| 1234567891011121314151617181920212223242526272829303132333435363738394041424344454647484950515253545556575859606162636465666768697071727374757677787980818283848586878889909192939495969798991001011021031041051061071081091101111121131141151161171181191201211221231241251261271281291301311321331341351361371381391401411421431441451461471481491501511521531541551561571581591601611621631641651661671681691701711721731741751761771781791801811821831841851861871881891901911921931941951961971981992002012022032042052062072082092102112122132142152162172182192202212222232242252262272282292302312322332342352362372382392402412422432442452462472482492502512522532542552562572582592602612622632642652662672682692702712722732742752762772782792802812822832842852862872882892902912922932942952962972982993003013023033043053063073083093103113123133143153163173183193203213223233243253263273283293303313323333343353363373383393403413423433443453463473483493503513523533543553563573583593603613623633643653663673683693703713723733743753763773783793803813823833843853863873883893903913923933943953963973983994004014024034044054064074084094104114124134144154164174184194204214224234244254264274284294304314324334344354364374384394404414424434444454464474484494504514524534544554564574584594604614624634644654664674684694704714724734744754764774784794804814824834844854864874884894904914924934944954964974984995005015025035045055065075085095105115125135145155165175185195205215225235245255265275285295305315325335345355365375385395405415425435445455465475485495505515525535545555565575585595605615625635645655665675685695705715725735745755765775785795805815825835845855865875885895905915925935945955965975985996006016026036046056066076086096106116126136146156166176186196206216226236246256266276286296306316326336346356366376386396406416426436446456466476486496506516526536546556566576586596606616626636646656666676686696706716726736746756766776786796806816826836846856866876886896906916926936946956966976986997007017027037047057067077087097107117127137147157167177187197207217227237247257267277287297307317327337347357367377387397407417427437447457467477487497507517527537547557567577587597607617627637647657667677687697707717727737747757767777787797807817827837847857867877887897907917927937947957967977987998008018028038048058068078088098108118128138148158168178188198208218228238248258268278288298308318328338348358368378388398408418428438448458468478488498508518528538548558568578588598608618628638648658668678688698708718728738748758768778788798808818828838848858868878888898908918928938948958968978988999009019029039049059069079089099109119129139149159169179189199209219229239249259269279289299309319329339349359369379389399409419429439449459469479489499509519529539549559569579589599609619629639649659669679689699709719729739749759769779789799809819829839849859869879889899909919929939949959969979989991000100110021003100410051006100710081009101010111012101310141015101610171018101910201021102210231024102510261027102810291030103110321033103410351036103710381039104010411042104310441045104610471048104910501051105210531054105510561057105810591060106110621063106410651066106710681069107010711072107310741075107610771078107910801081108210831084108510861087108810891090109110921093109410951096109710981099110011011102110311041105110611071108110911101111111211131114111511161117111811191120112111221123112411251126112711281129113011311132113311341135113611371138113911401141114211431144114511461147114811491150115111521153115411551156115711581159116011611162116311641165116611671168116911701171117211731174117511761177117811791180118111821183118411851186118711881189119011911192119311941195119611971198119912001201120212031204120512061207120812091210121112121213121412151216121712181219122012211222122312241225122612271228122912301231123212331234123512361237123812391240124112421243124412451246124712481249125012511252125312541255125612571258125912601261126212631264126512661267126812691270127112721273127412751276127712781279128012811282128312841285128612871288128912901291129212931294129512961297129812991300130113021303130413051306130713081309131013111312131313141315131613171318131913201321132213231324132513261327132813291330133113321333133413351336133713381339134013411342134313441345134613471348134913501351135213531354135513561357135813591360136113621363136413651366136713681369137013711372137313741375137613771378137913801381138213831384138513861387138813891390139113921393139413951396139713981399140014011402140314041405140614071408140914101411141214131414141514161417141814191420142114221423142414251426142714281429143014311432143314341435143614371438143914401441144214431444144514461447144814491450145114521453145414551456145714581459146014611462146314641465146614671468146914701471147214731474147514761477147814791480148114821483148414851486148714881489149014911492149314941495149614971498149915001501150215031504150515061507150815091510151115121513151415151516151715181519152015211522152315241525152615271528152915301531153215331534153515361537153815391540154115421543154415451546154715481549155015511552155315541555155615571558155915601561156215631564156515661567156815691570157115721573157415751576157715781579158015811582158315841585158615871588158915901591159215931594159515961597159815991600160116021603160416051606160716081609161016111612161316141615161616171618161916201621162216231624162516261627162816291630163116321633163416351636163716381639164016411642164316441645164616471648164916501651165216531654165516561657165816591660166116621663166416651666166716681669167016711672167316741675167616771678167916801681168216831684168516861687168816891690169116921693169416951696169716981699170017011702170317041705170617071708170917101711171217131714171517161717171817191720172117221723172417251726172717281729173017311732173317341735173617371738173917401741174217431744174517461747174817491750175117521753175417551756175717581759176017611762176317641765176617671768176917701771177217731774177517761777177817791780178117821783178417851786178717881789179017911792179317941795179617971798179918001801180218031804180518061807180818091810181118121813181418151816181718181819182018211822182318241825182618271828182918301831183218331834183518361837183818391840184118421843184418451846184718481849185018511852185318541855185618571858185918601861186218631864186518661867186818691870187118721873187418751876187718781879188018811882188318841885188618871888188918901891189218931894189518961897189818991900190119021903190419051906190719081909191019111912191319141915191619171918191919201921192219231924192519261927192819291930193119321933193419351936193719381939194019411942194319441945194619471948194919501951195219531954195519561957195819591960196119621963196419651966196719681969197019711972197319741975197619771978197919801981198219831984198519861987198819891990199119921993199419951996199