4get/scraper/yep.php

733 lines
16 KiB
PHP
Raw Normal View History

<?php
class yep{
public function __construct(){
2023-11-07 13:04:56 +00:00
include "lib/backend.php";
$this->backend = new backend("yep");
2024-05-23 12:58:46 +00:00
include "lib/fuckhtml.php";
$this->fuckhtml = new fuckhtml();
}
public function getfilters($page){
return [
"country" => [
"display" => "Country",
"option" => [
"all" => "All regions",
"af" => "Afghanistan",
"al" => "Albania",
"dz" => "Algeria",
"as" => "American Samoa",
"ad" => "Andorra",
"ao" => "Angola",
"ai" => "Anguilla",
"ag" => "Antigua and Barbuda",
"ar" => "Argentina",
"am" => "Armenia",
"aw" => "Aruba",
"au" => "Australia",
"at" => "Austria",
"az" => "Azerbaijan",
"bs" => "Bahamas",
"bh" => "Bahrain",
"bd" => "Bangladesh",
"bb" => "Barbados",
"by" => "Belarus",
"be" => "Belgium",
"bz" => "Belize",
"bj" => "Benin",
"bt" => "Bhutan",
"bo" => "Bolivia",
"ba" => "Bosnia and Herzegovina",
"bw" => "Botswana",
"br" => "Brazil",
"bn" => "Brunei Darussalam",
"bg" => "Bulgaria",
"bf" => "Burkina Faso",
"bi" => "Burundi",
"cv" => "Cabo Verde",
"kh" => "Cambodia",
"cm" => "Cameroon",
"ca" => "Canada",
"ky" => "Cayman Islands",
"cf" => "Central African Republic",
"td" => "Chad",
"cl" => "Chile",
"cn" => "China",
"co" => "Colombia",
"cg" => "Congo",
"cd" => "Congo, Democratic Republic",
"ck" => "Cook Islands",
"cr" => "Costa Rica",
"hr" => "Croatia",
"cu" => "Cuba",
"cy" => "Cyprus",
"cz" => "Czechia",
"ci" => "Côte d'Ivoire",
"dk" => "Denmark",
"dj" => "Djibouti",
"dm" => "Dominica",
"do" => "Dominican Republic",
"ec" => "Ecuador",
"eg" => "Egypt",
"sv" => "El Salvador",
"gq" => "Equatorial Guinea",
"ee" => "Estonia",
"et" => "Ethiopia",
"fo" => "Faroe Islands",
"fj" => "Fiji",
"fi" => "Finland",
"fr" => "France",
"gf" => "French Guiana",
"pf" => "French Polynesia",
"ga" => "Gabon",
"gm" => "Gambia",
"ge" => "Georgia",
"de" => "Germany",
"gh" => "Ghana",
"gi" => "Gibraltar",
"gr" => "Greece",
"gl" => "Greenland",
"gd" => "Grenada",
"gp" => "Guadeloupe",
"gu" => "Guam",
"gt" => "Guatemala",
"gg" => "Guernsey",
"gn" => "Guinea",
"gy" => "Guyana",
"ht" => "Haiti",
"hn" => "Honduras",
"hk" => "Hong Kong",
"hu" => "Hungary",
"is" => "Iceland",
"in" => "India",
"id" => "Indonesia",
"iq" => "Iraq",
"ie" => "Ireland",
"im" => "Isle of Man",
"il" => "Israel",
"it" => "Italy",
"jm" => "Jamaica",
"jp" => "Japan",
"je" => "Jersey",
"jo" => "Jordan",
"kz" => "Kazakhstan",
"ke" => "Kenya",
"ki" => "Kiribati",
"kw" => "Kuwait",
"kg" => "Kyrgyzstan",
"la" => "Lao People's Democratic Republic",
"lv" => "Latvia",
"lb" => "Lebanon",
"ls" => "Lesotho",
"ly" => "Libya",
"li" => "Liechtenstein",
"lt" => "Lithuania",
"lu" => "Luxembourg",
"mk" => "Macedonia",
"mg" => "Madagascar",
"mw" => "Malawi",
"my" => "Malaysia",
"mv" => "Maldives",
"ml" => "Mali",
"mt" => "Malta",
"mq" => "Martinique",
"mr" => "Mauritania",
"mu" => "Mauritius",
"yt" => "Mayotte",
"mx" => "Mexico",
"fm" => "Micronesia, Federated States of",
"md" => "Moldova",
"mc" => "Monaco",
"mn" => "Mongolia",
"me" => "Montenegro",
"ms" => "Montserrat",
"ma" => "Morocco",
"mz" => "Mozambique",
"mm" => "Myanmar",
"na" => "Namibia",
"nr" => "Nauru",
"np" => "Nepal",
"nl" => "Netherlands",
"nc" => "New Caledonia",
"nz" => "New Zealand",
"ni" => "Nicaragua",
