harper_core/
token_kind.rs

1use harper_brill::UPOS;
2use is_macro::Is;
3use serde::{Deserialize, Serialize};
4
5use crate::{DictWordMetadata, Number, Punctuation, Quote, TokenKind::Word};
6
7/// Generate wrapper code to pass a function call to the inner [`DictWordMetadata`],  
8/// if the token is indeed a word, while also emitting method-level documentation.
9macro_rules! delegate_to_metadata {
10    ($($method:ident),* $(,)?) => {
11        $(
12            #[doc = concat!(
13                "Delegates to [`DictWordMetadata::",
14                stringify!($method),
15                "`] when this token is a word.\n\n",
16                "Returns `false` if the token is not a word."
17            )]
18            pub fn $method(&self) -> bool {
19                let Word(Some(metadata)) = self else {
20                    return false;
21                };
22                metadata.$method()
23            }
24        )*
25    };
26}
27
28/// The parsed value of a [`Token`](crate::Token).
29/// Has a variety of queries available.
30/// If there is a query missing, it may be easy to implement by just calling the
31/// `delegate_to_metadata` macro.
32#[derive(Debug, Is, Clone, Serialize, Deserialize, Default, PartialOrd, Hash, Eq, PartialEq)]
33#[serde(tag = "kind", content = "value")]
34pub enum TokenKind {
35    /// `None` if the word does not exist in the dictionary.
36    Word(Option<DictWordMetadata>),
37    Punctuation(Punctuation),
38    Decade,
39    Number(Number),
40    /// A sequence of " " spaces.
41    Space(usize),
42    /// A sequence of "\n" newlines
43    Newline(usize),
44    EmailAddress,
45    Url,
46    Hostname,
47    /// A special token used for things like inline code blocks that should be
48    /// ignored by all linters.
49    #[default]
50    Unlintable,
51    ParagraphBreak,
52    Regexish,
53}
54
55impl TokenKind {
56    // DictWord metadata delegation methods grouped by part of speech
57    delegate_to_metadata! {
58        // Nominal methods (nouns and pronouns)
59        is_nominal,
60        is_noun,
61        is_pronoun,
62        is_proper_noun,
63        is_singular_nominal,
64        is_plural_nominal,
65        is_possessive_nominal,
66        is_non_plural_nominal,
67        is_singular_noun,
68        is_plural_noun,
69        is_non_plural_noun,
70        is_countable_noun,
71        is_non_countable_noun,
72        is_mass_noun,
73        is_mass_noun_only,
74        is_non_mass_noun,
75        is_singular_pronoun,
76        is_plural_pronoun,
77        is_non_plural_pronoun,
78        is_reflexive_pronoun,
79        is_personal_pronoun,
80        is_first_person_singular_pronoun,
81        is_first_person_plural_pronoun,
82        is_second_person_pronoun,
83        is_third_person_pronoun,
84        is_third_person_singular_pronoun,
85        is_third_person_plural_pronoun,
86        is_subject_pronoun,
87        is_object_pronoun,
88        is_possessive_noun,
89        is_possessive_pronoun,
90
91        // Verb methods
92        is_verb,
93        is_auxiliary_verb,
94        is_linking_verb,
95        is_verb_lemma,
96        is_verb_simple_past_form,
97        is_verb_past_participle_form,
98        is_verb_progressive_form,
99        is_verb_third_person_singular_present_form,
100
101        // Adjective methods
102        is_adjective,
103        is_comparative_adjective,
104        is_superlative_adjective,
105        is_positive_adjective,
106
107        // Adverb methods
108        is_adverb,
109
110        // Determiner methods
111        is_determiner,
112        is_demonstrative_determiner,
113        is_possessive_determiner,
114        is_quantifier,
115        is_non_quantifier_determiner,
116
117        // Conjunction methods
118        is_conjunction,
119
120        // Generic word methods
121        is_swear,
122        is_likely_homograph,
123
124        // Orthography methods
125        is_lowercase,
126        is_titlecase,
127        is_allcaps,
128        is_lower_camel,
129        is_upper_camel,
130        is_apostrophized,
131
132        is_roman_numerals
133    }
134
135    // DictWord metadata delegation methods not generated by macro
136    pub fn is_preposition(&self) -> bool {
137        let Word(Some(metadata)) = self else {
138            return false;
139        };
140        metadata.preposition
141    }
142
143    // Generic word is-methods
144
145    pub fn is_common_word(&self) -> bool {
146        let Word(Some(metadata)) = self else {
147            return true;
148        };
149        metadata.common
150    }
151
152    /// Checks whether the token is a member of a nominal phrase.
