-
Notifications
You must be signed in to change notification settings - Fork 2
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
WIP: refactor(metadata): use tokenization to get better parse diags.
TODO: - [ ] Actually get this working. Maybe not the right approach, if we can't stream tokens?
- Loading branch information
1 parent
9ed2b93
commit 97fe1ca
Showing
1 changed file
with
287 additions
and
0 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,287 @@ | ||
| use std::{cmp::Ordering, io, iter, str::Lines}; | ||
|
|
||
| use chumsky::{ | ||
| extra, | ||
| input::Stream, | ||
| prelude::{EmptyErr, Input, Rich}, | ||
| primitive::{any, choice, custom, just}, | ||
| span::SimpleSpan, | ||
| text::{ascii, inline_whitespace, newline}, | ||
| IterParser, Parser, | ||
| }; | ||
|
|
||
| #[derive(Clone, Debug)] | ||
| enum Event<'a> { | ||
| SingleLineComment { | ||
| span: SimpleSpan, | ||
| contents: &'a str, | ||
| contents_span: SimpleSpan, | ||
| }, | ||
| // Property { | ||
| // name: &'a str, | ||
| // name_span: SimpleSpan, | ||
| // value: &'a str, | ||
| // value_span: SimpleSpan, | ||
| // }, | ||
| EnterSection { | ||
| header_span: SimpleSpan, | ||
| name: UnescapedSectionHeader<'a>, | ||
| name_span: SimpleSpan, | ||
| }, | ||
| ExitSection { | ||
| end_point_span: SimpleSpan, | ||
| }, | ||
| } | ||
|
|
||
| // enum PropertyValue<'a> { | ||
| // Unconditional(&'a str), | ||
| // Conditional(ConditionalPropertyTokenizer<'a>), | ||
| // } | ||
| // | ||
| // struct ConditionalPropertyTokenizer<'a> {} | ||
|
|
||
| #[derive(Clone, Debug)] | ||
| struct UnescapedSectionHeader<'a> { | ||
| name: &'a str, | ||
| } | ||
|
|
||
| #[derive(Clone, Debug)] | ||
| struct Tokenizer<'a> { | ||
| input: &'a str, | ||
| current_idx: usize, | ||
| current_indent_level: u8, | ||
| } | ||
|
|
||
| #[derive(Clone, Debug)] | ||
| enum TokenizationError { | ||
| MisalignedIndent { span: SimpleSpan }, | ||
| UnexpectedIndent { indent_span: SimpleSpan }, | ||
| // TODO: populate | ||
| MalformedItem { skipped_span: SimpleSpan }, | ||
| } | ||
|
|
||
| fn lines<'a>(input: &'a str) -> impl Clone + Iterator<Item = (&'a str, SimpleSpan)> + 'a { | ||
| let parser = any::<_, extra::Err<EmptyErr>>() | ||
| .and_is(newline().not()) | ||
| .repeated() | ||
| .to_slice() | ||
| .then_ignore(newline().or_not()) | ||
| .map_with(|line, e| (e.span(), line)) | ||
| .lazy(); | ||
|
|
||
| let mut idx = 0; | ||
| iter::from_fn(move || loop { | ||
| if idx == input.len() { | ||
| return None; | ||
| } | ||
| let (span, line) = parser.parse(&input[idx..]).into_result().unwrap(); | ||
| idx += span.end; | ||
| if !line.trim().is_empty() { | ||
| break Some((line.trim_end(), span)); | ||
| } | ||
| }) | ||
| } | ||
|
|
||
| fn tokens<'a>(input: &'a str) -> impl Iterator<Item = Result<Event<'a>, TokenizationError>> + 'a { | ||
| const INDENT_SIZE: usize = 2; | ||
|
|
||
| let mut remaining_lines = Some(lines(input)); | ||
|
|
||
| iter::from_fn(move || { | ||
| let lines = remaining_lines.as_mut()?; | ||
|
|
||
| let (first_line, first_line_span) = lines.next()?; | ||
| let first_line = first_line.map_span(move |span| { | ||
| SimpleSpan::new( | ||
| span.start + first_line_span.start, | ||
| span.end + first_line_span.start, | ||
| ) | ||
| }); | ||
|
|
||
| let (indent_len, (unindented_span, unindented)) = | ||
| just::<'_, _, _, extra::Err<EmptyErr>>(" ") | ||
| .repeated() | ||
| .count() | ||
| .map_with(|indent, e| (e.span(), indent)) | ||
| .then( | ||
| any() | ||
| .repeated() | ||
| .to_slice() | ||
| .map_with(|line, e| (e.span(), line)), | ||
| ) | ||
| .parse(first_line) | ||
| .into_result() | ||
| .unwrap(); | ||
| let unindented = unindented.map_span(move |span| { | ||
| SimpleSpan::new( | ||
| span.start + unindented_span.start, | ||
| span.end + unindented_span.start, | ||
| ) | ||
| }); | ||
|
|
||
| if let Ok(comment) = just::<'_, _, _, extra::Err<EmptyErr>>('#') | ||
| .ignore_then(just(' ').or_not()) | ||
| .ignore_then( | ||
| any() | ||
| .repeated() | ||
| .to_slice() | ||
| .map_with(|contents, e| (e.span(), contents)), | ||
| ) | ||
| .map_with(|parsed, e| (e.span(), parsed)) | ||
| .parse(unindented.clone()) | ||
| .into_result() | ||
| { | ||
| let (span, (contents_span, contents)) = comment; | ||
| return Some(Ok(Event::SingleLineComment { | ||
| span, | ||
| contents, | ||