719981999200020012002200320042005200620072008200920102011201220132014201520162017201820192020202120222023202420252026202720282029203020312032203320342035203620372038203920402041204220432044204520462047204820492050205120522053205420552056205720582059206020612062206320642065206620672068206920702071207220732074207520762077207820792080208120822083208420852086208720882089209020912092209320942095209620972098209921002101210221032104210521062107210821092110211121122113211421152116211721182119212021212122212321242125212621272128212921302131213221332134213521362137213821392140214121422143214421452146214721482149215021512152215321542155215621572158215921602161216221632164216521662167216821692170217121722173217421752176217721782179218021812182218321842185218621872188218921902191219221932194219521962197219821992200220122022203220422052206220722082209221022112212221322142215221622172218221922202221222222232224222522262227222822292230223122322233223422352236223722382239224022412242224322442245224622472248224922502251225222532254225522562257225822592260226122622263226422652266226722682269227022712272227322742275227622772278227922802281228222832284228522862287228822892290229122922293229422952296229722982299230023012302230323042305230623072308230923102311231223132314231523162317231823192320232123222323232423252326232723282329233023312332233323342335233623372338233923402341234223432344234523462347234823492350235123522353235423552356235723582359236023612362236323642365236623672368236923702371237223732374237523762377237823792380238123822383238423852386238723882389239023912392239323942395239623972398239924002401240224032404240524062407240824092410241124122413241424152416241724182419242024212422242324242425242624272428242924302431243224332434243524362437243824392440244124422443244424452446244724482449245024512452245324542455245624572458245924602461246224632464246524662467246824692470247124722473247424752476247724782479248024812482248324842485248624872488248924902491249224932494249524962497249824992500250125022503250425052506250725082509251025112512251325142515251625172518251925202521252225232524252525262527252825292530253125322533253425352536253725382539254025412542254325442545254625472548254925502551255225532554255525562557255825592560256125622563256425652566256725682569257025712572257325742575257625772578257925802581258225832584258525862587258825892590259125922593259425952596259725982599260026012602260326042605260626072608260926102611261226132614261526162617261826192620262126222623262426252626262726282629263026312632263326342635263626372638263926402641264226432644264526462647264826492650265126522653265426552656265726582659266026612662266326642665266626672668266926702671267226732674267526762677267826792680268126822683268426852686268726882689269026912692269326942695269626972698269927002701270227032704270527062707270827092710271127122713271427152716271727182719272027212722272327242725272627272728272927302731273227332734273527362737273827392740274127422743274427452746274727482749275027512752275327542755275627572758275927602761276227632764276527662767276827692770277127722773277427752776277727782779278027812782278327842785278627872788278927902791279227932794279527962797279827992800280128022803280428052806280728082809281028112812281328142815281628172818281928202821282228232824282528262827282828292830283128322833283428352836283728382839284028412842284328442845284628472848284928502851285228532854285528562857285828592860286128622863286428652866286728682869287028712872287328742875287628772878287928802881288228832884288528862887288828892890289128922893289428952896289728982899290029012902290329042905290629072908290929102911291229132914291529162917291829192920292129222923292429252926292729282929293029312932293329342935293629372938293929402941294229432944294529462947294829492950295129522953295429552956295729582959296029612962296329642965296629672968296929702971297229732974297529762977297829792980298129822983298429852986298729882989299029912992299329942995299629972998299930003001300230033004300530063007300830093010301130123013301430153016301730183019302030213022302330243025302630273028302930303031303230333034303530363037303830393040304130423043304430453046304730483049305030513052305330543055305630573058305930603061306230633064306530663067306830693070307130723073307430753076307730783079308030813082308330843085308630873088308930903091309230933094309530963097309830993100310131023103310431053106310731083109311031113112311331143115311631173118311931203121312231233124312531263127312831293130313131323133313431353136313731383139314031413142314331443145314631473148314931503151315231533154315531563157315831593160316131623163316431653166316731683169317031713172317331743175317631773178317931803181318231833184318531863187318831893190319131923193319431953196319731983199320032013202320332043205320632073208320932103211321232133214321532163217321832193220322132223223322432253226322732283229323032313232323332343235323632373238323932403241324232433244324532463247324832493250325132523253325432553256325732583259326032613262326332643265326632673268326932703271327232733274327532763277327832793280328132823283328432853286328732883289329032913292329332943295329632973298329933003301330233033304330533063307330833093310331133123313331433153316331733183319332033213322332333243325332633273328332933303331333233333334333533363337333833393340334133423343334433453346334733483349335033513352335333543355335633573358335933603361336233633364336533663367336833693370337133723373337433753376337733783379338033813382338333843385338633873388338933903391339233933394339533963397339833993400340134023403340434053406340734083409341034113412341334143415341634173418341934203421342234233424342534263427342834293430343134323433343434353436343734383439344034413442344334443445344634473448344934503451345234533454345534563457345834593460346134623463346434653466346734683469347034713472347334743475347634773478347934803481348234833484348534863487348834893490349134923493349434953496349734983499350035013502350335043505350635073508350935103511351235133514351535163517351835193520352135223523352435253526352735283529353035313532353335343535353635373538353935403541354235433544354535463547354835493550355135523553355435553556355735583559356035613562356335643565356635673568356935703571357235733574357535763577357835793580358135823583358435853586358735883589359035913592359335943595359635973598359936003601360236033604360536063607 |
- <?php
- class google{
-
- private const is_class = ".";
- private const is_id = "#";
-
- public function __construct(){
-
- include "lib/fuckhtml.php";
- $this->fuckhtml = new fuckhtml();
-
- include "lib/backend.php";
- $this->backend = new backend("google");
- }
-
- public function getfilters($page){
-
- $base = [
- "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",
- "cd" => "Congo, the Democratic Republic",
