Skip to content

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

refactor(core): proper noun linters use canonical casing and JSON file

Elijah Potter3154473e3583
diff --git a/harper-core/dictionary.dict b/harper-core/dictionary.dict
index 538614fa2..1ed166762 100644
--- a/harper-core/dictionary.dict
+++ b/harper-core/dictionary.dict
@@ -27614,10 +27614,10 @@ hysterics/1M
i/18US
is/4813
iOS/2M
-iPad/1M
-iPhone/1M
-iPod/1M
-iTunes/M
+iPad/1MS
+iPhone/1MS
+iPod/1MS
+iTunes/MS
iamb/1MS
iambi/1
iambic/51SM
diff --git a/harper-core/proper_noun_rules.json b/harper-core/proper_noun_rules.json
new file mode 100644
index 000000000..a1576b17e
--- /dev/null
+++ b/harper-core/proper_noun_rules.json
@@ -0,0 +1,410 @@
+{
+ "Americas": {
+ "canonical": [
+ "South America",
+ "North America",
+ "Central America"
+ ],
+ "description": "When referring to North, Central, and South America, make sure to treat them as a proper noun."
+ },
+ "Australia": {
+ "canonical": [
+ "Australian Capital Territory",
+ "New South Wales",
+ "Northern Territory",
+ "South Australia",
+ "Western Australia",
+ "Alice Springs",
+ "Gold Coast",
+ "Sunshine Coast"
+ ],
+ "description": "When referring to the states of Australia, make sure to treat them as a proper noun."
+ },
+ "OceansAndSeas": {
+ "canonical": [
+ "Atlantic Ocean",
+ "Pacific Ocean",
+ "Indian Ocean",
+ "Southern Ocean",
+ "Arctic Ocean",
+ "Mediterranean Sea",
+ "Caribbean Sea",
+ "Baltic Sea",
+ "Red Sea",
+ "Black Sea",
+ "Caspian Sea",
+ "Coral Sea",
+ "Bering Sea",
+ "North Sea",
+ "South China Sea"
+ ],
+ "description": "When referring to the world's oceans and seas, ensure they are treated as proper nouns."
+ },
+ "Canada": {
+ "canonical": [
+ "British Columbia",
+ "New Brunswick",
+ "Northwest Territories",
+ "Nova Scotia",
+ "Prince Edward Island",
+ "Quebec City"
+ ],
+ "description": "When referring to the provinces of Canada, make sure to treat them as a proper noun."
+ },
+ "Malaysia": {
+ "canonical": [
+ "Negeri Sembilan",
+ "Alor Setar",
+ "George Town",
+ "Johor Bahru",
+ "Kota Bahru",
+ "Kota Kinabalu",
+ "Kuala Lumpur",
+ "Kuala Terengganu",
+ "Shah Alam"
+ ],
+ "description": "When referring to the states of Malaysia and their capitals, make sure to treat them as a proper noun."
+ },
+ "Countries": {
+ "canonical": [
+ "Equatorial Guinea",
+ "Papua New Guinea",
+ "Cayman Islands",
+ "Falkland Islands",
+ "Marshall Islands",
+ "Solomon Islands",
+ "British Virgin Islands",
+ "United States Virgin Islands",
+ "Northern Mariana Islands",
+ "New Caledonia",
+ "New Zealand",
+ "Northern Cyprus",
+ "Northern Ireland",
+ "Central African Republic",
+ "Czech Republic",
+ "Dominican Republic",
+ "Saint Helena",
+ "Saint Lucia",
+ "Saint Martin",
+ "South Africa",
+ "South Ossetia",
+ "South Sudan",
+ "American Samoa",
+ "Antigua and Barbuda",
+ "Bosnia and Herzegovina",
+ "Burkina Faso",
+ "Cape Verde",
+ "Costa Rica",
+ "Democratic Republic of the Congo",
+ "East Timor",
+ "El Salvador",
+ "French Polynesia",
+ "Guinea-Bissau",
+ "Isle of Man",
+ "Ivory Coast",
+ "North Macedonia",
+ "Puerto Rico",
+ "São Tomé and Príncipe",
+ "Saudi Arabia",
+ "Sierra Leone",
+ "Sint Maarten",
+ "Sri Lanka",
+ "Trinidad and Tobago",
+ "Western Sahara"
+ ],
+ "description": "When referring to Countries, make sure to treat it as a proper noun."
+ },
+ "NationalCapitals": {
+ "canonical": [
+ "Belize City",
+ "Guatemala City",
+ "Kuwait City",
+ "Mexico City",
+ "Panama City",
+ "Vatican City",
+ "Ho Chi Minh City",
+ "Saint Kitts and Nevis",
+ "Saint Pierre and Miquelon",
+ "Saint Vincent and the Grenadines",
+ "San José",
+ "San Juan",
+ "San Marino",
+ "San Salvador",
+ "Cape Town",
+ "George Town",
+ "Abu Dhabi",
+ "Addis Ababa",
+ "Bandar Seri Begawan",
+ "Buenos Aires",
+ "Diego Garcia",
+ "La Paz",
+ "New Delhi",
+ "Pago Pago",
+ "Phnom Penh",
+ "Port Louis",
+ "Port Moresby",
+ "Port Vila",
+ "Port of Spain",
+ "Porto-Novo",
+ "Santo Domingo",
+ "São Tomé",
+ "The Bahamas",
+ "The Hague"
+ ],
+ "description": "When referring to national capitals, make sure to treat it as a proper noun."
+ },
+ "ChineseCommunistParty": {
+ "canonical": [
+ "Chinese Communist Party"
+ ],
+ "description": "When referring to the political party, make sure to treat them as a proper noun."
+ },
+ "UnitedOrganizations": {
+ "canonical": [
+ "United Nations",
+ "United States",
+ "United Kingdom",
+ "United Airlines",
+ "United Arab Emirates"
+ ],
+ "description": "When referring to national or international organizations, make sure to treat them as a proper noun."
