Spaces:
Runtime error
Runtime error
Jann Marc Villablanca
mergify[bot]
commited on
Commit
•
9f23a1c
1
Parent(s):
9f5213c
✨ feat(engine): `mojeek` for the search engine (#464)
Browse files* 🧑💻 fix: add closing curly brace in the nix build step
* ✨ feat: add code to provide search results from `mojeek` engine
---------
Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>
- flake.nix +1 -1
- src/engines/mod.rs +1 -0
- src/engines/mojeek.rs +151 -0
- src/models/engine_models.rs +4 -0
- websurfx/config.lua +1 -0
flake.nix
CHANGED
@@ -60,4 +60,4 @@
|
|
60 |
# calls the build function
|
61 |
packages.websurfx = packages.default;
|
62 |
});
|
63 |
-
|
|
|
60 |
# calls the build function
|
61 |
packages.websurfx = packages.default;
|
62 |
});
|
63 |
+
}
|
src/engines/mod.rs
CHANGED
@@ -6,6 +6,7 @@
|
|
6 |
pub mod brave;
|
7 |
pub mod duckduckgo;
|
8 |
pub mod librex;
|
|
|
9 |
pub mod search_result_parser;
|
10 |
pub mod searx;
|
11 |
pub mod startpage;
|
|
|
6 |
pub mod brave;
|
7 |
pub mod duckduckgo;
|
8 |
pub mod librex;
|
9 |
+
pub mod mojeek;
|
10 |
pub mod search_result_parser;
|
11 |
pub mod searx;
|
12 |
pub mod startpage;
|
src/engines/mojeek.rs
ADDED
@@ -0,0 +1,151 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
//! The `mojeek` module handles the scraping of results from the mojeek search engine
|
2 |
+
//! by querying the upstream mojeek search engine with user provided query and with a page
|
3 |
+
//! number if provided.
|
4 |
+
|
5 |
+
use std::collections::HashMap;
|
6 |
+
|
7 |
+
use reqwest::header::HeaderMap;
|
8 |
+
use reqwest::Client;
|
9 |
+
use scraper::Html;
|
10 |
+
|
11 |
+
use crate::models::aggregation_models::SearchResult;
|
12 |
+
|
13 |
+
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
|
20 |
+
/// reduce code duplication as well as allows to create vector of different search engines easily.
|
21 |
+
pub struct Mojeek {
|
22 |
+
/// The parser, used to interpret the search result.
|
23 |
+
parser: SearchResultParser,
|
24 |
+
}
|
25 |
+
|
26 |
+
impl Mojeek {
|
27 |
+
/// Creates the Mojeek parser.
|
28 |
+
pub fn new() -> Result<Self, EngineError> {
|
29 |
+
Ok(Self {
|
30 |
+
parser: SearchResultParser::new(
|
31 |
+
".result-col",
|
32 |
+
".results-standard li",
|
33 |
+
"a span.url",
|
34 |
+
"h2 a.title",
|
35 |
+
"p.s",
|
36 |
+
)?,
|
37 |
+
})
|
38 |
+
}
|
39 |
+
}
|
40 |
+
|
41 |
+
#[async_trait::async_trait]
|
42 |
+
impl SearchEngine for Mojeek {
|
43 |
+
async fn results(
|
44 |
+
&self,
|
45 |
+
query: &str,
|
46 |
+
page: u32,
|
47 |
+
user_agent: &str,
|
48 |
+
client: &Client,
|
49 |
+
safe_search: u8,
|
50 |
+
) -> Result<HashMap<String, SearchResult>, EngineError> {
|
51 |
+
// Mojeek uses `start results from this number` convention
|
52 |
+
// So, for 10 results per page, page 0 starts at 1, page 1
|
53 |
+
// starts at 11, and so on.