"ne" => "Niger",
"ng" => "Nigeria",
"nu" => "Niue",
"no" => "Norway",
"om" => "Oman",
"pk" => "Pakistan",
"ps" => "Palestine, State of",
"pa" => "Panama",
"pg" => "Papua New Guinea",
"py" => "Paraguay",
"pe" => "Peru",
"ph" => "Philippines",
"pn" => "Pitcairn",
"pl" => "Poland",
"pt" => "Portugal",
"pr" => "Puerto Rico",
"qa" => "Qatar",
"ro" => "Romania",
"ru" => "Russian Federation",
"rw" => "Rwanda",
"re" => "Réunion",
"sh" => "Saint Helena",
"kn" => "Saint Kitts and Nevis",
"lc" => "Saint Lucia",
"vc" => "Saint Vincent and the Grenadines",
"ws" => "Samoa",
"sm" => "San Marino",
"st" => "Sao Tome and Principe",
"sa" => "Saudi Arabia",
"sn" => "Senegal",
"rs" => "Serbia",
"sc" => "Seychelles",
"sl" => "Sierra Leone",
"sg" => "Singapore",
"sk" => "Slovakia",
"si" => "Slovenia",
"sb" => "Solomon Islands",
"so" => "Somalia",
"kr" => "Sourth Korea",
"za" => "South Africa",
"es" => "Spain",
"lk" => "Sri Lanka",
"sr" => "Suriname",
"se" => "Sweden",
"ch" => "Switzerland",
"tw" => "Taiwan",
"tj" => "Tajikistan",
"tz" => "Tanzania",
"th" => "Thailand",
"tl" => "Timor-Leste",
"tg" => "Togo",
"tk" => "Tokelau",
"to" => "Tonga",
"tt" => "Trinidad and Tobago",
"tn" => "Tunisia",
"tr" => "Turkey",
"tm" => "Turkmenistan",
"ug" => "Uganda",
"ua" => "Ukraine",
"ae" => "United Arab Emirates",
"gb" => "United Kingdom",
"us" => "United States",
"uy" => "Uruguay",
"uz" => "Uzbekistan",
"vu" => "Vanuatu",
"ve" => "Venezuela",
"vn" => "Vietnam",
"vg" => "Virgin Islands, British",
"vi" => "Virgin Islands, U.S.",
"ye" => "Yemen",
"zm" => "Zambia",
"zw" => "Zimbabwe"
]
],
"nsfw" => [
"display" => "NSFW",
"option" => [
"yes" => "Yes",
"maybe" => "Maybe",
"no" => "No"
]
]
];
}
2023-11-07 13:04:56 +00:00
private function get($proxy, $url, $get = []){
$curlproc = curl_init();
if($get !== []){
$get = http_build_query($get);
$url .= "?" . $get;
}
curl_setopt($curlproc, CURLOPT_URL, $url);
curl_setopt($curlproc, CURLOPT_ENCODING, ""); // default encoding
curl_setopt($curlproc, CURLOPT_HTTPHEADER,
2023-11-07 13:04:56 +00:00
["User-Agent: " . config::USER_AGENT,
"Accept: */*",
"Accept-Language: en-US,en;q=0.5",
2024-05-23 12:58:46 +00:00
"Accept-Encoding: gzip, deflate, br, zstd",
"Connection: keep-alive",
"DNT: 1",
2024-05-23 12:58:46 +00:00
"Priority: u=1",
"Origin: https://yep.com",
"Referer: https://yep.com/",
"Connection: keep-alive",
"Sec-Fetch-Dest: empty",
"Sec-Fetch-Mode: cors",
2023-12-03 00:11:52 +00:00
"Sec-Fetch-Site: same-site",
"TE: trailers"]
);
curl_setopt($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);
2023-11-07 13:04:56 +00:00
$this->backend->assign_proxy($curlproc, $proxy);
$data = curl_exec($curlproc);
if(curl_errno($curlproc)){
throw new Exception(curl_error($curlproc));
}
curl_close($curlproc);
return $data;
}
2023-12-03 00:11:52 +00:00
public function web($get){
$search = $get["s"];
if(strlen($search) === 0){
throw new Exception("Search term is empty!");
}
$country = $get["country"];
$nsfw = $get["nsfw"];
switch($nsfw){
case "yes": $nsfw = "off"; break;
case "maybe": $nsfw = "moderate"; break;
case "no": $nsfw = "strict"; break;
}
$out = [
"status" => "ok",
"spelling" => [
"type" => "no_correction",
"using" => null,
"correction" => null
],
"npt" => null,
"answer" => [],
"web" => [],
"image" => [],
"video" => [],
"news" => [],
