Offline, privacy-first grammar checker. Fast, open-source, Rust-powered
fix: issues with C++ and TypeScript multiline comments
Elijah Potter9bde7339a773
diff --git a/Cargo.lock b/Cargo.lockindex 2c0ff9275..105af732e 100644--- a/Cargo.lock+++ b/Cargo.lock@@ -614,6 +614,7 @@ name = "harper-tree-sitter"version = "0.8.0"dependencies = ["harper-core",+ "paste","tree-sitter","tree-sitter-c","tree-sitter-c-sharp",@@ -1186,9 +1187,9 @@ checksum = "94143f37725109f92c262ed2cf5e59bce7498c01bcc1502d7b9afe439a4e9f49"[[package]]name = "serde"-version = "1.0.202"+version = "1.0.203"source = "registry+https://github.com/rust-lang/crates.io-index"-checksum = "226b61a0d411b2ba5ff6d7f73a476ac4f8bb900373459cd00fab8512828ba395"+checksum = "7253ab4de971e72fb7be983802300c30b5a7f0c2e56fab8abfc6a214307c0094"dependencies = ["serde_derive",]@@ -1206,9 +1207,9 @@ dependencies = [[[package]]name = "serde_derive"-version = "1.0.202"+version = "1.0.203"source = "registry+https://github.com/rust-lang/crates.io-index"-checksum = "6048858004bcff69094cd972ed40a32500f153bd3be9f716b2eed2e8217c4838"+checksum = "500cbc0ebeb6f46627f50f3f5811ccf6bf00643be300b4c3eabc0ef55dc5b5ba"dependencies = ["proc-macro2","quote",diff --git a/harper-cli/src/main.rs b/harper-cli/src/main.rsindex ddf58b99d..6fd172fc2 100644--- a/harper-cli/src/main.rs+++ b/harper-cli/src/main.rs@@ -7,90 +7,87 @@ use harper_core::{remove_overlaps, Document, FullDictionary, LintGroup, LintGrouuse harper_tree_sitter::TreeSitterParser;#[derive(Debug, Parser)]-struct Args {- /// The file you wish to grammar check.- file: PathBuf,- /// Whether to merely print out the number of errors encountered, without- /// further details.- #[arg(short, long)]- count: bool+enum Args {+ /// Lint a provided document.+ Lint {+ /// The file you wish to grammar check.+ file: PathBuf,+ /// Whether to merely print out the number of errors encountered,+ /// without further details.+ #[arg(short, long)]+ count: bool+ },+ /// Parse a provided document and print the detected symbols.+ Parse {+ /// The file you wish to parse.+ file: PathBuf+ }}fn main() -> anyhow::Result<()> {let args = Args::parse();- let source = std::fs::read_to_string(&args.file)?;+ match args {+ Args::Lint { file, count } => {+ let (doc, source) = load_file(&file)?;- let parser = TreeSitterParser::new_from_language_id(- filename_to_filetype(&args.file).ok_or(format_err!("Could not detect filetype."))?- )- .ok_or(format_err!("Could not detect language ID."))?;+ let mut linter = LintGroup::new(+ LintGroupConfig::default(),+ FullDictionary::create_from_curated()+ );+ let mut lints = linter.lint(&doc);- let doc = Document::new(&source, Box::new(parser));+ if count {+ println!("{}", lints.len());+ return Ok(());+ }- let mut linter = LintGroup::new(- LintGroupConfig::default(),- FullDictionary::create_from_curated()- );- let mut lints = linter.lint(&doc);+ if lints.is_empty() {+ println!("No lints found");+ return Ok(());+ }- if args.count {- println!("{}", lints.len());- return Ok(());- }+ remove_overlaps(&mut lints);- if lints.is_empty() {- println!