- "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",
- "ir" => "Iran, Islamic Republic",
- "iq" => "Iraq",
- "ie" => "Ireland",
- "il" => "Israel",
- "it" => "Italy",
- "jm" => "Jamaica",
- "jp" => "Japan",
- "jo" => "Jordan",
- "kz" => "Kazakhstan",
- "ke" => "Kenya",
- "ki" => "Kiribati",
- "kp" => "Korea, Democratic People's Republic",
- "kr" => "Korea, Republic",
- "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",
- "mk" => "Macedonia, the Former Yugosalv Republic",
- "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",
- "fm" => "Micronesia, Federated States",
- "md" => "Moldova, Republic",
- "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",
- "tz" => "Tanzania, United Republic",
- "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)"
- ]
- ]
- ];
-
- switch($page){
-
- case "web":
- return array_merge(
- $base,
- [
- "newer" => [ // &sort=review-date:r:20090301:20090430
- "display" => "Newer than",
- "option" => "_DATE"
- ],
- "older" => [
- "display" => "Older than",
- "option" => "_DATE"
- ]
- ]
- );
- break;
-
- case "images":
- return array_merge(
- $base,
- [
- "time" => [ // tbs=qrd:<size>
- "display" => "Time posted",
- "option" => [
- "any" => "Any time",
- "d" => "Past 24 hours",
- "w" => "Past week",
- "m" => "Past month",
- "y" => "Past year"
- ]
- ],
- "size" => [
- "display" => "Size",
- "option" => [
- // tbs=isz:<size>
- "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"
- ]
- ],
- "ratio" => [ // tbs=iar:<size>
- "display" => "Aspect ratio",
- "option" => [
- "any" => "Any ratio",
- "t" => "Tall",
- "s" => "Square",
- "w" => "Wide",
- "xw" => "Panoramic"
- ]
- ],
- "color" => [ // tbs=ic:<color>
- "display" => "Color",
- "option" => [
- "any" => "Any color",
- "color" => "Full color",
- "gray" => "Black & white",
- "trans" => "Transparent",
- // from there, its ic:specific,isc:<color>
- "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>
- "display" => "Type",
- "option" => [
- "any" => "Any type",
- "face" => "Faces",
- "clipart" => "Clip Art",
- "lineart" => "Line Drawing",
- "stock" => "Stock",
- "animated" => "Animated"
- ]
- ],
- "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"
- ]
- ],
- "rights" => [ // tbs=il:<rights>
- "display" => "Usage rights",
- "option" => [
- "any" => "Any license",
- "cl" => "Creative Commons licenses",
- "ol" => "Commercial & other licenses"
- ]
- ]
- ]
- );
- break;
-
- case "videos":
- return array_merge(
- $base,
- [
- "time" => [
- "display" => "Time posted",
- "option" => [ // tbs=qdr
- "any" => "Any time",
- "h" => "Past hour",
- "d" => "Past 24 hours",
- "w" => "Past week",
- "m" => "Past month",
- "y" => "Past year"
- ]
- ],
- "duration" => [
- "display" => "Duration",
- "option" => [
- "any" => "Any duration",
- "s" => "Short (0-4min)", // tbs=dur:s
- "m" => "Medium (4-20min)", // tbs=dur:m
- "l" => "Long (20+ min)" // tbs=dur:l
- ]
- ],
- "quality" => [
- "display" => "Quality",
- "option" => [
- "any" => "Any quality",
- "h" => "High quality" // tbs=hq:h
- ]
- ],
- "captions" => [
- "display" => "Captions",
- "option" => [
- "any" => "No preference",
- "yes" => "Closed captioned" // tbs=cc:1
- ]
- ]
- ]
- );
- break;
-
- case "news":
- return array_merge(
- $base,
- [
- "time" => [
- "display" => "Time posted",
- "option" => [ // tbs=qdr
- "any" => "Any time",
- "h" => "Past hour",
- "d" => "Past 24 hours",
- "w" => "Past week",
- "m" => "Past month",
- "y" => "Past year",
- "a" => "Archives" // tbs=ar:1
- ]
- ],
- "sort" => [
- "display" => "Sort",
- "option" => [
- "relevance" => "Relevance",
- "date" => "Date" // sbd:1
- ]
- ]
- ]
- );
- break;
- }
- }
-
- private function get($proxy, $url, $get = []){
-
- $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);
- $this->backend->assign_proxy($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($get["npt"]){
-
- [$req, $ip] = $this->backend->get($get["npt"], "web");
- parse_str(
- parse_url($req, PHP_URL_QUERY),
- $search
- );
-
- if(isset($search["q"])){
-
- $search = $search["q"];
- }else{
-
- $search = "a"; // lol
- }
-
- try{
- $html =
- $this->get(
- $ip,
- "https://www.google.com" . $req,
- []
- );
- }catch(Exception $error){
-
- throw new Exception("Failed to get HTML");
- }
- }else{
- $search = $get["s"];
- $country = $get["country"];
- $nsfw = $get["nsfw"];
- $lang = $get["lang"];
- $older = $get["older"];
- $newer = $get["newer"];
- $ip = $this->backend->get_ip();
-
- $params = [
- "q" => $search,
- "hl" => "en",
- "num" => 20 // get 20 results
- ];
-
- // country
- if($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;
- }
-
- try{
- $html =
- $this->get(
- $ip,
- "https://www.google.com/search",
- $params
- );
- }catch(Exception $error){
-
- throw new Exception("Failed to get HTML");
- }
- }
-
- return $this->parsepage($html, "web", $search, $ip);
- }
-
-
-
- public function video($get){
-
- if($get["npt"]){
-
- [$req, $ip] = $this->backend->get($get["npt"], "videos");
- parse_str(
- parse_url($req, PHP_URL_QUERY),
- $search
- );
-
- if(isset($search["q"])){
-
- $search = $search["q"];
- }else{
-
- $search = "a"; // lol
- }
-
- try{
-
- $html =
- $this->get(
- $ip,
- "https://www.google.com" . $req,
- []
- );
- }catch(Exception $error){
-
- throw new Exception("Failed to get HTML");
- }
-
- }else{
- $search = $get["s"];
- $country = $get["country"];
- $nsfw = $get["nsfw"];
- $lang = $get["lang"];
- $time = $get["time"];
- $duration = $get["duration"];
- $quality = $get["quality"];
- $captions = $get["captions"];
- $ip = $this->backend->get_ip();
-
- $params = [
- "q" => $search,
- "tbm" => "vid",
- "hl" => "en",
- "num" => "20"
- ];
-
- // country
- if($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[] = "qdr:" . $time;
- }
-
- // duration
- if($duration != "any"){
-
- $tbs[] = "dur:" . $duration;
- }
-
- // quality
- if($quality != "any"){
-
- $tbs[] = "hq:" . $quality;
- }
-
- // captions
- if($captions != "any"){
-
- $tbs[] = "cc:" . $captions;
- }
-
- // append tbs
- if(count($tbs) !== 0){
-
- $params["tbs"] =
- implode(",", $tbs);
- }
-
- try{
- $html =
- $this->get(
- $ip,
- "https://www.google.com/search",
- $params
- );
- }catch(Exception $error){
-
- throw new Exception("Failed to get HTML");
- }
- }
-
- $json = $this->parsepage($html, "videos", $search, $ip);
- $out = [
- "status" => "ok",
- "npt" => $json["npt"],
- "video" => [],
- "author" => [],
- "livestream" => [],
- "playlist" => [],
- "reel" => []
- ];
-
- foreach($json["web"] as $item){
-
- $out["video"][] = [
- "title" => $item["title"],
- "description" => $item["description"],
- "author" => [
- "name" => null,
- "url" => null,
- "avatar" => null
- ],
- "date" => isset($item["table"]["Posted"]) ? strtotime($item["table"]["Posted"]) : null,
- "duration" => isset($item["table"]["Duration"]) ? $this->hms2int($item["table"]["Duration"]) : null,
- "views" => null,
- "thumb" =>
- $item["thumb"]["url"] === null ?
