
}

fn digit_group_dec(input: &str) -> ParseResult<Vec<char>> {
    many1(one_of("0123456789"))(input)
}

fn digit_group_hex(input: &str) -> ParseResult<Vec<char>> {
    many1(one_of("0123456789abcdefABCDEF"))(input)
}

fn digit_group_bin(input: &str) -> ParseResult<Vec<char>> {
    many1(one_of("01"))(input)
}

fn parse_binary(digits: &str) -> Result<u64, &'static str> {
    let mut result: u64 = 0;
    let mut multiplier = 1;
    for d in digits.chars().rev() {
        match d {
            '1' => result += multiplier,
            '0' => (),
            '_' => continue,
            _ => unreachable!(),
        }
        multiplier = match multiplier.checked_mul(2) {
            Some(m) => m,
            None => return Err("Binary expression will overflow"),
        }
    }
    Ok(result)
}

fn parse_hex(digits: &str) -> Result<u64, &'static str> {
    let mut result: u64 = 0;
    let mut multiplier: u64 = 1;
    for d in digits.chars().rev() {
        if d == '_' {
            continue;
        }
        match d.to_digit(16) {
            Some(n) => result += n as u64 * multiplier,
            None => return Err("Internal parser error: invalid hex digit"),
        }
        multiplier = match multiplier.checked_mul(16) {
            Some(m) => m,
            None => return Err("Hexadecimal expression will overflow"),
        }
    }
    Ok(result)
}

#[cfg(test)]
mod test {
    use pretty_assertions::assert_eq;

    use super::*;

fn rc(s: &str) -> Rc<String> {
    Rc::new(s.to_owned())
}
macro_rules! qn {
  ( $( $component:ident),* ) => {
    {
        let mut components = vec![];
        $(
        components.push(rc(stringify!($component)));
        )*
        QualifiedName { components, id: Default::default() }
    }
  };
}

    #[test]
    fn combinator_test1() {
        assert_eq!(digits(digit_group_dec)("342").unwrap().1, vec!['3', '4', '2']);
        assert_eq!(bin_literal("0b1111qsdf"), Ok(("qsdf", 15)));
    }

    #[test]
    fn combinator_test_ws0() {
        assert_eq!(block_comment("/*yolo*/").unwrap(), ("", ()));
        assert_eq!(block_comment("/*yolo*/ jumpy /*nah*/").unwrap(), (" jumpy /*nah*/", ()));
        assert_eq!(ws0("/*    yolo  */ ").unwrap(), ("", ()));
        assert_eq!(ws0("/*  /* no */   yolo  */ ").unwrap(), ("", ()));
    }

    #[test]
    fn combinator_test2() {
        for s in ["  15", "   0b1111", "   1_5_", "0XF__", "0Xf"].iter() {
            assert_eq!(expression_kind(s).unwrap().1, ExpressionKind::NatLiteral(15));
        }

        assert_eq!(expression_kind(" /*gay*/  true").unwrap().1, ExpressionKind::BoolLiteral(true));
        assert_eq!(expression_kind(" /*yolo*/  barnaby").unwrap().1, ExpressionKind::Value(qn!(barnaby)));
    }

    #[test]
    fn combinator_test3() {
        let source = "{}";
        assert_eq!(block(source).unwrap().1, vec![].into());
        let source = r#"{

            //hella
            4_5 //bog
            11; /*chutney*/0xf
            }"#;
        let parsed = block(source).map_err(|err| match err {
            Err::Error(err) | Err::Failure(err) => nom::error::convert_error(source, err),
            _ => panic!()
        });

        if let Err(err) = parsed {
            println!("{}", err);
            panic!("parse error desu!");
        }

        assert_eq!(parsed.unwrap().1, vec![
            Statement { id: Default::default(), location:
            Default::default(), kind: StatementKind::Expression(Expression::new(Default::default(),
            ExpressionKind::NatLiteral(45))) },
            Statement { id: Default::default(), location:
            Default::default(), kind: StatementKind::Expression(Expression::new(Default::default(),
            ExpressionKind::NatLiteral(11))) },
            Statement { id: Default::default(), location:
            Default::default(), kind: StatementKind::Expression(Expression::new(Default::default(),
            ExpressionKind::NatLiteral(15))) },
        ].into());
    }
}