feat: created CharString and added config for diagnosticSeverity

Author: elijah-potter

Cargo.lock

@@ -31,4 +31,3 @@ Harper is even small enough to load via [WebAssembly.](https://writewithharper.c
 
 If you want to use Harper on your machine, you will want to look at the [documentation for
 `harper-ls`](./harper-ls/README.md), the Language Server Protocol implementation.
-

README.md

@@ -49622,3 +49622,4 @@ URI/s
 suprachiasmatic
 cannabinoids
 eudaimonia
+whitespace

harper-core/dictionary.dict

@@ -0,0 +1,17 @@
+use smallvec::SmallVec;
+
+pub type CharString = SmallVec<[char; 6]>;
+
+pub trait CharStringExt {
+    fn to_lower(&self) -> CharString;
+}
+
+impl CharStringExt for [char] {
+    fn to_lower(&self) -> CharString {
+        let mut out = CharString::with_capacity(self.len());
+
+        out.extend(self.iter().flat_map(|v| v.to_lowercase()));
+
+        out
+    }
+}

harper-core/src/char_string.rs

@@ -1,5 +1,6 @@
 #![allow(dead_code)]
 
+mod char_string;
 mod document;
 mod lexing;
 mod linting;
@@ -8,6 +9,7 @@ mod span;
 mod spell;
 mod token;
 
+pub use char_string::{CharString, CharStringExt};
 pub use document::Document;
 pub use linting::{Lint, LintGroup, LintGroupConfig, LintKind, Linter, Suggestion};
 pub use span::Span;

harper-core/src/lib.rs

@@ -1,6 +1,6 @@
 use itertools::Itertools;
 
-use crate::{Document, Lint, LintKind, Linter, Suggestion, TokenStringExt};
+use crate::{CharStringExt, Document, Lint, LintKind, Linter, Suggestion, TokenStringExt};
 
 #[derive(Debug, Default)]
 pub struct AnA;
@@ -23,7 +23,7 @@ impl Linter for AnA {
                 continue;
             };
 
-            let should_be_a_an = !starts_with_vowel(chars_second);
+            let should_be_a_an = !starts_with_vowel(&chars_second.to_lower());
 
             if a_an != should_be_a_an {
                 let replacement = match a_an {

harper-core/src/linting/an_a.rs

@@ -1,10 +1,20 @@
-use crate::spell::DictWord;
-use crate::{Document, Lint, LintKind, Linter, Punctuation, Span, Suggestion, Token, TokenKind};
+use crate::{
+    CharString,
+    Document,
+    Lint,
+    LintKind,
+    Linter,
+    Punctuation,
+    Span,
+    Suggestion,
+    Token,
+    TokenKind
+};
 
 #[derive(Debug, PartialEq, PartialOrd, Clone)]
 struct PatternToken {
     kind: TokenKind,
-    content: Option<DictWord>
+    content: Option<CharString>
 }
 
 impl PatternToken {

harper-core/src/linting/matcher.rs

@@ -1,14 +1,13 @@
 use hashbrown::HashSet;
 
 use super::{Lint, LintKind, Linter};
-use crate::spell::DictWord;
 use crate::token::{Token, TokenKind, TokenStringExt};
-use crate::{Document, Span, Suggestion};
+use crate::{CharString, Document, Span, Suggestion};
 
 #[derive(Debug, Clone)]
 pub struct RepeatedWords {
     /// The set of words that can be considered for repetition checking.
-    set: HashSet<DictWord>
+    set: HashSet<CharString>
 }
 
 impl RepeatedWords {

harper-core/src/linting/repeated_words.rs

@@ -4,7 +4,8 @@ use smallvec::{SmallVec, ToSmallVec};
 
 use super::dictionary::Dictionary;
 use super::hunspell::{parse_default_attribute_list, parse_default_word_list};
-use super::{seq_to_normalized, DictWord};
+use super::seq_to_normalized;
+use crate::CharString;
 
 /// A full, fat dictionary.
 /// All of the elements are stored in-memory.
@@ -16,13 +17,13 @@ pub struct FullDictionary {
     /// This is likely due to increased locality :shrug:.
     ///
     /// This list is sorted by word length (i.e. the shortest words are first).
-    words: Vec<DictWord>,
+    words: Vec<CharString>,
     /// A lookup list for each word length.
     /// Each index of this list will return the first index of [`Self::words`]
     /// that has a word whose index is that length.
     word_len_starts: Vec<usize>,
     /// All English words
-    word_set: HashSet<DictWord>
+    word_set: HashSet<CharString>
 }
 
 fn uncached_inner_new() -> FullDictionary {
@@ -81,7 +82,7 @@ impl FullDictionary {
     /// list. NOTE: This function will sort the original word list by its
     /// length. If the word list's order is changed after creating the
     /// lookup, it will no longer be valid.
-    fn create_len_starts(words: &mut [DictWord]) -> Vec<usize> {
+    fn create_len_starts(words: &mut [CharString]) -> Vec<usize> {
         words.sort_by_key(|a| a.len());
         let mut word_len_starts = vec![0, 0];
 

harper-core/src/spell/full_dictionary.rs

@@ -6,8 +6,7 @@ use smallvec::ToSmallVec;
 use super::matcher::Matcher;
 use super::word_list::MarkedWord;
 use super::Error;
-use crate::spell::DictWord;
-use crate::Span;
+use crate::{CharString, Span};
 
 #[derive(Debug, Clone)]
 struct AffixReplacement {
@@ -111,7 +110,7 @@ impl AttributeList {
     /// Will append to the given `dest`;
     ///
     /// In the future, I want to make this function cleaner and faster.
-    pub fn expand_marked_word(&self, word: MarkedWord, dest: &mut Vec<DictWord>) {
+    pub fn expand_marked_word(&self, word: MarkedWord, dest: &mut Vec<CharString>) {
         dest.reserve(word.attributes.len() + 1);
 
         let start_len = dest.len();
@@ -173,7 +172,7 @@ impl AttributeList {
     pub fn expand_marked_words(
         &self,
         words: impl IntoIterator<Item = MarkedWord>,
-        dest: &mut Vec<DictWord>
+        dest: &mut Vec<CharString>
     ) {
         for word in words {
             self.expand_marked_word(word, dest);
@@ -184,7 +183,7 @@ impl AttributeList {
         replacement: &AffixReplacement,
         letters: &[char],
         suffix: bool
-    ) -> Option<DictWord> {
+    ) -> Option<CharString> {
         if replacement.condition.len() > letters.len() {
             return None;
         }
@@ -199,7 +198,7 @@ impl AttributeList {
 
         if replacement.condition.matches(target_segment) {
             let mut replaced_segment = letters.to_smallvec();
-            let mut remove: DictWord = replacement.remove.to_smallvec();
+            let mut remove: CharString = replacement.remove.to_smallvec();
 
             if !suffix {
                 replaced_segment.reverse();