- [
- "url" => null,
- "ratio" => null
- ] :
- [
- "url" => $item["thumb"]["url"],
- "ratio" => "16:9"
- ],
- "url" => $item["url"]
- ];
- }
-
- return $out;
- }
-
-
-
- public function news($get){
-
- if($get["npt"]){
-
- [$req, $ip] = $this->backend->get($get["npt"], "news");
- parse_str(
- parse_url($req, PHP_URL_QUERY),
- $search
- );
-
- if(isset($search["q"])){
-
- $search = $search["q"];
- }else{
-
- $search = "a"; // lol
- }
-
- try{
-
- $html =
- $this->get(
- $ip,
- "https://www.google.com" . $req,
- []
- );
- }catch(Exception $error){
-
- throw new Exception("Failed to get HTML");
- }
-
- }else{
- $search = $get["s"];
- $country = $get["country"];
- $nsfw = $get["nsfw"];
- $lang = $get["lang"];
- $time = $get["time"];
- $sort = $get["sort"];
- $ip = $this->backend->get_ip();
-
- $params = [
- "q" => $search,
- "tbm" => "nws",
- "hl" => "en",
- "num" => "20"
- ];
-
- // country
- if($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"){
-
- if($time == "a"){
-
- $tbs[] = "ar:1";
- }else{
-
- $tbs[] = "qdr:" . $time;
- }
- }
-
- // relevance
- if($sort == "date"){
-
- $tbs[] = "sbd:1";
- }
-
- // append tbs
- if(count($tbs) !== 0){
-
- $params["tbs"] =
- implode(",", $tbs);
- }
-
- $html =
- $this->get(
- $ip,
- "https://www.google.com/search",
- $params
- );
- }
-
- $json = $this->parsepage($html, "news", $search, $ip);
- $out = [
- "status" => "ok",
- "npt" => $json["npt"],
- "news" => []
- ];
-
- foreach($json["web"] as $item){
-
- $description = array_key_first($item["table"]);
-
- if($description !== null){
-
- $date = $item["table"][$description];
- }else{
-
- $date = null;
- }
-
- $out["news"][] = [
- "title" => $item["title"],
- "author" => $item["author"],
- "description" => $description,
- "date" => strtotime($date),
- "thumb" =>
- $item["thumb"]["url"] === null ?
- [
- "url" => null,
- "ratio" => null
- ] :
- [
- "url" => $item["thumb"]["url"],
- "ratio" => "16:9"
- ],
- "url" => $item["url"]
- ];
- }
-
- return $out;
- }
-
-
-
- private function parsepage($html, $pagetype, $search, $ip){
- /*
- $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);
-
- //
- // parse accdef's
- //
- $has_appended_accdef = false;
-
- preg_match_all(
- '/window\.jsl\.dh\(\'(accdef_[0-9]+)\',\'(.*)\'\);/',
- $html,
- $accdefs_regex
- );
-
- $accdefs = [];
- for($i=0; $i<count($accdefs_regex[0]); $i++){
-
- // decode UTF-16 string
- $answer =
- $this->fuckhtml
- ->parseJsString(
- $accdefs_regex[2][$i]
- );
-
- $this->fuckhtml->load($answer);
-
- // get description
- $description =
- $this->fuckhtml
- ->getElementsByClassName(
- $this->findstyles(
- [
- "padding" => "12px 16px 12px",
- ],
- self::is_class
- ),
- "div"
- )[1];
-
- // get date (rare)
- $date =
- $this->fuckhtml
- ->getElementsByTagName("sub");
-
- if(count($date) !== 0){
-
- $description =
- str_replace(
- $date[0]["outerHTML"],
- "",
- $description["innerHTML"]
- );
-
- $date =
- strtotime(
- $this->fuckhtml
- ->getTextContent(
- $date[0]
- )
- );
- }else{
-
- $date = null;
- }
-
- // get information table
- $table = [];
-
- $tbody =
- $this->fuckhtml
- ->getElementsByTagName("tbody");
-
- if(count($tbody) !== 0){
-
- $this->fuckhtml->load($tbody[0]);
-
- $trs =
- $this->fuckhtml
- ->getElementsByTagName("tr");
-
- foreach($trs as $tr){
-
- $this->fuckhtml->load($tr);
-
- $tds =
- $this->fuckhtml
- ->getElementsByTagName("td");
-
- if(count($tds) === 2){
-
- $table[
- $this->fuckhtml
- ->getTextContent(
- $tds[0]
- )
- ] =
- $this->fuckhtml
- ->getTextContent(
- $tds[1]
- );
- }
- }
-
- // load back what we had
- $this->fuckhtml->load($answer);
- }
-
- // get title & link
- $a =
- $this->fuckhtml
- ->getElementsByTagName("a")[0];
-
- $this->fuckhtml->load($a);
-
- $title =
- $this->fuckhtml
- ->getElementsByTagName("span");
-
- if(count($title) === 0){
-
- continue;
- }
-
- $accdefs[] = [
- "title" =>
- $this->titledots(
- $this->fuckhtml
- ->getTextContent(
- $title[0]
- )
- ),
- "description" =>
- $this->fuckhtml
- ->getTextContent(
- $description
- ),
- "url" =>
- $this->unshiturl(
- $a["attributes"]["href"]
- ),
- "date" => $date,
- "type" => "web",
- "thumb" => [
- "url" => null,
- "ratio" => null
- ],
- "sublink" => [],
- "table" => $table
- ];
- }
-
- $this->fuckhtml->load($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);
-
- // detect spelling
- $spelling =
- $this->fuckhtml
- ->getElementsByClassName(
- $this->findstyles(
- [
- "font-size" => "20px",
- "line-height" => "26px",
- "padding-top" => "2px",
- "margin-bottom" => "1px"
- ],
- self::is_class
- ),
- "div"
- );
-
- if(count($spelling) !== 0){
-
- $a =
- $this->fuckhtml
- ->getElementsByTagName("a");
-
- if(count($a) !== 0){
-
- $scripts =
- $this->fuckhtml
- ->getElementsByTagName("script");
-
- foreach($scripts as $script){
-
- $container["innerHTML"] =
- str_replace(
- $script["outerHTML"],
- "",
- $container["innerHTML"]
- );
- }
-
- $container["innerHTML"] =
- $this->fuckhtml
- ->getTextContent(
- str_replace(
- $a[0]["outerHTML"],
- "",
- $container["innerHTML"]
- )
- );
-
- if(
- preg_match(
- '/^did you mean/i',
- $container["innerHTML"]
- )
- ){
-
- $out["spelling"] = [
- "type" => "not_many",
- "using" => $search,
- "correction" =>
- $this->fuckhtml
- ->getTextContent(
- $a[0]
- )
- ];
- }
-
- elseif(
- preg_match(
