Spaces:
Running
Running
Zsombor
mergify[bot]
commited on
Add Wikipedia as a search engine (#633)
Browse filesCurrently, it only search in the English wikipedia, but it can be
customized to use different ones.
Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>
- src/engines/bing.rs +2 -4
- src/engines/common.rs +23 -0
- src/engines/mod.rs +2 -0
- src/engines/mojeek.rs +3 -8
- src/engines/wikipedia.rs +101 -0
- src/models/engine_models.rs +4 -0
- websurfx/config.lua +1 -0
src/engines/bing.rs
CHANGED
@@ -15,6 +15,7 @@ use crate::models::engine_models::{EngineError, SearchEngine};
|
|
15 |
|
16 |
use error_stack::{Report, Result, ResultExt};
|
17 |
|
|
|
18 |
use super::search_result_parser::SearchResultParser;
|
19 |
|
20 |
/// A new Bing engine type defined in-order to implement the `SearchEngine` trait which allows to
|
@@ -73,10 +74,7 @@ impl SearchEngine for Bing {
|
|
73 |
("_UR=QS=0&TQS", "0"),
|
74 |
];
|
75 |
|
76 |
-
let
|
77 |
-
for (k, v) in &query_params {
|
78 |
-
cookie_string.push_str(&format!("{k}={v}; "));
|
79 |
-
}
|
80 |
|
81 |
let header_map = HeaderMap::try_from(&HashMap::from([
|
82 |
("User-Agent".to_string(), user_agent.to_string()),
|
|
|
15 |
|
16 |
use error_stack::{Report, Result, ResultExt};
|
17 |
|
18 |
+
use super::common::build_cookie;
|
19 |
use super::search_result_parser::SearchResultParser;
|
20 |
|
21 |
/// A new Bing engine type defined in-order to implement the `SearchEngine` trait which allows to
|
|
|
74 |
("_UR=QS=0&TQS", "0"),
|
75 |
];
|
76 |
|
77 |
+
let cookie_string = build_cookie(&query_params);
|
|
|
|
|
|
|
78 |
|
79 |
let header_map = HeaderMap::try_from(&HashMap::from([
|
80 |
("User-Agent".to_string(), user_agent.to_string()),
|
src/engines/common.rs
ADDED
@@ -0,0 +1,23 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
//! This module provides common functionalities for engines
|
2 |
+
|
3 |
+
/**
|
4 |
+
* Build a query from a list of key value pairs.
|
5 |
+
*/
|
6 |
+
pub fn build_query(query_params: &[(&str, &str)]) -> String {
|
7 |
+
let mut query_params_string = String::new();
|
8 |
+
for (k, v) in query_params {
|
9 |
+
query_params_string.push_str(&format!("&{k}={v}"));
|
10 |
+
}
|
11 |
+
query_params_string
|
12 |
+
}
|
13 |
+
|
14 |
+
/**
|
15 |
+
* Build a cookie from a list of key value pairs.
|
16 |
+
*/
|
17 |
+
pub fn build_cookie(cookie_params: &[(&str, &str)]) -> String {
|
18 |
+
let mut cookie_string = String::new();
|
19 |
+
for (k, v) in cookie_params {
|
20 |
+
cookie_string.push_str(&format!("{k}={v}; "));
|
21 |
+
}
|
22 |
+
cookie_string
|
23 |
+
}
|
src/engines/mod.rs
CHANGED
@@ -5,9 +5,11 @@
|
|
5 |
|
6 |
pub mod bing;
|
7 |
pub mod brave;
|
|
|
8 |
pub mod duckduckgo;
|
9 |
pub mod librex;
|
10 |
pub mod mojeek;
|
11 |
pub mod search_result_parser;
|
12 |
pub mod searx;
|
13 |
pub mod startpage;
|
|
|
|
5 |
|
6 |
pub mod bing;
|
7 |
pub mod brave;
|
8 |
+
pub mod common;
|
9 |
pub mod duckduckgo;
|
10 |
pub mod librex;
|
11 |
pub mod mojeek;
|
12 |
pub mod search_result_parser;
|
13 |
pub mod searx;
|
14 |
pub mod startpage;
|
15 |