+ },
+ "Holidays": {
+ "canonical": [
+ "Absolution Day",
+ "Admission Day",
+ "Alaska Day",
+ "Anzac Day",
+ "Arbor Day",
+ "Armistice Day",
+ "Ascension Day",
+ "Australia Day",
+ "Bastille Day",
+ "Boxing Day",
+ "Canada Day",
+ "Christmas Day",
+ "Columbus Day",
+ "Commonwealth Day",
+ "Darwin Day",
+ "Discovery Day",
+ "Dominion Day",
+ "Earth Day",
+ "Easter Day",
+ "Election Day",
+ "Emancipation Day",
+ "Empire Day",
+ "Father's Day",
+ "Flag Day",
+ "Freedom Day",
+ "Galentine's Day",
+ "Groundhog Day",
+ "Halloween",
+ "Independence Day",
+ "Jamhuri Day",
+ "Jubilee Day",
+ "Kamehameha Day",
+ "Kenyatta Day",
+ "Labor Day",
+ "Labour Day",
+ "Madaraka Day",
+ "Mashujaa Day",
+ "May Day",
+ "Memorial Day",
+ "Merdeka Day",
+ "Mother's Day",
+ "National Freedom Day",
+ "New Year's Day",
+ "Patrick's Day",
+ "Presidents' Day",
+ "Remembrance Day",
+ "Republic Day",
+ "Rizal Day",
+ "Thanksgiving Day",
+ "Ulster Day",
+ "Valentine's Day",
+ "Veterans Day",
+ "Victoria Day",
+ "Victory Day",
+ "Waitangi Day",
+ "Wattle Day",
+ "Year's Day",
+ "Youth Day",
+ "Black Friday",
+ "Cyber Monday"
+ ],
+ "description": "When referring to holidays, make sure to treat them as a proper noun."
+ },
+ "Koreas": {
+ "canonical": [
+ "South Korea",
+ "North Korea"
+ ],
+ "description": "When referring to the nations, make sure to treat them as a proper noun."
+ },
+ "AmazonNames": {
+ "canonical": [
+ "Amazon Shopping",
+ "Amazon Web Services",
+ "Amazon Lambda",
+ "Amazon RDS",
+ "Amazon DynamoDB",
+ "Amazon SageMaker",
+ "Amazon Rekognition",
+ "Amazon CloudFront",
+ "Amazon ECS",
+ "Amazon EKS",
+ "Amazon CloudWatch",
+ "Amazon IAM",
+ "Amazon Prime",
+ "Amazon Kindle"
+ ],
+ "description": "When referring to the various products of Amazon.com, make sure to treat them as a proper noun."
+ },
+ "GoogleNames": {
+ "canonical": [
+ "Google Search",
+ "Google Cloud",
+ "Google Maps",
+ "Google Docs",
+ "Google Sheets",
+ "Google Slides",
+ "Google Drive",
+ "Google Meet",
+ "Google Gmail",
+ "Google Calendar",
+ "Google Chrome",
+ "Google ChromeOS",
+ "Google Android",
+ "Google Play",
+ "Google Bard",
+ "Google Gemini",
+ "Google YouTube",
+ "Google Photos",
+ "Google Analytics",
+ "Google AdSense",
+ "Google Pixel",
+ "Google Nest",
+ "Google Workspace"
+ ],
+ "description": "When referring to Google products and services, make sure to treat them as proper nouns."
+ },
+ "AzureNames": {
+ "canonical": [
+ "Azure DevOps",
+ "Azure Functions",
+ "Azure Cosmos DB",
+ "Azure SQL Database",
+ "Azure Kubernetes Service",
+ "Azure Virtual Machines",
+ "Azure Monitor",
+ "Azure Storage",
+ "Azure Active Directory",
+ "Azure App Service",
+ "Azure Key Vault",
+ "Azure Cognitive Services",
+ "Azure Service Bus",
+ "Azure Event Hub"
+ ],
+ "description": "When referring to Azure cloud services, make sure to treat them as proper nouns."
+ },
+ "MicrosoftNames": {
+ "canonical": [
+ "Microsoft Windows",
+ "Microsoft Office",
+ "Microsoft Teams",
+ "Microsoft Excel",
+ "Microsoft PowerPoint",
+ "Microsoft Word",
+ "Microsoft Outlook",
+ "Microsoft OneDrive",
+ "Microsoft SharePoint",
+ "Microsoft Xbox",
+ "Microsoft Surface",
+ "Microsoft Edge",
+ "Microsoft Bing",
+ "Microsoft Dynamics",
+ "Microsoft Visual Studio"
+ ],
+ "description": "When referring to Microsoft products and services, make sure to treat them as proper nouns."
+ },
+ "AppleNames": {
+ "canonical": [
+ "Apple iPhone",
+ "Apple iPad",
+ "Apple iMac",
+ "Apple MacBook",
+ "Apple Watch",
+ "Apple TV",
+ "Apple Music",
+ "Apple Arcade",
+ "Apple iCloud",
+ "Apple Safari",
+ "Apple HomeKit",
+ "Apple CarPlay",
+ "Apple MacBook Pro",
+ "Apple MacBook Air",
+ "Apple Mac Pro",
+ "Apple Mac Mini",
+ "Apple AirPods",
+ "Apple AirPods Pro",
+ "Apple AirPods Max",
+ "Apple Vision Pro"
+ ],
+ "description": "When referring to Apple products and services, make sure to treat them as proper nouns."
+ },
+ "MetaNames": {
+ "canonical": [
+ "Meta Oculus",
+ "Meta Portals",
+ "Meta Quest",
+ "Meta Gaming",
+ "Meta Horizon",
+ "Meta Reality Labs"
+ ],
+ "description": "When referring to Meta products and services, make sure to treat them as proper nouns."
+ },
+ "JetpackNames": {
+ "canonical": [
+ "Jetpack VaultPress Backup",
+ "Jetpack VaultPress",
+ "Jetpack Scan",
+ "Jetpack Akismet Anti-spam",
+ "Jetpack Stats",
+ "Jetpack Social",
+ "Jetpack Blaze",
+ "Jetpack AI Assistant",
+ "Jetpack Site Search",
+ "Jetpack Boost",
+ "Jetpack VideoPress",
+ "Jetpack For Agencies",
+ "Jetpack CRM"
+ ],
+ "description": "Ensure proper capitalization of Jetpack-related terms."
+ },
+ "TumblrNames": {
+ "canonical": [
+ "Tumblr Blaze",
+ "Tumblr Pro",
+ "Tumblr Live",
+ "Tumblr Ads",
+ "Tumblr Communities",
+ "Tumblr Shop",
+ "Tumblr Dashboard"
+ ],
+ "description": "Ensure proper capitalization of Tumblr-related terms."
+ },
+ "PocketCastsNames": {
+ "canonical": [
+ "Pocket Casts",
+ "Pocket Casts Plus"
+ ],
+ "description": "Ensure proper capitalization of Pocket Casts and Pocket Casts Plus as brand names."
+ },
+ "DayOneNames": {
+ "canonical": [
+ "Day One",
+ "Day One Premium"
+ ],
+ "description": "Ensure proper capitalization of Day One and Day One Premium as brand names."