|
54 |
+
let results_per_page = 10;
|
55 |
+
let start_result = results_per_page * page + 1;
|
56 |
+
|
57 |
+
let results_per_page = results_per_page.to_string();
|
58 |
+
let start_result = start_result.to_string();
|
59 |
+
|
60 |
+
let search_engines = vec![
|
61 |
+
"Bing",
|
62 |
+
"Brave",
|
63 |
+
"DuckDuckGo",
|
64 |
+
"Ecosia",
|
65 |
+
"Google",
|
66 |
+
"Lilo",
|
67 |
+
"Metager",
|
68 |
+
"Qwant",
|
69 |
+
"Startpage",
|
70 |
+
"Swisscows",
|
71 |
+
"Yandex",
|
72 |
+
"Yep",
|
73 |
+
"You",
|
74 |
+
];
|
75 |
+
let qss = search_engines.join("%2C");
|
76 |
+
let safe = if safe_search == 0 { "0" } else { "1" };
|
77 |
+
|
78 |
+
// Mojeek detects automated requests, these are preferences that are
|
79 |
+
// able to circumvent the countermeasure. Some of these are
|
80 |
+
// not documented in their Search API
|
81 |
+
let query_params: Vec<(&str, &str)> = vec![
|
82 |
+
("t", results_per_page.as_str()),
|
83 |
+
("theme", "dark"),
|
84 |
+
("arc", "none"),
|
85 |
+
("date", "1"),
|
86 |
+
("cdate", "1"),
|
87 |
+
("tlen", "100"),
|
88 |
+
("ref", "1"),
|
89 |
+
("hp", "minimal"),
|
90 |
+
("lb", "en"),
|
91 |
+
("qss", &qss),
|
92 |
+
("safe", safe),
|
93 |
+
];
|
94 |
+
|
95 |
+
let mut query_params_string = String::new();
|
96 |
+
for (k, v) in &query_params {
|
97 |
+
query_params_string.push_str(&format!("&{k}={v}"));
|
98 |
+
}
|
99 |
+
|
100 |
+
let url: String = match page {
|
101 |
+
0 => {
|
102 |
+
format!("https://www.mojeek.com/search?q={query}{query_params_string}")
|
103 |
+
}
|
104 |
+
_ => {
|
105 |
+
format!(
|
106 |
+
"https://www.mojeek.com/search?q={query}&s={start_result}{query_params_string}"
|
107 |
+
)
|
108 |
+
}
|
109 |
+
};
|
110 |
+
|
111 |
+
let mut cookie_string = String::new();
|
112 |
+
for (k, v) in &query_params {
|
113 |
+
cookie_string.push_str(&format!("{k}={v}; "));
|
114 |
+
}
|
115 |
+
|
116 |
+
let header_map = HeaderMap::try_from(&HashMap::from([
|
117 |
+
("USER_AGENT".to_string(), user_agent.to_string()),
|
118 |
+
("REFERER".to_string(), "https://google.com/".to_string()),
|
119 |
+
(
|
120 |
+
"CONTENT_TYPE".to_string(),
|
121 |
+
"application/x-www-form-urlencoded".to_string(),
|
122 |
+
),
|
123 |
+
("COOKIE".to_string(), cookie_string),
|
124 |
+
]))
|
125 |
+
.change_context(EngineError::UnexpectedError)?;
|
126 |
+
|
127 |
+
let document: Html = Html::parse_document(
|
128 |
+
&Mojeek::fetch_html_from_upstream(self, &url, header_map, client).await?,
|
129 |
+
);
|
130 |
+
|
131 |
+
if let Some(no_result_msg) = self.parser.parse_for_no_results(&document).nth(0) {
|
132 |
+
if no_result_msg
|
133 |
+
.inner_html()
|
134 |
+
.contains("No pages found matching:")
|
135 |
+
{
|
136 |
+
return Err(Report::new(EngineError::EmptyResultSet));
|
137 |
+
}
|
138 |
+
}
|
139 |
+
|
140 |
+
// scrape all the results from the html
|
141 |
+
self.parser
|
142 |
+
.parse_for_results(&document, |title, url, desc| {
|
143 |
+
Some(SearchResult::new(
|
144 |
+
title.inner_html().trim(),
|
145 |
+
url.inner_html().trim(),
|
146 |
+
desc.inner_html().trim(),
|
147 |
+
&["mojeek"],
|
148 |
+
))
|
149 |
+
})
|
150 |
+
}
|
151 |
+
}
|
src/models/engine_models.rs
CHANGED
@@ -162,6 +162,10 @@ impl EngineHandler {
|
|
162 |
let engine = crate::engines::librex::LibreX::new()?;
|
163 |
("librex", Box::new(engine))
|
164 |
}
|
|
|
|
|
|
|
|
|
165 |
_ => {
|
166 |
return Err(Report::from(EngineError::NoSuchEngineFound(
|
167 |
engine_name.to_string(),
|
|
|
162 |
let engine = crate::engines::librex::LibreX::new()?;
|
163 |
("librex", Box::new(engine))
|
164 |
}
|
165 |
+
"mojeek" => {
|
166 |
+
let engine = crate::engines::mojeek::Mojeek::new()?;
|
167 |
+
("mojeek", Box::new(engine))
|
168 |
+
}
|
169 |
_ => {
|
170 |
return Err(Report::from(EngineError::NoSuchEngineFound(
|
171 |
engine_name.to_string(),
|
websurfx/config.lua
CHANGED
@@ -64,4 +64,5 @@ upstream_search_engines = {
|
|
64 |
Brave = false,
|
65 |
Startpage = false,
|
66 |
LibreX = false,
|
|
|
67 |
} -- select the upstream search engines from which the results should be fetched.
|
|
|
64 |
Brave = false,
|
65 |
Startpage = false,
|
66 |
LibreX = false,
|
67 |
+
Mojeek = false,
|
68 |
} -- select the upstream search engines from which the results should be fetched.
|