organic/src/parser/document.rs

use nom::branch::alt;
use nom::bytes::complete::tag;
use nom::character::complete::anychar;
use nom::character::complete::line_ending;
use nom::character::complete::space0;
use nom::character::complete::space1;
use nom::combinator::all_consuming;
use nom::combinator::eof;
use nom::combinator::map;
use nom::combinator::not;
use nom::combinator::opt;
use nom::combinator::recognize;
use nom::combinator::verify;
use nom::multi::many0;
use nom::multi::many1;
use nom::multi::many1_count;
use nom::multi::many_till;
use nom::multi::separated_list1;
use nom::sequence::tuple;

use super::in_buffer_settings::scan_for_in_buffer_settings;
use super::org_source::OrgSource;
use super::token::AllTokensIterator;
use super::token::Token;
use super::util::exit_matcher_parser;
use super::util::get_consumed;
use super::util::start_of_line;
use crate::context::parser_with_context;
use crate::context::Context;
use crate::context::ContextElement;
use crate::context::ExitClass;
use crate::context::ExitMatcherNode;
use crate::context::GlobalSettings;
use crate::context::List;
use crate::context::RefContext;
use crate::error::CustomError;
use crate::error::MyError;
use crate::error::Res;
use crate::parser::comment::comment;
use crate::parser::element_parser::element;
use crate::parser::object_parser::standard_set_object;
use crate::parser::org_source::convert_error;
use crate::parser::planning::planning;
use crate::parser::property_drawer::property_drawer;
use crate::parser::util::blank_line;
use crate::parser::util::maybe_consume_trailing_whitespace_if_not_exiting;
use crate::types::Document;
use crate::types::DocumentElement;
use crate::types::Element;
use crate::types::Heading;
use crate::types::Object;
use crate::types::Section;

/// Parse a full org-mode document.
///
/// This is the main entry point for Organic. It will parse the full contents of the input string as an org-mode document.
#[cfg_attr(feature = "tracing", tracing::instrument(ret, level = "debug"))]
#[allow(dead_code)]
pub fn parse<'s>(input: &'s str) -> Result<Document<'s>, String> {
    let global_settings = GlobalSettings::default();
    let initial_context = ContextElement::document_context();
    let initial_context = Context::new(&global_settings, List::new(&initial_context));
    let wrapped_input = OrgSource::new(input);
    let ret =
        all_consuming(parser_with_context!(document_org_source)(&initial_context))(wrapped_input)
            .map_err(|err| err.to_string())
            .map(|(_remaining, parsed_document)| parsed_document);
    ret
}

/// Parse a full org-mode document with starting settings.
///
/// This is the secondary entry point for Organic. It will parse the full contents of the input string as an org-mode document starting with the settings you supplied.
///
/// This will not prevent additional settings from being learned during parsing, for example when encountering a "#+TODO".
#[cfg_attr(feature = "tracing", tracing::instrument(ret, level = "debug"))]
#[allow(dead_code)]
pub fn parse_with_settings<'g, 's>(
    input: &'s str,
    global_settings: &'g GlobalSettings<'g, 's>,
) -> Result<Document<'s>, String> {
    let initial_context = ContextElement::document_context();
    let initial_context = Context::new(global_settings, List::new(&initial_context));
    let wrapped_input = OrgSource::new(input);
    let ret =
        all_consuming(parser_with_context!(document_org_source)(&initial_context))(wrapped_input)
            .map_err(|err| err.to_string())
            .map(|(_remaining, parsed_document)| parsed_document);
    ret
}

/// Parse a full org-mode document.
///
/// Use this entry point when you want to have direct control over the starting context or if you want to use this integrated with other nom parsers. For general-purpose usage, the `parse` and `parse_with_settings` functions are a lot simpler.
///
/// This will not prevent additional settings from being learned during parsing, for example when encountering a "#+TODO".
#[cfg_attr(feature = "tracing", tracing::instrument(ret, level = "debug"))]
#[allow(dead_code)]
pub fn document<'b, 'g, 'r, 's>(
    context: RefContext<'b, 'g, 'r, 's>,
    input: &'s str,
) -> Res<&'s str, Document<'s>> {
    let (remaining, doc) = document_org_source(context, input.into()).map_err(convert_error)?;
    Ok((Into::<&str>::into(remaining), doc))
}

