rustysearch/src/search/engine.rs

197 lines
6.6 KiB
Rust

use std::collections::HashMap;
use std::f64;
fn update_url_scores(old: &mut HashMap<String, f64>, new: &HashMap<String, f64>) {
for (url, score) in new {
old.entry(url.to_string()).and_modify(|e| *e += score).or_insert(*score);
}
}
fn normalize_string(input_string: &str) -> String {
let string_without_punc: String = input_string.chars().filter(|&c| !c.is_ascii_punctuation()).collect();
let string_without_double_spaces: String = string_without_punc.split_whitespace().collect::<Vec<&str>>().join(" ");
string_without_double_spaces.to_lowercase()
}
/// SearchEngine represents a search engine that indexes and searches documents based on the BM25 ranking algorithm.
///
/// The search engine maintains an index of words and their frequencies in each document, as well as the actual document content.
/// It provides methods to index documents, perform searches, and calculate relevance scores using the BM25 algorithm.
///
/// # Examples
///
/// ```
/// use std::collections::HashMap;
/// use rustysearch::search::engine::SearchEngine;
///
/// // Create a new search engine with k1 = 1.2 and b = 0.75
/// let mut engine = SearchEngine::new(1.2, 0.75);
///
/// // Index a document
/// engine.index("https://example.com/doc1", "This is the content of document 1");
///
/// // Perform a search
/// let results = engine.search("content");
///
/// // Print the search results
/// for (url, score) in results {
/// println!("{} - Relevance Score: {}", url, score);
/// }
/// ```
#[derive(Default, Debug, Clone)]
pub struct SearchEngine {
index: HashMap<String, HashMap<String, i32>>,
documents: HashMap<String, String>,
k1: f64,
b: f64,
}
impl SearchEngine {
/// Creates a new instance of SearchEngine with the given parameters.
///
/// # Arguments
///
/// * `k1` - The k1 parameter of the BM25 algorithm.
/// * `b` - The b parameter of the BM25 algorithm.
///
/// **Returns**
///
/// A new instance of SearchEngine.
pub fn new(k1: f64, b: f64) -> SearchEngine {
SearchEngine {
index: HashMap::new(),
documents: HashMap::new(),
k1,
b,
}
}
/// Returns a vector of all the document URLs in the search engine's index.
///
/// **Returns**
///
/// A vector of document URLs.
pub fn posts(&self) -> Vec<String> {
self.documents.keys().cloned().collect()
}
/// Returns the number of documents in the search engine's index.
///
/// **Returns**
///
/// The number of documents.
pub fn number_of_documents(&self) -> usize {
self.documents.len()
}
/// Returns the average document length in terms of number of words.
///
/// **Returns**
///
/// The average document length.
pub fn avdl(&self) -> f64 {
let total_length: usize = self.documents.values().map(|d| d.len()).sum();
total_length as f64 / self.documents.len() as f64
}
/// Calculates the inverse document frequency (IDF) score for a given keyword.
///
/// **Arguments**
///
/// * `kw` - The keyword for which to calculate the IDF score.
///
/// **Returns**
///
/// The IDF score.
pub fn idf(&self, kw: &str) -> f64 {
let n = self.number_of_documents() as f64;
let n_kw = self.get_urls(kw).len() as f64;
((n - n_kw + 0.5) / (n_kw + 0.5) + 1.0).ln()
}
/// Calculates the BM25 relevance scores for a given keyword.
///
/// **Arguments**
///
/// * `kw` - The keyword for which to calculate the relevance scores.
///
/// **Returns**
///
/// A HashMap containing the document URLs as keys and their relevance scores as values.
pub fn bm25(&self, kw: &str) -> HashMap<String, f64> {
let mut result = HashMap::new();
let idf_score = self.idf(kw);
let avdl = self.avdl();
for (url, freq) in self.get_urls(kw) {
let numerator = freq as f64 * (self.k1 + 1.0);
let denominator = freq as f64 + self.k1 * (1.0 - self.b + self.b * self.documents.get(&url).unwrap().len() as f64 / avdl);
result.insert(url.to_string(), idf_score * numerator / denominator);
}
result
}
/// Performs a search for the given query and returns the relevance scores for the matching documents.
///
/// **Arguments**
///
/// * `query` - The search query.
///
/// **Returns**
///
/// A HashMap containing the document URLs as keys and their relevance scores as values.
pub fn search(&mut self, query: &str) -> HashMap<String, f64> {
let keywords = normalize_string(query).split_whitespace().map(|s| s.to_string()).collect::<Vec<String>>();
let mut url_scores: HashMap<String, f64> = HashMap::new();
for kw in keywords {
let kw_urls_score = self.bm25(&kw);
update_url_scores(&mut url_scores, &kw_urls_score);
}
url_scores
}
/// Indexes a document with the given URL and content.
///
/// **Arguments**
///
/// * `url` - The URL of the document.
/// * `content` - The content of the document.
pub fn index(&mut self, url: &str, content: &str) {
self.documents.insert(url.to_string(), content.to_string());
let words = normalize_string(content).split_whitespace().map(|s| s.to_string()).collect::<Vec<String>>();
for word in words {
*self.index.entry(word).or_insert(HashMap::new()).entry(url.to_string()).or_insert(0) += 1;
}
}
/// Bulk indexes multiple documents.
///
/// **Arguments**
///
/// * `documents` - A vector of tuples containing the URL and content of each document.
pub fn bulk_index(&mut self, documents: Vec<(&str, &str)>) {
for (url, content) in documents {
self.index(url, content);
}
}
/// Returns the URLs and frequencies of a given keyword in the search engine's index.
///
/// **Arguments**
///
/// * `keyword` - The keyword to search for.
///
/// **Returns**
///
/// A HashMap containing the document URLs as keys and their frequencies as values.
pub fn get_urls(&self, keyword: &str) -> HashMap<String, i32> {
let keyword = normalize_string(keyword);
self.index.get(&keyword).cloned().unwrap_or(HashMap::new())
}
/// Prints the current state of the search engine's index and document collection for debugging purposes.
pub fn debug_index(&self) {
log::debug!("Index: {:?}", self.index);
log::debug!("Documents: {:?}", self.documents);
}
}