mirror-4get/scraper/google.php

1930 lines
42 KiB
PHP
Raw Normal View History

2023-07-22 20:41:14 +02:00
<?php
class google{
private const is_class = ".";
private const is_id = "#";
public function __construct(){
include "lib/fuckhtml.php";
$this->fuckhtml = new fuckhtml();
2023-11-07 14:04:56 +01:00
include "lib/backend.php";
$this->backend = new backend("google");
2023-07-22 20:41:14 +02:00
}
public function getfilters($page){
switch($page){
2023-07-27 01:03:06 +02:00
case "web":
case "videos":
case "news":
2023-07-22 20:41:14 +02:00
return [
2023-07-27 01:03:06 +02:00
"country" => [ // gl=<country>
2023-07-22 20:41:14 +02:00
"display" => "Country",
"option" => [
2023-07-27 01:03:06 +02:00
"any" => "Instance's country",
2023-07-22 20:41:14 +02:00
"af" => "Afghanistan",
"al" => "Albania",
"dz" => "Algeria",
"as" => "American Samoa",
"ad" => "Andorra",
"ao" => "Angola",
2023-07-27 01:03:06 +02:00
"ai" => "Anguilla",
"aq" => "Antarctica",
"ag" => "Antigua and Barbuda",
2023-07-22 20:41:14 +02:00
"ar" => "Argentina",
"am" => "Armenia",
2023-07-27 01:03:06 +02:00
"aw" => "Aruba",
2023-07-22 20:41:14 +02:00
"au" => "Australia",
"at" => "Austria",
"az" => "Azerbaijan",
"bs" => "Bahamas",
"bh" => "Bahrain",
"bd" => "Bangladesh",
2023-07-27 01:03:06 +02:00
"bb" => "Barbados",
2023-07-22 20:41:14 +02:00
"by" => "Belarus",
"be" => "Belgium",
"bz" => "Belize",
"bj" => "Benin",
2023-07-27 01:03:06 +02:00
"bm" => "Bermuda",
2023-07-22 20:41:14 +02:00
"bt" => "Bhutan",
"bo" => "Bolivia",
2023-07-27 01:03:06 +02:00
"ba" => "Bosnia and Herzegovina",
2023-07-22 20:41:14 +02:00
"bw" => "Botswana",
2023-07-27 01:03:06 +02:00
"bv" => "Bouvet Island",
2023-07-22 20:41:14 +02:00
"br" => "Brazil",
2023-07-27 01:03:06 +02:00
"io" => "British Indian Ocean Territory",
"bn" => "Brunei Darussalam",
2023-07-22 20:41:14 +02:00
"bg" => "Bulgaria",
"bf" => "Burkina Faso",
"bi" => "Burundi",
"kh" => "Cambodia",
"cm" => "Cameroon",
"ca" => "Canada",
"cv" => "Cape Verde",
2023-07-27 01:03:06 +02:00
"ky" => "Cayman Islands",
2023-07-22 20:41:14 +02:00
"cf" => "Central African Republic",
"td" => "Chad",
"cl" => "Chile",
2023-07-27 01:03:06 +02:00
"cn" => "China",
"cx" => "Christmas Island",
"cc" => "Cocos (Keeling) Islands",
2023-07-22 20:41:14 +02:00
"co" => "Colombia",
2023-07-27 01:03:06 +02:00
"km" => "Comoros",
"cg" => "Congo",
"cd" => "Congo, the Democratic Republic",
2023-07-22 20:41:14 +02:00
"ck" => "Cook Islands",
"cr" => "Costa Rica",
2023-07-27 01:03:06 +02:00
"ci" => "Cote D'ivoire",
2023-07-22 20:41:14 +02:00
"hr" => "Croatia",
"cu" => "Cuba",
"cy" => "Cyprus",
2023-07-27 01:03:06 +02:00
"cz" => "Czech Republic",
2023-07-22 20:41:14 +02:00
"dk" => "Denmark",
"dj" => "Djibouti",
"dm" => "Dominica",
"do" => "Dominican Republic",
"ec" => "Ecuador",
"eg" => "Egypt",
"sv" => "El Salvador",
2023-07-27 01:03:06 +02:00
"gq" => "Equatorial Guinea",
"er" => "Eritrea",
2023-07-22 20:41:14 +02:00
"ee" => "Estonia",
"et" => "Ethiopia",
2023-07-27 01:03:06 +02:00
"fk" => "Falkland Islands (Malvinas)",
"fo" => "Faroe Islands",
2023-07-22 20:41:14 +02:00
"fj" => "Fiji",
"fi" => "Finland",
"fr" => "France",
2023-07-27 01:03:06 +02:00
"gf" => "French Guiana",
"pf" => "French Polynesia",
"tf" => "French Southern Territories",
2023-07-22 20:41:14 +02:00
"ga" => "Gabon",
"gm" => "Gambia",
"ge" => "Georgia",
"de" => "Germany",
"gh" => "Ghana",
"gi" => "Gibraltar",
"gr" => "Greece",
"gl" => "Greenland",
2023-07-27 01:03:06 +02:00
"gd" => "Grenada",
"gp" => "Guadeloupe",
"gu" => "Guam",
2023-07-22 20:41:14 +02:00
"gt" => "Guatemala",
2023-07-27 01:03:06 +02:00
"gn" => "Guinea",
"gw" => "Guinea-Bissau",
2023-07-22 20:41:14 +02:00
"gy" => "Guyana",
"ht" => "Haiti",
2023-07-27 01:03:06 +02:00
"hm" => "Heard Island and Mcdonald Islands",
"va" => "Holy See (Vatican City State)",
2023-07-22 20:41:14 +02:00
"hn" => "Honduras",
"hk" => "Hong Kong",
"hu" => "Hungary",
"is" => "Iceland",
"in" => "India",
"id" => "Indonesia",
2023-07-27 01:03:06 +02:00
"ir" => "Iran, Islamic Republic",
2023-07-22 20:41:14 +02:00
"iq" => "Iraq",
"ie" => "Ireland",
"il" => "Israel",
"it" => "Italy",
"jm" => "Jamaica",
"jp" => "Japan",
"jo" => "Jordan",
"kz" => "Kazakhstan",
"ke" => "Kenya",
"ki" => "Kiribati",
2023-07-27 01:03:06 +02:00
"kp" => "Korea, Democratic People's Republic",
"kr" => "Korea, Republic",
2023-07-22 20:41:14 +02:00
"kw" => "Kuwait",
"kg" => "Kyrgyzstan",
2023-07-27 01:03:06 +02:00
"la" => "Lao People's Democratic Republic",
2023-07-22 20:41:14 +02:00
"lv" => "Latvia",
"lb" => "Lebanon",
"ls" => "Lesotho",
2023-07-27 01:03:06 +02:00
"lr" => "Liberia",
"ly" => "Libyan Arab Jamahiriya",
