organic/src/parser/plain_link.rs

use nom::branch::alt;
use nom::bytes::complete::tag;
use nom::bytes::complete::tag_no_case;
use nom::character::complete::anychar;
use nom::character::complete::none_of;
use nom::character::complete::one_of;
use nom::combinator::eof;
use nom::combinator::peek;
use nom::combinator::recognize;
use nom::multi::many_till;

use super::org_source::OrgSource;
use super::Context;
use crate::error::CustomError;
use crate::error::MyError;
use crate::error::Res;
use crate::parser::exiting::ExitClass;
use crate::parser::object::PlainLink;
use crate::parser::parser_context::ContextElement;
use crate::parser::parser_context::ExitMatcherNode;
use crate::parser::parser_with_context::parser_with_context;
use crate::parser::util::exit_matcher_parser;
use crate::parser::util::get_consumed;
use crate::parser::util::WORD_CONSTITUENT_CHARACTERS;

#[cfg_attr(feature = "tracing", tracing::instrument(ret, level = "debug"))]
pub fn plain_link<'r, 's>(
    context: Context<'r, 's>,
    input: OrgSource<'s>,
) -> Res<OrgSource<'s>, PlainLink<'s>> {
    let (remaining, _) = pre(context, input)?;
    let (remaining, proto) = protocol(context, remaining)?;
    let (remaining, _separator) = tag(":")(remaining)?;
    let (remaining, path) = path_plain(context, remaining)?;
    peek(parser_with_context!(post)(context))(remaining)?;
    let source = get_consumed(input, remaining);
    Ok((
        remaining,
        PlainLink {
            source: source.into(),
            link_type: proto.into(),
            path: path.into(),
        },
    ))
}

#[cfg_attr(feature = "tracing", tracing::instrument(ret, level = "debug"))]
fn pre<'r, 's>(_context: Context<'r, 's>, input: OrgSource<'s>) -> Res<OrgSource<'s>, ()> {
    let preceding_character = input.get_preceding_character();
    match preceding_character {
        // If None, we are at the start of the file which is fine
        None => {}
        Some(x) if !WORD_CONSTITUENT_CHARACTERS.contains(x) => {}
        Some(_) => {
            // Not at start of line, cannot be a heading
            return Err(nom::Err::Error(CustomError::MyError(MyError(
                "Not a valid pre character for plain link.".into(),
            ))));
        }
    };
    Ok((input, ()))
}

#[cfg_attr(feature = "tracing", tracing::instrument(ret, level = "debug"))]
fn post<'r, 's>(_context: Context<'r, 's>, input: OrgSource<'s>) -> Res<OrgSource<'s>, ()> {
    let (remaining, _) = alt((eof, recognize(none_of(WORD_CONSTITUENT_CHARACTERS))))(input)?;
    Ok((remaining, ()))
}

#[cfg_attr(feature = "tracing", tracing::instrument(ret, level = "debug"))]
pub fn protocol<'r, 's>(
    _context: Context<'r, 's>,
    input: OrgSource<'s>,
) -> Res<OrgSource<'s>, OrgSource<'s>> {
    // TODO: This should be defined by org-link-parameters
    let (remaining, proto) = alt((
        alt((
            tag_no_case("id"),
            tag_no_case("eww"),
            tag_no_case("rmail"),
            tag_no_case("mhe"),
            tag_no_case("irc"),
            tag_no_case("info"),
            tag_no_case("gnus"),
            tag_no_case("docview"),
            tag_no_case("bibtex"),
            tag_no_case("bbdb"),
            tag_no_case("w3m"),
        )),
        alt((
            tag_no_case("doi"),
            tag_no_case("file+sys"),
            tag_no_case("file+emacs"),
            tag_no_case("shell"),
            tag_no_case("news"),
            tag_no_case("mailto"),
            tag_no_case("https"),
            tag_no_case("http"),
            tag_no_case("ftp"),
            tag_no_case("help"),
            tag_no_case("file"),
            tag_no_case("elisp"),
        )),
    ))(input)?;
    Ok((remaining, proto))
}

#[cfg_attr(feature = "tracing", tracing::instrument(ret, level = "debug"))]
fn path_plain<'r, 's>(
    context: Context<'r, 's>,
    input: OrgSource<'s>,
) -> Res<OrgSource<'s>, OrgSource<'s>> {
    // TODO: "optionally containing parenthesis-wrapped non-whitespace non-bracket substrings up to a depth of two. The string must end with either a non-punctation non-whitespace character, a forwards slash, or a parenthesis-wrapped substring"
    let parser_context =
        context.with_additional_node(ContextElement::ExitMatcherNode(ExitMatcherNode {
            class: ExitClass::Beta,
            exit_matcher: &path_plain_end,
        }));

    let exit_matcher = parser_with_context!(exit_matcher_parser)(&parser_context);

    let (remaining, path) = recognize(many_till(anychar, peek(exit_matcher)))(input)?;
    Ok((remaining, path))
}

#[cfg_attr(feature = "tracing", tracing::instrument(ret, level = "debug"))]
fn path_plain_end<'r, 's>(
    _context: Context<'r, 's>,
    input: OrgSource<'s>,
) -> Res<OrgSource<'s>, OrgSource<'s>> {
    recognize(one_of(" \t\r\n()[]<>"))(input)
}