Skip to content

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

refactor: Rename `ACO` to `Word` (#1149)

Michael Schmidtf37fcd2cdef6
diff --git a/harper-core/src/linting/confident.rs b/harper-core/src/linting/confident.rs
index c4cb8aff2..7150131a0 100644
--- a/harper-core/src/linting/confident.rs
+++ b/harper-core/src/linting/confident.rs
@@ -1,7 +1,6 @@
use crate::{
Token,
- char_string::char_string,
- patterns::{AnyCapitalization, OwnedPatternExt, Pattern, SequencePattern},
+ patterns::{OwnedPatternExt, Pattern, SequencePattern, Word},
};
use super::{Lint, LintKind, PatternLinter, Suggestion};
@@ -15,7 +14,7 @@ impl Default for Confident {
let pattern = SequencePattern::default()
.then(
(|tok: &Token, _source: &[char]| tok.kind.is_verb() || tok.kind.is_determiner())
- .or(Box::new(AnyCapitalization::new(char_string!("very")))),
+ .or(Box::new(Word::new("very"))),
)
.then_whitespace()
.t_aco("confidant");
diff --git a/harper-core/src/linting/for_noun.rs b/harper-core/src/linting/for_noun.rs
index dbfb4eb12..6e17f710e 100644
--- a/harper-core/src/linting/for_noun.rs
+++ b/harper-core/src/linting/for_noun.rs
@@ -1,6 +1,6 @@
use crate::{
Token,
- patterns::{AnyCapitalization, NominalPhrase, OwnedPatternExt, Pattern, SequencePattern},
+ patterns::{NominalPhrase, OwnedPatternExt, Pattern, SequencePattern, Word},
};
use super::{Lint, LintKind, PatternLinter, Suggestion};
@@ -13,7 +13,7 @@ impl Default for ForNoun {
fn default() -> Self {
let pattern = SequencePattern::aco("fro")
.then_whitespace()
- .then(NominalPhrase.or(Box::new(AnyCapitalization::of("sure"))));
+ .then(NominalPhrase.or(Box::new(Word::new("sure"))));
Self {
pattern: Box::new(pattern),
diff --git a/harper-core/src/linting/the_my.rs b/harper-core/src/linting/the_my.rs
index 698dd453a..f1253e746 100644
--- a/harper-core/src/linting/the_my.rs
+++ b/harper-core/src/linting/the_my.rs
@@ -1,7 +1,7 @@
use super::{Lint, LintKind, PatternLinter, Suggestion};
use crate::{
CharStringExt, Token, TokenStringExt,
- patterns::{AnyCapitalization, EitherPattern, Pattern, SequencePattern, WordSet},
+ patterns::{EitherPattern, Pattern, SequencePattern, Word, WordSet},
};
pub struct TheMy {
@@ -10,7 +10,7 @@ pub struct TheMy {
impl Default for TheMy {
fn default() -> Self {
- let the = AnyCapitalization::of("the");
+ let the = Word::new("the");
let any_possessive = WordSet::new(&["my", "your", "his", "her", "its", "our", "their"]);
let the_poss = SequencePattern::default()
diff --git a/harper-core/src/linting/then_than.rs b/harper-core/src/linting/then_than.rs
index 4bf9f4fb2..5130b0b4f 100644
--- a/harper-core/src/linting/then_than.rs
+++ b/harper-core/src/linting/then_than.rs
@@ -2,8 +2,7 @@ use super::{Lint, LintKind, PatternLinter};
use crate::Token;
use crate::linting::Suggestion;
use crate::patterns::{
- All, AnyCapitalization, EitherPattern, Invert, OwnedPatternExt, Pattern, SequencePattern,
- WordSet,
+ All, EitherPattern, Invert, OwnedPatternExt, Pattern, SequencePattern, Word, WordSet,
};
#[doc = "Corrects the misuse of `then` to `than`."]