2023-07-22 20:41:14 +02:00
"li" => "Liechtenstein",
"lt" => "Lithuania",
"lu" => "Luxembourg",
2023-07-27 01:03:06 +02:00
"mo" => "Macao",
"mk" => "Macedonia, the Former Yugosalv Republic",
2023-07-22 20:41:14 +02:00
"mg" => "Madagascar",
"mw" => "Malawi",
"my" => "Malaysia",
"mv" => "Maldives",
"ml" => "Mali",
"mt" => "Malta",
2023-07-27 01:03:06 +02:00
"mh" => "Marshall Islands",
"mq" => "Martinique",
"mr" => "Mauritania",
2023-07-22 20:41:14 +02:00
"mu" => "Mauritius",
2023-07-27 01:03:06 +02:00
"yt" => "Mayotte",
2023-07-22 20:41:14 +02:00
"mx" => "Mexico",
2023-07-27 01:03:06 +02:00
"fm" => "Micronesia, Federated States",
"md" => "Moldova, Republic",
"mc" => "Monaco",
2023-07-22 20:41:14 +02:00
"mn" => "Mongolia",
2023-07-27 01:03:06 +02:00
"ms" => "Montserrat",
2023-07-22 20:41:14 +02:00
"ma" => "Morocco",
"mz" => "Mozambique",
2023-07-27 01:03:06 +02:00
"mm" => "Myanmar",
2023-07-22 20:41:14 +02:00
"na" => "Namibia",
"nr" => "Nauru",
"np" => "Nepal",
"nl" => "Netherlands",
2023-07-27 01:03:06 +02:00
"an" => "Netherlands Antilles",
"nc" => "New Caledonia",
2023-07-22 20:41:14 +02:00
"nz" => "New Zealand",
"ni" => "Nicaragua",
"ne" => "Niger",
"ng" => "Nigeria",
"nu" => "Niue",
2023-07-27 01:03:06 +02:00
"nf" => "Norfolk Island",
"mp" => "Northern Mariana Islands",
2023-07-22 20:41:14 +02:00
"no" => "Norway",
"om" => "Oman",
"pk" => "Pakistan",
2023-07-27 01:03:06 +02:00
"pw" => "Palau",
"ps" => "Palestinian Territory, Occupied",
2023-07-22 20:41:14 +02:00
"pa" => "Panama",
"pg" => "Papua New Guinea",
"py" => "Paraguay",
"pe" => "Peru",
"ph" => "Philippines",
2023-07-27 01:03:06 +02:00
"pn" => "Pitcairn",
2023-07-22 20:41:14 +02:00
"pl" => "Poland",
"pt" => "Portugal",
"pr" => "Puerto Rico",
"qa" => "Qatar",
2023-07-27 01:03:06 +02:00
"re" => "Reunion",
2023-07-22 20:41:14 +02:00
"ro" => "Romania",
2023-07-27 01:03:06 +02:00
"ru" => "Russian Federation",
2023-07-22 20:41:14 +02:00
"rw" => "Rwanda",
2023-07-27 01:03:06 +02:00
"sh" => "Saint Helena",
"kn" => "Saint Kitts and Nevis",
"lc" => "Saint Lucia",
"pm" => "Saint Pierre and Miquelon",
"vc" => "Saint Vincent and the Grenadines",
2023-07-22 20:41:14 +02:00
"ws" => "Samoa",
"sm" => "San Marino",
2023-07-27 01:03:06 +02:00
"st" => "Sao Tome and Principe",
2023-07-22 20:41:14 +02:00
"sa" => "Saudi Arabia",
"sn" => "Senegal",
2023-07-27 01:03:06 +02:00
"cs" => "Serbia and Montenegro",
2023-07-22 20:41:14 +02:00
"sc" => "Seychelles",
"sl" => "Sierra Leone",
"sg" => "Singapore",
"sk" => "Slovakia",
"si" => "Slovenia",
"sb" => "Solomon Islands",
"so" => "Somalia",
"za" => "South Africa",
2023-07-27 01:03:06 +02:00
"gs" => "South Georgia and the South Sandwich Islands",
2023-07-22 20:41:14 +02:00
"es" => "Spain",
"lk" => "Sri Lanka",
2023-07-27 01:03:06 +02:00
"sd" => "Sudan",
2023-07-22 20:41:14 +02:00
"sr" => "Suriname",
2023-07-27 01:03:06 +02:00
"sj" => "Svalbard and Jan Mayen",
"sz" => "Swaziland",
2023-07-22 20:41:14 +02:00
"se" => "Sweden",
"ch" => "Switzerland",
2023-07-27 01:03:06 +02:00
"sy" => "Syrian Arab Republic",
"tw" => "Taiwan, Province of China",
2023-07-22 20:41:14 +02:00
"tj" => "Tajikistan",
2023-07-27 01:03:06 +02:00
"tz" => "Tanzania, United Republic",
2023-07-22 20:41:14 +02:00
"th" => "Thailand",
"tl" => "Timor-Leste",
"tg" => "Togo",
2023-07-27 01:03:06 +02:00
"tk" => "Tokelau",
2023-07-22 20:41:14 +02:00
"to" => "Tonga",
2023-07-27 01:03:06 +02:00
"tt" => "Trinidad and Tobago",
2023-07-22 20:41:14 +02:00
"tn" => "Tunisia",
2023-07-27 01:03:06 +02:00
"tr" => "Turkey",
2023-07-22 20:41:14 +02:00
"tm" => "Turkmenistan",
2023-07-27 01:03:06 +02:00
"tc" => "Turks and Caicos Islands",
"tv" => "Tuvalu",
2023-07-22 20:41:14 +02:00
"ug" => "Uganda",
"ua" => "Ukraine",
"ae" => "United Arab Emirates",
2023-07-27 01:03:06 +02:00
"uk" => "United Kingdom",
2023-07-22 20:41:14 +02:00
"us" => "United States",
2023-07-27 01:03:06 +02:00
"um" => "United States Minor Outlying Islands",
2023-07-22 20:41:14 +02:00
"uy" => "Uruguay",
"uz" => "Uzbekistan",
"vu" => "Vanuatu",
"ve" => "Venezuela",
2023-07-27 01:03:06 +02:00
"vn" => "Viet Nam",
"vg" => "Virgin Islands, British",
"vi" => "Virgin Islands, U.S.",
"wf" => "Wallis and Futuna",
"eh" => "Western Sahara",
"ye" => "Yemen",
2023-07-22 20:41:14 +02:00
"zm" => "Zambia",
"zw" => "Zimbabwe"
]
],
"nsfw" => [
"display" => "NSFW",
"option" => [
2023-07-27 01:03:06 +02:00
"yes" => "Yes", // safe=active
"no" => "No" // safe=off
2023-07-22 20:41:14 +02:00
]
],
2023-07-27 01:03:06 +02:00
"lang" => [ // lr=<lang> (prefix lang with "lang_")
2023-07-22 20:41:14 +02:00
"display" => "Language",
"option" => [
"any" => "Any language",
2023-07-27 01:03:06 +02:00
