Compare commits

..

2 Commits

Author SHA1 Message Date
lolcat 063397dbd1 err 2024-09-01 10:52:46 -04:00
lolcat 1a5a653be3 added ghostery search 2024-09-01 10:52:28 -04:00
6 changed files with 327 additions and 7 deletions

View File

@ -1,4 +1,4 @@
[![ko-fi](https://ko-fi.com/img/githubbutton_sm.svg)](https://ko-fi.com/W7W2OZK5H)
<a href="https://4get.ca/donate">Donate to the project here!</a>
# 4get search
**4get** is a proxy search engine that doesn't suck.
@ -37,11 +37,12 @@ tl;dr the best way to actually browse for shit.
| Google | Google | Yandex | Startpage | | Google |
| Startpage | Startpage | Google | Qwant | | Startpage |
| Qwant | Qwant | Startpage | Mojeek | | Kagi |
| Yep | Yep | Qwant | | | Qwant |
| Greppr | Imgur | | | | Yep |
| Crowdview | FindThatMeme | | | | Marginalia |
| Mwmbl | | | | | YouTube |
| Mojeek | | | | | Soundcloud |
| Ghostery | Yep | Qwant | | | Qwant |
| Yep | Imgur | | | | Yep |
| Greppr | FindThatMeme | | | | Marginalia |
| Crowdview | | | | | YouTube |
| Mwmbl | | | | | Soundcloud |
| Mojeek | | | | | |
| Marginalia | | | | | |
| wiby | | | | | |
| Curlie | | | | | |

View File

@ -20,7 +20,8 @@ class autocomplete{
"yt" => "https://suggestqueries-clients6.youtube.com/complete/search?client=youtube&q={searchTerms}",
"sc" => "",
"startpage" => "https://www.startpage.com/suggestions?q={searchTerms}&format=opensearch&segment=startpage.defaultffx&lui=english",
"kagi" => "https://kagi.com/api/autosuggest?q={searchTerms}"
"kagi" => "https://kagi.com/api/autosuggest?q={searchTerms}",
"ghostery" => "https://ghosterysearch.com/suggest?q={searchTerms}"
];
/*

View File

@ -131,6 +131,7 @@ class config{
const PROXY_GOOGLE = false;
const PROXY_STARTPAGE = false;
const PROXY_QWANT = false;
const PROXY_GHOSTERY = false;
const PROXY_MARGINALIA = false;
const PROXY_MOJEEK = false;
const PROXY_SC = false; // soundcloud

View File

@ -941,6 +941,7 @@ class frontend{
"google" => "Google",
"startpage" => "Startpage",
"qwant" => "Qwant",
"ghostery" => "Ghostery",
"yep" => "Yep",
"greppr" => "Greppr",
"crowdview" => "Crowdview",