@@ -19,7 +18,7 @@ impl ThenThan {
// Comparative form of adjective
Box::new(
SequencePattern::default()
- .then(AnyCapitalization::of("other").or(Box::new(
+ .then(Word::new("other").or(Box::new(
|tok: &Token, source: &[char]| {
is_comparative_adjective(tok, source)
},
@@ -27,7 +26,7 @@ impl ThenThan {
.then_whitespace()
.then_any_capitalization_of("then")
.then_whitespace()
- .then(Invert::new(AnyCapitalization::of("that"))),
+ .then(Invert::new(Word::new("that"))),
),
// Positive form of adjective following "more" or "less"
Box::new(
@@ -38,7 +37,7 @@ impl ThenThan {
.then_whitespace()
.then_any_capitalization_of("then")
.then_whitespace()
- .then(Invert::new(AnyCapitalization::of("that"))),
+ .then(Invert::new(Word::new("that"))),
),
])),
// Exceptions to the rule.
diff --git a/harper-core/src/linting/widely_accepted.rs b/harper-core/src/linting/widely_accepted.rs
index 0d7554107..2bb36df61 100644
--- a/harper-core/src/linting/widely_accepted.rs
+++ b/harper-core/src/linting/widely_accepted.rs
@@ -1,7 +1,7 @@
use crate::{
Token,
linting::{Lint, LintKind, PatternLinter, Suggestion},
- patterns::{AnyCapitalization, Pattern, SequencePattern, WordSet},
+ patterns::{Pattern, SequencePattern, Word, WordSet},
};
pub struct WidelyAccepted {
@@ -11,7 +11,7 @@ pub struct WidelyAccepted {
impl Default for WidelyAccepted {
fn default() -> Self {
let pattern = SequencePattern::default()
- .then(AnyCapitalization::new("wide".chars().collect()))
+ .then(Word::new("wide"))
.then_whitespace()
.then(WordSet::new(&["accepted", "acceptable", "used"]));
diff --git a/harper-core/src/patterns/any_capitalization.rs b/harper-core/src/patterns/any_capitalization.rs
deleted file mode 100644
index 67cecab25..000000000
--- a/harper-core/src/patterns/any_capitalization.rs
+++ /dev/null
@@ -1,46 +0,0 @@
-use std::num::NonZeroUsize;
-
-use crate::{CharString, Token};
-
-use super::Pattern;
-
-/// A [`Pattern`] that matches any capitalization of a provided word.
-#[derive(Clone)]
-pub struct AnyCapitalization {
- word: CharString,
-}
-
-impl AnyCapitalization {
- pub fn new(word: CharString) -> Self {
- Self { word }
- }
-
- pub fn of(word: &str) -> Self {
- let chars = word.chars().collect();
-
- Self::new(chars)
- }
-}
-
-impl Pattern for AnyCapitalization {
- fn matches(&self, tokens: &[Token], source: &[char]) -> Option<NonZeroUsize> {
- let tok = tokens.first()?;
-
- if !tok.kind.is_word() {
- return None;
- }
-
- if tok.span.len() != self.word.len() {
- return None;
- }
-
- let tok_chars = tok.span.get_content(source);
-
- let partial_match = tok_chars
- .iter()
- .zip(&self.word)
- .all(|(a, b)| a.eq_ignore_ascii_case(b));
-
- NonZeroUsize::new(if partial_match { 1 } else { 0 })
- }
-}
diff --git a/harper-core/src/patterns/exact_phrase.rs b/harper-core/src/patterns/exact_phrase.rs
index 9eb543cd2..eb40061c9 100644
--- a/harper-core/src/patterns/exact_phrase.rs
+++ b/harper-core/src/patterns/exact_phrase.rs
@@ -2,7 +2,7 @@ use std::num::NonZeroUsize;
use crate::{Document, Token, TokenKind};
-use super::{AnyCapitalization, Pattern, SequencePattern};
+use super::{Pattern, SequencePattern, Word};
pub struct ExactPhrase {
inner: SequencePattern,
@@ -20,7 +20,7 @@ impl ExactPhrase {
for token in doc.fat_tokens() {
match token.kind {
TokenKind::Word(_word_metadata) => {
- phrase = phrase.then(AnyCapitalization::new(token.content.as_slice().into()));
+ phrase = phrase.then(Word::from_chars(token.content.as_slice()));
}
TokenKind::Space(_) => {
phrase = phrase.then_whitespace();
diff --git a/harper-core/src/patterns/mod.rs b/harper-core/src/patterns/mod.rs
index bc75e251e..3f5de927b 100644
--- a/harper-core/src/patterns/mod.rs
