mirror of
https://github.com/neon-mmd/websurfx.git
synced 2024-12-22 12:28:21 -05:00
Add Wikipedia as a search engine
Currently, it only search in the English wikipedia, but it can be customized to use different ones.
This commit is contained in:
parent
ef0ae2f0aa
commit
c23d5c99ff
@ -15,6 +15,7 @@ use crate::models::engine_models::{EngineError, SearchEngine};
|
||||
|
||||
use error_stack::{Report, Result, ResultExt};
|
||||
|
||||
use super::common::build_cookie;
|
||||
use super::search_result_parser::SearchResultParser;
|
||||
|
||||
/// A new Bing engine type defined in-order to implement the `SearchEngine` trait which allows to
|
||||
@ -73,10 +74,7 @@ impl SearchEngine for Bing {
|
||||
("_UR=QS=0&TQS", "0"),
|
||||
];
|
||||
|
||||
let mut cookie_string = String::new();
|
||||
for (k, v) in &query_params {
|
||||
cookie_string.push_str(&format!("{k}={v}; "));
|
||||
}
|
||||
let cookie_string = build_cookie(&query_params);
|
||||
|
||||
let header_map = HeaderMap::try_from(&HashMap::from([
|
||||
("User-Agent".to_string(), user_agent.to_string()),
|
||||
|
23
src/engines/common.rs
Normal file
23
src/engines/common.rs
Normal file
@ -0,0 +1,23 @@
|
||||
//! This module provides common functionalities for engines
|
||||
|
||||
/**
|
||||
* Build a query from a list of key value pairs.
|
||||
*/
|
||||
pub fn build_query(query_params: &[(&str, &str)]) -> String {
|
||||
let mut query_params_string = String::new();
|
||||
for (k, v) in query_params {
|
||||
query_params_string.push_str(&format!("&{k}={v}"));
|
||||
}
|
||||
query_params_string
|
||||
}
|
||||
|
||||
/**
|
||||
* Build a cookie from a list of key value pairs.
|
||||
*/
|
||||
pub fn build_cookie(cookie_params: &[(&str, &str)]) -> String {
|
||||
let mut cookie_string = String::new();
|
||||
for (k, v) in cookie_params {
|
||||
cookie_string.push_str(&format!("{k}={v}; "));
|
||||
}
|
||||
cookie_string
|
||||
}
|
@ -5,9 +5,11 @@
|
||||
|
||||
pub mod bing;
|
||||
pub mod brave;
|
||||
pub mod common;
|
||||
pub mod duckduckgo;
|
||||
pub mod librex;
|
||||
pub mod mojeek;
|
||||
pub mod search_result_parser;
|
||||
pub mod searx;
|
||||
pub mod startpage;
|
||||
pub mod wikipedia;
|
||||
|
@ -14,6 +14,7 @@ use crate::models::engine_models::{EngineError, SearchEngine};
|
||||
|
||||
use error_stack::{Report, Result, ResultExt};
|
||||
|
||||
use super::common::{build_cookie, build_query};
|
||||
use super::search_result_parser::SearchResultParser;
|
||||
|
||||
/// A new Mojeek engine type defined in-order to implement the `SearchEngine` trait which allows to
|
||||
@ -107,10 +108,7 @@ impl SearchEngine for Mojeek {
|
||||
("safe", &safe),
|
||||
];
|
||||
|
||||
let mut query_params_string = String::new();
|
||||
for (k, v) in &query_params {
|
||||
query_params_string.push_str(&format!("&{k}={v}"));
|
||||
}
|
||||
let query_params_string = build_query(&query_params);
|
||||
|
||||
let url: String = match page {
|
||||
0 => {
|
||||
@ -123,10 +121,7 @@ impl SearchEngine for Mojeek {
|
||||
}
|
||||
};
|
||||
|
||||
let mut cookie_string = String::new();
|
||||
for (k, v) in &query_params {
|
||||
cookie_string.push_str(&format!("{k}={v}; "));
|
||||
}
|
||||
let cookie_string = build_cookie(&query_params);
|
||||
|
||||
let header_map = HeaderMap::try_from(&HashMap::from([
|
||||
("User-Agent".to_string(), user_agent.to_string()),
|
||||
|
101
src/engines/wikipedia.rs
Normal file
101
src/engines/wikipedia.rs
Normal file
@ -0,0 +1,101 @@
|
||||
//! The `wikipedia` module handles the scraping of results from wikipedia
|
||||
//! with user provided query and with a page number if provided.
