Skip to content

Offline, privacy-first grammar checker. Fast, open-source, Rust-powered

fix: issues with structured newlines in JavaDoc lints

Elijah Potter04ea3e85dc72
diff --git a/harper-comments/src/comment_parser.rs b/harper-comments/src/comment_parser.rs
index 4062d5426..b46b38117 100644
--- a/harper-comments/src/comment_parser.rs
+++ b/harper-comments/src/comment_parser.rs
@@ -9,7 +9,7 @@ use tree_sitter::Node;
use crate::comment_parsers;
pub struct CommentParser {
- inner: parsers::Mask<TreeSitterMasker, Box<dyn Parser>>,
+ inner: parsers::Mask<TreeSitterMasker, Box<dyn Parser>>
}
impl CommentParser {
@@ -35,21 +35,21 @@ impl CommentParser {
"lua" => tree_sitter_lua::language(),
"sh" => tree_sitter_bash::language(),
"java" => tree_sitter_java::language(),
- _ => return None,
+ _ => return None
};
let comment_parser: Box<dyn Parser> = match language_id {
"javascriptreact" | "typescript" | "typescriptreact" | "javascript" => Box::new(JsDoc),
"java" => Box::new(JavaDoc::default()),
"go" => Box::new(Go),
- _ => Box::new(Unit),
+ _ => Box::new(Unit)
};
Some(Self {
inner: parsers::Mask::new(
TreeSitterMasker::new(language, Self::node_condition),
- comment_parser,
- ),
+ comment_parser
+ )
})
}
@@ -82,7 +82,7 @@ impl CommentParser {
"sh" => "sh",
"bash" => "sh",
"java" => "java",
- _ => return None,
+ _ => return None
})
}
diff --git a/harper-comments/src/comment_parsers/javadoc.rs b/harper-comments/src/comment_parsers/javadoc.rs
index 2a15521a4..d3e6d42e6 100644
--- a/harper-comments/src/comment_parsers/javadoc.rs
+++ b/harper-comments/src/comment_parsers/javadoc.rs
@@ -1,4 +1,7 @@
-use harper_core::{parsers::Parser, Token};
+use std::collections::VecDeque;
+
+use harper_core::parsers::Parser;
+use harper_core::{Punctuation, Token, TokenKind, VecExt};
use harper_html::HtmlParser;
use super::without_initiators;
@@ -15,6 +18,41 @@ impl Parser for JavaDoc {
let mut tokens = self.html_parser.parse(actual_source);
+ // We need to remove leading spaces and stars from the block of tokens.
+ let mut remove_these: VecDeque<usize> = VecDeque::new();
+
+ let mut cursor = 0;
+
+ while cursor < tokens.len() {
+ let maybe_newline = tokens[cursor];
+
+ if let TokenKind::Newline(_) = maybe_newline.kind {
+ cursor += 1;
+
+ loop {
+ if cursor >= tokens.len() {
+ break;
+ }
+
+ let maybe_removable = tokens[cursor];
+
+ if matches!(
+ maybe_removable.kind,
+ TokenKind::Punctuation(Punctuation::Star) | TokenKind::Space(_)
+ ) {
+ remove_these.push_back(cursor);
+ cursor += 1;
+ } else {
+ break;
+ }
+ }
+ } else {
+ cursor += 1;
+ }
+ }
+
+ tokens.remove_indices(remove_these);
+
for token in tokens.iter_mut() {
token.span.push_by(actual.start);
}
diff --git a/harper-comments/src/comment_parsers/jsdoc.rs b/harper-comments/src/comment_parsers/jsdoc.rs
index 3e9b42135..0948008aa 100644
--- a/harper-comments/src/comment_parsers/jsdoc.rs
+++ b/harper-comments/src/comment_parsers/jsdoc.rs
@@ -17,7 +17,7 @@ impl Parser for JsDoc {
new_tokens.push(Token::new(
Span::new_with_len(line.len(), 1),
- harper_core::TokenKind::Newline(1),