+++ b/harper-core/src/patterns/mod.rs
@@ -10,7 +10,6 @@ use std::{collections::VecDeque, num::NonZeroUsize};
use crate::{Document, Span, Token, VecExt};
mod all;
-mod any_capitalization;
mod any_pattern;
mod either_pattern;
mod exact_phrase;
@@ -27,11 +26,11 @@ mod similar_to_phrase;
mod split_compound_word;
mod whitespace_pattern;
mod within_edit_distance;
+mod word;
mod word_pattern_group;
mod word_set;
pub use all::All;
-pub use any_capitalization::AnyCapitalization;
pub use any_pattern::AnyPattern;
use blanket::blanket;
pub use either_pattern::EitherPattern;
@@ -48,6 +47,7 @@ pub use sequence_pattern::SequencePattern;
pub use similar_to_phrase::SimilarToPhrase;
pub use split_compound_word::SplitCompoundWord;
pub use whitespace_pattern::WhitespacePattern;
+pub use word::Word;
pub use word_pattern_group::WordPatternGroup;
pub use word_set::WordSet;
diff --git a/harper-core/src/patterns/sequence_pattern.rs b/harper-core/src/patterns/sequence_pattern.rs
index abfd2817d..959fc45e1 100644
--- a/harper-core/src/patterns/sequence_pattern.rs
+++ b/harper-core/src/patterns/sequence_pattern.rs
@@ -3,7 +3,7 @@ use std::num::NonZeroUsize;
use paste::paste;
use super::whitespace_pattern::WhitespacePattern;
-use super::{AnyCapitalization, AnyPattern, IndefiniteArticle, Pattern, RepeatingPattern};
+use super::{AnyPattern, IndefiniteArticle, Pattern, RepeatingPattern, Word};
use crate::{Token, TokenKind};
/// A pattern that checks that a sequence of other patterns match.
@@ -94,25 +94,7 @@ impl SequencePattern {
}
pub fn then_exact_word(mut self, word: &'static str) -> Self {
- self.token_patterns
- .push(Box::new(|tok: &Token, source: &[char]| {
- if !tok.kind.is_word() {
- return false;
- }
-
- let tok_chars = tok.span.get_content(source);
-
- let mut w_char_count = 0;
- for (i, w_char) in word.chars().enumerate() {
- w_char_count += 1;
-
- if tok_chars.get(i).cloned() != Some(w_char) {
- return false;
- }
- }
-
- w_char_count == tok_chars.len()
- }));
+ self.token_patterns.push(Box::new(Word::new_exact(word)));
self
}
@@ -132,8 +114,7 @@ impl SequencePattern {
/// Match examples of `word` that have any capitalization.
pub fn then_any_capitalization_of(mut self, word: &'static str) -> Self {
- self.token_patterns
- .push(Box::new(AnyCapitalization::of(word)));
+ self.token_patterns.push(Box::new(Word::new(word)));
self
}
diff --git a/harper-core/src/patterns/similar_to_phrase.rs b/harper-core/src/patterns/similar_to_phrase.rs
index 10fdd1c31..1577a4388 100644
--- a/harper-core/src/patterns/similar_to_phrase.rs
+++ b/harper-core/src/patterns/similar_to_phrase.rs
@@ -2,9 +2,7 @@ use std::num::NonZeroUsize;
use crate::{Document, Token, TokenKind};
-use super::{
- AnyCapitalization, Pattern, SequencePattern, within_edit_distance::WithinEditDistance,
-};
+use super::{Pattern, SequencePattern, Word, within_edit_distance::WithinEditDistance};
pub struct SimilarToPhrase {
phrase: SequencePattern,
@@ -37,7 +35,7 @@ impl SimilarToPhrase {
for token in document.fat_tokens() {
match token.kind {
TokenKind::Word(_word_metadata) => {
- phrase = phrase.then(AnyCapitalization::new(token.content.as_slice().into()));
+ phrase = phrase.then(Word::from_chars(token.content.as_slice()));
fuzzy_phrase = fuzzy_phrase
.then(WithinEditDistance::new(token.content.into(), max_edit_dist));
}
diff --git a/harper-core/src/patterns/word.rs b/harper-core/src/patterns/word.rs
new file mode 100644
index 000000000..53a098e77
--- /dev/null
+++ b/harper-core/src/patterns/word.rs
@@ -0,0 +1,98 @@
+use std::num::NonZeroUsize;
+
+use super::Pattern;
+
+use crate::{CharString, Token};
+
+/// Matches a predefined word.