+
pub mod wikipedia;
|
src/engines/mojeek.rs
CHANGED
@@ -14,6 +14,7 @@ use crate::models::engine_models::{EngineError, SearchEngine};
|
|
14 |
|
15 |
use error_stack::{Report, Result, ResultExt};
|
16 |
|
|
|
17 |
use super::search_result_parser::SearchResultParser;
|
18 |
|
19 |
/// A new Mojeek engine type defined in-order to implement the `SearchEngine` trait which allows to
|
@@ -107,10 +108,7 @@ impl SearchEngine for Mojeek {
|
|
107 |
("safe", &safe),
|
108 |
];
|
109 |
|
110 |
-
let
|
111 |
-
for (k, v) in &query_params {
|
112 |
-
query_params_string.push_str(&format!("&{k}={v}"));
|
113 |
-
}
|
114 |
|
115 |
let url: String = match page {
|
116 |
0 => {
|
@@ -123,10 +121,7 @@ impl SearchEngine for Mojeek {
|
|
123 |
}
|
124 |
};
|
125 |
|
126 |
-
let
|
127 |
-
for (k, v) in &query_params {
|
128 |
-
cookie_string.push_str(&format!("{k}={v}; "));
|
129 |
-
}
|
130 |
|
131 |
let header_map = HeaderMap::try_from(&HashMap::from([
|
132 |
("User-Agent".to_string(), user_agent.to_string()),
|
|
|
14 |
|
15 |
use error_stack::{Report, Result, ResultExt};
|
16 |
|
17 |
+
use super::common::{build_cookie, build_query};
|
18 |
use super::search_result_parser::SearchResultParser;
|
19 |
|
20 |
/// A new Mojeek engine type defined in-order to implement the `SearchEngine` trait which allows to
|
|
|
108 |
("safe", &safe),
|
109 |
];
|
110 |
|
111 |
+
let query_params_string = build_query(&query_params);
|
|
|
|
|
|
|
112 |
|
113 |
let url: String = match page {
|
114 |
0 => {
|
|
|
121 |
}
|
122 |
};
|
123 |
|
124 |
+
let cookie_string = build_cookie(&query_params);
|
|
|
|
|
|
|
125 |
|
126 |
let header_map = HeaderMap::try_from(&HashMap::from([
|
127 |
("User-Agent".to_string(), user_agent.to_string()),
|
src/engines/wikipedia.rs
ADDED
@@ -0,0 +1,101 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
//! The `wikipedia` module handles the scraping of results from wikipedia
|
2 |
+
//! with user provided query and with a page number if provided.
|
3 |
+
|
4 |
+
use std::collections::HashMap;
|
5 |
+
|
6 |
+
use reqwest::header::HeaderMap;
|
7 |
+
use reqwest::Client;
|
8 |
+
use scraper::Html;
|
9 |
+
|
10 |
+
use crate::models::aggregation_models::SearchResult;
|
11 |
+
|
12 |
+
use crate::models::engine_models::{EngineError, SearchEngine};
|
13 |
+
|
14 |
+
use error_stack::{Report, Result, ResultExt};
|
15 |
+
|
16 |
+
use super::common::build_query;
|
17 |
+
use super::search_result_parser::SearchResultParser;
|
18 |
+
|
19 |
+
/// A new Wikipedia engine type defined in-order to implement the `SearchEngine` trait which allows to
|
20 |
+
/// reduce code duplication as well as allows to create vector of different search engines easily.
|
21 |
+
pub struct Wikipedia {
|
22 |
+
/// The parser, used to interpret the search result.
|
23 |
+
parser: SearchResultParser,
|
24 |
+
/// The id of the engine, equals to 'wikipedia-' + language
|
25 |
+
id: String,
|
26 |
+
/// The host where wikipedia can be accessed.
|
27 |
+
host: String,
|
28 |
+
}
|
29 |
+
|
30 |
+
impl Wikipedia {
|
31 |
+
/// Creates the Wikipedia parser.
