Skip to content

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

refactor(core)!: `Lrc<[char]>` instead of `Lrc<Vec<char>>` (#3060)

Steve Kurch2b5c8cd352a1
diff --git a/harper-core/src/document.rs b/harper-core/src/document.rs
index 34adf59e5..f9bc348fe 100644
--- a/harper-core/src/document.rs
+++ b/harper-core/src/document.rs
@@ -17,7 +17,7 @@ use crate::{OrdinalSuffix, Span};
/// A document containing some amount of lexed and parsed English text.
#[derive(Debug, Clone)]
pub struct Document {
- source: Lrc<Vec<char>>,
+ source: Lrc<[char]>,
tokens: Vec<Token>,
}
@@ -53,23 +53,23 @@ impl Document {
/// Lexes and parses text to produce a document using a provided language
/// parser and dictionary.
pub fn new(text: &str, parser: &impl Parser, dictionary: &impl Dictionary) -> Self {
- let source: Vec<_> = text.chars().collect();
+ let source: Lrc<_> = text.chars().collect();
- Self::new_from_vec(Lrc::new(source), parser, dictionary)
+ Self::new_from_chars(source, parser, dictionary)
}
/// Lexes and parses text to produce a document using a provided language
/// parser and the included curated dictionary.
pub fn new_curated(text: &str, parser: &impl Parser) -> Self {
- let source: Vec<_> = text.chars().collect();
+ let source: Lrc<_> = text.chars().collect();
- Self::new_from_vec(Lrc::new(source), parser, &FstDictionary::curated())
+ Self::new_from_chars(source, parser, &FstDictionary::curated())
}
/// Lexes and parses text to produce a document using a provided language
/// parser and dictionary.
- pub fn new_from_vec(
- source: Lrc<Vec<char>>,
+ pub fn new_from_chars(
+ source: Lrc<[char]>,
parser: &impl Parser,
dictionary: &impl Dictionary,
) -> Self {
@@ -84,11 +84,7 @@ impl Document {
/// Create a new document from character data using the built-in [`PlainEnglish`]
/// parser and curated dictionary. This avoids string-to-char conversions.
pub fn new_plain_english_curated_chars(source: &[char]) -> Self {
- Self::new_from_vec(
- Lrc::new(source.to_vec()),
- &PlainEnglish,
- &FstDictionary::curated(),
- )
+ Self::new_from_chars(Lrc::from(source), &PlainEnglish, &FstDictionary::curated())
}
/// Parse text to produce a document using the built-in [`PlainEnglish`]
@@ -103,7 +99,7 @@ impl Document {
/// This avoids running potentially expensive metadata generation code, so this is more
/// efficient if you don't need that information.
pub(crate) fn new_basic_tokenize(text: &str, parser: &impl Parser) -> Self {
- let source = Lrc::new(text.chars().collect_vec());
+ let source: Lrc<_> = text.chars().collect();
let tokens = parser.parse(&source);
let mut document = Self { source, tokens };
document.apply_fixups();
@@ -129,7 +125,7 @@ impl Document {
/// Create a new document from character data using the built-in [`Markdown`] parser
/// and curated dictionary. This avoids string-to-char conversions.
pub fn new_markdown_default_curated_chars(chars: &[char]) -> Self {
- Self::new_from_vec(
+ Self::new_from_chars(
chars.to_vec().into(),
&Markdown::default(),
&FstDictionary::curated(),
diff --git a/harper-core/src/title_case.rs b/harper-core/src/title_case.rs
index 925800803..0db00b039 100644
--- a/harper-core/src/title_case.rs
+++ b/harper-core/src/title_case.rs
@@ -12,20 +12,20 @@ use crate::{CharStringExt, Document, TokenStringExt, parsers::Parser};
/// A helper function for [`make_title_case`] that uses Strings instead of char buffers.
pub fn make_title_case_str(source: &str, parser: &impl Parser, dict: &impl Dictionary) -> String {
- let source: Vec<char> = source.chars().collect();
+ let source: Lrc<_> = source.chars().collect();
- make_title_case_chars(Lrc::new(source), parser, dict).to_string()
+ make_title_case_chars(source, parser, dict).to_string()
}
// Make a given string [title case](https://en.wikipedia.org/wiki/Title_case) following the Chicago Manual of Style.
pub fn make_title_case_chars(
- source: Lrc<Vec<char>>,
+ source: Lrc<[char]>,
parser: &impl Parser,
dict: &impl Dictionary,
) -> Vec<char> {
- let document = Document::new_from_vec(source.clone(), parser, dict);
+ let document = Document::new_from_chars(source.clone(), parser, dict);