308
scraper/ghostery.php Normal file
View File

@ -0,0 +1,308 @@
<?php
class ghostery{
public function __construct(){
include "lib/backend.php";
$this->backend = new backend("ghostery");
include "lib/fuckhtml.php";
$this->fuckhtml = new fuckhtml();
}
public function getfilters($page){
if($page != "web"){
return [];
}
return [
"country" => [
"display" => "Country",
"option" => [
"any" => "All regions",
"AR" => "Argentina",
"AU" => "Australia",
"AT" => "Austria",
"BE" => "Belgium",
"BR" => "Brazil",
"CA" => "Canada",
"CL" => "Chile",
"DK" => "Denmark",
"FI" => "Finland",
"FR" => "France",
"DE" => "Germany",
"HK" => "Hong Kong",
"IN" => "India",
"ID" => "Indonesia",
"IT" => "Italy",
"JP" => "Japan",
"KR" => "Korea",
"MY" => "Malaysia",
"MX" => "Mexico",
"NL" => "Netherlands",
"NZ" => "New Zealand",
"NO" => "Norway",
"CN" => "People's Republic of China",
"PL" => "Poland",
"PT" => "Portugal",
"PH" => "Republic of the Philippines",
"RU" => "Russia",
"SA" => "Saudi Arabia",
"ZA" => "South Africa",
"ES" => "Spain",
"SE" => "Sweden",
"CH" => "Switzerland",
"TW" => "Taiwan",
"TR" => "Turkey",
"GB" => "United Kingdom",
"US" => "United States"
]
]
];
}
private function get($proxy, $url, $get = [], $country){
$curlproc = curl_init();
if($get !== []){
$get = http_build_query($get);
$url .= "?" . $get;
}
curl_setopt($curlproc, CURLOPT_URL, $url);
curl_setopt($curlproc, CURLOPT_ENCODING, ""); // default encoding
curl_setopt($curlproc, CURLOPT_HTTPHEADER,
["User-Agent: " . config::USER_AGENT,
"Accept: text/html,application/xhtml+xml,application/xml;q=0.9,image/avif,image/webp,*/*;q=0.8",
"Accept-Language: en-US,en;q=0.5",
"Accept-Encoding: gzip",
"Referer: https://ghosterysearch.com",
"DNT: 1",
"Sec-GPC: 1",
"Connection: keep-alive",
"Cookie: ctry=" . ($country == "any" ? "--" : $country) . "; noads=true",
"Upgrade-Insecure-Requests: 1",
"Sec-Fetch-Dest: document",
"Sec-Fetch-Mode: navigate",
"Sec-Fetch-Site: same-origin",
"Sec-Fetch-User: ?1",
"Priority: u=0, i"]
);
// http2 bypass
curl_setopt($curlproc, CURLOPT_HTTP_VERSION, CURL_HTTP_VERSION_2_0);
curl_setopt($curlproc, CURLOPT_RETURNTRANSFER, true);
curl_setopt($curlproc, CURLOPT_SSL_VERIFYHOST, 2);
curl_setopt($curlproc, CURLOPT_SSL_VERIFYPEER, true);
curl_setopt($curlproc, CURLOPT_CONNECTTIMEOUT, 30);
curl_setopt($curlproc, CURLOPT_TIMEOUT, 30);
$this->backend->assign_proxy($curlproc, $proxy);
$data = curl_exec($curlproc);
if(curl_errno($curlproc)){
throw new Exception(curl_error($curlproc));
}
curl_close($curlproc);
return $data;
}
public function web($get){
if($get["npt"]){
[$query, $proxy] = $this->backend->get($get["npt"], "web");
parse_str($query, $query);
// country
$country = $query["c"];
unset($query["c"]);
$query = http_build_query($query);
$html =
$this->get(
$proxy,
"https://ghosterysearch.com/search?" . $query,
[],
$country
);
}else{
$proxy = $this->backend->get_ip();
$html =
$this->get(
$proxy,
"https://ghosterysearch.com/search",
[
"q" => $get["s"]
],
$get["country"]
);
}
$out = [
"status" => "ok",
"spelling" => [
"type" => "no_correction",
"using" => null,
"correction" => null
],
"npt" => null,
"answer" => [],
"web" => [],
"image" => [],
"video" => [],
"news" => [],
"related" => []
];
$this->fuckhtml->load($html);
$results_wrapper =
$this->fuckhtml
->getElementsByClassName(
"results",
"section"
);
if(count($results_wrapper) === 0){
throw new Exception("Failed to grep result section");
}
$this->fuckhtml->load($results_wrapper[0]);
// get search results
$results =
$this->fuckhtml
->getElementsByClassName(
"result",
"li"
);
if(count($results) === 0){
return $out;
}
foreach($results as $result){
$this->fuckhtml->load($result);
$a =
$this->fuckhtml
->getElementsByClassName(
"url",
"a"
);
if(count($a) === 0){
continue;
}
$a = $a[0];
$out["web"][] = [
"title" =>
$this->titledots(
$this->fuckhtml
->getTextContent(
$this->fuckhtml
->getElementsByTagName(
"h2"
)[0]
)
),
"description" =>
$this->titledots(
$this->fuckhtml
->getTextContent(
$this->fuckhtml
->getElementsByTagName(
"p"
)[0]
)
),
"url" =>
$this->fuckhtml
->getTextContent(
$a
["attributes"]
["href"]
),
"date" => null,
"type" => "web",
"thumb" => [
"url" => null,
"ratio" => null
],
"sublink" => [],
"table" => []
];
}
$this->fuckhtml->load($html);
// get pagination token
$pagination_wrapper =
$this->fuckhtml
->getElementsByClassName(
"pagination",
"div"
);
if(count($pagination_wrapper) !== 0){
// found next page!
$this->fuckhtml->load($pagination_wrapper[0]);
$a =
$this->fuckhtml
->getElementsByTagName(
"a"
);
if(count($a) !== 0){
$q =
parse_url(
$this->fuckhtml
->getTextContent(
$a[count($a) - 1]
["attributes"]
["href"]
),
PHP_URL_QUERY
);
$out["npt"] =
$this->backend
->store(
$q . "&c=" . $get["country"],
"web",
$proxy
);
}
}
return $out;
}
private function titledots($title){
return trim($title, " .\t\n\r\0\x0B");
}
}

View File

@ -91,6 +91,10 @@ $settings = [
"value" => "qwant",
"text" => "Qwant"
],
[
"value" => "ghostery",
"text" => "Ghostery"
],
[
"value" => "yep",
"text" => "Yep"
@ -137,6 +141,10 @@ $settings = [
"value" => "qwant",
"text" => "Qwant"
],
[
"value" => "ghostery",
"text" => "Ghostery"
],
[
"value" => "yep",
"text" => "Yep"