+ }
+}
diff --git a/harper-core/src/linting/proper_noun_capitalization_linters.rs b/harper-core/src/linting/proper_noun_capitalization_linters.rs
index 004318dfa..352e898d8 100644
--- a/harper-core/src/linting/proper_noun_capitalization_linters.rs
+++ b/harper-core/src/linting/proper_noun_capitalization_linters.rs
@@ -1,47 +1,117 @@
+use hashbrown::HashMap;
+use serde::{Deserialize, Serialize};
+
use super::{Lint, LintKind, Suggestion};
use super::{LintGroup, PatternLinter};
-use crate::patterns::{EitherPattern, IsNotTitleCase, Pattern, SequencePattern, WordSet};
-use crate::{Dictionary, make_title_case};
+use crate::parsers::PlainEnglish;
+use crate::patterns::{AnyCapitalization, Pattern, PatternMap, SequencePattern};
+use crate::{Dictionary, Document, TokenKind};
use crate::{Token, TokenStringExt};
use std::sync::Arc;
+/// A linter that corrects the capitalization of multi-word proper nouns.
+/// They are corrected to a "canonical capitalization" provided at construction.
+///
+/// If you would like to add a proper noun to Harper, see `proper_noun_rules.json`.
pub struct ProperNounCapitalizationLinter<D: Dictionary + 'static> {
- pattern: Box<dyn Pattern>,
+ pattern_map: PatternMap<Document>,
description: String,
dictionary: Arc<D>,
}
impl<D: Dictionary + 'static> ProperNounCapitalizationLinter<D> {
+ /// Wrapper function around [`Self::new`] that allows construction with Strings.
+ pub fn new_strs(
+ canonical_versions: impl IntoIterator<Item = impl AsRef<str>>,
+ description: impl ToString,
+ dictionary: D,
+ ) -> Self {
+ Self::new(
+ canonical_versions
+ .into_iter()
+ .map(|s| s.as_ref().chars().collect::<Vec<_>>()),
+ description,
+ dictionary,
+ )
+ }
+
+ /// Create a linter that corrects the capitalization of phrases provided.
pub fn new(
- search_for: impl Pattern + 'static,
+ canonical_versions: impl IntoIterator<Item = impl AsRef<[char]>>,
description: impl ToString,
dictionary: D,
) -> Self {
let dictionary = Arc::new(dictionary);
+ let mut pattern_map = PatternMap::default();
+
+ for can_vers in canonical_versions {
+ let (pattern, document) = Self::create_pattern(can_vers.as_ref().to_vec(), &dictionary);
+ pattern_map.insert(pattern, document);
+ }
+
Self {
- pattern: Box::new(IsNotTitleCase::new(
- Box::new(search_for),
- dictionary.clone(),
- )),
+ pattern_map,
dictionary: dictionary.clone(),
description: description.to_string(),
}
}
+
+ /// Given the canonical casing of a phrase, create a [`Pattern`] that matches
+ /// _incorrect_ casing of that phrase.
+ fn create_pattern(
+ canonical_case: Vec<char>,
+ dictionary: &impl Dictionary,
+ ) -> (SequencePattern, Document) {
+ // Split into tokens.
+ let doc = Document::new_from_vec(canonical_case.into(), &PlainEnglish, dictionary);
+
+ let mut pattern = SequencePattern::default();
+
+ for token in doc.tokens() {
+ match token.kind {
+ TokenKind::Word(_) => {
+ let word_canon_chars = doc.get_span_content(token.span);
+
+ pattern = pattern.then(AnyCapitalization::new(word_canon_chars.into()));
+ }
+ TokenKind::Space(_) | TokenKind::Newline(_) => pattern = pattern.then_whitespace(),
+ _ => (),
+ }
+ }
+
+ (pattern, doc)
+ }
}
impl<D: Dictionary + 'static> PatternLinter for ProperNounCapitalizationLinter<D> {
fn pattern(&self) -> &dyn Pattern {
- self.pattern.as_ref()
+ &self.pattern_map
}
fn match_to_lint(&self, matched_tokens: &[Token], source: &[char]) -> Option<Lint> {
- let proper = make_title_case(matched_tokens, source, &self.dictionary);
+ let canonical_case = self.pattern_map.lookup(matched_tokens, source).unwrap();
+
+ let mut broken = false;
+
+ for (err_token, correct_token) in matched_tokens.iter().zip(canonical_case.fat_tokens()) {
+ let err_chars = err_token.span.get_content(source);
+ if err_chars != correct_token.content {
+ broken = true;
+ break;
+ }
+ }
+
+ if !broken {
+ return None;
+ }
Some(Lint {
span: matched_tokens.span()?,
lint_kind: LintKind::Capitalization,
- suggestions: vec![Suggestion::ReplaceWith(proper)],
+ suggestions: vec![Suggestion::ReplaceWith(
+ canonical_case.get_source().to_vec(),
+ )],
message: self.description.to_string(),
priority: 31,
})
@@ -52,1190 +122,39 @@ impl<D: Dictionary + 'static> PatternLinter for ProperNounCapitalizationLinter<D
}
}
-pub fn lint_group(dictionary: Arc<impl Dictionary + 'static>) -> LintGroup {
- let mut group = LintGroup::empty();
-
- group.add(
- "Americas",
- Box::new(ProperNounCapitalizationLinter::new(
- SequencePattern::default()
- .then(WordSet::new(&["South", "North", "Central"]))
- .then_whitespace()
- .t_aco("America"),
- "When referring to North, Central, and South America, make sure to treat them as a proper noun.",
- dictionary.clone()))
-);
-
- group.add(
- "Australia",
- Box::new(ProperNounCapitalizationLinter::new(
- EitherPattern::new(vec![
- // the states and territories
- Box::new(
- SequencePattern::default()
- .t_aco("Australian")
- .then_whitespace()
- .t_aco("Capital")
- .then_whitespace()
- .t_aco("Territory"),
- ),
- Box::new(
- SequencePattern::default()
- .t_aco("New")
- .then_whitespace()
- .t_aco("South")
- .then_whitespace()
- .t_aco("Wales"),
- ),
- Box::new(
- SequencePattern::default()
- .t_aco("Northern")
- .then_whitespace()
- .t_aco("Territory"),
- ),
- Box::new(
- SequencePattern::default()
- .t_aco("South")
- .then_whitespace()
- .t_aco("Australia"),
- ),
- Box::new(
- SequencePattern::default()
- .t_aco("Western")