- '/^showing results for/i',
- $container["innerHTML"]
- )
- ){
-
- $out["spelling"] = [
- "type" => "including",
- "using" =>
- $this->fuckhtml
- ->getTextContent(
- $a[0]
- ),
- "correction" => $search
- ];
- }
- }
-
- continue;
- }
-
- $title =
- $this->fuckhtml
- ->getElementsByClassName(
- $this->findstyles(
- [
- "color" => "#1967d2",
- "font-size" => "20px",
- "line-height" => "26px"
- ],
- self::is_class
- ),
- "div"
- );
-
- if(count($title) !== 0){
-
- //
- // Container is a web link
- //
- $web = [
- "title" =>
- $this->titledots(
- $this->fuckhtml
- ->getTextContent(
- $title[0]
- )
- ),
- "description" => null,
- "url" => null,
- "date" => null,
- "type" => "web",
- "thumb" => [
- "url" => null,
- "ratio" => null
- ],
- "sublink" => [],
- "table" => []
- ];
-
- // get link
- $web["url"] =
- $this->unshiturl(
- $this->fuckhtml
- ->getElementsByTagName("a")
- [0]
- ["attributes"]
- ["href"]
- );
-
- //
- // check if link contains a carousel
- //
- $carousels = $this->parsecarousels();
- if(count($carousels) !== 0){
-
- $first = true;
- foreach($carousels as $carousel_cat){
-
- foreach($carousel_cat as $carousel){
-
- if($first === true){
-
- $first = false;
- }elseif($carousel["image"] !== null){
-
- $out["image"][] = [
- "title" => $carousel["title"],
- "source" => [
- [
- "url" => $carousel["image"],
- "width" => null,
- "height" => null
- ]
- ],
- "url" => $carousel["url"]
- ];
- }
-
- $web["sublink"][] = [
- "title" => $carousel["title"],
- "date" => $carousel["date"],
- "description" => $carousel["description"],
- "url" => $carousel["url"]
- ];
- }
- }
-
- if($carousels[0][0]["image"] !== null){
- $web["thumb"] = [
- "url" => $carousels[0][0]["image"],
- "ratio" => "16:9"
- ];
- }
-
- $out["web"][] = $web;
- continue;
- }
-
- //
- // no carousel entries, parse as normal link
- //
- $this->fuckhtml->load($container);
-
- // parse URL
- $web["url"] =
- $this->unshiturl(
- $this->fuckhtml
- ->getElementsByTagName("a")
- [0]
- ["attributes"]
- ["href"]
- );
-
- $container = $container["innerHTML"];
-
- $line_detect =
- $this->fuckhtml
- ->getElementsByClassName(
- $this->findstyles(
- [
- "height" => "1px",
- "background-color" => "#dadce0",
- "margin" => "0 16px"
- ],
- self::is_class
- ),
- "div"
- );
-
- if(count($line_detect) !== 0){
-
- // we found a line, this means we're dealing with a
- // "featured snippet"
- $featured = true;
-
- $description_container =
- $this->fuckhtml
- ->getElementsByClassName(
- $this->findstyles(
- [
- "white-space" => "pre-line",
- "word-wrap" => "break-word"
- ],
- self::is_class
- ),
- "div"
- )[1];
-
- // get date node for it
- $date =
- $this->fuckhtml
- ->getElementsByTagName("sub");
-
- if(count($date) !== 0){
- $web["date"] =
- strtotime(
- $this->fuckhtml
- ->getTextContent(
- $date[0]
- )
- );
- }
- }else{
-
- // we're dealing with a normal link
- $featured = false;
-
- $description_container =
- $this->fuckhtml
- ->getElementsByClassName(
- $this->findstyles(
- [
- "padding" => "12px 16px 12px"
- ],
- self::is_class
- ),
- "div"
- )[1];
- }
-
- //
- // Get author if we're parsing news
- //
- if($pagetype == "news"){
-
- $author =
- $this->fuckhtml
- ->getElementsByClassName(
- $this->findstyles(
- [
- "position" => "absolute",
- "width" => "100%",
- "top" => "0",
- "left" => "0",
- "padding-top" => "1px",
- "margin-bottom" => "-1px"
- ],
- self::is_class
- ),
- "div"
- );
-
- if(count($author) !== 0){
-
- $web["author"] =
- $this->fuckhtml
- ->getTextContent(
- $author[0]
- );
- }else{
-
- $web["author"] = null;
- }
- }
-
- $description =
- $description_container["innerHTML"];
-
- $this->fuckhtml->load($description);
-
- //
- // get thumbnail before we call loadhtml again
- //
- $img =
- $this->fuckhtml
- ->getElementsByTagName("img");
-
- if(count($img) !== 0){
-
- $skip = true;
-
- if(
- isset($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"]
- );
- }else{
-
- $skip = false;
- }
-
- //
- // get sublinks
- //
- $links =
- $this->fuckhtml
- ->getElementsByTagName("a");
-
- foreach($links as $link){
-
- if($skip === true){
-
- $skip = false;
- continue;
- }
-
- $description =
- str_replace(
- $link["outerHTML"],
- "",
- $description
- );
-
- $sublink = [
- "title" => null,
- "description" => null,
- "url" => null,
- "date" => null
- ];
-
- $sublink["title"] =
- $this->titledots(
- $this->fuckhtml
- ->getTextContent(
- $link
- )
- );
-
- $sublink["url"] =
- $this->unshiturl(
- $link
- ["attributes"]
- ["href"]
- );
-
- if(parse_url($sublink["url"], PHP_URL_HOST) !== null){
-
- $web["sublink"][] = $sublink;
- }
- }
-
- //
- // Parse spans in description
- //
- $this->fuckhtml->load($description);
-
- if($featured === false){
-
- $levels =
- $this->fuckhtml
- ->getElementsByClassName(
- $this->findstyles(
- [
- "padding-bottom" => "8px"
- ],
- self::is_class
- ),
- "div"
- );
-
- // oh my god yes, fucking great, sometimes there are NO levels
- // hahahahahhahahahahahahahahahhahaa
- if(count($levels) === 0){
-
- $levels = [$description];
- }
-
- foreach($levels as $level){
-
- $this->fuckhtml->load($level);
-
- $spans =
- $this->fuckhtml
- ->getElementsByTagName(
- "span"
- );
-
- $is_rating = -1;
-
- foreach($spans as $span){
-
- $innertext =
- trim(
- $this->fuckhtml
- ->getTextContent(
- $span
- ),
- " ·."