+ harper_core::TokenKind::Newline(1)
));
new_tokens
diff --git a/harper-comments/src/comment_parsers/unit.rs b/harper-comments/src/comment_parsers/unit.rs
index 07eb144f6..2cd505401 100644
--- a/harper-comments/src/comment_parsers/unit.rs
+++ b/harper-comments/src/comment_parsers/unit.rs
@@ -22,7 +22,7 @@ impl Parser for Unit {
new_tokens.push(Token::new(
Span::new_with_len(line.len(), 1),
- harper_core::TokenKind::Newline(1),
+ harper_core::TokenKind::Newline(1)
));
new_tokens
diff --git a/harper-core/src/document.rs b/harper-core/src/document.rs
index 1865c48b2..892854220 100644
--- a/harper-core/src/document.rs
+++ b/harper-core/src/document.rs
@@ -10,6 +10,7 @@ use crate::parsers::{Markdown, Parser, PlainEnglish};
use crate::punctuation::Punctuation;
use crate::span::Span;
use crate::token::NumberSuffix;
+use crate::vec_ext::VecExt;
use crate::{FatToken, Lrc, Token, TokenKind, TokenStringExt};
pub struct Document {
@@ -376,7 +377,7 @@ impl Document {
cursor += 1;
}
- remove_indices(&mut self.tokens, remove_these);
+ self.tokens.remove_indices(remove_these);
}
/// Searches for multiple sequential newline tokens and condenses them down
@@ -414,7 +415,7 @@ impl Document {
cursor += 1;
}
- remove_indices(&mut self.tokens, remove_these);
+ self.tokens.remove_indices(remove_these);
}
/// Searches for contractions and condenses them down into single
@@ -531,35 +532,9 @@ fn is_sentence_terminator(token: &TokenKind) -> bool {
}
}
-/// Removes a list of indices from a Vector.
-/// Assumes that the provided indices are already in sorted order.
-fn remove_indices<T>(vec: &mut Vec<T>, mut to_remove: VecDeque<usize>) {
- let mut i = 0;
-
- let mut next_remove = to_remove.pop_front();
-
- vec.retain(|_| {
- let keep = if let Some(next_remove) = next_remove {
- i != next_remove
- } else {
- true
- };
-
- if !keep {
- next_remove = to_remove.pop_front();
- }
-
- i += 1;
- keep
- });
-}
-
#[cfg(test)]
mod tests {
- use std::collections::VecDeque;
-
use super::Document;
- use crate::document::remove_indices;
use crate::parsers::{Markdown, PlainEnglish};
use crate::token::TokenStringExt;
use crate::{Span, Token, TokenKind};
@@ -651,14 +626,4 @@ mod tests {
18,
);
}
-
- #[test]
- fn removes_requested_indices() {
- let mut data: Vec<i32> = (0..10).collect();
- let remove: VecDeque<usize> = vec![1, 4, 6].into_iter().collect();
-
- remove_indices(&mut data, remove);
-
- assert_eq!(data, vec![0, 2, 3, 5, 7, 8, 9])
- }
}
diff --git a/harper-core/src/lib.rs b/harper-core/src/lib.rs
index c6f18df42..d69e87d8e 100644
--- a/harper-core/src/lib.rs
+++ b/harper-core/src/lib.rs
@@ -14,6 +14,7 @@ mod span;
mod spell;
mod sync;
mod token;
+mod vec_ext;
pub use char_string::{CharString, CharStringExt};
pub use document::Document;
@@ -24,6 +25,7 @@ pub use span::Span;
pub use spell::{Dictionary, FullDictionary, MergedDictionary};
pub use sync::Lrc;
pub use token::{FatToken, Token, TokenKind, TokenStringExt};
+pub use vec_ext::VecExt;
/// A utility function that removes overlapping lints in a vector,
/// keeping the more important ones.