"ar" => "Arabic",
"bg" => "Bulgarian",
"ca" => "Catalan",
"cs" => "Czech",
"da" => "Danish",
"de" => "German",
"el" => "Greek",
2023-07-22 20:41:14 +02:00
"en" => "English",
2023-07-27 01:03:06 +02:00
"es" => "Spanish",
"et" => "Estonian",
"fi" => "Finnish",
"fr" => "French",
"hr" => "Croatian",
"hu" => "Hungarian",
"id" => "Indonesian",
"is" => "Icelandic",
"it" => "Italian",
"iw" => "Hebrew",
"ja" => "Japanese",
"ko" => "Korean",
"lt" => "Lithuanian",
"lv" => "Latvian",
"nl" => "Dutch",
"no" => "Norwegian",
"pl" => "Polish",
"pt" => "Portuguese",
"ro" => "Romanian",
"ru" => "Russian",
"sk" => "Slovak",
"sl" => "Slovenian",
"sr" => "Serbian",
"sv" => "Swedish",
"tr" => "Turkish",
"zh-CN" => "Chinese (Simplified)",
"zh-TW" => "Chinese (Traditional)"
2023-07-22 20:41:14 +02:00
]
],
2023-07-27 01:03:06 +02:00
"newer" => [ // &sort=review-date:r:20090301:20090430
"display" => "Newer than",
"option" => "_DATE"
2023-07-22 20:41:14 +02:00
],
2023-07-27 01:03:06 +02:00
"older" => [
"display" => "Older than",
"option" => "_DATE"
2023-07-22 20:41:14 +02:00
]
2023-07-27 01:03:06 +02:00
];
2023-07-22 20:41:14 +02:00
break;
case "images":
return [
"country" => [ // gl=<country>
"display" => "Country",
"option" => [
"any" => "Instance's country",
"af" => "Afghanistan",
"al" => "Albania",
"dz" => "Algeria",
"as" => "American Samoa",
"ad" => "Andorra",
"ao" => "Angola",
"ai" => "Anguilla",
"aq" => "Antarctica",
"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",
"bm" => "Bermuda",
"bt" => "Bhutan",
"bo" => "Bolivia",
"ba" => "Bosnia and Herzegovina",
"bw" => "Botswana",
"bv" => "Bouvet Island",
"br" => "Brazil",
"io" => "British Indian Ocean Territory",
"bn" => "Brunei Darussalam",
"bg" => "Bulgaria",
"bf" => "Burkina Faso",
"bi" => "Burundi",
"kh" => "Cambodia",
"cm" => "Cameroon",
"ca" => "Canada",
"cv" => "Cape Verde",
"ky" => "Cayman Islands",
"cf" => "Central African Republic",
"td" => "Chad",
"cl" => "Chile",
"cn" => "China",
"cx" => "Christmas Island",
"cc" => "Cocos (Keeling) Islands",
"co" => "Colombia",
"km" => "Comoros",
"cg" => "Congo",
2023-07-27 01:03:06 +02:00
"cd" => "Congo, the Democratic Republic",
2023-07-22 20:41:14 +02:00
"ck" => "Cook Islands",
"cr" => "Costa Rica",
"ci" => "Cote D'ivoire",
"hr" => "Croatia",
"cu" => "Cuba",
"cy" => "Cyprus",
"cz" => "Czech Republic",
"dk" => "Denmark",
"dj" => "Djibouti",
"dm" => "Dominica",
"do" => "Dominican Republic",
"ec" => "Ecuador",
"eg" => "Egypt",
"sv" => "El Salvador",
"gq" => "Equatorial Guinea",
"er" => "Eritrea",
"ee" => "Estonia",
"et" => "Ethiopia",
"fk" => "Falkland Islands (Malvinas)",
"fo" => "Faroe Islands",
"fj" => "Fiji",
"fi" => "Finland",
"fr" => "France",
"gf" => "French Guiana",
"pf" => "French Polynesia",
"tf" => "French Southern Territories",
"ga" => "Gabon",
"gm" => "Gambia",
"ge" => "Georgia",
"de" => "Germany",
"gh" => "Ghana",
"gi" => "Gibraltar",
"gr" => "Greece",
"gl" => "Greenland",
"gd" => "Grenada",
"gp" => "Guadeloupe",
"gu" => "Guam",
"gt" => "Guatemala",
"gn" => "Guinea",
"gw" => "Guinea-Bissau",
"gy" => "Guyana",
"ht" => "Haiti",
"hm" => "Heard Island and Mcdonald Islands",
"va" => "Holy See (Vatican City State)",
"hn" => "Honduras",
"hk" => "Hong Kong",
"hu" => "Hungary",
"is" => "Iceland",
"in" => "India",
"id" => "Indonesia",
2023-07-27 01:03:06 +02:00
"ir" => "Iran, Islamic Republic",
2023-07-22 20:41:14 +02:00
"iq" => "Iraq",
"ie" => "Ireland",
"il" => "Israel",
"it" => "Italy",
"jm" => "Jamaica",
"jp" => "Japan",
"jo" => "Jordan",
"kz" => "Kazakhstan",
"ke" => "Kenya",
"ki" => "Kiribati",
2023-07-27 01:03:06 +02:00
"kp" => "Korea, Democratic People's Republic",
"kr" => "Korea, Republic",
2023-07-22 20:41:14 +02:00
"kw" => "Kuwait",
"kg" => "Kyrgyzstan",
"la" => "Lao People's Democratic Republic",
"lv" => "Latvia",
"lb" => "Lebanon",
"ls" => "Lesotho",
"lr" => "Liberia",
"ly" => "Libyan Arab Jamahiriya",
"li" => "Liechtenstein",
"lt" => "Lithuania",
"lu" => "Luxembourg",
"mo" => "Macao",
2023-07-27 01:03:06 +02:00
"mk" => "Macedonia, the Former Yugosalv Republic",
2023-07-22 20:41:14 +02:00
"mg" => "Madagascar",
"mw" => "Malawi",
"my" => "Malaysia",
"mv" => "Maldives",
"ml" => "Mali",
"mt" => "Malta",
"mh" => "Marshall Islands",
"mq" => "Martinique",
"mr" => "Mauritania",
"mu" => "Mauritius",
"yt" => "Mayotte",
"mx" => "Mexico",
2023-07-27 01:03:06 +02:00
"fm" => "Micronesia, Federated States",
"md" => "Moldova, Republic",
2023-07-22 20:41:14 +02:00
"mc" => "Monaco",
"mn" => "Mongolia",
"ms" => "Montserrat",
"ma" => "Morocco",
"mz" => "Mozambique",
"mm" => "Myanmar",
"na" => "Namibia",
"nr" => "Nauru",
"np" => "Nepal",
"nl" => "Netherlands",