("No lints found");- return Ok(());- }+ let primary_color = Color::Magenta;- remove_overlaps(&mut lints);+ let filename = file+ .file_name()+ .map(|s| s.to_string_lossy().into())+ .unwrap_or("<file>".to_string());- let primary_color = Color::Magenta;+ let mut report_builder = Report::build(ReportKind::Advice, &filename, 0);- let filename = args- .file- .file_name()- .map(|s| s.to_string_lossy().into())- .unwrap_or("<file>".to_string());+ for lint in lints {+ report_builder = report_builder.with_label(+ Label::new((&filename, lint.span.into()))+ .with_message(lint.message)+ .with_color(primary_color)+ );+ }- let mut report_builder = Report::build(ReportKind::Advice, &filename, 0);+ let report = report_builder.finish();+ report.print((&filename, Source::from(source)))?;- for lint in lints {- report_builder = report_builder.with_label(- Label::new((&filename, lint.span.into()))- .with_message(lint.message)- .with_color(primary_color)- );- }+ std::process::exit(1);+ }+ Args::Parse { file } => {+ let (doc, _) = load_file(&file)?;- let report = report_builder.finish();- report.print((&filename, Source::from(source)))?;+ for token in doc.tokens() {+ dbg!(token);+ }- std::process::exit(1);+ Ok(())+ }+ }}-/// Convert a provided path to a corresponding Language Server Protocol file-/// type.-///-/// Note to contributors: try to keep this in sync with-/// [`TreeSitterParser::new_from_language_id`]-fn filename_to_filetype(path: &Path) -> Option<&'static str> {- Some(match path.extension()?.to_str()? {- "rs" => "rust",- "ts" => "typescript",- "tsx" => "typescriptreact",- "js" => "javascript",- "jsx" => "javascriptreact",- "go" => "go",- "c" => "c",- "cpp" => "cpp",- "h" => "cpp",- "rb" => "ruby",- "swift" => "swift",- "cs" => "csharp",- "toml" => "toml",- "lua" => "lua",- _ => return None- })+fn load_file(file: &Path) -> anyhow::Result<(Document, String)> {+ let source = std::fs::read_to_string(file)?;++ let parser = TreeSitterParser::new_from_filename(file)+ .ok_or(format_err!("Could not detect language ID."))?;++ Ok((Document::new(&source, Box::new(parser)), source))}diff --git a/harper-tree-sitter/Cargo.toml b/harper-tree-sitter/Cargo.tomlindex e8fb59583..3406eb547 100644--- a/harper-tree-sitter/Cargo.toml+++ b/harper-tree-sitter/Cargo.toml@@ -18,3 +18,7 @@ tree-sitter-swift = "=0.4.0"tree-sitter-c-sharp = "0.20.0"tree-sitter-toml = "0.20.0"tree-sitter-lua = "0.0.19"++[dev-dependencies]+paste = "1.0.15"+diff --git a/harper-tree-sitter/src/tree_sitter_parser.rs b/harper-tree-sitter/src/tree_sitter_parser.rsindex 47cfb39b8..d8115b14d 100644--- a/harper-tree-sitter/src/tree_sitter_parser.rs+++ b/harper-tree-sitter/src/tree_sitter_parser.rs@@ -1,12 +1,13 @@use std::collections::HashSet;+use std::path::Path;use harper_core::parsers::Parser;-use harper_core::{FullDictionary, Span, Token};+use harper_core::{FullDictionary, Span, Token, TokenKind};use tree_sitter::{Language, Node, Tree, TreeCursor};use super::comment_parsers::{Go, Unit};-/// A Harper parser that wraps various [`super::comment_parsers`] that+/// A Harper parser that wraps various comment parsers that/// exclusively parses comments in any language supported by [`tree_sitter`].pub struct TreeSitterParser {language: Language,@@ -44,6 +45,36 @@ impl TreeSitterParser {})}+ /// Infer the programming language from a provided filename.+ pub fn new_from_filename(filename: &Path) -> Option<Self> {+ Self::new_from_language_id(Self::filename_to_filetype(filename)?)+ }++ /// Convert a provided path to a corresponding Language Server Protocol file+ /// type.