organic/src/parser/document.rs

use nom::combinator::all_consuming;
use nom::combinator::opt;
use nom::multi::many0;

use super::headline::heading;
use super::in_buffer_settings::apply_in_buffer_settings;
use super::in_buffer_settings::scan_for_in_buffer_settings;
use super::org_source::OrgSource;
use super::section::zeroth_section;
use super::token::AllTokensIterator;
use super::token::Token;
use super::util::get_consumed;
use crate::context::parser_with_context;
use crate::context::Context;
use crate::context::ContextElement;
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::org_source::convert_error;
use crate::parser::util::blank_line;
use crate::types::Document;
use crate::types::Object;

/// 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.
#[allow(dead_code)]
pub fn parse<'s>(input: &'s str) -> Result<Document<'s>, String> {
    parse_with_settings(input, &GlobalSettings::default())
}

/// 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".
#[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".
#[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);
    let new_settings = apply_in_buffer_settings(final_settings, context.get_global_settings())
        .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(),
            )))
        })?;
    let new_context = context.with_global_settings(&new_settings);
    let context = &new_context;

    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,
        },
    ))
}

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