Merge branch 'bullshitium'

build.rs

@@ -66,10 +66,6 @@ fn write_test(test_file: &mut File, test: &walkdir::DirEntry) {
 }
 
 #[cfg(feature = "compare")]
-fn is_expect_fail(name: &str) -> Option<&str> {
+fn is_expect_fail(_name: &str) -> Option<&str> {
-    match name {
+    None
-        "greater_element_drawer_drawer_with_headline_inside" => Some("Apparently lines with :end: become their own paragraph. This odd behavior needs to be investigated more."),
-        "element_container_priority_footnote_definition_dynamic_block" => Some("Apparently broken begin lines become their own paragraph."),
-        _ => None,
-    }
 }

org_mode_samples/bullshitium/fake_paragraph/broken_end.org

@@ -0,0 +1,3 @@
+foo
+:end:
+bar

org_mode_samples/bullshitium/fake_paragraph/broken_end_isolated.org

@@ -0,0 +1,2 @@
+foo
+:end:

src/parser/bullshitium.rs

@@ -0,0 +1,165 @@
+use nom::branch::alt;
+use nom::bytes::complete::tag_no_case;
+use nom::character::complete::anychar;
+use nom::character::complete::space0;
+use nom::multi::many_till;
+use nom::sequence::tuple;
+
+use super::paragraph::paragraph;
+use super::util::maybe_consume_trailing_whitespace_if_not_exiting;
+use super::util::org_line_ending;
+use super::util::start_of_line;
+use super::OrgSource;
+use crate::context::bind_context;
+use crate::context::RefContext;
+use crate::error::CustomError;
+use crate::error::Res;
+use crate::parser::macros::element;
+use crate::types::AffiliatedKeywords;
+use crate::types::Object;
+use crate::types::Paragraph;
+use crate::types::PlainText;
+
+#[cfg_attr(
+    feature = "tracing",
+    tracing::instrument(ret, level = "debug", skip(context))
+)]
+pub(crate) fn bullshitium<'b, 'g, 'r, 's>(
+    context: RefContext<'b, 'g, 'r, 's>,
+    input: OrgSource<'s>,
+) -> Res<OrgSource<'s>, Paragraph<'s>> {
+    alt((
+        bind_context!(broken_end, context),
+        bind_context!(broken_dynamic_block, context),
+    ))(input)
+}
+
+#[cfg_attr(
+    feature = "tracing",
+    tracing::instrument(ret, level = "debug", skip(context))
+)]
+pub(crate) fn detect_bullshitium<'b, 'g, 'r, 's>(
+    context: RefContext<'b, 'g, 'r, 's>,
+    input: OrgSource<'s>,
+) -> Res<OrgSource<'s>, ()> {
+    element!(detect_broken_end, context, input);
+    element!(detect_broken_dynamic_block, context, input);
+    Err(nom::Err::Error(CustomError::Static("No bullshitium.")))
+}
+
+#[cfg_attr(
+    feature = "tracing",
+    tracing::instrument(ret, level = "debug", skip(context))
+)]
+pub(crate) fn broken_end<'b, 'g, 'r, 's>(
+    context: RefContext<'b, 'g, 'r, 's>,
+    input: OrgSource<'s>,
+) -> Res<OrgSource<'s>, Paragraph<'s>> {
+    start_of_line(input)?;
+    let (remaining, _) = space0(input)?;
+    let (remaining, _) = tag_no_case(":end:")(remaining)?;
+    let (lead_in_remaining, _) = tuple((space0, org_line_ending))(remaining)?;
+    if let Ok((remaining, mut paragraph)) =
+        paragraph(std::iter::empty(), lead_in_remaining, context, input)
+    {
+        match paragraph.children.first_mut() {
+            Some(Object::PlainText(plain_text)) => {
+                plain_text.source = input.get_until_end_of_str(plain_text.source).into();
+            }
+            Some(obj) => {
+                panic!("Unhandled first object type inside bullshitium {:?}", obj);
+            }
+            None => {
+                unreachable!("Paragraph must have children.");
+            }
+        };
+        Ok((remaining, paragraph))
+    } else {
+        let (remaining, _trailing_ws) =
+            maybe_consume_trailing_whitespace_if_not_exiting(context, lead_in_remaining)?;
+
+        Ok((
+            remaining,
+            Paragraph {
+                source: input.get_until(remaining).into(),
+                affiliated_keywords: AffiliatedKeywords::default(),
+                children: vec![Object::PlainText(PlainText {
+                    source: input.get_until(lead_in_remaining).into(),
+                })],
+            },
+        ))
+    }
+}
+
+#[cfg_attr(
+    feature = "tracing",
+    tracing::instrument(ret, level = "debug", skip(_context))
+)]
+pub(crate) fn detect_broken_end<'b, 'g, 'r, 's>(
+    _context: RefContext<'b, 'g, 'r, 's>,
+    input: OrgSource<'s>,
+) -> Res<OrgSource<'s>, ()> {
+    start_of_line(input)?;
+    let (remaining, _) = space0(input)?;
+    let (remaining, _) = tag_no_case(":end:")(remaining)?;
+    let (_remaining, _) = tuple((space0, org_line_ending))(remaining)?;
+    Ok((input, ()))
+}
+
+#[cfg_attr(
+    feature = "tracing",
+    tracing::instrument(ret, level = "debug", skip(context))
+)]
+pub(crate) fn broken_dynamic_block<'b, 'g, 'r, 's>(
+    context: RefContext<'b, 'g, 'r, 's>,
+    input: OrgSource<'s>,
+) -> Res<OrgSource<'s>, Paragraph<'s>> {
+    start_of_line(input)?;
+    let (remaining, _) = space0(input)?;
+    let (remaining, _) = tag_no_case("#+BEGIN:")(remaining)?;
+    let (lead_in_remaining, _) = many_till(anychar, org_line_ending)(remaining)?;
+    if let Ok((remaining, mut paragraph)) =
+        paragraph(std::iter::empty(), lead_in_remaining, context, input)
+    {
+        match paragraph.children.first_mut() {
+            Some(Object::PlainText(plain_text)) => {
+                plain_text.source = input.get_until_end_of_str(plain_text.source).into();
+            }
+            Some(obj) => {
+                panic!("Unhandled first object type inside bullshitium {:?}", obj);
+            }
+            None => {
+                unreachable!("Paragraph must have children.");
+            }
+        };
+        Ok((remaining, paragraph))
+    } else {
+        let (remaining, _trailing_ws) =
+            maybe_consume_trailing_whitespace_if_not_exiting(context, lead_in_remaining)?;
+
+        Ok((
+            remaining,
+            Paragraph {
+                source: input.get_until(remaining).into(),
+                affiliated_keywords: AffiliatedKeywords::default(),
+                children: vec![Object::PlainText(PlainText {
+                    source: input.get_until(lead_in_remaining).into(),
+                })],
+            },
+        ))
+    }
+}
+
+#[cfg_attr(
+    feature = "tracing",
+    tracing::instrument(ret, level = "debug", skip(_context))
+)]
+pub(crate) fn detect_broken_dynamic_block<'b, 'g, 'r, 's>(
+    _context: RefContext<'b, 'g, 'r, 's>,
+    input: OrgSource<'s>,
+) -> Res<OrgSource<'s>, ()> {
+    start_of_line(input)?;
+    let (remaining, _) = space0(input)?;
+    let (_remaining, _) = tag_no_case("#+BEGIN:")(remaining)?;
+    Ok((input, ()))
+}