- .then_whitespace()
- .t_aco("Australia"),
- ),
- // major cities
- Box::new(
- SequencePattern::default()
- .t_aco("Alice")
- .then_whitespace()
- .t_aco("Springs"),
- ),
- Box::new(
- SequencePattern::default()
- .t_aco("Gold")
- .then_whitespace()
- .t_aco("Coast"),
- ),
- Box::new(
- SequencePattern::default()
- .t_aco("Sunshine")
- .then_whitespace()
- .t_aco("Coast"),
- ),
- ]),
- "When referring to the states of Australia, make sure to treat them as a proper noun.",
- dictionary.clone(),
- )),
- );
-
- group.add(
- "OceansAndSeas",
- Box::new(ProperNounCapitalizationLinter::new(
- EitherPattern::new(vec![
- Box::new(
- SequencePattern::default()
- .then(WordSet::new(&[
- "Atlantic",
- "Pacific",
- "Indian",
- "Southern",
- "Arctic",
- ]))
- .then_whitespace()
- .t_aco("Ocean")
- ),
- Box::new(
- SequencePattern::default()
- .then(WordSet::new(&[
- "Mediterranean",
- "Caribbean",
- "Baltic",
- "Red",
- "Black",
- "Caspian",
- "Coral",
- "Bering",
- "North",
- ]))
- .then_whitespace()
- .t_aco("Sea")
- ),
- Box::new(
- SequencePattern::default()
- .t_aco("South")
- .then_whitespace()
- .t_aco("China")
- .then_whitespace()
- .t_aco("Sea")
- ),
- ]),
- "When referring to the world's oceans and seas, ensure they are treated as proper nouns.",
- dictionary.clone()
- ))
-);
-
- group.add(
- "Canada",
- Box::new(ProperNounCapitalizationLinter::new(
- EitherPattern::new(vec![
- // the provinces and territories
- Box::new(
- SequencePattern::default()
- .t_aco("British")
- .then_whitespace()
- .t_aco("Columbia"),
- ),
- Box::new(
- SequencePattern::default()
- .t_aco("New")
- .then_whitespace()
- .t_aco("Brunswick"),
- ),
- Box::new(
- SequencePattern::default()
- .t_aco("Northwest")
- .then_whitespace()
- .t_aco("Territories"),
- ),
- Box::new(
- SequencePattern::default()
- .t_aco("Nova")
- .then_whitespace()
- .t_aco("Scotia"),
- ),
- Box::new(
- SequencePattern::default()
- .t_aco("Prince")
- .then_whitespace()
- .t_aco("Edward")
- .then_whitespace()
- .t_aco("Island"),
- ),
- // major cities
- Box::new(
- SequencePattern::default()
- .t_aco("Quebec")
- .then_whitespace()
- .t_aco("City"),
- ),
- ]),
- "When referring to the provinces of Canada, make sure to treat them as a proper noun.",
- dictionary.clone(),
- )),
- );
-
- group.add(
- "Koreas",
- Box::new(ProperNounCapitalizationLinter::new(
- SequencePattern::default()
- .then(WordSet::new(&["South", "North"]))
- .then_whitespace()
- .t_aco("Korea"),
- "When referring to the nations, make sure to treat them as a proper noun.",
- dictionary.clone(),
- )),
- );
-
- group.add(
- "Malaysia",
- Box::new(ProperNounCapitalizationLinter::new(
- EitherPattern::new(vec![
- // multi-word states
- Box::new(SequencePattern::default()
- .t_aco("Negeri")
- .then_whitespace()
- .t_aco("Sembilan")
- ),
- // multi-word state capitals
- Box::new(SequencePattern::default()
- .t_aco("Alor")
- .then_whitespace()
- .t_aco("Setar")
- ),
- Box::new(SequencePattern::default()
- .t_aco("George")
- .then_whitespace()
- .t_aco("Town")
- ),
- Box::new(SequencePattern::default()
- .then(EitherPattern::new(vec![
- Box::new(WordSet::new(&[
- "Johor",
- "Kota"
- ])),
- ]))
- .then_whitespace()
- .t_aco("Bahru")
- ),
- Box::new(SequencePattern::default()
- .t_aco("Kota")
- .then_whitespace()
- .t_aco("Kinabalu")
- ),
- Box::new(SequencePattern::default()
- .t_aco("Kuala")
- .then_whitespace()
- .then(EitherPattern::new(vec![
- Box::new(WordSet::new(&[
- "Lumpur",
- "Terengganu"
- ])),
- ]))
- ),
- Box::new(SequencePattern::default()
- .t_aco("Shah")
- .then_whitespace()
- .t_aco("Alam")
- )
- ]),
- "When referring to the states of Malaysia and their capitals, make sure to treat them as a proper noun.",
- dictionary.clone()))
-);
-
- group.add(
- "Countries",
- Box::new(ProperNounCapitalizationLinter::new(
- EitherPattern::new(vec![
- // Grouped country names
- // ... Guinea
- Box::new(
- SequencePattern::default()
- .then(EitherPattern::new(vec![
- Box::new(SequencePattern::aco("Equatorial")),
- Box::new(SequencePattern::aco("Papua").then_whitespace().t_aco("New")),
- ]))
- .then_whitespace()
- .t_aco("Guinea"),
- ),
- // ... Islands
- Box::new(
- SequencePattern::default()
- .then(EitherPattern::new(vec![
- Box::new(WordSet::new(&["Cayman", "Falkland", "Marshall", "Solomon"])),
- Box::new(
- SequencePattern::default()
- .then(EitherPattern::new(vec![
- Box::new(SequencePattern::aco("British")),
- Box::new(
- SequencePattern::aco("United")
- .then_whitespace()
- .t_aco("States"),
- ),
- ]))
- .then_whitespace()
- .t_aco("Virgin"),
- ),
- Box::new(
- SequencePattern::aco("Northern")
- .then_whitespace()
- .t_aco("Mariana"),
- ),
- ]))
- .then_whitespace()
- .t_aco("Islands"),
- ),
- // New ...
- Box::new(
- SequencePattern::aco("New")
- .then_whitespace()
- .then(WordSet::new(&["Caledonia", "Zealand"])),
- ),
- // Northern ...
- Box::new(
- SequencePattern::aco("Northern")
- .then_whitespace()
- .then(WordSet::new(&["Cyprus", "Ireland"])),
- ),
- // ... Republic
- Box::new(
- SequencePattern::default()
- .then(EitherPattern::new(vec![
- Box::new(
- SequencePattern::aco("Central")
- .then_whitespace()
- .t_aco("African"),
- ),
- Box::new(WordSet::new(&["Czech", "Dominican"])),
- ]))
- .then_whitespace()
- .t_aco("Republic"),
- ),
- // Saint ...
- Box::new(
- SequencePattern::aco("Saint")
- .then_whitespace()
- .then(WordSet::new(&["Helena", "Lucia", "Martin"])),
- ),
- // South ...