diff --git a/harper-core/src/parsers/mask.rs b/harper-core/src/parsers/mask.rs
index 2bd21aae6..fd1b21a2d 100644
--- a/harper-core/src/parsers/mask.rs
+++ b/harper-core/src/parsers/mask.rs
@@ -6,16 +6,16 @@ use crate::{Span, Token, TokenKind};
pub struct Mask<M, P>
where
M: Masker,
- P: Parser,
+ P: Parser
{
pub masker: M,
- pub parser: P,
+ pub parser: P
}
impl<M, P> Mask<M, P>
where
M: Masker,
- P: Parser,
+ P: Parser
{
pub fn new(masker: M, parser: P) -> Self {
Self { masker, parser }
@@ -25,7 +25,7 @@ where
impl<M, P> Parser for Mask<M, P>
where
M: Masker,
- P: Parser,
+ P: Parser
{
fn parse(&mut self, source: &[char]) -> Vec<Token> {
let mask = self.masker.create_mask(source);
diff --git a/harper-core/src/token.rs b/harper-core/src/token.rs
index 13b62ca69..a4f84db6a 100644
--- a/harper-core/src/token.rs
+++ b/harper-core/src/token.rs
@@ -10,7 +10,7 @@ use crate::Quote;
#[derive(Debug, Clone, Copy, Serialize, Deserialize, PartialEq, Default)]
pub struct Token {
pub span: Span,
- pub kind: TokenKind,
+ pub kind: TokenKind
}
impl Token {
@@ -24,7 +24,7 @@ impl Token {
FatToken {
content,
- kind: self.kind,
+ kind: self.kind
}
}
}
@@ -34,7 +34,7 @@ impl Token {
#[derive(Debug, Clone, Serialize, Deserialize, PartialEq, PartialOrd)]
pub struct FatToken {
pub content: Vec<char>,
- pub kind: TokenKind,
+ pub kind: TokenKind
}
#[derive(Debug, Is, Clone, Copy, Serialize, Deserialize, PartialEq, Default, PartialOrd)]
@@ -54,7 +54,7 @@ pub enum TokenKind {
/// A special token used for things like inline code blocks that should be
/// ignored by all linters.
Unlintable,
- ParagraphBreak,
+ ParagraphBreak
}
#[derive(Debug, Serialize, Deserialize, Default, PartialEq, PartialOrd, Clone, Copy, Is)]
@@ -63,7 +63,7 @@ pub enum NumberSuffix {
Th,
St,
Nd,
- Rd,
+ Rd
}
impl NumberSuffix {
@@ -89,7 +89,7 @@ impl NumberSuffix {
7 => Some(Self::Th),
8 => Some(Self::Th),
9 => Some(Self::Th),
- _ => None,
+ _ => None
}
}
@@ -98,7 +98,7 @@ impl NumberSuffix {
NumberSuffix::Th => vec!['t', 'h'],
NumberSuffix::St => vec!['s', 't'],
NumberSuffix::Nd => vec!['n', 'd'],
- NumberSuffix::Rd => vec!['r', 'd'],
+ NumberSuffix::Rd => vec!['r', 'd']
}
}
@@ -126,7 +126,7 @@ impl NumberSuffix {
('R', 'd') => Some(NumberSuffix::Rd),
('r', 'D') => Some(NumberSuffix::Rd),
('R', 'D') => Some(NumberSuffix::Rd),
- _ => None,
+ _ => None
}
}
}
diff --git a/harper-core/src/vec_ext.rs b/harper-core/src/vec_ext.rs
new file mode 100644
index 000000000..930913a4f
--- /dev/null
+++ b/harper-core/src/vec_ext.rs
@@ -0,0 +1,47 @@
+use std::collections::VecDeque;
+
+pub trait VecExt {
+ /// Removes a list of indices from a Vector.
+ /// Assumes that the provided indices are already in sorted order.
