language_text_analysis/
term.rs

1//! The normalized [`Term`] token type.
2
3use std::fmt;
4
5use serde::{Deserialize, Serialize};
6
7/// A single normalized search term — the output unit of the analysis
8/// pipeline and the key under which the full-text index stores
9/// postings.
10///
11/// A `Term` is always the product of [`Analyzer::analyze`]: normalized
12/// (lowercased, accent-stripped), tokenized, stop-word-filtered, and
13/// optionally stemmed. The same pipeline runs at index time and query
14/// time, so a query `Term` is byte-for-byte comparable with the indexed
15/// `Term` it should match.
16///
17/// Construct one only from already-analyzed text. [`from_normalized`]
18/// exists for the pipeline itself and for callers that have run the
19/// identical analysis; it does **not** normalize on your behalf.
20///
21/// [`Analyzer::analyze`]: crate::Analyzer::analyze
22/// [`from_normalized`]: Term::from_normalized
23#[derive(Clone, Debug, PartialEq, Eq, Hash, PartialOrd, Ord, Serialize, Deserialize)]
24pub struct Term(String);
25
26impl Term {
27    /// Wrap an already-normalized token.
28    ///
29    /// The pipeline guarantees the input has been normalized; this
30    /// constructor performs no normalization itself.
31    #[must_use]
32    pub fn from_normalized(token: impl Into<String>) -> Self {
33        Self(token.into())
34    }
35
36    /// The token text.
37    #[must_use]
38    pub fn as_str(&self) -> &str {
39        &self.0
40    }
41
42    /// Consume the term, yielding its inner string.
43    #[must_use]
44    pub fn into_inner(self) -> String {
45        self.0
46    }
47}
48
49impl fmt::Display for Term {
50    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
51        f.write_str(&self.0)
52    }
53}
language_text_analysis/term.rs

language_text_analysis/
term.rs
