Merge branch 'rolling' into rolling

src/engines/brave.rs

@@ -4,7 +4,7 @@
 
 use std::collections::HashMap;
 
-use reqwest::header::HeaderMap;
+use reqwest::{header::HeaderMap, Client};
 use scraper::Html;
 
 use crate::models::aggregation_models::SearchResult;
@@ -42,7 +42,7 @@ impl SearchEngine for Brave {
         query: &str,
         page: u32,
         user_agent: &str,
-        request_timeout: u8,
+        client: &Client,
         safe_search: u8,
     ) -> Result<HashMap<String, SearchResult>, EngineError> {
         let url = format!("https://search.brave.com/search?q={query}&offset={page}");
@@ -68,7 +68,7 @@ impl SearchEngine for Brave {
         .change_context(EngineError::UnexpectedError)?;
 
         let document: Html = Html::parse_document(
-            &Brave::fetch_html_from_upstream(self, &url, header_map, request_timeout).await?,
+            &Brave::fetch_html_from_upstream(self, &url, header_map, client).await?,
         );
 
         if let Some(no_result_msg) = self.parser.parse_for_no_results(&document).nth(0) {

src/engines/duckduckgo.rs

@@ -5,6 +5,7 @@
 use std::collections::HashMap;
 
 use reqwest::header::HeaderMap;
+use reqwest::Client;
 use scraper::Html;
 
 use crate::models::aggregation_models::SearchResult;
@@ -44,7 +45,7 @@ impl SearchEngine for DuckDuckGo {
         query: &str,
         page: u32,
         user_agent: &str,
-        request_timeout: u8,
+        client: &Client,
         _safe_search: u8,
     ) -> Result<HashMap<String, SearchResult>, EngineError> {
         // Page number can be missing or empty string and so appropriate handling is required
@@ -76,7 +77,7 @@ impl SearchEngine for DuckDuckGo {
         .change_context(EngineError::UnexpectedError)?;
 
         let document: Html = Html::parse_document(
-            &DuckDuckGo::fetch_html_from_upstream(self, &url, header_map, request_timeout).await?,
+            &DuckDuckGo::fetch_html_from_upstream(self, &url, header_map, client).await?,
         );
 
         if self.parser.parse_for_no_results(&document).next().is_some() {

src/engines/searx.rs

@@ -3,6 +3,7 @@
 //! number if provided.
 
 use reqwest::header::HeaderMap;
+use reqwest::Client;
 use scraper::Html;
 use std::collections::HashMap;
 
@@ -40,7 +41,7 @@ impl SearchEngine for Searx {
         query: &str,
         page: u32,
         user_agent: &str,
-        request_timeout: u8,
+        client: &Client,
         mut safe_search: u8,
     ) -> Result<HashMap<String, SearchResult>, EngineError> {
         // Page number can be missing or empty string and so appropriate handling is required
@@ -68,7 +69,7 @@ impl SearchEngine for Searx {
         .change_context(EngineError::UnexpectedError)?;
 
         let document: Html = Html::parse_document(
-            &Searx::fetch_html_from_upstream(self, &url, header_map, request_timeout).await?,
+            &Searx::fetch_html_from_upstream(self, &url, header_map, client).await?,
         );
 
         if let Some(no_result_msg) = self.parser.parse_for_no_results(&document).nth(1) {

src/models/engine_models.rs

@@ -3,7 +3,8 @@
 
 use super::aggregation_models::SearchResult;
 use error_stack::{Report, Result, ResultExt};
-use std::{collections::HashMap, fmt, time::Duration};
+use reqwest::Client;
+use std::{collections::HashMap, fmt};
 
 /// A custom error type used for handle engine associated errors.
 #[derive(Debug)]
@@ -71,12 +72,11 @@ pub trait SearchEngine: Sync + Send {
         &self,
         url: &str,
         header_map: reqwest::header::HeaderMap,
-        request_timeout: u8,
+        client: &Client,
     ) -> Result<String, EngineError> {
         // fetch the html from upstream search engine
-        Ok(reqwest::Client::new()
+        Ok(client
             .get(url)
-            .timeout(Duration::from_secs(request_timeout as u64)) // Add timeout to request to avoid DDOSing the server
             .headers(header_map) // add spoofed headers to emulate human behavior
             .send()
             .await
@@ -109,7 +109,7 @@ pub trait SearchEngine: Sync + Send {
         query: &str,
         page: u32,
         user_agent: &str,
-        request_timeout: u8,
+        client: &Client,
         safe_search: u8,
     ) -> Result<HashMap<String, SearchResult>, EngineError>;
 }

src/results/aggregator.rs

@@ -9,6 +9,7 @@ use crate::models::{
 };
 use error_stack::Report;
 use regex::Regex;
+use reqwest::{Client, ClientBuilder};
 use std::time::{SystemTime, UNIX_EPOCH};
 use std::{
     collections::HashMap,
@@ -18,6 +19,9 @@ use std::{
 use std::{fs::File, io::BufRead};
 use tokio::task::JoinHandle;
 
+/// A constant for holding the prebuilt Client globally in the app.
+static CLIENT: std::sync::OnceLock<Client> = std::sync::OnceLock::new();
+
 /// Aliases for long type annotations
 type FutureVec = Vec<JoinHandle<Result<HashMap<String, SearchResult>, Report<EngineError>>>>;
 
@@ -68,6 +72,16 @@ pub async fn aggregate(
     request_timeout: u8,
     safe_search: u8,
 ) -> Result<SearchResults, Box<dyn std::error::Error>> {
+    let client = CLIENT.get_or_init(|| {
+        ClientBuilder::new()
+            .timeout(Duration::from_secs(request_timeout as u64)) // Add timeout to request to avoid DDOSing the server
+            .https_only(true)
+            .gzip(true)
+            .brotli(true)
+            .build()
+            .unwrap()
+    });
+
     let user_agent: &str = random_user_agent();
 
     // Add a random delay before making the request.
@@ -88,7 +102,7 @@ pub async fn aggregate(
         let query: String = query.to_owned();
         tasks.push(tokio::spawn(async move {
             search_engine
-                .results(&query, page, user_agent, request_timeout, safe_search)
+                .results(&query, page, user_agent, client, safe_search)
                 .await
         }));
     }