|
32 |
+
pub fn new(language: &str) -> Result<Self, EngineError> {
|
33 |
+
let host = format!("https://{}.wikipedia.org", &language);
|
34 |
+
let id = format!("wikipedia-{}", &language);
|
35 |
+
Ok(Self {
|
36 |
+
parser: SearchResultParser::new(
|
37 |
+
"p.mw-search-nonefound",
|
38 |
+
".mw-search-results li.mw-search-result",
|
39 |
+
".mw-search-result-heading a",
|
40 |
+
".mw-search-result-heading a",
|
41 |
+
".searchresult",
|
42 |
+
)?,
|
43 |
+
id,
|
44 |
+
host,
|
45 |
+
})
|
46 |
+
}
|
47 |
+
}
|
48 |
+
|
49 |
+
#[async_trait::async_trait]
|
50 |
+
impl SearchEngine for Wikipedia {
|
51 |
+
async fn results(
|
52 |
+
&self,
|
53 |
+
query: &str,
|
54 |
+
page: u32,
|
55 |
+
user_agent: &str,
|
56 |
+
client: &Client,
|
57 |
+
_safe_search: u8,
|
58 |
+
) -> Result<Vec<(String, SearchResult)>, EngineError> {
|
59 |
+
let header_map = HeaderMap::try_from(&HashMap::from([
|
60 |
+
("User-Agent".to_string(), user_agent.to_string()),
|
61 |
+
("Referer".to_string(), self.host.to_string()),
|
62 |
+
]))
|
63 |
+
.change_context(EngineError::UnexpectedError)?;
|
64 |
+
|
65 |
+
let offset = (page * 20).to_string();
|
66 |
+
let query_params: Vec<(&str, &str)> = vec![
|
67 |
+
("limit", "20"),
|
68 |
+
("offset", &offset),
|
69 |
+
("profile", "default"),
|
70 |
+
("search", query),
|
71 |
+
("title", "Special:Search"),
|
72 |
+
("ns0", "1"),
|
73 |
+
];
|
74 |
+
|
75 |
+
let query_params_string = build_query(&query_params);
|
76 |
+
|
77 |
+
let url: String = format!("{}/w/index.php?{}", self.host, query_params_string);
|
78 |
+
|
79 |
+
let document: Html = Html::parse_document(
|
80 |
+
&Wikipedia::fetch_html_from_upstream(self, &url, header_map, client).await?,
|
81 |
+
);
|
82 |
+
|
83 |
+
if self.parser.parse_for_no_results(&document).next().is_some() {
|
84 |
+
return Err(Report::new(EngineError::EmptyResultSet));
|
85 |
+
}
|
86 |
+
|
87 |
+
// scrape all the results from the html
|
88 |
+
self.parser
|
89 |
+
.parse_for_results(&document, |title, url, desc| {
|
90 |
+
let found_url = url.attr("href");
|
91 |
+
found_url.map(|relative_url| {
|
92 |
+
SearchResult::new(
|
93 |
+
title.inner_html().trim(),
|
94 |
+
&format!("{}{relative_url}", self.host),
|
95 |
+
desc.inner_html().trim(),
|
96 |
+
&[&self.id],
|
97 |
+
)
|
98 |
+
})
|
99 |
+
})
|
100 |
+
}
|
101 |
+
}
|
src/models/engine_models.rs
CHANGED
@@ -206,6 +206,10 @@ impl EngineHandler {
|
|
206 |
let engine = crate::engines::bing::Bing::new()?;
|
207 |
("bing", Box::new(engine))
|
208 |
}
|
|
|
|
|
|
|
|
|
209 |
_ => {
|
210 |
return Err(Report::from(EngineError::NoSuchEngineFound(
|
211 |
engine_name.to_string(),
|
|
|
206 |
let engine = crate::engines::bing::Bing::new()?;
|
207 |
("bing", Box::new(engine))
|
208 |
}
|
209 |
+
"wikipedia" => {
|
210 |
+
let engine = crate::engines::wikipedia::Wikipedia::new("en")?;
|
211 |
+
("wikipedia", Box::new(engine))
|
212 |
+
}
|
213 |
_ => {
|
214 |
return Err(Report::from(EngineError::NoSuchEngineFound(
|
215 |
engine_name.to_string(),
|
websurfx/config.lua
CHANGED
@@ -76,6 +76,7 @@ upstream_search_engines = {
|
|
76 |
LibreX = false,
|
77 |
Mojeek = false,
|
78 |
Bing = false,
|
|
|
79 |
} -- select the upstream search engines from which the results should be fetched.
|
80 |
|
81 |
proxy = nil -- Proxy to send outgoing requests through. Set to nil to disable.
|
|
|
76 |
LibreX = false,
|
77 |
Mojeek = false,
|
78 |
Bing = false,
|
79 |
+
Wikipedia = true,
|
80 |
} -- select the upstream search engines from which the results should be fetched.
|
81 |
|
82 |
proxy = nil -- Proxy to send outgoing requests through. Set to nil to disable.
|