"an" => "Netherlands Antilles",
"nc" => "New Caledonia",
"nz" => "New Zealand",
"ni" => "Nicaragua",
"ne" => "Niger",
"ng" => "Nigeria",
"nu" => "Niue",
"nf" => "Norfolk Island",
"mp" => "Northern Mariana Islands",
"no" => "Norway",
"om" => "Oman",
"pk" => "Pakistan",
"pw" => "Palau",
"ps" => "Palestinian Territory, Occupied",
"pa" => "Panama",
"pg" => "Papua New Guinea",
"py" => "Paraguay",
"pe" => "Peru",
"ph" => "Philippines",
"pn" => "Pitcairn",
"pl" => "Poland",
"pt" => "Portugal",
"pr" => "Puerto Rico",
"qa" => "Qatar",
"re" => "Reunion",
"ro" => "Romania",
"ru" => "Russian Federation",
"rw" => "Rwanda",
"sh" => "Saint Helena",
"kn" => "Saint Kitts and Nevis",
"lc" => "Saint Lucia",
"pm" => "Saint Pierre and Miquelon",
"vc" => "Saint Vincent and the Grenadines",
"ws" => "Samoa",
"sm" => "San Marino",
"st" => "Sao Tome and Principe",
"sa" => "Saudi Arabia",
"sn" => "Senegal",
"cs" => "Serbia and Montenegro",
"sc" => "Seychelles",
"sl" => "Sierra Leone",
"sg" => "Singapore",
"sk" => "Slovakia",
"si" => "Slovenia",
"sb" => "Solomon Islands",
"so" => "Somalia",
"za" => "South Africa",
"gs" => "South Georgia and the South Sandwich Islands",
"es" => "Spain",
"lk" => "Sri Lanka",
"sd" => "Sudan",
"sr" => "Suriname",
"sj" => "Svalbard and Jan Mayen",
"sz" => "Swaziland",
"se" => "Sweden",
"ch" => "Switzerland",
"sy" => "Syrian Arab Republic",
"tw" => "Taiwan, Province of China",
"tj" => "Tajikistan",
2023-07-27 01:03:06 +02:00
"tz" => "Tanzania, United Republic",
2023-07-22 20:41:14 +02:00
"th" => "Thailand",
"tl" => "Timor-Leste",
"tg" => "Togo",
"tk" => "Tokelau",
"to" => "Tonga",
"tt" => "Trinidad and Tobago",
"tn" => "Tunisia",
"tr" => "Turkey",
"tm" => "Turkmenistan",
"tc" => "Turks and Caicos Islands",
"tv" => "Tuvalu",
"ug" => "Uganda",
"ua" => "Ukraine",
"ae" => "United Arab Emirates",
"uk" => "United Kingdom",
"us" => "United States",
"um" => "United States Minor Outlying Islands",
"uy" => "Uruguay",
"uz" => "Uzbekistan",
"vu" => "Vanuatu",
"ve" => "Venezuela",
"vn" => "Viet Nam",
"vg" => "Virgin Islands, British",
"vi" => "Virgin Islands, U.S.",
"wf" => "Wallis and Futuna",
"eh" => "Western Sahara",
"ye" => "Yemen",
"zm" => "Zambia",
"zw" => "Zimbabwe"
]
],
"nsfw" => [
"display" => "NSFW",
"option" => [
"yes" => "Yes", // safe=active
"no" => "No" // safe=off
]
],
"lang" => [ // lr=<lang> (prefix lang with "lang_")
"display" => "Language",
"option" => [
"any" => "Any language",
"ar" => "Arabic",
"bg" => "Bulgarian",
"ca" => "Catalan",
"cs" => "Czech",
"da" => "Danish",
"de" => "German",
"el" => "Greek",
"en" => "English",
"es" => "Spanish",
"et" => "Estonian",
"fi" => "Finnish",
"fr" => "French",
"hr" => "Croatian",
"hu" => "Hungarian",
"id" => "Indonesian",
"is" => "Icelandic",
"it" => "Italian",
"iw" => "Hebrew",
"ja" => "Japanese",
"ko" => "Korean",
"lt" => "Lithuanian",
"lv" => "Latvian",
"nl" => "Dutch",
"no" => "Norwegian",
"pl" => "Polish",
"pt" => "Portuguese",
"ro" => "Romanian",
"ru" => "Russian",
"sk" => "Slovak",
"sl" => "Slovenian",
"sr" => "Serbian",
"sv" => "Swedish",
"tr" => "Turkish",
"zh-CN" => "Chinese (Simplified)",
"zh-TW" => "Chinese (Traditional)"
]
],
"time" => [ // tbs=qrd:<size>
"display" => "Time posted",
"option" => [
"any" => "Any time",
"d" => "Past 24 hours",
"w" => "Past week",
"m" => "Past month",
"y" => "Past year"
]
2023-07-22 20:41:14 +02:00
],
"size" => [
2023-07-22 20:41:14 +02:00
"display" => "Size",
"option" => [
// tbs=isz:<size>
2023-07-22 20:41:14 +02:00
"any" => "Any size",
"l" => "Large",
"m" => "Medium",
"i" => "Icon",
// from here
// tbz:lt,islt:<size>
"qsvga" => "Larger than 400x300",
"vga" => "Larger than 640x480",
"qsvga" => "Larger than 800x600",
"xga" => "Larger than 1024x768",
"2mp" => "Larger than 2MP",
"4mp" => "Larger than 4MP",
"6mp" => "Larger than 6MP",
"8mp" => "Larger than 8MP",
"10mp" => "Larger than 10MP",
"12mp" => "Larger than 12MP",
"15mp" => "Larger than 15MP",
"20mp" => "Larger than 20MP",
"40mp" => "Larger than 40MP",
"70mp" => "Larger than 70MP"
2023-07-22 20:41:14 +02:00
]
],
"ratio" => [ // tbs=iar:<size>
"display" => "Aspect ratio",
2023-07-27 01:03:06 +02:00
"option" => [
"any" => "Any ratio",
"t" => "Tall",
"s" => "Square",
"w" => "Wide",
"xw" => "Panoramic"
2023-07-27 01:03:06 +02:00
]
],
"color" => [ // tbs=ic:<color>
2023-07-22 20:41:14 +02:00
"display" => "Color",
"option" => [
"any" => "Any color",
"color" => "Full color",
"gray" => "Black & white",
"trans" => "Transparent",
// from there, its ic:specific,isc:<color>
2023-07-22 20:41:14 +02:00
"red" => "Red",
"orange" => "Orange",
"yellow" => "Yellow",
"green" => "Green",
"teal" => "Teal",