153    pub fn is_np_member(&self) -> bool {
154        let Word(Some(metadata)) = self else {
155            return false;
156        };
157        metadata.np_member.unwrap_or(false)
158    }
159
160    /// Checks whether a word token is out-of-vocabulary (not found in the dictionary).
161    ///
162    /// Returns `true` if the token is a word that was not found in the dictionary,
163    /// `false` if the token is a word found in the dictionary or is not a word token.
164    pub fn is_oov(&self) -> bool {
165        matches!(self, TokenKind::Word(None))
166    }
167
168    // Punctuation and symbol is-methods
169
170    pub fn is_open_square(&self) -> bool {
171        matches!(self, TokenKind::Punctuation(Punctuation::OpenSquare))
172    }
173
174    pub fn is_close_square(&self) -> bool {
175        matches!(self, TokenKind::Punctuation(Punctuation::CloseSquare))
176    }
177
178    pub fn is_open_round(&self) -> bool {
179        matches!(self, TokenKind::Punctuation(Punctuation::OpenRound))
180    }
181
182    pub fn is_close_round(&self) -> bool {
183        matches!(self, TokenKind::Punctuation(Punctuation::CloseRound))
184    }
185
186    pub fn is_pipe(&self) -> bool {
187        matches!(self, TokenKind::Punctuation(Punctuation::Pipe))
188    }
189
190    pub fn is_currency(&self) -> bool {
191        matches!(self, TokenKind::Punctuation(Punctuation::Currency(..)))
192    }
193
194    pub fn is_ellipsis(&self) -> bool {
195        matches!(self, TokenKind::Punctuation(Punctuation::Ellipsis))
196    }
197
198    pub fn is_hyphen(&self) -> bool {
199        matches!(self, TokenKind::Punctuation(Punctuation::Hyphen))
200    }
201
202    pub fn is_quote(&self) -> bool {
203        matches!(self, TokenKind::Punctuation(Punctuation::Quote(_)))
204    }
205
206    pub fn is_apostrophe(&self) -> bool {
207        matches!(self, TokenKind::Punctuation(Punctuation::Apostrophe))
208    }
209
210    pub fn is_period(&self) -> bool {
211        matches!(self, TokenKind::Punctuation(Punctuation::Period))
212    }
213
214    pub fn is_at(&self) -> bool {
215        matches!(self, TokenKind::Punctuation(Punctuation::At))
216    }
217
218    pub fn is_comma(&self) -> bool {
219        matches!(self, TokenKind::Punctuation(Punctuation::Comma))
220    }
221
222    pub fn is_semicolon(&self) -> bool {
223        matches!(self, TokenKind::Punctuation(Punctuation::Semicolon))
224    }
225
226    pub fn is_ampersand(&self) -> bool {
227        matches!(self, TokenKind::Punctuation(Punctuation::Ampersand))
228    }
229
230    pub fn is_slash(&self) -> bool {
231        matches!(self, TokenKind::Punctuation(Punctuation::ForwardSlash))
232    }
233
234    // Miscellaneous is-methods
235
236    /// Checks whether a token is word-like--meaning it is more complex than punctuation and can
237    /// hold semantic meaning in the way a word does.
238    pub fn is_word_like(&self) -> bool {
239        matches!(
240            self,
241            TokenKind::Word(..)
242                | TokenKind::EmailAddress
243                | TokenKind::Hostname
244                | TokenKind::Decade
245                | TokenKind::Number(..)
246        )
247    }
248
249    pub(crate) fn is_chunk_terminator(&self) -> bool {
250        if self.is_sentence_terminator() {
251            return true;
252        }
253
254        match self {
255            TokenKind::Punctuation(punct) => {
256                matches!(
257                    punct,
258                    Punctuation::Comma | Punctuation::Quote { .. } | Punctuation::Colon
259                )
260            }
261            _ => false,
262        }
263    }
264
265    pub(crate) fn is_sentence_terminator(&self) -> bool {
266        match self {
267            TokenKind::Punctuation(punct) => [
268                Punctuation::Period,
269                Punctuation::Bang,
270                Punctuation::Question,
271            ]
272            .contains(punct),
273            TokenKind::ParagraphBreak => true,
274            _ => false,
275        }
276    }
277
278    /// Used by `crate::parsers::CollapseIdentifiers`
279    /// TODO: Separate this into two functions and add OR functionality to
280    /// pattern matching
281    pub fn is_case_separator(&self) -> bool {
282        matches!(self, TokenKind::Punctuation(Punctuation::Underscore))
283            || matches!(self, TokenKind::Punctuation(Punctuation::Hyphen))
284    }
285
286    /// Checks whether the token is whitespace.