| contents_span, | ||
| })); | ||
| } | ||
|
|
||
| let parsed_indent_level = { | ||
| let (span, indent_len) = indent_len; | ||
| if indent_len % INDENT_SIZE != 0 { | ||
| remaining_lines.take(); | ||
| return Some(Err(TokenizationError::MisalignedIndent { span })); | ||
| } | ||
| indent_len / INDENT_SIZE | ||
| }; | ||
|
|
||
| enum Parsed<'a> { | ||
| SectionHeader { | ||
| span: &'a str, | ||
| unescaped_name: &'a str, | ||
| unescaped_name_span: SimpleSpan, | ||
| }, | ||
| } | ||
|
|
||
|
|
||
| .map_err(|_e| TokenizationError::MalformedItem { | ||
| skipped_span: first_line_span, | ||
| }) | ||
| .and_then(|res| res) | ||
| .map(Some) | ||
| .transpose() | ||
| }) | ||
| // let parser = custom(|input| { | ||
| // let indent_len = input.next().unwrap(); | ||
| // | ||
| // let comment = inline_whitespace() | ||
| // .ignore_then(just('#')) | ||
| // .ignore_then(any()); | ||
| // input.check(parser); | ||
| // | ||
| // choice((comment, section_header)) | ||
| // }) | ||
| // .lazy(); | ||
|
|
||
| // let mut current_indent_level = 0; | ||
| // iter::from_fn(move || { | ||
| // const INDENT_SIZE: usize = 2; | ||
| // | ||
| // let fuse = |current_idx: &mut _, input: &str| { | ||
| // *current_idx = input.len(); | ||
| // }; | ||
| // | ||
| // let mut next_nonempty_line = || loop { | ||
| // let (line, line_span) = lines.as_mut()?.next()?; | ||
| // | ||
| // let unindented = line.trim_start_matches(' '); | ||
| // if unindented.is_empty() { | ||
| // continue; | ||
| // } | ||
| // | ||
| // let indent_len = line.len() - unindented.len(); | ||
| // | ||
| // break Some((line_span, indent_len, unindented)); | ||
| // }; | ||
| // let (parsed_indent_level, unindented) = { | ||
| // let (line_span, indent_len, unindented) = next_nonempty_line()?; | ||
| // | ||
| // if unindented.starts_with('#') { | ||
| // return Some(Ok(Event::SingleLineComment { | ||
| // contents: fully_trimmed, | ||
| // contents_span: todo!(), | ||
| // })); | ||
| // } | ||
| // | ||
| // if indent_len % INDENT_SIZE != 0 { | ||
| // fuse(current_idx, input); | ||
| // return Some(Err(TokenizationError::MisalignedIndent { | ||
| // span: SimpleSpan::new(*current_idx, *current_idx + indent_len), | ||
| // })); | ||
| // } | ||
| // (indent_len / INDENT_SIZE, unindented) | ||
| // }; | ||
| // | ||
| // match parsed_indent_level.cmp(&usize::from(*current_indent_level)) { | ||
| // Ordering::Equal => (), | ||
| // Ordering::Less => { | ||
| // *current_indent_level = *current_indent_level - 1; | ||
| // // Deliberate: don't update `current_idx`. We'll emit another event for the next | ||
| // // level of indentation. | ||
| // return Some(Ok(Event::ExitSection { | ||
| // end_point_span: todo!(), | ||
| // })); | ||
| // } | ||
| // Ordering::Greater => { | ||
| // return Some(Err(TokenizationError::UnexpectedIndent { | ||
| // indent_span: todo!(), | ||
| // })) | ||
| // } | ||
| // }; | ||
| // | ||
| // // TODO: check for a property or section | ||
| // if let Some(rest) = unindented.strip_prefix('[') { | ||
| // // TODO: escaping | ||
| // } else if let Ok(property_ident) = ascii::ident::<_, _, extra::Err<EmptyErr>>() | ||
| // .lazy() | ||
| // .parse(*input) | ||
| // .into_result() | ||
| // { | ||
| // if let Some(property_val_on_line) = input[property_ident.len()..].strip_prefix(':') { | ||
| // let property_val_on_line = property_val_on_line.trim_start(); | ||
| // if property_val_on_line.is_empty() { | ||
| // while let Some((indent_len, unindented)) = next_nonempty_line() { | ||
| // if indent_len / INDENT_SIZE <= *current_indent_level { | ||
| // break; | ||
| // } | ||
| // } | ||
| // | ||
| // // TODO: slurp up indented | ||
| // } | ||
| // } | ||
| // } | ||
| // | ||
| // // TODO: recover by skipping indented items, probably relevant to this malformed | ||
| // // whatever-it-is | ||
| // Some(Err(TokenizationError::MalformedItem { | ||
| // skipped_span: todo!(), | ||
| // })) | ||
| // }) | ||
| } | ||
|
|
||
| fn main() { | ||
| dbg!(tokens( | ||
| r#" | ||
| # hay sup | ||
| [thing] | ||
| [thang] | ||
| stuff: things | ||
| [blarg] "#, | ||
| ) | ||
| .collect::<Vec<_>>()); | ||
|
|
||
| // dbg!(Tokenizer::new( | ||
| // r#" | ||
| // [thing] | ||
| // [thang] | ||
| // | ||
| // stuff: things | ||
| // | ||
| // [blarg] "# | ||
| // ) | ||
| // .collect::<Vec<_>>()); | ||
| } |