"blue" => "Blue",
"purple" => "Purple",
"pink" => "Pink",
"white" => "White",
"gray" => "Gray",
"black" => "Black",
"brown" => "Brown"
]
],
"type" => [ // tbs=itp:<type>
2023-07-22 20:41:14 +02:00
"display" => "Type",
"option" => [
"any" => "Any type",
2023-07-27 01:03:06 +02:00
"face" => "Faces",
2023-07-22 20:41:14 +02:00
"clipart" => "Clip Art",
"lineart" => "Line Drawing",
2023-07-27 01:03:06 +02:00
"stock" => "Stock",
"animated" => "Animated"
2023-07-22 20:41:14 +02:00
]
],
"format" => [ // tbs=ift:<format>
"display" => "Format",
"option" => [
"any" => "Any format",
"jpg" => "JPG",
"gif" => "GIF",
"png" => "PNG",
"bmp" => "BMP",
"svg" => "SVG",
"webp" => "WEBP",
"ico" => "ICO",
"craw" => "RAW"
]
],
2023-07-22 20:41:14 +02:00
"rights" => [ // tbs=il:<rights>
"display" => "Usage rights",
"option" => [
"any" => "Any license",
2023-07-22 20:41:14 +02:00
"cl" => "Creative Commons licenses",
"ol" => "Commercial & other licenses"
]
]
];
break;
}
}
2023-11-07 14:04:56 +01:00
private function get($proxy, $url, $get = []){
2023-07-22 20:41:14 +02:00
$headers = [
"User-Agent: Mozilla/5.0 (Linux; U; Android 2.3.3; pt-pt; LG-P500h-parrot Build/GRI40) AppleWebKit/533.1 (KHTML, like Gecko) Version/4.0 Mobile Safari/533.1 MMS/LG-Android-MMS-V1.0/1.2",
"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",
"DNT: 1",
"Connection: keep-alive",
"Upgrade-Insecure-Requests: 1",
"Sec-Fetch-Dest: document",
"Sec-Fetch-Mode: navigate",
"Sec-Fetch-Site: none",
"Sec-Fetch-User: ?1"
];
$curlproc = curl_init();
if($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, $headers);
curl_setopt($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 14:04:56 +01:00
$this->backend->assign_proxy($curlproc, $proxy);
2023-07-22 20:41:14 +02:00
$data = curl_exec($curlproc);
if(curl_errno($curlproc)){
throw new Exception(curl_error($curlproc));
}
curl_close($curlproc);
return $data;
}
2023-11-07 14:04:56 +01:00
/*
2023-07-22 20:41:14 +02:00
public function web($get){
2023-07-27 01:03:06 +02:00
$search = $get["s"];
$country = $get["country"];
$nsfw = $get["nsfw"];
$lang = $get["lang"];
$older = $get["older"];
$newer = $get["newer"];
$params = [
"num" => 20 // get 20 results
];
// country
if($country != "any"){
$params["gl"] = $country;
}
// nsfw
$params["safe"] = $nsfw == "yes" ? "off" : "active";
// language
if($lang != "any"){
$params["lr"] = "lang_" . $lang;
}
// &sort=review-date:r:20090301:20090430
$older = $older === false ? false : date("Ymd", $older);
$newer = $newer === false ? false : date("Ymd", $newer);
if(
$older !== false &&
$newer === false
){
$newer = date("Ymd", time());
}
if(
$older !== false ||
$newer !== false
){
$params["sort"] = "review-date:r:" . $older . ":" . $newer;
}
2023-07-22 20:41:14 +02:00
$handle = fopen("scraper/google.html", "r");
$html = fread($handle, filesize("scraper/google.html"));
fclose($handle);
$out = [
"status" => "ok",
"spelling" => [
"type" => "no_correction",
"using" => null,
"correction" => null
],
"npt" => null,
"answer" => [],
"web" => [],
"image" => [],
"video" => [],
"news" => [],
"related" => []
];
$this->parsejavascript($html);
$containers =
$this->fuckhtml
->getElementsByClassName(
$this->findstyles(
[
"background-color" => "#fff",
"margin-bottom" => "10px",
"-webkit-box-shadow" => "0 1px 6px rgba(32,33,36,0.28)",
"border-radius" => "8px"
],
self::is_class
),
"div"
);
foreach($containers as $container){
$this->fuckhtml->load($container);
$title =
$this->fuckhtml
->getElementsByClassName(
$this->findstyles(
[
"color" => "#1967d2",
"font-size" => "20px",
"line-height" => "26px"
],
self::is_class
),
"div"
);
if(count($title) !== 0){
2023-11-07 14:04:56 +01:00
//
// Container is a web link
//
$web = [
"title" =>
$this->titledots(
$this->fuckhtml
->getTextContent(
$title[0]
)
),
"description" => null,
"url" =>
$this->decodeurl(
$this->fuckhtml
->getElementsByTagName("a")
[0]
["attributes"]
["href"]
),
"date" => null,
"type" => "web",
"thumb" => [
"url" => null,
"ratio" => null
],
"sublink" => [],
"table" => []
];
$container = $container["innerHTML"];
$description_container =
$this->fuckhtml
->getElementsByClassName(
$this->findstyles(
[
"padding" => "12px 16px 12px"
],
self::is_class
),
"div"
)[1];
$description =
$description_container["innerHTML"];
// get sublinks
$this->fuckhtml->load($description);
$links =
$this->fuckhtml
->getElementsByTagName("a");
$skip = true;
foreach($links as $link){
$description =
str_replace(
$link["outerHTML"],
"",
$description
);
if($skip){
$skip = false;
continue;
}
$sublink = [
"title" => null,
"description" => null,
"url" => null,
"date" => null
];
$sublink["title"] =
$this->fuckhtml
->getTextContent(