#[cfg_attr(feature = "tracing", tracing::instrument(ret, level = "debug"))]
#[allow(dead_code)]
fn document_org_source<'b, 'g, 'r, 's>(
    context: RefContext<'b, 'g, 'r, 's>,
    input: OrgSource<'s>,
) -> Res<OrgSource<'s>, Document<'s>> {
    let mut final_settings = Vec::new();
    let (_, document_settings) = scan_for_in_buffer_settings(input)?;
    let setup_files: Vec<String> = document_settings
        .iter()
        .filter(|kw| kw.key.eq_ignore_ascii_case("setupfile"))
        .map(|kw| kw.value)
        .map(|setup_file| {
            context
                .get_global_settings()
                .file_access
                .read_file(setup_file)
                .map_err(|err| nom::Err::<CustomError<OrgSource<'_>>>::Failure(err.into()))
        })
        .collect::<Result<Vec<_>, _>>()?;
    for setup_file in setup_files.iter().map(String::as_str) {
        let (_, setup_file_settings) =
            scan_for_in_buffer_settings(setup_file.into()).map_err(|_err| {
                nom::Err::Error(CustomError::MyError(MyError(
                    "TODO: make this take an owned string so I can dump err.to_string() into it."
                        .into(),
                )))
            })?;
        final_settings.extend(setup_file_settings);
    }
    final_settings.extend(document_settings);
    // TODO: read the keywords into settings and apply them to the GlobalSettings.

    let (remaining, document) =
        _document(context, input).map(|(rem, out)| (Into::<&str>::into(rem), out))?;
    {
        // If there are radio targets in this document then we need to parse the entire document again with the knowledge of the radio targets.
        let all_radio_targets: Vec<&Vec<Object<'_>>> = document
            .iter_tokens()
            .filter_map(|tkn| match tkn {
                Token::Object(obj) => Some(obj),
                _ => None,
            })
            .filter_map(|obj| match obj {
                Object::RadioTarget(rt) => Some(rt),
                _ => None,
            })
            .map(|rt| &rt.children)
            .collect();
        if !all_radio_targets.is_empty() {
            let mut new_global_settings = context.get_global_settings().clone();
            new_global_settings.radio_targets = all_radio_targets;
            let parser_context = context.with_global_settings(&new_global_settings);
            let (remaining, document) = _document(&parser_context, input)
                .map(|(rem, out)| (Into::<&str>::into(rem), out))?;
            return Ok((remaining.into(), document));
        }
    }
    Ok((remaining.into(), document))
}

#[cfg_attr(feature = "tracing", tracing::instrument(ret, level = "debug"))]
fn _document<'b, 'g, 'r, 's>(
    context: RefContext<'b, 'g, 'r, 's>,
    input: OrgSource<'s>,
) -> Res<OrgSource<'s>, Document<'s>> {
    let zeroth_section_matcher = parser_with_context!(zeroth_section)(context);
    let heading_matcher = parser_with_context!(heading(0))(context);
    let (remaining, _blank_lines) = many0(blank_line)(input)?;
    let (remaining, zeroth_section) = opt(zeroth_section_matcher)(remaining)?;
    let (remaining, children) = many0(heading_matcher)(remaining)?;
    let source = get_consumed(input, remaining);
    Ok((
        remaining,
        Document {
            source: source.into(),
            zeroth_section,
            children,
        },
    ))
}

#[cfg_attr(feature = "tracing", tracing::instrument(ret, level = "debug"))]
fn zeroth_section<'b, 'g, 'r, 's>(
    context: RefContext<'b, 'g, 'r, 's>,
    input: OrgSource<'s>,
) -> Res<OrgSource<'s>, Section<'s>> {
    // TODO: The zeroth section is specialized so it probably needs its own parser
    let contexts = [
        ContextElement::ConsumeTrailingWhitespace(true),
        ContextElement::Context("section"),
        ContextElement::ExitMatcherNode(ExitMatcherNode {
            class: ExitClass::Document,
            exit_matcher: &section_end,
        }),
    ];
    let parser_context = context.with_additional_node(&contexts[0]);
    let parser_context = parser_context.with_additional_node(&contexts[1]);
    let parser_context = parser_context.with_additional_node(&contexts[2]);
    let without_consuming_whitespace_context = ContextElement::ConsumeTrailingWhitespace(false);
    let without_consuming_whitespace_context =
        parser_context.with_additional_node(&without_consuming_whitespace_context);

    let element_matcher = parser_with_context!(element(true))(&parser_context);
    let exit_matcher = parser_with_context!(exit_matcher_parser)(&parser_context);

    let (remaining, comment_and_property_drawer_element) = opt(tuple((
        opt(parser_with_context!(comment)(
            &without_consuming_whitespace_context,
        )),
        parser_with_context!(property_drawer)(context),
        many0(blank_line),
    )))(input)?;