+///
+/// Note that any capitalization of the contained words will result in a match.
+#[derive(Clone)]
+pub struct Word {
+ word: CharString,
+ exact: bool,
+}
+
+impl Word {
+ /// Matches the provided word, ignoring case.
+ pub fn new(word: &'static str) -> Self {
+ Self {
+ word: word.chars().collect(),
+ exact: false,
+ }
+ }
+ /// Matches the provided word, ignoring case.
+ pub fn from_chars(word: &[char]) -> Self {
+ Self {
+ word: word.iter().copied().collect(),
+ exact: false,
+ }
+ }
+
+ /// Matches the provided word, case-sensitive.
+ pub fn new_exact(word: &'static str) -> Self {
+ Self {
+ word: word.chars().collect(),
+ exact: true,
+ }
+ }
+}
+
+impl Pattern for Word {
+ fn matches(&self, tokens: &[Token], source: &[char]) -> Option<NonZeroUsize> {
+ let tok = tokens.first()?;
+ if !tok.kind.is_word() {
+ return None;
+ }
+ if tok.span.len() != self.word.len() {
+ return None;
+ }
+
+ let chars = tok.span.get_content(source);
+ let eq = if self.exact {
+ chars == self.word.as_slice()
+ } else {
+ chars
+ .iter()
+ .zip(&self.word)
+ .all(|(a, b)| a.eq_ignore_ascii_case(b))
+ };
+
+ NonZeroUsize::new(if eq { 1 } else { 0 })
+ }
+}
+
+#[cfg(test)]
+mod tests {
+ use crate::{Document, Span, patterns::DocPattern};
+
+ use super::Word;
+
+ #[test]
+ fn fruit() {
+ let doc = Document::new_markdown_default_curated("I ate a banana and an apple today.");
+
+ assert_eq!(
+ Word::new("banana").find_all_matches_in_doc(&doc),
+ vec![Span::new(6, 7)]
+ );
+ assert_eq!(
+ Word::new_exact("banana").find_all_matches_in_doc(&doc),
+ vec![Span::new(6, 7)]
+ );
+ }
+
+ #[test]
+ fn fruit_whack_capitalization() {
+ let doc = Document::new_markdown_default_curated("I Ate A bAnaNa And aN apPlE today.");
+
+ assert_eq!(
+ Word::new("banana").find_all_matches_in_doc(&doc),
+ vec![Span::new(6, 7)]
+ );
+ assert_eq!(
+ Word::new_exact("banana").find_all_matches_in_doc(&doc),
+ vec![]
+ );
+ }
+}
diff --git a/harper-core/src/patterns/word_set.rs b/harper-core/src/patterns/word_set.rs
index cbb892004..e60961b80 100644
--- a/harper-core/src/patterns/word_set.rs
+++ b/harper-core/src/patterns/word_set.rs
@@ -5,10 +5,10 @@ use smallvec::SmallVec;
use crate::{CharString, Token};
-// A [`Pattern`] that matches against any of a set of provided words.
-// For small sets of short words, it doesn't allocate.
-//
-// Note that any capitalization of the contained words will result in a match.
+/// A [`Pattern`] that matches against any of a set of provided words.
+/// For small sets of short words, it doesn't allocate.
+///
+/// Note that any capitalization of the contained words will result in a match.
#[derive(Debug, Default, Clone)]
pub struct WordSet {
words: SmallVec<[CharString; 4]>,