$link
);
$sublink["url"] =
$this->decodeurl(
$link
["attributes"]
["href"]
);
$web["sublink"][] = $sublink;
}
// get thumbnail before we call loadhtml again
$img =
$this->fuckhtml
->getElementsByTagName("img");
if(count($img) !== 0){
if(
isset($img[0]["attributes"]["alt"]) &&
stripos($img[0]["attributes"]["alt"], "Video for") !== false
){
// is a video thumbnail
$web["thumb"]["ratio"] = "16:9";
}else{
// is a google thumbnail
$web["thumb"]["ratio"] = "1:1";
}
$web["thumb"]["url"] =
$this->getimage(
$img[0]["attributes"]["id"]
);
}
// get table elements
$this->fuckhtml->load($description);
$levels =
$this->fuckhtml
->getElementsByClassName(
$this->findstyles(
[
"padding-bottom" => "8px"
],
self::is_class
),
"div"
);
$additional_info = [];
foreach($levels as $level){
$this->fuckhtml->load($level);
$spans =
$this->fuckhtml
->getElementsByTagName(
"span"
);
$is_rating = -2;
foreach($spans as $span){
// clean up description
$description =
str_replace(
$span["outerHTML"],
"",
$description
);
$innertext =
$this->fuckhtml
->getTextContent(
$span
);
if($innertext == ""){ continue; }
if(
strtolower($innertext)
== "rating"
){
$is_rating = -1;
continue;
}
2023-11-07 14:04:56 +01:00
//
// Parse rating object
//
if($is_rating >= -1){
if($span["level"] !== 1){ continue; }
$is_rating++;
// 10/10 (123)
if($is_rating === 0){
$innertext = explode(" ", $innertext, 2);
$web["table"]["Rating"] = $innertext[0];
$web["table"]["Hits"] =
trim(
str_replace(
[
"(",
")"
],
"",
$innertext[1]
)
);
continue;
}
// US$4.99
// MYR 50.00
// $38.34
// JP¥6,480
if($is_rating === 2){
$web["table"]["Price"] = $innertext;
continue;
}
// Android / In stock
if($is_rating === 4){
$web["table"]["Support"] = $innertext;
continue;
}
// ignore the rest
continue;
}
2023-11-07 14:04:56 +01:00
//
// Parse standalone text
//
$additional_info[] = $innertext;
}
}
for($i=0; $i<count($additional_info); $i++){
// @TODO
// generate better node names
$web["table"]["Info node #$i"] = $additional_info[$i];
}
$this->fuckhtml->load($description);
// get date node
$span =
$this->fuckhtml
->getElementsByTagName(
"span"
);
if(count($span) !== 0){
$description =
str_replace(
$span[0]["outerHTML"],
"",
$description
);
$span =
strtotime(
$this->fuckhtml
->getTextContent(
$span[0]
)
);
if($span){
$web["date"] = $span;
}
}
$web["description"] =
trim(
$this->fuckhtml
->getTextContent(
$description
),
" ·."
);
$out["web"][] = $web;
continue;
}
// check for container title header
$container_title =
$this->fuckhtml
->getElementsByClassName(
$this->findstyles(
[
"font-weight" => "bold",
"font-size" => "16px",
"color" => "#000",
"margin" => "0",
"padding" => "12px 16px 0 16px"
],
self::is_class
),
"div"
);
if(count($container_title) !== 0){
$container_title =
strtolower(
$this->fuckhtml
->getTextContent(
$container_title[0]
)
);
if(
$container_title == "related searches" ||
$container_title == "people also search for"
){
2023-11-07 14:04:56 +01:00
//
// Parse related searches
//
$as =
$this->fuckhtml
->getElementsByTagName("a");
foreach($as as $a){
$out["related"][] =
$this->fuckhtml
->getTextContent($a);
}
}
continue;
}
2023-11-07 14:04:56 +01:00
//
// Parse image carousel
//
$title_container =
$this->fuckhtml
->getElementsByClassName(
$this->findstyles(
[
"padding" => "12px 16px 12px"
],
self::is_class
),
"div"
);
if(count($title_container) !== 0){
$title_container =
strtolower(
$this->fuckhtml
->getTextContent(
$title_container[0]
)
);
if($title_container == "imagesview all"){
2023-11-07 14:04:56 +01:00
//
// Image carousel
//
$pcitem =
$this->fuckhtml
->getElementsByClassName(
"pcitem",
"div"
);
foreach($pcitem as $item){
$this->fuckhtml->load($item);
$link =
$this->fuckhtml
->getElementsByTagName(
"a"
)[0];
parse_str(
parse_url(
$this->fuckhtml
->getTextContent(
$link
["attributes"]
["href"]
),
PHP_URL_QUERY
),
$link
);
if(isset($link["tbm"])){
continue;
}
$image =
$this->fuckhtml
->getElementsByTagName("img")[0];
$title =
$this->fuckhtml
->getTextContent(
$image
["attributes"]
["alt"]
);
$image =
$this->getimage(
$image
["attributes"]
["id"]
);
$out["image"][] = [
"title" => $title,
"source" => [
[
"url" => $link["imgurl"],
"width" => (int)$link["w"],
"height" => (int)$link["h"]
],
[
"url" => $image,
"width" => (int)$link["tbnw"],
"height" => (int)$link["tbnh"]
]
],
"url" => $link["imgrefurl"]
];
}
}
}
2023-11-07 14:04:56 +01:00
//
// Get next page
//
$as =
$this->fuckhtml
->getElementsByTagName("a");
foreach($as as $a){
if(
isset($a["attributes"]["aria-label"]) &&
strtolower($a["attributes"]["aria-label"]) == "next page"
){
$out["npt"] =
$this->fuckhtml
->getTextContent(
$a["attributes"]["href"]