    let (remaining, (mut children, _exit_contents)) = verify(
        many_till(element_matcher, exit_matcher),
        |(children, _exit_contents)| {
            !children.is_empty() || comment_and_property_drawer_element.is_some()
        },
    )(remaining)?;

    comment_and_property_drawer_element.map(|(comment, property_drawer, _ws)| {
        children.insert(0, Element::PropertyDrawer(property_drawer));
        comment
            .map(Element::Comment)
            .map(|ele| children.insert(0, ele));
    });

    let (remaining, _trailing_ws) =
        maybe_consume_trailing_whitespace_if_not_exiting(context, remaining)?;

    let source = get_consumed(input, remaining);
    Ok((
        remaining,
        Section {
            source: source.into(),
            children,
        },
    ))
}

#[cfg_attr(feature = "tracing", tracing::instrument(ret, level = "debug"))]
fn section<'b, 'g, 'r, 's>(
    context: RefContext<'b, 'g, 'r, 's>,
    mut input: OrgSource<'s>,
) -> Res<OrgSource<'s>, Section<'s>> {
    // TODO: The zeroth section is specialized so it probably needs its own parser
    let contexts = [
        ContextElement::ConsumeTrailingWhitespace(true),
        ContextElement::Context("section"),
        ContextElement::ExitMatcherNode(ExitMatcherNode {
            class: ExitClass::Document,
            exit_matcher: &section_end,
        }),
    ];
    let parser_context = context.with_additional_node(&contexts[0]);
    let parser_context = parser_context.with_additional_node(&contexts[1]);
    let parser_context = parser_context.with_additional_node(&contexts[2]);
    let element_matcher = parser_with_context!(element(true))(&parser_context);
    let exit_matcher = parser_with_context!(exit_matcher_parser)(&parser_context);
    let (mut remaining, (planning_element, property_drawer_element)) = tuple((
        opt(parser_with_context!(planning)(&parser_context)),
        opt(parser_with_context!(property_drawer)(&parser_context)),
    ))(input)?;
    if planning_element.is_none() && property_drawer_element.is_none() {
        let (remain, _ws) = many0(blank_line)(remaining)?;
        remaining = remain;
        input = remain;
    }
    let (remaining, (mut children, _exit_contents)) = verify(
        many_till(element_matcher, exit_matcher),
        |(children, _exit_contents)| {
            !children.is_empty() || property_drawer_element.is_some() || planning_element.is_some()
        },
    )(remaining)?;
    property_drawer_element
        .map(Element::PropertyDrawer)
        .map(|ele| children.insert(0, ele));
    planning_element
        .map(Element::Planning)
        .map(|ele| children.insert(0, ele));

    let (remaining, _trailing_ws) =
        maybe_consume_trailing_whitespace_if_not_exiting(context, remaining)?;

    let source = get_consumed(input, remaining);
    Ok((
        remaining,
        Section {
            source: source.into(),
            children,
        },
    ))
}

#[cfg_attr(feature = "tracing", tracing::instrument(ret, level = "debug"))]
fn section_end<'b, 'g, 'r, 's>(
    _context: RefContext<'b, 'g, 'r, 's>,
    input: OrgSource<'s>,
) -> Res<OrgSource<'s>, OrgSource<'s>> {
    recognize(detect_headline)(input)
}

const fn heading(
    parent_stars: usize,
) -> impl for<'b, 'g, 'r, 's> Fn(
    RefContext<'b, 'g, 'r, 's>,
    OrgSource<'s>,
) -> Res<OrgSource<'s>, Heading<'s>> {
    move |context, input: OrgSource<'_>| _heading(context, input, parent_stars)
}

#[cfg_attr(feature = "tracing", tracing::instrument(ret, level = "debug"))]
fn _heading<'b, 'g, 'r, 's>(
    context: RefContext<'b, 'g, 'r, 's>,
    input: OrgSource<'s>,
    parent_stars: usize,
) -> Res<OrgSource<'s>, Heading<'s>> {
    not(|i| context.check_exit_matcher(i))(input)?;
    let (remaining, (star_count, _ws, maybe_todo_keyword, title, heading_tags)) =
        headline(context, input, parent_stars)?;
    let section_matcher = parser_with_context!(section)(context);
    let heading_matcher = parser_with_context!(heading(star_count))(context);
    let (remaining, maybe_section) =
        opt(map(section_matcher, DocumentElement::Section))(remaining)?;
    let (remaining, mut children) =
        many0(map(heading_matcher, DocumentElement::Heading))(remaining)?;
    if let Some(section) = maybe_section {
        children.insert(0, section);
    }
    let remaining = if children.is_empty() {
        // Support empty headings
        let (remain, _ws) = many0(blank_line)(remaining)?;
        remain
    } else {
        remaining
    };

    let source = get_consumed(input, remaining);
    Ok((
        remaining,
        Heading {
            source: source.into(),
            stars: star_count,
            todo_keyword: maybe_todo_keyword
                .map(|(todo_keyword, _ws)| Into::<&str>::into(todo_keyword)),
            title,
            tags: heading_tags,
            children,
        },
    ))
}