- );
-
- if($innertext == ""){ continue; }
-
- if(
- strtolower($innertext)
- == "rating"
- ){
-
- $is_rating = 0;
-
- // clean up before we go
- $description =
- str_replace(
- $span["outerHTML"],
- "",
- $description
- );
- continue;
- }
-
- //
- // Parse rating object
- //
- if($is_rating >= 0){
-
- // clean up description
- $description =
- str_replace(
- $span["outerHTML"],
- "",
- $description
- );
-
- if($span["level"] !== 1){ continue; }
- $is_rating++;
-
- // 10/10 (123)
- if($is_rating === 1){
-
- $innertext = explode(" ", $innertext, 2);
-
- $web["table"]["Rating"] = $innertext[0];
-
- if(count($innertext) === 2){
- $web["table"]["Hits"] =
- trim(
- str_replace(
- [
- "(",
- ")"
- ],
- "",
- $innertext[1]
- )
- );
-
- if($web["table"]["Hits"] == ""){
-
- unset($web["table"]["Hits"]);
- }
- }
- continue;
- }
-
- // US$4.99
- // MYR 50.00
- // $38.34
- // JP¥6,480
- // Reviewed by your mom
- if($is_rating === 2){
-
- if(
- preg_match(
- '/^Review by (.+)/',
- $innertext,
- $match
- )
- ){
-
- $web["table"]["Author"] = $match[1];
- continue;
- }
-
- $web["table"]["Price"] = $innertext;
- continue;
- }
-
- // Android / In stock
- if($is_rating === 3){
-
- $web["table"]["Support"] = $innertext;
- continue;
- }
-
- // ignore the rest
- continue;
- }
-
- //
- // Parse standalone text
- //
-
- // If we reach this point:
- // 1. Ratings have been parsed
- // 2. We're parsing a WEB link, not some shitty piece of shit
-
- // check for date
- // if span has no text before it, assume it's a date
- $desc_split =
- explode(
- $span["outerHTML"],
- $description,
- 2
- );
-
- if(
- $this->fuckhtml
- ->getTextContent(
- $desc_split[0]
- ) == ""
- ){
-
- // has no text before
- $date = strtotime($innertext);
- if($date){
-
- $web["date"] = $date;
- }
-
- // cleanup
- $description =
- str_replace(
- $span["outerHTML"],
- "",
- $description
- );
-
- continue;
- }
-
- // Ready to parse table
- if(count($desc_split) === 2){
- $this->fuckhtml->load($desc_split[1]);
-
- $web["table"][
- $this->fuckhtml
- ->getTextContent(
- trim($desc_split[0], ": ")
- )
- ] = $innertext;
-
- // cleanup
- $description =
- str_replace(
- $desc_split[0] . $span["outerHTML"],
- "",
- $description
- );
- }
- }
- }
- }
-
- $web["description"] =
- trim(
- $this->fuckhtml
- ->getTextContent(
- $description
- ),
- " ·."
- );
-
- if($web["description"] == ""){
-
- $web["description"] = null;
- }
-
- $out["web"][] = $web;
-
- continue;
- }
-
- //
- // Detect wikipedia shit
- //
- $wiki_title =
- $this->fuckhtml
- ->getElementsByTagName("h3");
-
- if(count($wiki_title) !== 0){
-
- $description_after = [];
- $description = [];
- $table = [];
- $sublink = [];
-
- $as =
- $this->fuckhtml
- ->getElementsByTagName("a");
-
- foreach($as as $a){
-
- if(
- isset($a["attributes"]["href"]) &&
- parse_url($a["attributes"]["href"], PHP_URL_HOST) == "maps.google.com"
- ){
-
- // detected maps embed, ignore
- continue 2;
- }
- }
-
- // get carousels and remove them from container for image grepper
- $carousels = $this->parsecarousels($container["innerHTML"]);
- $this->fuckhtml->load($container);
-
- // add images to image tab, if applicable
- for($i=0; $i<count($carousels); $i++){
-
- foreach($carousels[$i] as $item){
-
- if(
- $item["url"] !== null &&
- $item["ref"] !== null &&
- $item["image"] !== null &&
- $item["title"] !== null
- ){
-
- $out["image"][] = [
- "title" => $item["title"],
- "source" => [
- [
- "url" => $item["url"],
- "width" => $item["image_width"],
- "height" => $item["image_height"]
- ],
- [
- "url" => $item["image"],
- "width" => $item["thumb_width"],
- "height" => $item["thumb_height"]
- ]
- ],
- "url" => $item["ref"]
- ];
-
- unset($carousels[$i]);
- }
- }
- }
-
- $carousels = array_values($carousels);
-
- // interpret remaining carousels as title + carousel
- $titles =
- $this->fuckhtml
- ->getElementsByClassName(
- $this->findstyles(
- [
- "font-weight" => "700",
- "letter-spacing" => "0.75px",
- "text-transform" => "uppercase"
- ],
- self::is_class
- )
- );
-
- for($i=0; $i<count($titles); $i++){
-
- if(!isset($carousels[$i])){
-
- break;
- }
-
- $description_after[] = [
- "type" => "title",
- "value" =>
- $this->fuckhtml
- ->getTextContent(
- $titles[$i]
- )
- ];
-
- foreach($carousels[$i] as $carousel){
-
- $description_after[] = [
- "type" => "link",
- "url" => "web?s=" . urlencode($carousel["description"]) . "&scraper=google",
- "value" => $carousel["description"]
- ];
-
- if($carousel["subtext"] !== null){
-
- $description_after[] = [
- "type" => "quote",
- "value" => $carousel["subtext"]
- ];
- }
-
- $description_after[] = [
- "type" => "image",
- "url" => $carousel["image"]
- ];
- }
- }
-
- $categories =
- $this->fuckhtml
- ->getElementsByClassName(
- $this->findstyles(
- [
- "padding" => "12px 16px 12px"
- ],
- self::is_class
- )
- );
-
- $image =
- $this->fuckhtml
- ->getElementsByTagName("img");
-
- if(count($image) !== 0){
-
- $image = $this->getimage($image[0]["attributes"]["id"]);
- }else{
-
- $image = null;
- }
-
- $url = null;
-
- for($i=0; $i<count($categories); $i++){
-
- $this->fuckhtml->load($categories[$i]);
-
- if($i === 0){
- // first node. this should be the header with the small
- // information snippet
-
- $url =
- $this->fuckhtml
- ->getElementsByTagName("a");
-
- if(count($url) !== 0){
-
- $url =
- $this->unshiturl(
- $url[0]["attributes"]["href"]
- );
-
- if(parse_url($url, PHP_URL_HOST) == "encrypted-tbn0.gstatic.com"){
-
- $image = $url;
- $url = null;
- }
- }else{
-
- $url = null;
- }
-
- $categories[$i]["innerHTML"] =
- str_replace(
- $wiki_title[0]["outerHTML"],
- "",
- $categories[$i]["innerHTML"]
- );
-
- $subtext =
- $this->fuckhtml
- ->getTextContent(
- $categories[$i]["innerHTML"]
- );
-
- if(strlen($subtext) !== 0){
-
- $description[] = [
- "type" => "quote",
- "value" =>
- $this->fuckhtml
- ->getTextContent(
- $categories[$i]["innerHTML"]
- )
- ];
- }
-
- // detect audio file
- $audio =
- $this->fuckhtml
- ->getElementsByTagName("audio");
-
- if(count($audio) !== 0){
-
- $description[] = [
- "type" => "audio",
- "url" =>
- $this->fuckhtml
- ->getTextContent(
- $audio[0]["attributes"]["src"]
- )
- ];
- }
- }else{
-
- // check for separator elements IN THERE
- $separators =
- $this->fuckhtml
- ->getElementsByClassName(
- $this->findstyles(
- [
- "white-space" => "pre-line",
- "word-wrap" => "break-word"
- ],
- self::is_class
- ),
- "div"
- );
-
- // detect container type
- foreach($separators as $separator){
-
- $this->fuckhtml->load($separator);
-
- // ignore wrong levels
- if($separator["level"] !== 2){
-
- continue;
- }
-
- //
- // Detect word definition
- //
- $wordwraps =
- $this->fuckhtml
- ->getElementsByClassName(
- $this->findstyles(
- [
- "padding-bottom" => "12px"
- ],
- self::is_class
- ),
- "div"
- );
-
- if(count($wordwraps) !== 0){
-
- foreach($wordwraps as $word){
-
- $this->fuckhtml->load($word);
-
- // detect title
- $span =
- $this->fuckhtml
- ->getElementsByTagName(
- "span"
- );
-
- if(
- count($span) === 1 &&
- $this->fuckhtml
- ->getTextContent(
- str_replace(
- $span[0]["outerHTML"],
- "",
- $word["innerHTML"]
- )
- ) == ""
- ){
-
- $description[] = [
- "type" => "title",
- "value" =>
- $this->fuckhtml
- ->getTextContent(
- $span[0]
- )
- ];
- continue;
- }
-
- // detect list element
- $lists =
- $this->fuckhtml
- ->getElementsByTagName("ol");
-
- if(count($lists) !== 0){
- foreach($lists as $list){
-
- $this->fuckhtml->load($list);
-
- $items =
- $this->fuckhtml
- ->getElementsByTagName("li");
-
- $w = 0;
- foreach($items as $item){
-
- $w++;
- $this->fuckhtml->load($item);
-
- // get subnodes
- $subnodes =
- $this->fuckhtml
- ->getElementsByClassName(
- $this->findstyles(
- [
- "white-space" => "pre-line",
- "word-wrap" => "break-word"
- ],
- self::is_class
- ),
- "div"
- );
-
- foreach($subnodes as $subnode){
-
- $this->fuckhtml->load($subnode);
-
- $spans =
- $this->fuckhtml
- ->getElementsByTagName("span");
-
- if(count($spans) !== 0){
-
- // append quote
- $description[] = [
- "type" => "quote",
- "value" =>
- $this->fuckhtml
- ->getTextContent(
- $subnode
- )
- ];
- }else{
-
- // append text
- $description[] = [
- "type" => "text",
- "value" =>
- $w . ". " .