);
}
}
}
return $out;
2023-11-07 14:04:56 +01:00
}*/
public function image($get){
// generate parameters
if($get["npt"]){
2023-11-07 14:04:56 +01:00
[$params, $proxy] =
$this->backend->get(
$get["npt"],
"images"
);
2023-11-07 14:04:56 +01:00
$params = json_decode($params, true);
}else{
$search = $get["s"];
2023-11-07 14:04:56 +01:00
if(strlen($search) === 0){
throw new Exception("Search term is empty!");
}
$proxy = $this->backend->get_ip();
$country = $get["country"];
$nsfw = $get["nsfw"];
$lang = $get["lang"];
$time = $get["time"];
$size = $get["size"];
$ratio = $get["ratio"];
$color = $get["color"];
$type = $get["type"];
$format = $get["format"];
$rights = $get["rights"];
$params = [
"q" => $search,
"tbm" => "isch"
];
// country
if($country != "any"){
$params["gl"] = $country;
}
// nsfw
$params["safe"] = $nsfw == "yes" ? "off" : "active";
// language
if($lang != "any"){
$params["lr"] = "lang_" . $lang;
}
$tbs = [];
// time
if($time != "any"){
$tbs[] = "qrd:" . $time;
}
// size
if($size != "any"){
if(
in_array(
$size,
["l", "s", "i"]
)
){
$tbs[] = "isz:" . $size;
}else{
$tbs[] = "tbz:lt";
$tbs[] = "islt:" . $size;
}
}
// ratio
if($ratio != "any"){
$tbs[] = "iar:" . $ratio;
}
// color
if($color != "any"){
if(
in_array(
$color,
["color", "gray", "trans"]
)
){
$tbs[] = "ic:" . $color;
}else{
$tbs[] = "ic:specific";
$tbs[] = "isc:" . $color;
}
}
// type
if($type != "any"){
$tbs[] = "itp:" . $type;
}
// format
if($format != "any"){
$tbs[] = "ift:" . $format;
}
// rights
if($rights != "any"){
$tbs[] = "il:" . $rights;
}
// append tbs
if(count($tbs) !== 0){
$params["tbs"] =
implode(",", $tbs);
}
}
/*
$handle = fopen("scraper/google-img.html", "r");
$html = fread($handle, filesize("scraper/google-img.html"));
fclose($handle);*/
// scrape images
try{
$html =
$this->get(
2023-11-07 14:04:56 +01:00
$proxy,
"https://www.google.com/search",
$params
);
}catch(Exception $error){
throw new Exception("Failed to get search page");
}
$this->fuckhtml->load($html);
$out = [
"status" => "ok",
"npt" => null,
"image" => []
];
$images =
$this->fuckhtml
->getElementsByClassName(
"islrtb isv-r",
"div"
);
foreach($images as $image){
$this->fuckhtml->load($image);
$img =
$this->fuckhtml
->getElementsByTagName("img")[0];
$og_width = (int)$image["attributes"]["data-ow"];
$og_height = (int)$image["attributes"]["data-oh"];
$thumb_width = (int)$image["attributes"]["data-tw"];
$ratio = $og_width / $og_height;
if(isset($img["attributes"]["data-src"])){
$src = &$img["attributes"]["data-src"];
}else{
$src = &$img["attributes"]["src"];
}
$thumb_height = floor($thumb_width / $ratio);
$out["image"][] = [
"title" =>
$this->titledots(
$this->fuckhtml
->getTextContent(
$image["attributes"]["data-pt"]
)
),
"source" => [
[
"url" =>
$this->fuckhtml
->getTextContent(
$image["attributes"]["data-ou"]
),
"width" => $og_width,
"height" => $og_height
],
[
"url" =>
$this->fuckhtml
->getTextContent(
$src
),
"width" => $thumb_width,
"height" => $thumb_height
]
],
"url" =>
$this->fuckhtml
->getTextContent(
$image["attributes"]["data-ru"]
)
];
}
// get next page
// https://www.google.com/search
// ?q=higurashi
// &tbm=isch
// &async=_id%3Aislrg_c%2C_fmt%3Ahtml
// &asearch=ichunklite
// &ved=0ahUKEwidjYXJqJSAAxWrElkFHZ07CDwQtDIIQygA
if(count($out["image"]) !== 100){
// no more results
return $out;
}
if($get["npt"]){
// update nextpage information
$params["start"] = (int)$params["start"] + count($out["image"]);
$params["ijn"] = (int)$params["ijn"] + 1;
$out["npt"] =
2023-11-07 14:04:56 +01:00
$this->backend->store(
json_encode($params),
2023-11-07 14:04:56 +01:00
"images",
$proxy
);
}else{
// scrape nextpage information
$this->fuckhtml->load($html);
$ved =
$this->fuckhtml
->getElementById("islrg", "div");
if($ved){
$ved =
$this->fuckhtml
->getTextContent(
$ved["attributes"]["data-ved"]
);
// &vet=1{$ved}..i (10ahUKEwidjYXJqJSAAxWrElkFHZ07CDwQtDIIQygA..i)
/*
These 2 are handled by us
start = start + number of results
ijn = current page number
*/
// &start=100
// &ijn=1
// &imgvl=CAEY7gQgBSj3Aji8VTjXVUC4AUC3AUgAYNdV
preg_match(
'/var e=\'([A-z0-9]+)\';/',
$html,
$imgvl
);
2023-09-13 15:01:23 +02:00
if(isset($imgvl[1])){
$imgvl = $imgvl[1];
$params["async"] = "_id:islrg_c,_fmt:html";
$params["asearch"] = "ichunklite";
$params["ved"] = $ved;
$params["vet"] = "1" . $ved . "..i";
$params["start"] = 100;
$params["ijn"] = 1;
$params["imgvl"] = $imgvl;
$out["npt"] =
2023-11-07 14:04:56 +01:00
$this->backend->store(
2023-09-13 15:01:23 +02:00
json_encode($params),
2023-11-07 14:04:56 +01:00
"images",
$proxy
2023-09-13 15:01:23 +02:00
);
}
}
}
return $out;
}
private function hms2int($time){
