lolcat
/
4get
6
13
Fork 4
Code Issues 13 Pull Requests Packages Projects Releases Wiki Activity

yep scraper cloudflare error handling

This commit is contained in:
lolcat 2024-05-23 08:58:46 -04:00
parent bcb5c4d519
commit 92d0102738
1 changed files with 38 additions and 16 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

54
scraper/yep.php
View File

@ -6,6 +6,9 @@ class yep{
include "lib/backend.php"; include "lib/backend.php";
$this->backend = new backend("yep"); $this->backend = new backend("yep");
include "lib/fuckhtml.php";
$this->fuckhtml = new fuckhtml();
} }
public function getfilters($page){ public function getfilters($page){
@ -254,8 +257,10 @@ class yep{
["User-Agent: " . config::USER_AGENT, ["User-Agent: " . config::USER_AGENT,
"Accept: */*", "Accept: */*",
"Accept-Language: en-US,en;q=0.5", "Accept-Language: en-US,en;q=0.5",
"Accept-Encoding: gzip", "Accept-Encoding: gzip, deflate, br, zstd",
"Connection: keep-alive",
"DNT: 1", "DNT: 1",
"Priority: u=1",
"Origin: https://yep.com", "Origin: https://yep.com",
"Referer: https://yep.com/", "Referer: https://yep.com/",
"Connection: keep-alive", "Connection: keep-alive",
@ -265,6 +270,9 @@ class yep{
"TE: trailers"] "TE: trailers"]
); );
// http3 bypass
curl_setopt($curlproc, CURLOPT_HTTP_VERSION, 30);
curl_setopt($curlproc, CURLOPT_RETURNTRANSFER, true); curl_setopt($curlproc, CURLOPT_RETURNTRANSFER, true);
curl_setopt($curlproc, CURLOPT_SSL_VERIFYHOST, 2); curl_setopt($curlproc, CURLOPT_SSL_VERIFYHOST, 2);
curl_setopt($curlproc, CURLOPT_SSL_VERIFYPEER, true); curl_setopt($curlproc, CURLOPT_SSL_VERIFYPEER, true);
@ -324,27 +332,41 @@ class yep{
// https://api.yep.com/fs/2/search?client=web&gl=CA&no_correct=false&q=undefined+variable+javascript&safeSearch=off&type=web // https://api.yep.com/fs/2/search?client=web&gl=CA&no_correct=false&q=undefined+variable+javascript&safeSearch=off&type=web
$json = $json =
json_decode( $this->get(
$this->get( $this->backend->get_ip(),
$this->backend->get_ip(), "https://api.yep.com/fs/2/search",
"https://api.yep.com/fs/2/search", [
[ "client" => "web",
"client" => "web", "gl" => $country == "all" ? $country : strtoupper($country),
"gl" => $country == "all" ? $country : strtoupper($country), "limit" => "99999",
"limit" => "99999", "no_correct" => "false",
"no_correct" => "false", "q" => $search,
"q" => $search, "safeSearch" => $nsfw,
"safeSearch" => $nsfw, "type" => "web"
"type" => "web" ]
]
),
true
); );
}catch(Exception $error){ }catch(Exception $error){
throw new Exception("Failed to fetch JSON"); throw new Exception("Failed to fetch JSON");
} }
// detect cloudflare page
$this->fuckhtml->load($json);
if(
count(
$this->fuckhtml
->getElementsByClassName(
"cf-wrapper",
"div"
)
) !== 0
){
throw new Exception("Blocked by Cloudflare");
}
$json = json_decode($json, true);
//$json = json_decode(file_get_contents("scraper/yep.json"), true); //$json = json_decode(file_get_contents("scraper/yep.json"), true);
if($json === null){ if($json === null){