"related" => []
];
try{
// https://api.yep.com/fs/2/search?client=web&gl=CA&no_correct=false&q=undefined+variable+javascript&safeSearch=off&type=web
$json =
2024-05-23 12:58:46 +00:00
$this->get(
$this->backend->get_ip(),
"https://api.yep.com/fs/2/search",
[
"client" => "web",
"gl" => $country == "all" ? $country : strtoupper($country),
"limit" => "99999",
"no_correct" => "false",
"q" => $search,
"safeSearch" => $nsfw,
"type" => "web"
]
2023-12-03 00:11:52 +00:00
);
}catch(Exception $error){
throw new Exception("Failed to fetch JSON");
}
2024-05-23 13:05:04 +00:00
$this->detect_cf($json);
2024-05-23 12:58:46 +00:00
$json = json_decode($json, true);
2023-12-03 00:11:52 +00:00
//$json = json_decode(file_get_contents("scraper/yep.json"), true);
if($json === null){
throw new Exception("Failed to decode JSON");
}
if(isset($json[1]["correction"])){
$out["spelling"] = [
"type" => "not_many",
"using" => $search,
"correction" => $json[1]["correction"][1]
];
}
if(isset($json[1]["results"])){
foreach($json[1]["results"] as $item){
switch(strtolower($item["type"])){
case "organic":
$sublinks = [];
if(isset($item["sitelinks"]["full"])){
foreach($item["sitelinks"]["full"] as $link){
$sublinks[] = [
"title" => $link["title"],
"date" => null,
"description" =>
$this->titledots(
strip_tags(
html_entity_decode(
$link["snippet"]
)
)
),
"url" => $link["url"]
];
}
}
$out["web"][] = [
"title" => $item["title"],
"description" =>
$this->titledots(
strip_tags(
html_entity_decode(
$item["snippet"]
)
)
),
"url" => $item["url"],
"date" => strtotime($item["first_seen"]),
"type" => "web",
"thumb" => [
"url" => null,
"ratio" => null
],
"sublink" => $sublinks,
"table" => []
];
break;
}
}
}
if(isset($json[1]["featured_news"])){
foreach($json[1]["featured_news"] as $news){
$out["news"][] = [
"title" => $news["title"],
"description" =>
$this->titledots(
strip_tags(
html_entity_decode(
$news["snippet"]
)
)
),
"date" => strtotime($news["first_seen"]),
"thumb" =>
isset($news["img"]) ?
[
"url" => $this->unshiturl($news["img"]),
"ratio" => "16:9"
] :
[
"url" => null,
"ratio" => null
],
"url" => $news["url"]
];
}
}
if(isset($json[1]["featured_images"])){
foreach($json[1]["featured_images"] as $image){
if(
$image["width"] !== 0 &&
$image["height"] !== 0
){
$thumb_width = $image["width"] >= 260 ? 260 : $image["width"];
$thumb_height = ceil($image["height"] * ($thumb_width / $image["width"]));
$width = $image["width"];
$height = $image["height"];
}else{
$thumb_width = null;
$thumb_height = null;
$width = null;
$height = null;
}
$out["image"][] = [
"title" => $image["title"],
"source" => [
[
"url" => $image["image_id"],
"width" => $width,
"height" => $height
],
[
"url" => $image["src"],
"width" => $thumb_width,
"height" => $thumb_height
]
],
"url" => $image["host_page"]
];
}
}
return $out;
}
public function image($get){
$search = $get["s"];
2023-11-07 13:04:56 +00:00
if(strlen($search) === 0){
throw new Exception("Search term is empty!");
}
$country = $get["country"];
$nsfw = $get["nsfw"];
switch($nsfw){
case "yes": $nsfw = "off"; break;
case "maybe": $nsfw = "moderate"; break;
case "no": $nsfw = "strict"; break;
}
$out = [
"status" => "ok",
"npt" => null,
"image" => []
];
try{
$json =
2024-05-23 13:05:04 +00:00
$this->get(
$this->backend->get_ip(), // no nextpage!