- Box::new(
- SequencePattern::aco("South")
- .then_whitespace()
- .then(WordSet::new(&["Africa", "Ossetia", "Sudan"])),
- ),
- // South Korea is under "Koreas"
- // One-off country names
- Box::new(
- SequencePattern::aco("American")
- .then_whitespace()
- .t_aco("Samoa"),
- ),
- Box::new(
- SequencePattern::aco("Antigua")
- .then_whitespace()
- .t_aco("and")
- .then_whitespace()
- .t_aco("Barbuda"),
- ),
- // United Arab Emirates is under "United Organizations"
- Box::new(
- SequencePattern::aco("Bosnia")
- .then_whitespace()
- .t_aco("and")
- .then_whitespace()
- .t_aco("Herzegovina"),
- ),
- Box::new(
- SequencePattern::aco("Burkina")
- .then_whitespace()
- .t_aco("Faso"),
- ),
- Box::new(
- SequencePattern::aco("Cape")
- .then_whitespace()
- .t_aco("Verde"),
- ),
- Box::new(
- SequencePattern::aco("Costa")
- .then_whitespace()
- .t_aco("Rica"),
- ),
- Box::new(
- SequencePattern::aco("Democratic")
- .then_whitespace()
- .t_aco("Republic")
- .then_whitespace()
- .t_aco("of")
- .then_whitespace()
- .t_aco("the")
- .then_whitespace()
- .t_aco("Congo"),
- ),
- Box::new(
- SequencePattern::aco("East")
- .then_whitespace()
- .t_aco("Timor"),
- ),
- Box::new(
- SequencePattern::aco("El")
- .then_whitespace()
- .t_aco("Salvador"),
- ),
- Box::new(
- SequencePattern::aco("French")
- .then_whitespace()
- .t_aco("Polynesia"),
- ),
- Box::new(SequencePattern::aco("Guinea").then_hyphen().t_aco("Bissau")),
- Box::new(
- SequencePattern::aco("Isle")
- .then_whitespace()
- .t_aco("of")
- .then_whitespace()
- .t_aco("Man"),
- ),
- Box::new(
- SequencePattern::aco("Ivory")
- .then_whitespace()
- .t_aco("Coast"),
- ),
- Box::new(
- SequencePattern::aco("North")
- .then_whitespace()
- .t_aco("Macedonia"),
- ),
- Box::new(
- SequencePattern::aco("Puerto")
- .then_whitespace()
- .t_aco("Rico"),
- ),
- Box::new(
- SequencePattern::aco("São")
- .then_whitespace()
- .t_aco("Tomé")
- .then_whitespace()
- .t_aco("and")
- .then_whitespace()
- .t_aco("Príncipe"),
- ),
- Box::new(
- SequencePattern::aco("Saudi")
- .then_whitespace()
- .t_aco("Arabia"),
- ),
- Box::new(
- SequencePattern::aco("Sierra")
- .then_whitespace()
- .t_aco("Leone"),
- ),
- Box::new(
- SequencePattern::aco("Sint")
- .then_whitespace()
- .t_aco("Maarten"),
- ),
- Box::new(SequencePattern::aco("Sri").then_whitespace().t_aco("Lanka")),
- Box::new(
- SequencePattern::aco("Trinidad")
- .then_whitespace()
- .t_aco("and")
- .then_whitespace()
- .t_aco("Tobago"),
- ),
- Box::new(
- SequencePattern::aco("Western")
- .then_whitespace()
- .t_aco("Sahara"),
- ),
- ]),
- "When referring to Countries, make sure to treat it as a proper noun.",
- dictionary.clone(),
- )),
- );
-
- group.add(
- "NationalCapitals",
- Box::new(ProperNounCapitalizationLinter::new(
- EitherPattern::new(vec![
- // Grouped capital names
- // ... City
- Box::new(
- SequencePattern::default()
- .then(EitherPattern::new(vec![
- Box::new(WordSet::new(&[
- "Belize",
- "Guatemala",
- "Kuwait",
- "Mexico",
- "Panama",
- "Vatican",
- ])),
- Box::new(
- SequencePattern::aco("Ho")
- .then_whitespace()
- .t_aco("Chi")
- .then_whitespace()
- .t_aco("Minh"),
- ),
- ]))
- .then_whitespace()
- .t_aco("City"),
- ),
- // Saint ...
- Box::new(
- SequencePattern::aco("Saint")
- .then_whitespace()
- .then(EitherPattern::new(vec![
- Box::new(
- SequencePattern::aco("Kitts")
- .then_whitespace()
- .t_aco("and")
- .then_whitespace()
- .t_aco("Nevis"),
- ),
- Box::new(
- SequencePattern::aco("Pierre")
- .then_whitespace()
- .t_aco("and")
- .then_whitespace()
- .t_aco("Miquelon"),
- ),
- Box::new(
- SequencePattern::aco("Vincent")
- .then_whitespace()
- .t_aco("and")
- .then_whitespace()
- .t_aco("the")
- .then_whitespace()
- .t_aco("Grenadines"),
- ),
- ])),
- ),
- // San ...
- Box::new(
- SequencePattern::aco("San")
- .then_whitespace()
- .then(WordSet::new(&["José", "Juan", "Marino", "Salvador"])),
- ),
- // St. ... TODO the period should be optional but this doesn't match even when it's not optional
- // Box::new(
- // SequencePattern::aco("St")
- // .then_period()
- // .then_whitespace()
- // .then(Box::new(WordSet::new(&["Helier", "John's", "Pierre"])))
- // ),
- // ... Town
- Box::new(
- SequencePattern::default()
- .then(WordSet::new(&[
- "Cape", "George", // Cayman Islands
- ]))
- .then_whitespace()
- .t_aco("Town"),
- ),
- // One-off capital names
- Box::new(SequencePattern::aco("Abu").then_whitespace().t_aco("Dhabi")),
- Box::new(
- SequencePattern::aco("Addis")
- .then_whitespace()
- .t_aco("Ababa"),
- ),
- // Andorra la Vella can't be done here because "la" must not be capitalized
- Box::new(
- SequencePattern::aco("Bandar")
- .then_whitespace()
- .t_aco("Seri")
- .then_whitespace()
- .t_aco("Begawan"),
- ),
- Box::new(
- SequencePattern::aco("Buenos")
- .then_whitespace()
- .t_aco("Aires"),
- ),
- // Dar es Salaam can't be done here because "es" must not be capitalized
- Box::new(
- SequencePattern::aco("Diego")
- .then_whitespace()
- .t_aco("Garcia"),
- ),
- // Kuala Lumpur is under "Malaysia"
- Box::new(SequencePattern::aco("La").then_whitespace().t_aco("Paz")),
- Box::new(SequencePattern::aco("New").then_whitespace().t_aco("Delhi")),
- Box::new(SequencePattern::aco("Pago").then_whitespace().t_aco("Pago")),
- Box::new(
- SequencePattern::aco("Phnom")
- .then_whitespace()
- .t_aco("Penh"),
- ),
- // Port-au-Prince can't be done here because "au" must not be capitalized
- Box::new(
- SequencePattern::aco("Port")
- .then_whitespace()
- .then(EitherPattern::new(vec![
- Box::new(WordSet::new(&["Louis", "Moresby", "Vila"])),