- $this->fuckhtml
- ->getTextContent(
- $subnode
- )
- ];
- }
- }
- }
- }
- }else{
-
- // parse without list
- // get subnodes
- $subnodes =
- $this->fuckhtml
- ->getElementsByClassName(
- $this->findstyles(
- [
- "white-space" => "pre-line",
- "word-wrap" => "break-word"
- ],
- self::is_class
- ),
- "div"
- );
-
- foreach($subnodes as $subnode){
-
- $this->fuckhtml->load($subnode);
-
- $spans =
- $this->fuckhtml
- ->getElementsByTagName("span");
-
- if(count($spans) !== 0){
-
- // append quote
- $description[] = [
- "type" => "quote",
- "value" =>
- $this->fuckhtml
- ->getTextContent(
- $subnode
- )
- ];
- }else{
-
- // append text
- $description[] = [
- "type" => "text",
- "value" =>
- $this->fuckhtml
- ->getTextContent(
- $subnode
- )
- ];
- }
- }
- }
- }
- }else{
-
- //
- // Parse table
- //
- $spans =
- $this->fuckhtml
- ->getElementsByTagName("span");
-
- foreach($spans as $span){
-
- if(!isset($span["attributes"]["class"])){
-
- // found table
- $row =
- explode(
- ":",
- $this->fuckhtml
- ->getTextContent(
- $separator
- ),
- 2
- );
-
- if(count($row) === 2){
-
- $table[rtrim($row[0])] =
- ltrim($row[1]);
-
- }
- continue 2;
- }
- }
-
- //
- // Parse normal description
- //
- $links_rem =
- $this->fuckhtml
- ->getElementsByTagName("a");
-
- foreach($links_rem as $rem){
-
- $separator["innerHTML"] =
- str_replace(
- $rem["outerHTML"],
- "",
- $separator["innerHTML"]
- );
- }
-
- $description[] = [
- "type" => "text",
- "value" =>
- rtrim(
- $this->fuckhtml
- ->getTextContent(
- $separator
- ),
- " .,"
- )
- ];
- }
- }
- }
-
- // detect huge buttons
- $buttons =
- $this->fuckhtml
- ->getElementsByClassName(
- $this->findstyles(
- [
- "display" => "table-cell",
- "vertical-align" => "middle",
- "height" => "52px",
- "text-align" => "center"
- ],
- self::is_class
- ),
- "a"
- );
-
- if(count($buttons) !== 0){
-
- foreach($buttons as $button){
-
- if(isset($button["attributes"]["href"])){
-
- $sublink[
- $this->fuckhtml
- ->getTextContent(
- $button
- )
- ] =
- $this->unshiturl(
- $button["attributes"]["href"]
- );
- }
- }
- }
- }
-
- // append description_after (contains carousel info)
- $description = array_merge(
- $description,
- $description_after
- );
-
- $out["answer"][] = [
- "title" =>
- $this->fuckhtml
- ->getTextContent(
- $wiki_title[0]
- ),
- "description" => $description,
- "url" => $url,
- "thumb" => $image,
- "table" => $table,
- "sublink" => $sublink
- ];
-
- continue;
- }
-
- //
- // Detect related searches containers
- //
- $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){
-
- // get carousel entries
- $carousels = $this->parsecarousels($container["innerHTML"]);
- $this->fuckhtml->load($container);
-
- foreach($carousels as $carousel){
-
- foreach($carousel as $item){
-
- if($item["url"] !== null){
-
- $out["related"][] = $item["url"];
- }
- }
- }
-
- $container_title =
- strtolower(
- $this->fuckhtml
- ->getTextContent(
- $container_title[0]
- )
- );
-
- switch($container_title){
-
- case "related searches":
- case "people also search for":
- //
- // Parse related searches
- //
- $as =
- $this->fuckhtml
- ->getElementsByTagName("a");
-
- foreach($as as $a){
-
- $out["related"][] =
- $this->fuckhtml
- ->getTextContent($a);
- }
- break;
-
- case "people also ask":
- // get related queries
- $divs =
- $this->fuckhtml
- ->getElementsByTagName("div");
-
- foreach($divs as $div){
-
- // add accdef's here
- if($has_appended_accdef === false){
-
- $out["web"] = array_merge($out["web"], $accdefs);
- $has_appended_accdef = true;
- }
-
- // add accdef's questions
- if(isset($div["attributes"]["role"])){
-
- $out["related"][] =
- $this->fuckhtml
- ->getTextContent($div);
-
- continue;
- }
- }
- break;
- }
-
- continue;
- }
-
- //
- // Parse news
- //
- $title =
- $this->fuckhtml
- ->getElementsByClassName(
- $this->findstyles(
- [
- "font-size" => "16px",
- "line-height" => "20px",
- "font-weight" => "400"
- ],
- self::is_class
- ),
- "div"
- );
-
- if(count($title) !== 0){
-
- $carousels = $this->parsecarousels();
- $this->fuckhtml->load($container);
-
- if(count($carousels) === 0){
-
- // no carousels found
- continue;
- }
-
- $title =
- strtolower(
- $this->fuckhtml
- ->getTextContent(
- $title[0]
- )
- );
-
- if(
- preg_match(
- '/^latest from|^top stories/',
- $title
- )
- ){
-
- // Found news article
- foreach($carousels[0] as $carousel){
-
- if($carousel["image"] !== null){
-
- $thumb = [
- "url" => $carousel["image"],
- "ratio" => "16:9"
- ];
- }else{
-
- $thumb = [
- "url" => null,
- "ratio" => null
- ];
- }
-
- $out["news"][] = [
- "title" => $carousel["title"],
- "description" => $carousel["description"],
- "date" => $carousel["date"],
- "thumb" => $thumb,
- "url" => $carousel["url"]
- ];
- }
- }
-
- elseif(
- $title == "images"
- ){
-
- foreach($carousels as $carousel){
-
- foreach($carousel as $item){