#[cfg_attr(feature = "tracing", tracing::instrument(ret, level = "debug"))]
fn detect_headline<'s>(input: OrgSource<'s>) -> Res<OrgSource<'s>, ()> {
    tuple((start_of_line, many1(tag("*")), space1))(input)?;
    Ok((input, ()))
}

#[cfg_attr(feature = "tracing", tracing::instrument(ret, level = "debug"))]
fn headline<'b, 'g, 'r, 's>(
    context: RefContext<'b, 'g, 'r, 's>,
    input: OrgSource<'s>,
    parent_stars: usize,
) -> Res<
    OrgSource<'s>,
    (
        usize,
        OrgSource<'s>,
        Option<(OrgSource<'s>, OrgSource<'s>)>,
        Vec<Object<'s>>,
        Vec<&'s str>,
    ),
> {
    let parser_context = ContextElement::ExitMatcherNode(ExitMatcherNode {
        class: ExitClass::Document,
        exit_matcher: &headline_title_end,
    });
    let parser_context = context.with_additional_node(&parser_context);
    let standard_set_object_matcher = parser_with_context!(standard_set_object)(&parser_context);

    let (
        remaining,
        (_sol, star_count, ws, maybe_todo_keyword, title, maybe_tags, _ws, _line_ending),
    ) = tuple((
        start_of_line,
        verify(many1_count(tag("*")), |star_count| {
            *star_count > parent_stars
        }),
        space1,
        opt(tuple((heading_keyword, space1))),
        many1(standard_set_object_matcher),
        opt(tuple((space0, tags))),
        space0,
        alt((line_ending, eof)),
    ))(input)?;
    Ok((
        remaining,
        (
            star_count,
            ws,
            maybe_todo_keyword,
            title,
            maybe_tags
                .map(|(_ws, tags)| {
                    tags.into_iter()
                        .map(|single_tag| Into::<&str>::into(single_tag))
                        .collect()
                })
                .unwrap_or(Vec::new()),
        ),
    ))
}

#[cfg_attr(feature = "tracing", tracing::instrument(ret, level = "debug"))]
fn headline_title_end<'b, 'g, 'r, 's>(
    _context: RefContext<'b, 'g, 'r, 's>,
    input: OrgSource<'s>,
) -> Res<OrgSource<'s>, OrgSource<'s>> {
    recognize(tuple((
        opt(tuple((space0, tags, space0))),
        alt((line_ending, eof)),
    )))(input)
}

#[cfg_attr(feature = "tracing", tracing::instrument(ret, level = "debug"))]
fn tags<'s>(input: OrgSource<'s>) -> Res<OrgSource<'s>, Vec<OrgSource<'s>>> {
    let (remaining, (_open, tags, _close)) =
        tuple((tag(":"), separated_list1(tag(":"), single_tag), tag(":")))(input)?;
    Ok((remaining, tags))
}

#[cfg_attr(feature = "tracing", tracing::instrument(ret, level = "debug"))]
fn single_tag<'r, 's>(input: OrgSource<'s>) -> Res<OrgSource<'s>, OrgSource<'s>> {
    recognize(many1(verify(anychar, |c| {
        c.is_alphanumeric() || "_@#%".contains(*c)
    })))(input)
}

#[cfg_attr(feature = "tracing", tracing::instrument(ret, level = "debug"))]
fn heading_keyword<'s>(input: OrgSource<'s>) -> Res<OrgSource<'s>, OrgSource<'s>> {
    // TODO: This should take into account the value of "#+TODO:" ref https://orgmode.org/manual/Per_002dfile-keywords.html and possibly the configurable variable org-todo-keywords ref https://orgmode.org/manual/Workflow-states.html. Case is significant.
    alt((tag("TODO"), tag("DONE")))(input)
}

impl<'s> Document<'s> {
    pub fn iter_tokens<'r>(&'r self) -> impl Iterator<Item = Token<'r, 's>> {
        AllTokensIterator::new(Token::Document(self))
    }
}