- Box::new(SequencePattern::aco("of").then_whitespace().t_aco("Spain")),
- ])),
- ),
- Box::new(SequencePattern::aco("Porto").then_hyphen().t_aco("Novo")),
- Box::new(
- SequencePattern::aco("Santo")
- .then_whitespace()
- .t_aco("Domingo"),
- ),
- Box::new(SequencePattern::aco("São").then_whitespace().t_aco("Tomé")),
- Box::new(
- SequencePattern::aco("The")
- .then_whitespace()
- .then(WordSet::new(&["Bahamas", "Hague"])),
- ),
- ]),
- "When referring to national capitals, make sure to treat it as a proper noun.",
- dictionary.clone(),
- )),
- );
-
- group.add(
- "ChineseCommunistParty",
- Box::new(ProperNounCapitalizationLinter::new(
- SequencePattern::aco("Chinese")
- .then_whitespace()
- .t_aco("Communist")
- .then_whitespace()
- .t_aco("Party"),
- "When referring to the political party, make sure to treat them as a proper noun.",
- dictionary.clone(),
- )),
- );
-
- group.add(
- "UnitedOrganizations",
- Box::new(ProperNounCapitalizationLinter::new(
- SequencePattern::default()
- .t_aco("United")
- .then_whitespace()
- .then(EitherPattern::new(vec![
- Box::new(SequencePattern::aco("Nations")),
- Box::new(SequencePattern::aco("States")),
- Box::new(SequencePattern::aco("Kingdom")),
- Box::new(SequencePattern::aco("Airlines")),
- Box::new(
- SequencePattern::default()
- .t_aco("Arab")
- .then_whitespace()
- .t_aco("Emirates")
- )
- ])),
- "When referring to national or international organizations, make sure to treat them as a proper noun.",
- dictionary.clone()))
-);
-
- group.add(
- "Holidays",
- Box::new(ProperNounCapitalizationLinter::new(
- EitherPattern::new(vec![
- Box::new(
- SequencePattern::default()
- .then(EitherPattern::new(vec![
- Box::new(WordSet::new(&[
- "Absolution",
- "Admission",
- "Alaska",
- "Anzac",
- "ANZAC",
- "Arbor",
- "Armistice",
- "Ascension",
- "Australia",
- "Ayurveda",
- "Bastille",
- "Bonifacio",
- "Boxing",
- "Canada",
- "Career",
- "Chewidden",
- "Christmas",
- "Class",
- "Columbus",
- "Commonwealth",
- "D",
- "Darwin",
- "Discovery",
- "Distaff",
- "Dominion",
- "Earth",
- "Easter",
- "Election",
- "Emancipation",
- "Empire",
- "Evolution",
- "Family",
- "Father's",
- "Fathers'",
- "Flag",
- "Forefathers'",
- "Foundation",
- "Freedom",
- "Galentine's",
- "Groundhog",
- "Gypsy",
- "Halloween",
- "Independence",
- "Invasion",
- "Ivy",
- "Jamhuri",
- "Jubilee",
- "Kamehameha",
- "Kenyatta",
- "Labor",
- "Labour",
- "Lady",
- "Land",
- "Lei",
- "Madaraka",
- "Mashujaa",
- "May",
- "Memorial",
- "Merdeka",
- "Midsummer",
- "Midsummer's",
- "Mother's",
- "Mothers'",
- "Nakba",
- "Nevada",
- "Occupation",
- "Parents",
- "Patrick's",
- "Patriots'",
- "Pi",
- "Picrous",
- "Pioneer",
- "Presidents'",
- "Remembrance",
- "Republic",
- "Restoration",
- "Rizal",
- "Roc",
- "Rock",
- "Seward's",
- "Singles'",
- "Statehood",
- "Tax",
- "Thanksgiving",
- "Treason",
- "Ulster",
- "Valentine's",
- "VE",
- "VJ",
- "VP",
- "Veterans",
- "Victoria",
- "Victory",
- "Waffle",
- "Waitangi",
- "Wattle",
- "White",
- "Wren",
- "Years",
- "Year's",
- "Youth",
- ])),
- Box::new(
- SequencePattern::default()
- .t_aco("National")
- .then_whitespace()
- .t_aco("Freedom"),
- ),
- Box::new(
- SequencePattern::default()
- .t_aco("All")
- .then_whitespace()
- .t_aco("Saints"),
- ),
- Box::new(
- SequencePattern::default()
- .t_aco("All")
- .then_whitespace()
- .t_aco("Souls"),
- ),
- ]))
- .then_whitespace()
- .t_aco("Day"),
- ),
- Box::new(
- SequencePattern::default()
- .t_aco("Black")
- .then_whitespace()
- .t_aco("Friday"),
- ),
- Box::new(
- SequencePattern::default()
- .t_aco("Cyber")
- .then_whitespace()
- .t_aco("Monday"),
- ),
- ]),
- "When referring to holidays, make sure to treat them as a proper noun.",
- dictionary.clone(),
- )),
- );
-
- group.add(
- "AmazonNames",
- Box::new(ProperNounCapitalizationLinter::new(
- SequencePattern::default()
- .t_aco("Amazon")
- .then_whitespace()
- .then(EitherPattern::new(vec![
- Box::new(
- SequencePattern::default()
- .t_aco("Shopping")
- ),
- Box::new(
- SequencePattern::default()
- .t_aco("Web")
- .then_whitespace()
- .t_aco("Services")
- ),
- Box::new(
- SequencePattern::default()
- .t_aco("Lambda")
- ),
- Box::new(
- SequencePattern::default()
- .t_aco("RDS")
- ),
- Box::new(
- SequencePattern::default()
- .t_aco("DynamoDB")
- ),
- Box::new(
- SequencePattern::default()
- .t_aco("SageMaker")
- ),
- Box::new(
- SequencePattern::default()
- .t_aco("Rekognition")
- ),
- Box::new(
- SequencePattern::default()
- .t_aco("CloudFront")
- ),
- Box::new(
- SequencePattern::default()
- .t_aco("ECS")
- ),
- Box::new(
- SequencePattern::default()
- .t_aco("EKS")
- ),
- Box::new(
- SequencePattern::default()
- .t_aco("CloudWatch")
- ),
- Box::new(
- SequencePattern::default()
- .t_aco("IAM")
- ),
- Box::new(
- SequencePattern::default()
- .t_aco("Prime")
- ),
- Box::new(
- SequencePattern::default()
- .t_aco("Kindle")
- )
- ])),
- "When referring to the various products of Amazon.com, make sure to treat them as a proper noun.",
- dictionary.clone()))
-);
-
- group.add(
- "GoogleNames",
- Box::new(ProperNounCapitalizationLinter::new(
- SequencePattern::default()
- .t_aco("Google")
- .then_whitespace()
- .then(WordSet::new(&[
- "Search",
- "Cloud",
- "Maps",
- "Docs",
- "Sheets",
- "Slides",
- "Drive",
- "Meet",
- "Gmail",
- "Calendar",
- "Chrome",
- "ChromeOS",
- "Android",
- "Play",
- "Bard",
- "Gemini",
- "YouTube",
- "Photos",
- "Analytics",
- "AdSense",
- "Pixel",
- "Nest",
- "Workspace",
- ])),
- "When referring to Google products and services, make sure to treat them as proper nouns."
