Skip to content

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

feat(core): Lint for those who don't like Oxford commas

hippietrail2733b59f3407
diff --git a/harper-core/src/document.rs b/harper-core/src/document.rs
index 7b5bf493a..fd9fe6a3d 100644
--- a/harper-core/src/document.rs
+++ b/harper-core/src/document.rs
@@ -542,6 +542,7 @@ impl TokenStringExt for Document {
create_fns_on_doc!(punctuation);
create_fns_on_doc!(currency);
create_fns_on_doc!(likely_homograph);
+ create_fns_on_doc!(comma);
fn first_sentence_word(&self) -> Option<Token> {
self.tokens.first_sentence_word()
diff --git a/harper-core/src/linting/lint_group.rs b/harper-core/src/linting/lint_group.rs
index 4f2563c67..3b1dcdadf 100644
--- a/harper-core/src/linting/lint_group.rs
+++ b/harper-core/src/linting/lint_group.rs
@@ -33,7 +33,7 @@ use super::that_which::ThatWhich;
use super::unclosed_quotes::UnclosedQuotes;
use super::use_genitive::UseGenitive;
use super::wrong_quotes::WrongQuotes;
-use super::{CurrencyPlacement, Lint, Linter, OxfordComma};
+use super::{CurrencyPlacement, Lint, Linter, NoOxfordComma, OxfordComma};
use crate::{Dictionary, Document};
macro_rules! create_lint_group_config {
@@ -190,6 +190,7 @@ create_lint_group_config!(
MergeWords => true,
PluralConjugate => false,
OxfordComma => true,
+ NoOxfordComma => true,
PronounContraction => true,
CurrencyPlacement => true,
SomewhatSomething => true,
diff --git a/harper-core/src/linting/mod.rs b/harper-core/src/linting/mod.rs
index 48f26dd41..8d406eed1 100644
--- a/harper-core/src/linting/mod.rs
+++ b/harper-core/src/linting/mod.rs
@@ -22,6 +22,7 @@ mod matcher;
mod merge_linters;
mod merge_words;
mod multiple_sequential_pronouns;
+mod no_oxford_comma;
mod number_suffix_capitalization;
mod oxford_comma;
mod pattern_linter;
@@ -59,6 +60,7 @@ pub use long_sentences::LongSentences;
pub use matcher::Matcher;
pub use merge_words::MergeWords;
pub use multiple_sequential_pronouns::MultipleSequentialPronouns;
+pub use no_oxford_comma::NoOxfordComma;
pub use number_suffix_capitalization::NumberSuffixCapitalization;
pub use oxford_comma::OxfordComma;
pub use pattern_linter::PatternLinter;
diff --git a/harper-core/src/linting/no_oxford_comma.rs b/harper-core/src/linting/no_oxford_comma.rs
new file mode 100644
index 000000000..cf1b08095
--- /dev/null
+++ b/harper-core/src/linting/no_oxford_comma.rs
@@ -0,0 +1,162 @@
+use crate::{
+ patterns::{Pattern, SequencePattern, WordSet},
+ Document, Token, TokenStringExt,
+};
+
+use super::{Lint, LintKind, Linter, Suggestion};
+
+pub struct NoOxfordComma {
+ pattern: SequencePattern,
+}
+
+impl NoOxfordComma {
+ pub fn new() -> Self {
+ Self {
+ pattern: SequencePattern::default()
+ .then_noun_phrase()
+ .then_comma()
+ .then_whitespace()
+ .then_noun_phrase()
+ .then_comma()
+ .then_whitespace()
+ .then(Box::new(WordSet::all(&["and", "or", "nor"]))),
+ }
+ }
+
+ fn match_to_lint(&self, matched_toks: &[Token], _source: &[char]) -> Lint {
+ let last_comma_index = matched_toks.last_comma_index().unwrap();
+ let offender = matched_toks[last_comma_index];
+
+ Lint {
+ span: offender.span,
+ lint_kind: LintKind::Style,
+ suggestions: vec![Suggestion::Remove],
+ message: "Remove the Oxford comma here.".to_owned(),
+ priority: 31,
+ }
+ }
+}
+
+impl Default for NoOxfordComma {
+ fn default() -> Self {
+ Self::new()
+ }
+}
+
+impl Linter for NoOxfordComma {
+ fn lint(&mut self, document: &Document) -> Vec<Lint> {
