Compare heading tags.

org_mode_samples/greater_element/dynamic_block/simple.org

@@ -0,0 +1,25 @@
+#+BEGIN: clocktable :scope file :maxlevel 2
+#+CAPTION: Clock summary at [2023-08-25 Fri 05:34]
+| Headline     | Time   |
+|--------------+--------|
+| *Total time* | *0:00* |
+#+END:
+
+#+BEGIN: columnview :hlines 1 :id global
+| ITEM  | TODO | PRIORITY | TAGS                         |
+|-------+------+----------+------------------------------|
+| Foo   |      | B        |                              |
+|-------+------+----------+------------------------------|
+| Bar   | TODO | B        |                              |
+|-------+------+----------+------------------------------|
+| Baz   |      | B        | :thisisatag:                 |
+| Lorem |      | B        | :thisshouldinheritfromabove: |
+| Ipsum |      | B        | :multiple:tags:              |
+#+END:
+* Foo
+* TODO Bar
+* Baz :thisisatag:
+** Lorem :thisshouldinheritfromabove:
+*** Ipsum                                                   :multiple:tags:
+* Dolar ::
+* cat :dog: bat

src/compare/diff.rs

@@ -1,3 +1,5 @@
+use std::collections::HashSet;
+
 use super::util::assert_bounds;
 use super::util::assert_name;
 use crate::parser::sexp::Token;
@@ -58,6 +60,7 @@ use crate::parser::Timestamp;
 use crate::parser::Underline;
 use crate::parser::Verbatim;
 use crate::parser::VerseBlock;
+use crate::parser::sexp::unquote;
 
 #[derive(Debug)]
 pub struct DiffResult {
@@ -323,6 +326,7 @@ fn compare_heading<'s>(
     let children = emacs.as_list()?;
     let mut child_status = Vec::new();
     let mut this_status = DiffStatus::Good;
+    let mut message = None;
     let emacs_name = "headline";
     if assert_name(emacs, emacs_name).is_err() {
         this_status = DiffStatus::Bad;
@@ -347,6 +351,19 @@ fn compare_heading<'s>(
     for (emacs_child, rust_child) in title.as_list()?.iter().zip(rust.title.iter()) {
         child_status.push(compare_object(source, emacs_child, rust_child)?);
     }
+    let emacs_tags = get_tags_from_heading(emacs)?;
+    let emacs_tags: HashSet<_> = emacs_tags.iter().map(|val| val.as_str()).collect();
+    let rust_tags: HashSet<&str> = rust.tags.iter().map(|val| *val).collect();
+    let difference: Vec<&str> = emacs_tags
+        .symmetric_difference(&rust_tags)
+        .map(|val| *val)
+        .collect();
+    if !difference.is_empty() {
+        this_status = DiffStatus::Bad;
+        message = Some(format!("Mismatched tags: {}", difference.join(", ")));
+    }
+
+    // TODO: Compare todo-keyword, level, priority
 
     for (emacs_child, rust_child) in children.iter().skip(2).zip(rust.children.iter()) {
         match rust_child {
@@ -362,11 +379,38 @@ fn compare_heading<'s>(
     Ok(DiffResult {
         status: this_status,
         name: emacs_name.to_owned(),
-        message: None,
+        message,
         children: child_status,
     })
 }
 
+fn get_tags_from_heading<'s>(
+    emacs: &'s Token<'s>,
+) -> Result<HashSet<String>, Box<dyn std::error::Error>> {
+    let children = emacs.as_list()?;
+    let attributes_child = children
+        .iter()
+        .nth(1)
+        .ok_or("Should have an attributes child.")?;
+    let attributes_map = attributes_child.as_map()?;
+    let tags = attributes_map
+        .get(":tags")
+        .ok_or("Missing :tags attribute.")?;
+    match tags.as_atom() {
+        Ok("nil") => {
+            return Ok(HashSet::new());
+        }
+        Ok(val) => panic!("Unexpected value for tags: {:?}", val),
+        Err(_) => {}
+    };
+    let tags = {
+        let tags = tags.as_list()?;
+        let strings = tags.iter().map(Token::as_atom).collect::<Result<Vec<&str>, _>>()?;
+        strings.into_iter().map(unquote).collect::<Result<HashSet<String>, _>>()?
+    };
+    Ok(tags)
+}
+
 fn compare_paragraph<'s>(
     source: &'s str,
     emacs: &'s Token<'s>,
@@ -1025,7 +1069,7 @@ fn compare_plain_text<'s>(
             rust.source.len()
         ));
     }
-    let unquoted_text = text.unquote()?;
+    let unquoted_text = unquote(text.text)?;
     if unquoted_text != rust.source {
         this_status = DiffStatus::Bad;
         message = Some(format!(

src/parser/document.rs

@@ -1,6 +1,8 @@
 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::eof;
 use nom::combinator::map;
@@ -12,6 +14,7 @@ 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::element::Element;
@@ -51,6 +54,7 @@ pub struct Heading<'s> {
     pub source: &'s str,
     pub stars: usize,
     pub title: Vec<Object<'s>>,
+    pub tags: Vec<&'s str>,
     pub children: Vec<DocumentElement<'s>>,
 }
 