- ,dictionary.clone()))
- );
-
- group.add(
- "AzureNames",
- Box::new(ProperNounCapitalizationLinter::new(
- SequencePattern::default()
- .t_aco("Azure")
- .then_whitespace()
- .then(EitherPattern::new(vec![
- Box::new(SequencePattern::aco("DevOps")),
- Box::new(SequencePattern::aco("Functions")),
- Box::new(
- SequencePattern::default()
- .t_aco("Cosmos")
- .then_whitespace()
- .t_aco("DB"),
- ),
- Box::new(
- SequencePattern::default()
- .t_aco("SQL")
- .then_whitespace()
- .t_aco("Database"),
- ),
- Box::new(
- SequencePattern::default()
- .t_aco("Kubernetes")
- .then_whitespace()
- .t_aco("Service"),
- ),
- Box::new(
- SequencePattern::default()
- .t_aco("Virtual")
- .then_whitespace()
- .t_aco("Machines"),
- ),
- Box::new(SequencePattern::aco("Monitor")),
- Box::new(SequencePattern::aco("Storage")),
- Box::new(
- SequencePattern::default()
- .t_aco("Active")
- .then_whitespace()
- .t_aco("Directory"),
- ),
- Box::new(
- SequencePattern::default()
- .t_aco("App")
- .then_whitespace()
- .t_aco("Service"),
- ),
- Box::new(
- SequencePattern::default()
- .t_aco("Key")
- .then_whitespace()
- .t_aco("Vault"),
- ),
- Box::new(
- SequencePattern::default()
- .t_aco("Cognitive")
- .then_whitespace()
- .t_aco("Services"),
- ),
- Box::new(
- SequencePattern::default()
- .t_aco("Service")
- .then_whitespace()
- .t_aco("Bus"),
- ),
- Box::new(
- SequencePattern::default()
- .t_aco("Event")
- .then_whitespace()
- .t_aco("Hub"),
- ),
- ])),
- "When referring to Azure cloud services, make sure to treat them as proper nouns.",
- dictionary.clone(),
- )),
- );
-
- group.add(
- "MicrosoftNames",
- Box::new(ProperNounCapitalizationLinter::new(
- SequencePattern::default()
- .t_aco("Microsoft")
- .then_whitespace()
- .then(EitherPattern::new(vec![
- Box::new(WordSet::new(&[
- "Windows",
- "Office",
- "Teams",
- "Excel",
- "PowerPoint",
- "Word",
- "Outlook",
- "OneDrive",
- "SharePoint",
- "Xbox",
- "Surface",
- "Edge",
- "Bing",
- "Dynamics",
- ])),
- Box::new(
- SequencePattern::default()
- .t_aco("Visual")
- .then_whitespace()
- .t_aco("Studio")
- )
- ])),
- "When referring to Microsoft products and services, make sure to treat them as proper nouns.",
- dictionary.clone()))
-);
-
- group.add(
- "AppleNames",
- Box::new(ProperNounCapitalizationLinter::new(
- SequencePattern::default()
- .t_aco("Apple")
- .then_whitespace()
- .then(EitherPattern::new(vec![
- Box::new(WordSet::new(&[
- "iPhone", "iPad", "iMac", "MacBook", "Watch", "TV", "Music", "Arcade",
- "iCloud", "Safari", "HomeKit", "CarPlay",
- ])),
- Box::new(
- SequencePattern::aco("MacBook")
- .then_whitespace()
- .t_aco("Pro")
- ),
- Box::new(
- SequencePattern::aco("MacBook")
- .then_whitespace()
- .t_aco("Air")
- ),
- Box::new(SequencePattern::aco("Mac").then_whitespace().t_aco("Pro")),
- Box::new(SequencePattern::aco("Mac").then_whitespace().t_aco("Mini")),
- Box::new(SequencePattern::aco("AirPods")),
- Box::new(
- SequencePattern::aco("AirPods")
- .then_whitespace()
- .t_aco("Pro")
- ),
- Box::new(
- SequencePattern::aco("AirPods")
- .then_whitespace()
- .t_aco("Max")
- ),
- Box::new(
- SequencePattern::default()
- .t_aco("Vision")
- .then_whitespace()
- .t_aco("Pro")
- )
- ])),
- "When referring to Apple products and services, make sure to treat them as proper nouns.",
- dictionary.clone()))
-
- );
-
- group.add(
- "MetaNames",
- Box::new(ProperNounCapitalizationLinter::new(SequencePattern::aco("Meta")
- .then_whitespace()
- .then(EitherPattern::new(vec![
- Box::new(WordSet::new(&[
- "Oculus", "Portals", "Quest", "Gaming", "Horizon",
- ])),
- Box::new(
- SequencePattern::default()
- .t_aco("Reality")
- .then_whitespace()
- .t_aco("Labs")
- ),
- ])),
- "When referring to Meta products and services, make sure to treat them as proper nouns."
- , dictionary.clone()
- ))
- );
-
- group.add(
- "JetpackNames",
- Box::new(ProperNounCapitalizationLinter::new(
- SequencePattern::default()
- .t_aco("Jetpack")
- .then_whitespace()
- .then(EitherPattern::new(vec![
- Box::new(
- SequencePattern::default()
- .t_aco("VaultPress")
- .then_whitespace()
- .t_aco("Backup"),
- ),
- Box::new(SequencePattern::default().t_aco("VaultPress")),
- Box::new(SequencePattern::default().t_aco("Scan")),
- Box::new(
- SequencePattern::default()
- .t_aco("Akismet")
- .then_whitespace()
- .t_aco("Anti-spam"),
- ),
- Box::new(SequencePattern::default().t_aco("Stats")),
- Box::new(SequencePattern::default().t_aco("Social")),
- Box::new(SequencePattern::default().t_aco("Blaze")),
- Box::new(
- SequencePattern::default()
- .t_aco("AI")
- .then_whitespace()
- .t_aco("Assistant"),
- ),
- Box::new(
- SequencePattern::default()
- .t_aco("Site")
- .then_whitespace()
- .t_aco("Search"),
- ),
- Box::new(SequencePattern::default().t_aco("Boost")),
- Box::new(SequencePattern::default().t_aco("VideoPress")),
- Box::new(
- SequencePattern::default()
- .t_aco("For")
- .then_whitespace()
- .t_aco("Agencies"),
- ),
- Box::new(SequencePattern::default().t_aco("CRM")),
- ])),
- "Ensure proper capitalization of Jetpack-related terms.",
- dictionary.clone(),
- )),
- );
+#[derive(Serialize, Deserialize)]
+struct RuleEntry {
+ canonical: Vec<String>,
+ description: String,
+}
- group.add(
- "TumblrNames",
- Box::new(ProperNounCapitalizationLinter::new(
- SequencePattern::default()
- .t_aco("Tumblr")
- .then_whitespace()
- .then(EitherPattern::new(vec![
- Box::new(SequencePattern::default().t_aco("Blaze")),
- Box::new(SequencePattern::default().t_aco("Pro")),
- Box::new(SequencePattern::default().t_aco("Live")),
- Box::new(SequencePattern::default().t_aco("Ads")),
- Box::new(SequencePattern::default().t_aco("Communities")),
- Box::new(SequencePattern::default().t_aco("Shop")),
- Box::new(SequencePattern::default().t_aco("Dashboard")),
- ])),
- "Ensure proper capitalization of Tumblr-related terms.",
- dictionary.clone(),
- )),
- );
+/// For the time being, this panics on invalid JSON.