+ ///+ /// Note to contributors: try to keep this in sync with+ /// [`Self::new_from_language_id`]+ fn filename_to_filetype(path: &Path) -> Option<&'static str> {+ Some(match path.extension()?.to_str()? {+ "rs" => "rust",+ "ts" => "typescript",+ "tsx" => "typescriptreact",+ "js" => "javascript",+ "jsx" => "javascriptreact",+ "go" => "go",+ "c" => "c",+ "cpp" => "cpp",+ "h" => "cpp",+ "rb" => "ruby",+ "swift" => "swift",+ "cs" => "csharp",+ "toml" => "toml",+ "lua" => "lua",+ _ => return None+ })+ }+fn parse_root(&self, text: &str) -> Option<Tree> {let mut parser = tree_sitter::Parser::new();parser.set_language(self.language).unwrap();@@ -127,6 +158,11 @@ impl Parser for TreeSitterParser {byte_spans_to_char_spans(&mut comments_spans, &text);+ dbg!(comments_spans+ .iter()+ .map(|span| span.get_content_string(text.chars().collect::<Vec<_>>().as_slice()))+ .collect::<Vec<_>>());+let mut tokens = Vec::new();for (s_index, span) in comments_spans.iter().enumerate() {@@ -144,6 +180,13 @@ impl Parser for TreeSitterParser {}}+ // Same goes for newlines within the comment+ for t in new_tokens.iter_mut() {+ if let TokenKind::Newline(n) = &mut t.kind {+ *n = 1;+ }+ }+tokens.append(&mut new_tokens);}diff --git a/harper-tree-sitter/tests/language_support.rs b/harper-tree-sitter/tests/language_support.rsnew file mode 100644index 000000000..d730f1e75--- /dev/null+++ b/harper-tree-sitter/tests/language_support.rs@@ -0,0 +1,38 @@+use std::path::Path;++use harper_core::{Document, FullDictionary, LintGroup, LintGroupConfig, Linter};+use harper_tree_sitter::TreeSitterParser;++/// Creates a unit test checking that the linting of a source file in+/// `language_support_sources` produces the expected number of lints.+macro_rules! create_test {+ ($filename:ident.$ext:ident, $correct_expected:expr) => {+ paste::paste! {+ #[test]+ fn [<lints_$ext _ $filename _correctly>](){+ let filename = concat!(stringify!($filename), ".", stringify!($ext));+ let source = include_str!(+ concat!(+ "./language_support_sources/",+ concat!(+ stringify!($filename), ".", stringify!($ext))+ )+ );++ let parser = TreeSitterParser::new_from_filename(Path::new(filename)).unwrap();+ let document = Document::new(&source, Box::new(parser));++ let mut linter = LintGroup::new(+ LintGroupConfig::default(),+ FullDictionary::create_from_curated()+ );+ let lints = linter.lint(&document);++ assert_eq!(lints.len(), $correct_expected);+ }+ }+ };+}++create_test!(multiline_comments.cpp, 3);+create_test!(multiline_comments.ts, 3);diff --git a/harper-tree-sitter/tests/language_support_sources/multiline_comments.cpp b/harper-tree-sitter/tests/language_support_sources/multiline_comments.cppnew file mode 100644index 000000000..5eeeffcb0--- /dev/null+++ b/harper-tree-sitter/tests/language_support_sources/multiline_comments.cpp@@ -0,0 +1,12 @@+/// This is an example of an problematic comment.+/// It should produce one error.+int test() {}++/***+ * This is an example of a possible error:+ * these subsequent lines should not be considered a new sentence and should+ * produce no errors.+ */+int arbitrary() {}++/// Let's aadd a cuple spelling errors for good measure.diff --git a/harper-tree-sitter/tests/language_support_sources/multiline_comments.ts b/harper-tree-sitter/tests/language_support_sources/multiline_comments.tsnew file mode 100644index 000000000..51d87c35e--- /dev/null+++ b/harper-tree-sitter/tests/language_support_sources/multiline_comments.ts@@ -0,0 +1,12 @@+/// This is an example of an problematic comment.+/// It should produce one error.+function test() {}++/***+ * This is an example of a possible error:+ * these subsequent lines should not be considered a new sentence and should+ * produce no errors.+ */+function arbitrary() {}++/// Let's aadd a cuple spelling errors for good measure.