$parts = explode(":", $time, 3);
$time = 0;
if(count($parts) === 3){
// hours
$time = $time + ((int)$parts[0] * 3600);
array_shift($parts);
}
if(count($parts) === 2){
// minutes
$time = $time + ((int)$parts[0] * 60);
array_shift($parts);
}
// seconds
$time = $time + (int)$parts[0];
return $time;
}
private function parsejavascript($html){
$this->fuckhtml->load($html);
2023-07-22 20:41:14 +02:00
$styles =
$this->fuckhtml
->getElementsByTagName("style");
$this->computedstyle = [];
2023-08-08 09:09:47 +02:00
$this->ask = [];
2023-07-22 20:41:14 +02:00
foreach($styles as $style){
$this->computedstyle =
array_merge(
$this->computedstyle,
$this->parsestyles($style["innerHTML"])
);
}
// get images in javascript var
preg_match(
'/google\.ldi=({[^}]+})/',
$html,
2023-07-27 01:03:06 +02:00
$this->js_image
2023-07-22 20:41:14 +02:00
);
2023-07-27 01:03:06 +02:00
if(count($this->js_image) !== 0){
2023-07-22 20:41:14 +02:00
2023-07-27 01:03:06 +02:00
$this->js_image = json_decode($this->js_image[1], true);
2023-07-22 20:41:14 +02:00
}else{
2023-07-27 01:03:06 +02:00
$this->js_image = [];
}
// additional js_images present in <script> tags
// ugh i fucking hate you
$scripts =
$this->fuckhtml
->getElementsByTagName("script");
foreach($scripts as $script){
if(!isset($script["innerHTML"])){
continue;
}
preg_match_all(
'/var s=\'(data:image[^\']+)\';var i=\[\'([^\']+)\'];/',
$script["innerHTML"],
$image_grep
);
if(count($image_grep[0]) !== 0){
$this->js_image[trim($image_grep[2][0])] =
$this->fuckhtml
->getTextContent(
$image_grep[1][0]
);
}
2023-08-08 09:09:47 +02:00
// even more javascript crap
// "People also ask" node is loaded trough javascript
preg_match_all(
'/window\.jsl\.dh\(\'([^\']+)\',\'(.+)\'\);/',
$script["innerHTML"],
$ask_grep
);
for($i=0; $i<count($ask_grep[0]); $i++){
$this->ask[trim($ask_grep[1][$i])] =
stripcslashes(
$ask_grep[2][$i]
);
}
2023-07-22 20:41:14 +02:00
}
}
private function findstyles($rules, $is){
ksort($rules);
foreach($this->computedstyle as $stylename => $styles){
if($styles == $rules){
preg_match(
'/\\' . $is . '([^ .]+)/',
$stylename,
$out
);
if(count($out) === 2){
return $out[1];
}
return false;
}
}
return false;
}
private function parsestyles($style){
// get style tags
preg_match_all(
'/([^{]+){([^}]+)}/',
$style,
$tags_regex
);
$tags = [];
for($i=0; $i<count($tags_regex[0]); $i++){
$tagnames = explode(",", trim($tags_regex[1][$i]));
foreach($tagnames as $tagname){
$tagname = trim($tagname);
if(!isset($tags[$tagname])){
$tags[$tagname] = [];
}
$values = explode(";", $tags_regex[2][$i]);
foreach($values as $value){
$value = explode(":", $value, 2);
if(count($value) !== 2){
continue;
}
$tags[$tagname][trim($value[0])] =
trim($value[1]);
}
}
}
foreach($tags as &$value){
ksort($value);
}
return $tags;
}
2023-07-27 01:03:06 +02:00
private function getimage($id){
if(
isset($this->js_image[$id]) &&
$this->js_image[$id] != "data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAABQAAAAUCAYAAACNiR0NAAABAUlEQVR4AWMYesChoYElLjkzPj4lY3d8csZjIL4MxPNjUzPcSTYsISFLAqj5NBD/h+LPQPwbiT87NCuLh2gDgRr2QzXuT0jNMoBYksARn5zuHJ+UcR0kB6RXE2VYXHJGOlTDZmzyIJcB5e+D1CSkZDgQNBAaZv+jU1JkcKpJygiGeZ0I76a/Byq8jU9NZFqaCNTA48SE33/iDcw8TIyBt0GKQTFN0Msp6f2EIyUpo57YSIlLSrMhIg0WCIBcCfXSdlzJBsheTHQ6jEnOUgEFOLaEDbMIlhZBOYrorAdJk+nroVnvPsSgdGdoOF7HZyhZ2XPoGQoqjbCpIbt0AiejIQMArVLI7k/DXFkAAAAASUVORK5CYII="
){
if(stripos($this->js_image[$id], "data:image") !== false){
return
explode(
"\\x3d",
$this->js_image[$id],
2
)[0];
}
return $this->js_image[$id];
}
return null;
}
2023-07-22 20:41:14 +02:00
private function decodeurl($url){
preg_match(
'/^\/url\?q=([^&]+)|^\/interstitial\?url=([^&]+)/',
$this->fuckhtml
->getTextContent($url),
$match
);
2023-08-08 09:09:47 +02:00
if(count($match) === 0){
2023-07-22 20:41:14 +02:00
2023-08-08 09:09:47 +02:00
return null;
}
$url = empty($match[1]) ? urldecode($match[2]) : urldecode($match[1]);
$domain = parse_url($url, PHP_URL_HOST);
if(
preg_match(
'/wikipedia.org$/',
$domain
)
){
2023-07-22 20:41:14 +02:00
2023-08-08 09:09:47 +02:00
// rewrite wikipedia mobile URLs to desktop
$url =
$this->replacedomain(
$url,
preg_replace(
'/([a-z0-9]+)(\.m\.)/',
'$1.',
$domain
)
);
2023-07-22 20:41:14 +02:00
}
2023-08-08 09:09:47 +02:00
if(
preg_match(
'/imdb\.com$|youtube\.[^.]+$/',
$domain
)
){
// rewrite imdb and youtube mobile URLs too
$url =
$this->replacedomain(
$url,
preg_replace(
'/^m\./',
"",
$domain
)
);
}
return $url;
}
private function replacedomain($url, $domain){
return
preg_replace(
'/(https?:\/\/)([^\/]+)/',
'$1' . $domain,
$url
);
2023-07-22 20:41:14 +02:00
}
private function titledots($title){
return rtrim($title, ".… \t\n\r\0\x0B");
}
}