+/// Do not use with user provided JSON.
+fn lint_group_from_json(json: &str, dictionary: Arc<impl Dictionary + 'static>) -> LintGroup {
+ let mut group = LintGroup::empty();
- group.add(
- "PocketCastsNames",
- Box::new(ProperNounCapitalizationLinter::new(
- EitherPattern::new(vec![
- Box::new(
- SequencePattern::default()
- .t_aco("Pocket")
- .then_whitespace()
- .t_aco("Casts"),
- ),
- Box::new(
- SequencePattern::default()
- .t_aco("Pocket")
- .then_whitespace()
- .t_aco("Casts")
- .then_whitespace()
- .t_aco("Plus"),
- ),
- ]),
- "Ensure proper capitalization of Pocket Casts and Pocket Casts Plus as brand names.",
- dictionary.clone(),
- )),
- );
+ let rules: HashMap<String, RuleEntry> = serde_json::from_str(json).unwrap();
- group.add(
- "DayOneNames",
- Box::new(ProperNounCapitalizationLinter::new(
- EitherPattern::new(vec![
- Box::new(
- SequencePattern::default()
- .t_aco("Day")
- .then_whitespace()
- .t_aco("One"),
- ),
- Box::new(
- SequencePattern::default()
- .t_aco("Day")
- .then_whitespace()
- .t_aco("One")
- .then_whitespace()
- .t_aco("Premium"),
- ),
- ]),
- "Ensure proper capitalization of Day One and Day One Premium as brand names.",
- dictionary.clone(),
- )),
- );
+ for (key, rule) in rules.into_iter() {
+ group.add(
+ key,
+ Box::new(ProperNounCapitalizationLinter::new_strs(
+ rule.canonical,
+ rule.description,
+ dictionary.clone(),
+ )),
+ );
+ }
group.set_all_rules_to(Some(true));
group
}
+pub fn lint_group(dictionary: Arc<impl Dictionary + 'static>) -> LintGroup {
+ lint_group_from_json(include_str!("../../proper_noun_rules.json"), dictionary)
+}
+
#[cfg(test)]
mod tests {
use crate::{
@@ -1364,8 +283,6 @@ mod tests {
assert_suggestion_result("arctic ocean", lint_group(dictionary), "Arctic Ocean");
}
- // Lowercase tests for seas
-
#[test]
fn test_mediterranean_sea_lowercase() {
let dictionary = FstDictionary::curated();
diff --git a/harper-core/src/patterns/mod.rs b/harper-core/src/patterns/mod.rs
index 738397930..ad7ebc677 100644
--- a/harper-core/src/patterns/mod.rs
+++ b/harper-core/src/patterns/mod.rs
@@ -21,6 +21,7 @@ mod invert;
mod is_not_title_case;
mod naive_pattern_group;
mod noun_phrase;
+mod pattern_map;
mod repeating_pattern;
mod sequence_pattern;
mod similar_to_phrase;
@@ -45,6 +46,7 @@ pub use invert::Invert;
pub use is_not_title_case::IsNotTitleCase;
pub use naive_pattern_group::NaivePatternGroup;
pub use noun_phrase::NounPhrase;
+pub use pattern_map::PatternMap;
pub use repeating_pattern::RepeatingPattern;
pub use sequence_pattern::SequencePattern;
pub use similar_to_phrase::SimilarToPhrase;
diff --git a/harper-core/src/patterns/pattern_map.rs b/harper-core/src/patterns/pattern_map.rs
new file mode 100644
index 000000000..89d3fd0bf
--- /dev/null
+++ b/harper-core/src/patterns/pattern_map.rs
@@ -0,0 +1,80 @@
+use crate::Token;
+
+use super::Pattern;
+
+/// A map from [`Pattern`] to arbitrary data.
+///
+/// When used as a [`Pattern`] in of itself, it simply iterates through
+/// all contained patterns, returning the first match found.
+/// You should not assume this search is deterministic.
+///
+/// If you'd like to use this structure in a [`PatternLinter`](crate::linting::PatternLinter), you may want to provide
+/// the map as the search pattern, then use a pattern look-up once more to determine
+/// the corresponding key.
+pub struct PatternMap<T>
+where
+ T: Send + Sync,
+{
+ rows: Vec<Row<T>>,
+}
+
+struct Row<T>
+where
+ T: Send + Sync,
+{
+ pub key: Box<dyn Pattern>,
+ pub element: T,
+}
+
+impl<T> Default for PatternMap<T>
+where
+ T: Send + Sync,
+{
+ fn default() -> Self {
+ Self {
+ rows: Default::default(),
+ }
+ }
+}
+
+impl<T> PatternMap<T>
+where
+ T: Send + Sync,
+{
+ pub fn insert(&mut self, pattern: impl Pattern + 'static, value: T) {
+ self.rows.push(Row {
+ key: Box::new(pattern),
+ element: value,
+ });
+ }
+
+ /// Lookup the corresponding value for the given map.
+ pub fn lookup(&self, tokens: &[Token], source: &[char]) -> Option<&T> {
+ for row in &self.rows {
+ let len = row.key.matches(tokens, source);
+
+ if len != 0 {
+ return Some(&row.element);
+ }
+ }
+
+ None
+ }
+}
+
+impl<T> Pattern for PatternMap<T>
+where
+ T: Send + Sync,
+{
+ fn matches(&self, tokens: &[Token], source: &[char]) -> usize {
+ for row in &self.rows {
+ let len = row.key.matches(tokens, source);
+
+ if len != 0 {
+ return len;
+ }
+ }
+
+ 0
+ }
+}