@@ -267,7 +271,7 @@ fn heading<'r, 's>(
     input: OrgSource<'s>,
 ) -> Res<OrgSource<'s>, Heading<'s>> {
     not(|i| context.check_exit_matcher(i))(input)?;
-    let (remaining, (star_count, _ws, title)) = headline(context, input)?;
+    let (remaining, (star_count, _ws, title, heading_tags)) = headline(context, input)?;
     let section_matcher = parser_with_context!(section)(context);
     let heading_matcher = parser_with_context!(heading)(context);
     let (remaining, children) = many0(alt((
@@ -284,6 +288,7 @@ fn heading<'r, 's>(
             source: source.into(),
             stars: star_count,
             title,
+            tags: heading_tags,
             children,
         },
     ))
@@ -293,30 +298,63 @@ fn heading<'r, 's>(
 fn headline<'r, 's>(
     context: Context<'r, 's>,
     input: OrgSource<'s>,
-) -> Res<OrgSource<'s>, (usize, OrgSource<'s>, Vec<Object<'s>>)> {
+) -> Res<OrgSource<'s>, (usize, OrgSource<'s>, Vec<Object<'s>>, Vec<&'s str>)> {
     let parser_context =
         context.with_additional_node(ContextElement::ExitMatcherNode(ExitMatcherNode {
             class: ExitClass::Document,
-            exit_matcher: &headline_end,
+            exit_matcher: &headline_title_end,
         }));
     let standard_set_object_matcher = parser_with_context!(standard_set_object)(&parser_context);
 
-    let (remaining, (_sol, star_count, ws, title, _line_ending)) = tuple((
+    let (remaining, (_sol, star_count, ws, title, maybe_tags, _ws, _line_ending)) = tuple((
         start_of_line,
         many1_count(tag("*")),
         space1,
         many1(standard_set_object_matcher),
+        opt(tuple((space0, tags))),
+        space0,
         alt((line_ending, eof)),
     ))(input)?;
-    Ok((remaining, (star_count, ws, title)))
+    Ok((
+        remaining,
+        (
+            star_count,
+            ws,
+            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_end<'r, 's>(
+fn headline_title_end<'r, 's>(
     _context: Context<'r, 's>,
     input: OrgSource<'s>,
 ) -> Res<OrgSource<'s>, OrgSource<'s>> {
-    line_ending(input)
+    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)
 }
 
 impl<'s> Document<'s> {

src/parser/sexp.rs

@@ -35,44 +35,6 @@ pub struct TextWithProperties<'s> {
     pub properties: Vec<Token<'s>>,
 }
 
-impl<'s> TextWithProperties<'s> {
-    pub fn unquote(&self) -> Result<String, Box<dyn std::error::Error>> {
-        let mut out = String::with_capacity(self.text.len());
-        if !self.text.starts_with(r#"""#) {
-            return Err("Quoted text does not start with quote.".into());
-        }
-        if !self.text.ends_with(r#"""#) {
-            return Err("Quoted text does not end with quote.".into());
-        }
-        let interior_text = &self.text[1..(self.text.len() - 1)];
-        let mut state = ParseState::Normal;
-        for current_char in interior_text.chars().into_iter() {
-            state = match (state, current_char) {
-                (ParseState::Normal, '\\') => ParseState::Escape,
-                (ParseState::Normal, _) => {
-                    out.push(current_char);
-                    ParseState::Normal
-                }
-                (ParseState::Escape, 'n') => {
-                    out.push('\n');
-                    ParseState::Normal
-                }
-                (ParseState::Escape, '\\') => {
-                    out.push('\\');
-                    ParseState::Normal
-                }
-                (ParseState::Escape, '"') => {
-                    out.push('"');
-                    ParseState::Normal
-                }
-                _ => todo!(),
-            };
-        }
-
-        Ok(out)
-    }
-}
-
 enum ParseState {
     Normal,
     Escape,
@@ -133,6 +95,42 @@ impl<'s> Token<'s> {
     }
 }
 
+pub fn unquote(text: &str) -> Result<String, Box<dyn std::error::Error>> {
+    let mut out = String::with_capacity(text.len());
+    if !text.starts_with(r#"""#) {
+        return Err("Quoted text does not start with quote.".into());
+    }
+    if !text.ends_with(r#"""#) {
+        return Err("Quoted text does not end with quote.".into());
+    }
+    let interior_text = &text[1..(text.len() - 1)];
+    let mut state = ParseState::Normal;
+    for current_char in interior_text.chars().into_iter() {
+        state = match (state, current_char) {
+            (ParseState::Normal, '\\') => ParseState::Escape,
+            (ParseState::Normal, _) => {
+                out.push(current_char);
+                ParseState::Normal
+            }
+            (ParseState::Escape, 'n') => {
+                out.push('\n');
+                ParseState::Normal
+            }
+            (ParseState::Escape, '\\') => {
+                out.push('\\');
+                ParseState::Normal
+            }
+            (ParseState::Escape, '"') => {
+                out.push('"');
+                ParseState::Normal
+            }
+            _ => todo!(),
+        };
+    }
+
+    Ok(out)
+}
+
 #[cfg_attr(feature = "tracing", tracing::instrument(ret, level = "debug"))]
 pub fn sexp_with_padding<'s>(input: &'s str) -> Res<&'s str, Token<'s>> {
     let (remaining, _) = multispace0(input)?;