- make_title_case(document.get_tokens(), source.as_slice(), dict)
+ make_title_case(document.get_tokens(), &source, dict)
}
pub fn try_make_title_case(
diff --git a/harper-core/src/weir/mod.rs b/harper-core/src/weir/mod.rs
index fb4caf682..a3e0a6a7a 100644
--- a/harper-core/src/weir/mod.rs
+++ b/harper-core/src/weir/mod.rs
@@ -171,7 +171,7 @@ impl WeirLinter {
continue;
}
- let doc = Document::new_from_vec(
+ let doc = Document::new_from_chars(
current.chars().collect::<Vec<_>>().into(),
&Markdown::default(),
&FstDictionary::curated(),
@@ -197,7 +197,7 @@ impl WeirLinter {
let mut iter_count = 0;
loop {
- let test = Document::new_from_vec(
+ let test = Document::new_from_chars(
text_chars.clone().into(),
&Markdown::default(),
&FstDictionary::curated(),
@@ -224,7 +224,7 @@ impl WeirLinter {
}
fn lint_count(text: &str, linter: &mut impl Linter) -> usize {
- let document = Document::new_from_vec(
+ let document = Document::new_from_chars(
text.chars().collect::<Vec<_>>().into(),
&Markdown::default(),
&FstDictionary::curated(),
diff --git a/harper-wasm/src/lib.rs b/harper-wasm/src/lib.rs
index fb4b3bb6a..6e5350f87 100644
--- a/harper-wasm/src/lib.rs
+++ b/harper-wasm/src/lib.rs
@@ -247,13 +247,10 @@ impl Linter {
}
pub fn ignore_lint(&mut self, source_text: String, lint: Lint) {
- let source: Vec<_> = source_text.chars().collect();
+ let source: Lrc<_> = source_text.chars().collect();
- let document = Document::new_from_vec(
- source.into(),
- &lint.language.create_parser(),
- &self.dictionary,
- );
+ let document =
+ Document::new_from_chars(source, &lint.language.create_parser(), &self.dictionary);
self.ignored_lints.ignore_lint(&lint.inner, &document);
}
@@ -267,7 +264,7 @@ impl Linter {
pub fn context_hash(&self, source_text: String, lint: &Lint) -> u64 {
let source: Vec<_> = source_text.chars().collect();
- let document = Document::new_from_vec(
+ let document = Document::new_from_chars(
source.into(),
&lint.language.create_parser(),
&self.dictionary,
@@ -284,8 +281,7 @@ impl Linter {
all_headings: bool,
regex_mask: Option<String>,
) -> Vec<OrganizedGroup> {
- let source: Vec<_> = text.chars().collect();
- let source = Lrc::new(source);
+ let source: Lrc<_> = text.chars().collect();
let mut parser = language.create_parser();
@@ -302,7 +298,7 @@ impl Linter {
parser = Box::new(OopsAllHeadings::new(parser));
}
- let document = Document::new_from_vec(source.clone(), &parser, &self.dictionary);
+ let document = Document::new_from_chars(source.clone(), &parser, &self.dictionary);
let temp = self.lint_group.config.clone();
self.lint_group.config.fill_with_curated();
@@ -325,7 +321,7 @@ impl Linter {
.into_iter()
.map(|l| {
let problem_text = l.get_str(&source);
- let span = Into::<Span>::into(l.span).to_js_indices(source.as_slice());
+ let span = Into::<Span>::into(l.span).to_js_indices(&source);
Lint::new(l, span, problem_text, language)
})
@@ -344,8 +340,7 @@ impl Linter {
all_headings: bool,
regex_mask: Option<String>,
) -> Vec<Lint> {
- let source: Vec<_> = text.chars().collect();
- let source = Lrc::new(source);
+ let source: Lrc<_> = text.chars().collect();
let mut parser = language.create_parser();
@@ -362,7 +357,7 @@ impl Linter {
parser = Box::new(OopsAllHeadings::new(parser));
}
- let document = Document::new_from_vec(source.clone(), &parser, &self.dictionary);
+ let document = Document::new_from_chars(source.clone(), &parser, &self.dictionary);
let temp = self.lint_group.config.clone();
self.lint_group.config.fill_with_curated();
@@ -378,7 +373,7 @@ impl Linter {
.into_iter()
.map(|l| {
let problem_text = l.get_str(&source);
- let span = Into::<Span>::into(l.span).to_js_indices(source.as_slice());
+ let span = Into::<Span>::into(l.span).to_js_indices(&source);
Lint::new(l, span, problem_text, language)
})
.collect()
@@ -453,7 +448,7 @@ impl Linter {
) -> Result<String, String> {
let mut source: Vec<_> = source_text.chars().collect();
- let doc = Document::new_from_vec(
+ let doc = Document::new_from_chars(
source.clone().into(),
&lint.language.create_parser(),
&self.dictionary,