feat(dict): Override builtin dictionary

Sometimes you just have to live with a typo or its done intentionally
(like weird company names).  With this commit, a user can now identifier
blessed identifiers and words.

This is ostly what is needed for #9 but sometimes people will have
common typos that they'll want to provide corrections for.

docs/reference.md

@@ -29,4 +29,6 @@ Configuration is read from the following (in precedence order)
 | default.identifier-include-digits   | \-   | bool   | Allow identifiers to include digits, in addition to letters. |
 | default.identifier-leading-chars    | \-   | string | Allow identifiers to start with one of these characters. |
 | default.identifier-include-chars    | \-   | string | Allow identifiers to include these characters. |
-| default.locale         | \-                | en, en-us, en-gb, en-ca, en-au   |  |
+| default.locale         | \-                | en, en-us, en-gb, en-ca, en-au   | English dialect to correct to. |
+| default.extend-valid-identifiers    | \-   | list of strings | Identifiers to presume as correct, skipping spell checking.  This extends the list when layering configuration, rather than replacing it. |
+| default.extend-valid-words          | \-   | list of strings | Words to presume as correct, skipping spell checking.  This extends the list when layering configuration, rather than replacing it. |

src/config.rs

@@ -96,6 +96,14 @@ pub trait FileSource {
     fn locale(&self) -> Option<Locale> {
         None
     }
+
+    fn extend_valid_identifiers(&self) -> &[String] {
+        &[]
+    }
+
+    fn extend_valid_words(&self) -> &[String] {
+        &[]
+    }
 }
 
 #[derive(Debug, Clone, Default, serde::Serialize, serde::Deserialize)]
@@ -287,6 +295,8 @@ pub struct FileConfig {
     pub identifier_include_digits: Option<bool>,
     pub identifier_include_chars: Option<String>,
     pub locale: Option<Locale>,
+    pub extend_valid_identifiers: Vec<String>,
+    pub extend_valid_words: Vec<String>,
 }
 
 impl FileConfig {
@@ -315,6 +325,10 @@ impl FileConfig {
         if let Some(source) = source.locale() {
             self.locale = Some(source);
         }
+        self.extend_valid_identifiers
+            .extend(source.extend_valid_identifiers().iter().cloned());
+        self.extend_valid_words
+            .extend(source.extend_valid_words().iter().cloned());
     }
 
     pub fn check_filename(&self) -> bool {
@@ -348,6 +362,14 @@ impl FileConfig {
     pub fn locale(&self) -> Locale {
         self.locale.unwrap_or_default()
     }
+
+    pub fn extend_valid_identifiers(&self) -> &[String] {
+        self.extend_valid_identifiers.as_slice()
+    }
+
+    pub fn extend_valid_words(&self) -> &[String] {
+        self.extend_valid_words.as_slice()
+    }
 }
 
 impl FileSource for FileConfig {
@@ -382,6 +404,14 @@ impl FileSource for FileConfig {
     fn locale(&self) -> Option<Locale> {
         self.locale
     }
+
+    fn extend_valid_identifiers(&self) -> &[String] {
+        self.extend_valid_identifiers.as_slice()
+    }
+
+    fn extend_valid_words(&self) -> &[String] {
+        self.extend_valid_words.as_slice()
+    }
 }
 
 fn find_project_file(dir: std::path::PathBuf, name: &str) -> Option<std::path::PathBuf> {

src/dict.rs

@@ -1,4 +1,5 @@
 use std::borrow::Cow;
+use std::collections::HashSet;
 
 use unicase::UniCase;
 
@@ -130,6 +131,49 @@ fn case_correct(correction: &str, case: Case) -> Cow<'_, str> {
     }
 }
 
+pub struct Override<'i, 'w, D> {
+    valid_identifiers: HashSet<&'i str>,
+    valid_words: HashSet<unicase::UniCase<&'w str>>,
+    inner: D,
+}
+
+impl<'i, 'w, D: typos::Dictionary> Override<'i, 'w, D> {
+    pub fn new(inner: D) -> Self {
+        Self {
+            valid_identifiers: Default::default(),
+            valid_words: Default::default(),
+            inner,
+        }
+    }
+
+    pub fn valid_identifiers<I: Iterator<Item = &'i str>>(&mut self, valid_identifiers: I) {
+        self.valid_identifiers = valid_identifiers.collect();
+    }
+
+    pub fn valid_words<I: Iterator<Item = &'w str>>(&mut self, valid_words: I) {
+        self.valid_words = valid_words.map(UniCase::new).collect();
+    }
+}
+
+impl<'i, 'w, D: typos::Dictionary> typos::Dictionary for Override<'i, 'w, D> {
+    fn correct_ident<'s, 't>(&'s self, ident: typos::tokens::Identifier<'t>) -> Vec<Cow<'s, str>> {
+        if self.valid_identifiers.contains(ident.token()) {
+            Vec::new()
+        } else {
+            self.inner.correct_ident(ident)
+        }
+    }
+
+    fn correct_word<'s, 't>(&'s self, word: typos::tokens::Word<'t>) -> Vec<Cow<'s, str>> {
+        let w = UniCase::new(word.token());
+        if self.valid_words.contains(&w) {
+            Vec::new()
+        } else {
+            self.inner.correct_word(word)
+        }
+    }
+}
+
 #[cfg(test)]
 mod test {
     use super::*;

src/main.rs

@@ -57,6 +57,21 @@ fn run() -> Result<i32, anyhow::Error> {
             .build();
 
         let dictionary = crate::dict::BuiltIn::new(config.default.locale());
+        let mut dictionary = crate::dict::Override::new(dictionary);
+        dictionary.valid_identifiers(
+            config
+                .default
+                .extend_valid_identifiers()
+                .iter()
+                .map(|s| s.as_str()),
+        );
+        dictionary.valid_words(
+            config
+                .default
+                .extend_valid_words()
+                .iter()
+                .map(|s| s.as_str()),
+        );
 
         let mut settings = typos::checks::TyposSettings::new();
         settings