+ fn remove_indices(&mut self, to_remove: VecDeque<usize>);
+}
+
+impl<T> VecExt for Vec<T> {
+ fn remove_indices(&mut self, mut to_remove: VecDeque<usize>) {
+ let mut i = 0;
+
+ let mut next_remove = to_remove.pop_front();
+
+ self.retain(|_| {
+ let keep = if let Some(next_remove) = next_remove {
+ i != next_remove
+ } else {
+ true
+ };
+
+ if !keep {
+ next_remove = to_remove.pop_front();
+ }
+
+ i += 1;
+ keep
+ });
+ }
+}
+
+#[cfg(test)]
+mod tests {
+ use std::collections::VecDeque;
+
+ use crate::vec_ext::VecExt;
+
+ #[test]
+ fn removes_requested_indices() {
+ let mut data: Vec<i32> = (0..10).collect();
+ let remove: VecDeque<usize> = vec![1, 4, 6].into_iter().collect();
+
+ data.remove_indices(remove);
+
+ assert_eq!(data, vec![0, 2, 3, 5, 7, 8, 9])
+ }
+}
diff --git a/harper-html/src/lib.rs b/harper-html/src/lib.rs
index b4aefcf94..33dcfb90a 100644
--- a/harper-html/src/lib.rs
+++ b/harper-html/src/lib.rs
@@ -4,7 +4,7 @@ use harper_tree_sitter::TreeSitterMasker;
use tree_sitter::Node;
pub struct HtmlParser {
- inner: parsers::Mask<TreeSitterMasker, PlainEnglish>,
+ inner: parsers::Mask<TreeSitterMasker, PlainEnglish>
}
impl HtmlParser {
@@ -18,8 +18,8 @@ impl Default for HtmlParser {
Self {
inner: parsers::Mask::new(
TreeSitterMasker::new(tree_sitter_html::language(), Self::node_condition),
- PlainEnglish,
- ),
+ PlainEnglish
+ )
}
}
}
diff --git a/harper-ls/src/backend.rs b/harper-ls/src/backend.rs
index 507ac9656..068709137 100644
--- a/harper-ls/src/backend.rs
+++ b/harper-ls/src/backend.rs
@@ -5,7 +5,14 @@ use std::sync::Arc;
use harper_comments::CommentParser;
use harper_core::parsers::{Markdown, PlainEnglish};
use harper_core::{
- Dictionary, Document, FullDictionary, LintGroup, Linter, MergedDictionary, Token, TokenKind,
+ Dictionary,
+ Document,
+ FullDictionary,
+ LintGroup,
+ Linter,
+ MergedDictionary,
+ Token,
+ TokenKind
};
use harper_html::HtmlParser;
use serde_json::Value;
@@ -13,13 +20,31 @@ use tokio::sync::{Mutex, RwLock};
use tower_lsp::jsonrpc::Result;
use tower_lsp::lsp_types::notification::PublishDiagnostics;
use tower_lsp::lsp_types::{
- CodeActionOrCommand, CodeActionParams, CodeActionProviderCapability, CodeActionResponse,
- Command, Diagnostic, DidChangeConfigurationParams, DidChangeTextDocumentParams,
- DidCloseTextDocumentParams, DidOpenTextDocumentParams, DidSaveTextDocumentParams,
- ExecuteCommandOptions, ExecuteCommandParams, InitializeParams, InitializeResult,
- InitializedParams, MessageType, PublishDiagnosticsParams, Range, ServerCapabilities,
- TextDocumentSyncCapability, TextDocumentSyncKind, TextDocumentSyncOptions,
- TextDocumentSyncSaveOptions, Url,
+ CodeActionOrCommand,
+ CodeActionParams,
+ CodeActionProviderCapability,
+ CodeActionResponse,
+ Command,
+ Diagnostic,
+ DidChangeConfigurationParams,
+ DidChangeTextDocumentParams,
+ DidCloseTextDocumentParams,
+ DidOpenTextDocumentParams,
+ DidSaveTextDocumentParams,
+ ExecuteCommandOptions,
+ ExecuteCommandParams,
+ InitializeParams,
+ InitializeResult,
+ InitializedParams,
+ MessageType,
+ PublishDiagnosticsParams,
+ Range,
+ ServerCapabilities,
+ TextDocumentSyncCapability,
+ TextDocumentSyncKind,
+ TextDocumentSyncOptions,
+ TextDocumentSyncSaveOptions,
+ Url
};
use tower_lsp::{Client, LanguageServer};
use tracing::{error, info, instrument};
@@ -35,7 +60,7 @@ struct DocumentState {
document: Document,
ident_dict: Arc<FullDictionary>,
linter: LintGroup<MergedDictionary<FullDictionary>>,