"https://api.yep.com/fs/2/search",
[
"client" => "web",
"gl" => $country == "all" ? $country : strtoupper($country),
"no_correct" => "false",
"q" => $search,
"safeSearch" => $nsfw,
"type" => "images"
]
);
}catch(Exception $error){
throw new Exception("Failed to fetch JSON");
}
2024-05-23 13:05:04 +00:00
$this->detect_cf($json);
$json = json_decode($json, true);
if($json === null){
throw new Exception("Failed to decode JSON");
}
2023-12-03 00:11:52 +00:00
if(isset($json[1]["results"])){
foreach($json[1]["results"] as $item){
2023-12-03 00:11:52 +00:00
if(
$item["width"] !== 0 &&
$item["height"] !== 0
){
$thumb_width = $item["width"] >= 260 ? 260 : $item["width"];
$thumb_height = ceil($item["height"] * ($thumb_width / $item["width"]));
$width = $item["width"];
$height = $item["height"];
}else{
$thumb_width = null;
$thumb_height = null;
$width = null;
$height = null;
}
2023-12-03 00:11:52 +00:00
$out["image"][] = [
"title" => $item["title"],
"source" => [
[
"url" => $item["image_id"],
"width" => $width,
"height" => $height
],
[
"url" => $item["src"],
"width" => $thumb_width,
"height" => $thumb_height
]
],
"url" => $item["host_page"]
];
}
2023-12-03 00:11:52 +00:00
}
return $out;
}
public function news($get){
$search = $get["s"];
if(strlen($search) === 0){
2023-12-03 00:11:52 +00:00
throw new Exception("Search term is empty!");
}
$country = $get["country"];
$nsfw = $get["nsfw"];
switch($nsfw){
case "yes": $nsfw = "off"; break;
case "maybe": $nsfw = "moderate"; break;
case "no": $nsfw = "strict"; break;
}
$out = [
"status" => "ok",
"npt" => null,
"news" => []
];
try{
// https://api.yep.com/fs/2/search?client=web&gl=CA&no_correct=false&q=undefined+variable+javascript&safeSearch=off&type=web
$json =
2024-05-23 13:05:04 +00:00
$this->get(
$this->backend->get_ip(),
"https://api.yep.com/fs/2/search",
[
"client" => "web",
"gl" => $country == "all" ? $country : strtoupper($country),
"limit" => "99999",
"no_correct" => "false",
"q" => $search,
"safeSearch" => $nsfw,
"type" => "news"
]
2023-12-03 00:11:52 +00:00
);
}catch(Exception $error){
throw new Exception("Failed to fetch JSON");
}
2024-05-23 13:05:04 +00:00
$this->detect_cf($json);
$json = json_decode($json, true);
2023-12-03 00:11:52 +00:00
//$json = json_decode(file_get_contents("scraper/yep.json"), true);
if($json === null){
throw new Exception("Failed to decode JSON");
}
if(isset($json[1]["results"])){
foreach($json[1]["results"] as $item){
$out["news"][] = [
"title" => $item["title"],
"author" => null,
"description" =>
$this->titledots(
strip_tags(
html_entity_decode(
$item["snippet"]
)
)
),
"date" => strtotime($item["first_seen"]),
"thumb" =>
isset($item["img"]) ?
[
"url" => $this->unshiturl($item["img"]),
"ratio" => "16:9"
] :
[
"url" => null,
"ratio" => null
],
"url" => $item["url"]
];
}
}
return $out;
}
2023-12-03 00:11:52 +00:00
2024-05-23 13:05:04 +00:00
private function detect_cf($payload){
// detect cloudflare page
$this->fuckhtml->load($payload);
if(
count(
$this->fuckhtml
->getElementsByClassName(
"cf-wrapper",
"div"
)
) !== 0
){
2024-05-25 00:35:12 +00:00
throw new Exception("Blocked by Cloudflare. Please follow curl-impersonate installation instructions");
2024-05-23 13:05:04 +00:00
}
}
2023-12-03 00:11:52 +00:00
private function titledots($title){
$substr = substr($title, -4);
if(
strpos($substr, "...") !== false ||
strpos($substr, "") !== false
){
return trim(substr($title, 0, -4));
}
return trim($title);
}
private function unshiturl($url){
$newurl = parse_url($url, PHP_URL_QUERY);
parse_str($newurl, $newurl);
if(isset($newurl["url"])){
return $newurl["url"];
}
return $url;
}
}