yonderly.org/4get
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Compare commits

base: yonderly.org:2b8d90af12b5b3900633ae86643d09188ad14df6
Branches Tags
yonderly.org:master
..
compare: yonderly.org:0bdd5e73df28c570acec49754e2433ec9ac40281
Branches Tags
yonderly.org:master

No commits in common. "2b8d90af12b5b3900633ae86643d09188ad14df6" and "0bdd5e73df28c570acec49754e2433ec9ac40281" have entirely different histories.
2b8d90af12 ... 0bdd5e73df

5 changed files with 4 additions and 2266 deletions
Show Stats Expand all files Collapse all files

3
data/config.php
View File

@ -119,7 +119,7 @@ class config{
// Default user agent to use for scraper requests. Sometimes ignored to get specific webpages
// Changing this might break things.
const USER_AGENT = "Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:141.0) Gecko/20100101 Firefox/141.0";
const USER_AGENT = "Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:134.0) Gecko/20100101 Firefox/134.0";
// Proxy pool assignments for each scraper
// false = Use server's raw IP
@ -133,7 +133,6 @@ class config{
const PROXY_GOOGLE_CSE = false;
const PROXY_STARTPAGE = false;
const PROXY_QWANT = false;
const PROXY_BAIDU = false;
const PROXY_GHOSTERY = false;
const PROXY_MARGINALIA = false;
const PROXY_MOJEEK = false;

6
lib/frontend.php
View File

@ -949,7 +949,6 @@ class frontend{
"crowdview" => "Crowdview",
"mwmbl" => "Mwmbl",
"mojeek" => "Mojeek",
"baidu" => "Baidu",
"solofield" => "Solofield",
"marginalia" => "Marginalia",
"wiby" => "wiby",
@ -970,7 +969,6 @@ class frontend{
"startpage" => "Startpage",
"qwant" => "Qwant",
"yep" => "Yep",
"baidu" => "Baidu",
"solofield" => "Solofield",
"pinterest" => "Pinterest",
"flickr" => "Flickr",
@ -995,7 +993,6 @@ class frontend{
"google" => "Google",
"startpage" => "Startpage",
"qwant" => "Qwant",
"baidu" => "Baidu",
"solofield" => "Solofield"
]
];
@ -1011,8 +1008,7 @@ class frontend{
"startpage" => "Startpage",
"qwant" => "Qwant",
"yep" => "Yep",
"mojeek" => "Mojeek",
"baidu" => "Baidu"
"mojeek" => "Mojeek"
]
];
break;

16
lib/fuckhtml.php
View File

@ -240,13 +240,12 @@ class fuckhtml{
public function getElementsByFuzzyAttributeValue(string $name, string $value, $collection = null){
$elems = $this->getElementsByAttributeName($name, $collection);
$value =
explode(
" ",
trim(
preg_replace(
'/\s+/',
'/ +/',
" ",
$value
)
@ -259,18 +258,7 @@ class fuckhtml{
foreach($elem["attributes"] as $attrib_name => $attrib_value){
$attrib_value =
explode(
" ",
trim(
preg_replace(
'/\s+/',
" ",
$attrib_value
)
)
);
$attrib_value = explode(" ", $attrib_value);
$ac = count($attrib_value);
$nc = count($value);
$cr = 0;

2229
scraper/baidu.php
View File

File diff suppressed because it is too large Load Diff

16
settings.php
View File

@ -169,10 +169,6 @@ $settings = [
"value" => "mojeek",
"text" => "Mojeek"
],
[
"value" => "baidu",
"text" => "Baidu"
],
[
"value" => "solofield",
"text" => "Solofield"
@ -227,10 +223,6 @@ $settings = [
"value" => "yep",
"text" => "Yep"
],
[
"value" => "baidu",
"text" => "Baidu"
],
[
"value" => "solofield",
"text" => "Solofield"
@ -293,10 +285,6 @@ $settings = [
"value" => "qwant",
"text" => "Qwant"
],
[
"value" => "baidu",
"text" => "Baidu"
],
[
"value" => "solofield",
"text" => "Solofield"
@ -334,10 +322,6 @@ $settings = [
[
"value" => "mojeek",
"text" => "Mojeek"
],
[
"value" => "baidu",
"text" => "Baidu"
]
]
],