|
||||
|
||||
use std::collections::HashMap;
|
||||
|
||||
use reqwest::header::HeaderMap;
|
||||
use reqwest::Client;
|
||||
use scraper::Html;
|
||||
|
||||
use crate::models::aggregation_models::SearchResult;
|
||||
|
||||
use crate::models::engine_models::{EngineError, SearchEngine};
|
||||
|
||||
use error_stack::{Report, Result, ResultExt};
|
||||
|
||||
use super::common::build_query;
|
||||
use super::search_result_parser::SearchResultParser;
|
||||
|
||||
/// A new Wikipedia engine type defined in-order to implement the `SearchEngine` trait which allows to
|
||||
/// reduce code duplication as well as allows to create vector of different search engines easily.
|
||||
pub struct Wikipedia {
|
||||
/// The parser, used to interpret the search result.
|
||||
parser: SearchResultParser,
|
||||
/// The id of the engine, equals to 'wikipedia-' + language
|
||||
id: String,
|
||||
/// The host where wikipedia can be accessed.
|
||||
host: String,
|
||||
}
|
||||
|
||||
impl Wikipedia {
|
||||
/// Creates the Wikipedia parser.
|
||||
pub fn new(language: &str) -> Result<Self, EngineError> {
|
||||
let host = format!("https://{}.wikipedia.org", &language);
|
||||
let id = format!("wikipedia-{}", &language);
|
||||
Ok(Self {
|
||||
parser: SearchResultParser::new(
|
||||
"p.mw-search-nonefound",
|
||||
".mw-search-results li.mw-search-result",
|
||||
".mw-search-result-heading a",
|
||||
".mw-search-result-heading a",
|
||||
".searchresult",
|
||||
)?,
|
||||
id,
|
||||
host,
|
||||
})
|
||||
}
|
||||
}
|
||||
|
||||
#[async_trait::async_trait]
|
||||
impl SearchEngine for Wikipedia {
|
||||
async fn results(
|
||||
&self,
|
||||
query: &str,
|
||||
page: u32,
|
||||
user_agent: &str,
|
||||
client: &Client,
|
||||
_safe_search: u8,
|
||||
) -> Result<Vec<(String, SearchResult)>, EngineError> {
|
||||
let header_map = HeaderMap::try_from(&HashMap::from([
|
||||
("User-Agent".to_string(), user_agent.to_string()),
|
||||
("Referer".to_string(), self.host.to_string()),
|
||||
]))
|
||||
.change_context(EngineError::UnexpectedError)?;
|
||||
|
||||
let offset = (page * 20).to_string();
|
||||
let query_params: Vec<(&str, &str)> = vec![
|
||||
("limit", "20"),
|
||||
("offset", &offset),
|
||||
("profile", "default"),
|
||||
("search", query),
|
||||
("title", "Special:Search"),
|
||||
("ns0", "1"),
|
||||
];
|
||||
|
||||
let query_params_string = build_query(&query_params);
|
||||
|
||||
let url: String = format!("{}/w/index.php?{}", self.host, query_params_string);
|
||||
|
||||
let document: Html = Html::parse_document(
|
||||
&Wikipedia::fetch_html_from_upstream(self, &url, header_map, client).await?,
|
||||
);
|
||||
|
||||
if self.parser.parse_for_no_results(&document).next().is_some() {
|
||||
return Err(Report::new(EngineError::EmptyResultSet));
|
||||
}
|
||||
|
||||
// scrape all the results from the html
|
||||
self.parser
|
||||
.parse_for_results(&document, |title, url, desc| {
|
||||
let found_url = url.attr("href");
|
||||
found_url.map(|relative_url| {
|
||||
SearchResult::new(
|
||||
title.inner_html().trim(),
|
||||
&format!("{}{relative_url}", self.host),
|
||||
desc.inner_html().trim(),
|
||||
&[&self.id],
|
||||
)
|
||||
})
|
||||
})
|
||||
}
|
||||
}
|
@ -206,6 +206,10 @@ impl EngineHandler {
|
||||
let engine = crate::engines::bing::Bing::new()?;
|
||||
("bing", Box::new(engine))
|
||||
}
|
||||
"wikipedia" => {
|
||||
let engine = crate::engines::wikipedia::Wikipedia::new("en")?;
|
||||
("wikipedia", Box::new(engine))
|
||||
}
|
||||
_ => {
|
||||
return Err(Report::from(EngineError::NoSuchEngineFound(
|
||||
engine_name.to_string(),
|
||||
|
@ -74,6 +74,7 @@ upstream_search_engines = {
|
||||
LibreX = false,
|
||||
Mojeek = false,
|
||||
Bing = false,
|
||||
Wikipedia = true,
|
||||
} -- select the upstream search engines from which the results should be fetched.
|
||||
|
||||
proxy = nil -- Proxy to send outgoing requests through. Set to nil to disable.
|
Loading…
Reference in New Issue
Block a user