287    pub fn is_whitespace(&self) -> bool {
288        matches!(self, TokenKind::Space(_) | TokenKind::Newline(_))
289    }
290
291    pub fn is_upos(&self, upos: UPOS) -> bool {
292        let Some(Some(meta)) = self.as_word() else {
293            return false;
294        };
295
296        meta.pos_tag == Some(upos)
297    }
298
299    // Miscellaneous non-is methods
300
301    /// Checks that `self` is the same enum variant as `other`, regardless of
302    /// whether the inner metadata is also equal.
303    pub fn matches_variant_of(&self, other: &Self) -> bool {
304        self.with_default_data() == other.with_default_data()
305    }
306
307    /// Produces a copy of `self` with any inner data replaced with its default
308    /// value. Useful for making comparisons on just the variant of the
309    /// enum.
310    pub fn with_default_data(&self) -> Self {
311        match self {
312            TokenKind::Word(_) => TokenKind::Word(Default::default()),
313            TokenKind::Punctuation(_) => TokenKind::Punctuation(Default::default()),
314            TokenKind::Number(..) => TokenKind::Number(Default::default()),
315            TokenKind::Space(_) => TokenKind::Space(Default::default()),
316            TokenKind::Newline(_) => TokenKind::Newline(Default::default()),
317            _ => self.clone(),
318        }
319    }
320
321    /// Construct a [`TokenKind::Word`] with no metadata.
322    pub fn blank_word() -> Self {
323        Self::Word(None)
324    }
325
326    // Punctuation and symbol non-is methods
327
328    pub fn as_mut_quote(&mut self) -> Option<&mut Quote> {
329        self.as_mut_punctuation()?.as_mut_quote()
330    }
331
332    pub fn as_quote(&self) -> Option<&Quote> {
333        self.as_punctuation()?.as_quote()
334    }
335}
336
337#[cfg(test)]
338mod tests {
339    use crate::Document;
340
341    #[test]
342    fn car_is_singular_noun() {
343        let doc = Document::new_plain_english_curated("car");
344        let tk = &doc.tokens().next().unwrap().kind;
345        assert!(tk.is_singular_noun());
346    }
347
348    #[test]
349    fn traffic_is_mass_noun_only() {
350        let doc = Document::new_plain_english_curated("traffic");
351        let tk = &doc.tokens().next().unwrap().kind;
352        assert!(tk.is_mass_noun_only());
353    }
354
355    #[test]
356    fn equipment_is_mass_noun() {
357        let doc = Document::new_plain_english_curated("equipment");
358        let tk = &doc.tokens().next().unwrap().kind;
359        assert!(tk.is_mass_noun());
360    }
361
362    #[test]
363    fn equipment_is_non_countable_noun() {
364        let doc = Document::new_plain_english_curated("equipment");
365        let tk = &doc.tokens().next().unwrap().kind;
366        assert!(tk.is_non_countable_noun());
367    }
368
369    #[test]
370    fn equipment_isnt_countable_noun() {
371        let doc = Document::new_plain_english_curated("equipment");
372        let tk = &doc.tokens().next().unwrap().kind;
373        assert!(!tk.is_countable_noun());
374    }
375
376    #[test]
377    fn oov_word_is_oov() {
378        let doc = Document::new_plain_english_curated("nonexistentword");
379        let tk = &doc.tokens().next().unwrap().kind;
380        assert!(tk.is_oov());
381    }
382
383    #[test]
384    fn known_word_is_not_oov() {
385        let doc = Document::new_plain_english_curated("car");
386        let tk = &doc.tokens().next().unwrap().kind;
387        assert!(!tk.is_oov());
388    }
389
390    #[test]
391    fn non_word_tokens_are_not_oov() {
392        let doc = Document::new_plain_english_curated("Hello, world!");
393        let tokens: Vec<_> = doc.tokens().collect();
394
395        // Comma should not be OOV
396        assert!(!tokens[1].kind.is_oov());
397        // Exclamation mark should not be OOV
398        assert!(!tokens[3].kind.is_oov());
399    }
400}
harper_core/token_kind.rs

harper_core/
token_kind.rs