- language_id: Option<String>,
+ language_id: Option<String>
}
/// Deallocate
@@ -43,7 +68,7 @@ pub struct Backend {
client: Client,
static_dictionary: Arc<FullDictionary>,
config: RwLock<Config>,
- doc_state: Mutex<HashMap<Url, DocumentState>>,
+ doc_state: Mutex<HashMap<Url, DocumentState>>
}
impl Backend {
@@ -54,7 +79,7 @@ impl Backend {
client,
static_dictionary: dictionary.into(),
doc_state: Mutex::new(HashMap::new()),
- config: RwLock::new(config),
+ config: RwLock::new(config)
}
}
@@ -149,7 +174,7 @@ impl Backend {
#[instrument(skip(self))]
async fn generate_file_dictionary(
&self,
- url: &Url,
+ url: &Url
) -> anyhow::Result<MergedDictionary<FullDictionary>> {
let (global_dictionary, file_dictionary) = tokio::join!(
self.generate_global_dictionary(),
@@ -166,11 +191,11 @@ impl Backend {
async fn update_document_from_file(
&self,
url: &Url,
- language_id: Option<&str>,
+ language_id: Option<&str>
) -> anyhow::Result<()> {
let content = match tokio::fs::read_to_string(
url.to_file_path()
- .map_err(|_| anyhow::format_err!("Could not extract file path."))?,
+ .map_err(|_| anyhow::format_err!("Could not extract file path."))?
)
.await
{
@@ -189,7 +214,7 @@ impl Backend {
&self,
url: &Url,
text: &str,
- language_id: Option<&str>,
+ language_id: Option<&str>
) -> anyhow::Result<()> {
let mut doc_lock = self.doc_state.lock().await;
let config_lock = self.config.read().await;
@@ -201,7 +226,7 @@ impl Backend {
let mut doc_state = DocumentState {
linter: LintGroup::new(
config_lock.lint_config,
- self.generate_file_dictionary(url).await?,
+ self.generate_file_dictionary(url).await?
),
language_id: language_id
.map(|v| v.to_string())
@@ -254,7 +279,7 @@ impl Backend {
async fn generate_code_actions(
&self,
url: &Url,
- range: Range,
+ range: Range
) -> Result<Vec<CodeActionOrCommand>> {
let (config, mut doc_states) = tokio::join!(self.config.read(), self.doc_state.lock());
let Some(doc_state) = doc_states.get_mut(url) else {
@@ -286,7 +311,7 @@ impl Backend {
actions.push(CodeActionOrCommand::Command(Command::new(
"Open URL".to_string(),
"HarperOpen".to_string(),
- Some(vec![doc_state.document.get_span_content_str(span).into()]),
+ Some(vec![doc_state.document.get_span_content_str(span).into()])
)))
}
@@ -306,7 +331,7 @@ impl Backend {
lints_to_diagnostics(
doc_state.document.get_full_content(),
&lints,
- config.diagnostic_severity,
+ config.diagnostic_severity
)
}
@@ -317,7 +342,7 @@ impl Backend {
let result = PublishDiagnosticsParams {
uri: url.clone(),
diagnostics,
- version: None,
+ version: None
};
self.client
@@ -347,11 +372,11 @@ impl LanguageServer for Backend {
change: Some(TextDocumentSyncKind::FULL),
will_save: None,
will_save_wait_until: None,
- save: Some(TextDocumentSyncSaveOptions::Supported(true)),
- },
+ save: Some(TextDocumentSyncSaveOptions::Supported(true))
+ }
)),
..Default::default()
- },
+ }
})
}
@@ -382,7 +407,7 @@ impl LanguageServer for Backend {
.update_document(
&params.text_document.uri,
&params.text_document.text,
- Some(&params.text_document.language_id),
+ Some(&params.text_document.language_id)
)
.await;
@@ -471,7 +496,7 @@ impl LanguageServer for Backend {
error!("Unable to open URL: {}", err);
}
},
- _ => (),
+ _ => ()
}
Ok(None)
@@ -502,7 +527,7 @@ impl LanguageServer for Backend {
// Update documents with new config
futures::future::join_all(
keys.iter()
- .map(|key| self.update_document_from_file(key, None)),
+ .map(|key| self.update_document_from_file(key, None))
)
.await;
}