src/parser/element_parser.rs

@@ -32,6 +32,8 @@ use crate::event_count::record_event;
 #[cfg(feature = "event_count")]
 use crate::event_count::EventType;
 use crate::parser::affiliated_keyword::affiliated_keywords;
+use crate::parser::bullshitium::bullshitium;
+use crate::parser::bullshitium::detect_bullshitium;
 use crate::parser::macros::ak_element;
 use crate::parser::macros::element;
 use crate::parser::table::org_mode_table;
@@ -242,6 +244,9 @@ fn _element<'b, 'g, 'r, 's>(
     );
 
     if can_be_paragraph {
+        // Fake paragraphs
+        element!(bullshitium, context, input, Element::Paragraph);
+
         // Paragraph without affiliated keyword
         ak_element!(
             paragraph,
@@ -319,6 +324,11 @@ fn _detect_element<'b, 'g, 'r, 's>(
         input
     );
 
+    // Fake paragraphs
+    if !can_be_paragraph {
+        element!(detect_bullshitium, context, input);
+    }
+
     if _element(context, input, can_be_paragraph).is_ok() {
         return Ok((input, ()));
     }

src/parser/mod.rs

@@ -1,6 +1,7 @@
 mod affiliated_keyword;
 mod angle_link;
 mod babel_call;
+mod bullshitium;
 mod citation;
 mod citation_reference;
 mod clock;

src/parser/org_source.rs

@@ -82,6 +82,15 @@ impl<'s> OrgSource<'s> {
         self.slice(..(other.end - self.start))
     }
 
+    pub(crate) fn get_until_end_of_str(&self, other: &'s str) -> OrgSource<'s> {
+        let full_source_start = self.full_source.as_ptr() as usize;
+        let other_start = other.as_ptr() as usize - full_source_start;
+        let other_end = other_start + other.len();
+        debug_assert!(other_start >= self.start);
+        debug_assert!(other_end <= self.end);
+        self.slice(..(other_end - self.start))
+    }
+
     pub(crate) fn get_start_of_line(&self) -> OrgSource<'s> {
         let skipped_text = self.text_since_line_break();
         let mut bracket_depth = self.bracket_depth;