+ let mut lints = Vec::new();
+
+ for sentence in document.iter_sentences() {
+ let mut tok_cursor = 0;
+
+ loop {
+ if tok_cursor >= sentence.len() {
+ break;
+ }
+
+ let match_len = self
+ .pattern
+ .matches(&sentence[tok_cursor..], document.get_source());
+
+ if match_len != 0 {
+ let lint = self.match_to_lint(
+ &sentence[tok_cursor..tok_cursor + match_len],
+ document.get_source(),
+ );
+
+ lints.push(lint);
+ tok_cursor += match_len;
+ } else {
+ tok_cursor += 1;
+ }
+ }
+ }
+
+ lints
+ }
+
+ fn description(&self) -> &str {
+ "The Oxford comma is one of the more controversial rules in common use today. Enabling this lint checks that there is no comma before `and`, `or` or `nor` when listing out more than two ideas."
+ }
+}
+
+#[cfg(test)]
+mod tests {
+ use crate::linting::tests::{assert_lint_count, assert_suggestion_result};
+
+ use super::NoOxfordComma;
+
+ #[test]
+ fn fruits() {
+ assert_lint_count(
+ "An apple, a banana, and a pear",
+ NoOxfordComma::default(),
+ 1,
+ );
+ }
+
+ #[test]
+ fn people() {
+ assert_suggestion_result(
+ "Nancy, Steve, and Carl are going to the coffee shop.",
+ NoOxfordComma::default(),
+ "Nancy, Steve and Carl are going to the coffee shop.",
+ );
+ }
+
+ #[test]
+ fn places() {
+ assert_suggestion_result(
+ "I've always wanted to visit Paris, Tokyo, and Rome.",
+ NoOxfordComma::default(),
+ "I've always wanted to visit Paris, Tokyo and Rome.",
+ );
+ }
+
+ #[test]
+ fn foods() {
+ assert_suggestion_result(
+ "My favorite foods are pizza, sushi, tacos, and burgers.",
+ NoOxfordComma::default(),
+ "My favorite foods are pizza, sushi, tacos and burgers.",
+ );
+ }
+
+ #[test]
+ fn allows_clean_music() {
+ assert_lint_count(
+ "I enjoy listening to pop music, rock, hip-hop, electronic dance and classical music.",
+ NoOxfordComma::default(),
+ 0,
+ );
+ }
+
+ #[test]
+ fn allows_clean_nations() {
+ assert_lint_count("The team consists of players from different countries: France, Germany, Italy and Spain.", NoOxfordComma::default(), 0);
+ }
+
+ #[test]
+ fn or_writing() {
+ assert_suggestion_result("Harper can be a lifesaver when writing technical documents, emails, or other formal forms of communication.", NoOxfordComma::default(), "Harper can be a lifesaver when writing technical documents, emails or other formal forms of communication.",);
+ }
+
+ #[test]
+ fn sports() {
+ assert_suggestion_result(
+ "They enjoy playing soccer, basketball, or tennis.",
+ NoOxfordComma::default(),
+ "They enjoy playing soccer, basketball or tennis.",
+ );
+ }
+
+ #[test]
+ fn nor_vegetables() {
+ assert_suggestion_result(
+ "I like carrots, kale, nor broccoli.",
+ NoOxfordComma::default(),
+ "I like carrots, kale nor broccoli.",
+ );
+ }
+}
diff --git a/harper-core/src/token.rs b/harper-core/src/token.rs
index 83eb246f1..af8892a01 100644
--- a/harper-core/src/token.rs
+++ b/harper-core/src/token.rs
@@ -96,6 +96,7 @@ pub trait TokenStringExt {
create_decl_for!(punctuation);
create_decl_for!(currency);
create_decl_for!(likely_homograph);
+ create_decl_for!(comma);
fn iter_linking_verb_indices(&self) -> impl Iterator<Item = usize> + '_;
fn iter_linking_verbs(&self) -> impl Iterator<Item = Token> + '_;
@@ -137,6 +138,7 @@ impl TokenStringExt for [Token] {
create_fns_for!(chunk_terminator);
create_fns_for!(currency);
create_fns_for!(likely_homograph);
+ create_fns_for!(comma);
fn first_non_whitespace(&self) -> Option<Token> {
self.iter().find(|t| !t.kind.is_whitespace()).copied()