-
- $out["image"][] = [
- "title" => $item["title"],
- "source" => [
- [
- "url" => $item["url"],
- "width" => $item["image_width"],
- "height" => $item["image_height"]
- ],
- [
- "url" => $item["image"],
- "width" => $item["thumb_width"],
- "height" => $item["thumb_height"]
- ]
- ],
- "url" => $item["ref"]
- ];
- }
- }
- }
-
- continue;
- }
-
- //
- // Detect nodes with only text + links
- //
-
- // ignore elements with <style> tags
- $style =
- $this->fuckhtml
- ->getElementsByTagName("style");
-
- if(count($style) !== 0){
-
- continue;
- }
-
- $as =
- $this->fuckhtml
- ->getElementsByTagName("a");
-
- $description = [];
-
- foreach($as as $a){
-
- //
- // Detect next page
- //
- if(
- isset($a["attributes"]["aria-label"]) &&
- strtolower($a["attributes"]["aria-label"]) == "next page"
- ){
-
- $out["npt"] =
- $this->backend->store(
- $this->fuckhtml
- ->getTextContent(
- $a["attributes"]["href"]
- ),
- $pagetype,
- $ip
- );
- continue 2;
- }
-
- //
- // Parse as text node
- //
- $container["innerHTML"] =
- explode(
- $a["outerHTML"],
- $container["innerHTML"],
- 2
- );
-
- $before =
- $this->fuckhtml
- ->getTextContent(
- $container["innerHTML"][0],
- false,
- false
- );
-
- // set after
- if(count($container["innerHTML"]) === 2){
-
- $container["innerHTML"] =
- $container["innerHTML"][1];
- }else{
-
- $container["innerHTML"] = "";
- }
-
- if($before != ""){
-
- $description[] = [
- "type" => "text",
- "value" => $before
- ];
- }
-
- // add link
- $description[] = [
- "type" => "link",
- "url" =>
- $this->unshiturl(
- $a["attributes"]
- ["href"]
- ),
- "value" =>
- $this->fuckhtml
- ->getTextContent(
- $a
- )
- ];
- }
-
- if($container["innerHTML"] != ""){
-
- $description[] = [
- "type" => "text",
- "value" =>
- $this->fuckhtml
- ->getTextContent(
- $container["innerHTML"]
- )
- ];
- }
-
- $out["answer"][] = [
- "title" => "Notice",
- "description" => $description,
- "url" => null,
- "thumb" => null,
- "table" => [],
- "sublink" => []
- ];
- }
-
- //
- // remove duplicate web links cause instant answers
- // sometimes contains duplicates
- //
- $c = count($out["web"]);
- $links = [];
-
- for($i=0; $i<$c; $i++){
-
- foreach($links as $link){
-
- if($out["web"][$i]["url"] == $link){
-
- unset($out["web"][$i]);
- continue 2;
- }
- }
-
- $links[] = $out["web"][$i]["url"];
- }
-
- $out["web"] = array_values($out["web"]);
-
- return $out;
- }
-
-
-
-
- public function image($get){
-
- // generate parameters
- if($get["npt"]){
-
- [$params, $proxy] =
- $this->backend->get(
- $get["npt"],
- "images"
- );
-
- $params = json_decode($params, true);
- }else{
-
- $search = $get["s"];
- if(strlen($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(
- $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"] =
- $this->backend->store(
- json_encode($params),
- "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,
- $imgvl
- );
-
- if(isset($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"] =
- $this->backend->store(
- json_encode($params),
- "images",
- $proxy
- );
- }
- }
- }
-
- 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);
-
- $styles =
- $this->fuckhtml
- ->getElementsByTagName("style");
-
- $this->computedstyle = [];
- $this->ask = [];
-
- foreach($styles as $style){
-
- $this->computedstyle =
- array_merge(
- $this->computedstyle,
- $this->parsestyles($style["innerHTML"])
- );
- }
-
- // get images in javascript var
- preg_match(
- '/google\.ldi=({[^}]+})/',
- $html,
- $this->js_image
- );
-
- if(count($this->js_image) !== 0){
-
- $this->js_image = json_decode($this->js_image[1], true);
- }else{
-
- $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){
-
- $items = explode(",", $image_grep[2][0]);
- $value =
- $this->fuckhtml
- ->getTextContent(
- $image_grep[1][0]
- );
-
- foreach($items as $item){
-
- $this->js_image[trim($item, "' ")] = $value;
- }
- }
-
- // 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])] =
- $this->fuckhtml->parseJsString(
- $ask_grep[2][$i]
- );
- }
- }
- }
-
- 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 $tags;
- }
-
- private function getimage($id){
-
- if(isset($this->js_image[$id])){
-
- $return = $this->fuckhtml->parseJsString($this->js_image[$id]);
-
- if(
- $return != "data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAABQAAAAUCAYAAACNiR0NAAABAUlEQVR4AWMYesChoYElLjkzPj4lY3d8csZjIL4MxPNjUzPcSTYsISFLAqj5NBD/h+LPQPwbiT87NCuLh2gDgRr2QzXuT0jNMoBYksARn5zuHJ+UcR0kB6RXE2VYXHJGOlTDZmzyIJcB5e+D1CSkZDgQNBAaZv+jU1JkcKpJygiGeZ0I76a/Byq8jU9NZFqaCNTA48SE33/iDcw8TIyBt0GKQTFN0Msp6f2EIyUpo57YSIlLSrMhIg0WCIBcCfXSdlzJBsheTHQ6jEnOUgEFOLaEDbMIlhZBOYrorAdJk+nroVnvPsSgdGdoOF7HZyhZ2XPoGQoqjbCpIbt0AiejIQMArVLI7k/DXFkAAAAASUVORK5CYII=" &&
- $return != "data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAABgAAAAYCAYAAADgdz34AAAA6ElEQVR4Ae2UvQ7BYBSGW4uFxlVICLsYDA0D14NEunRxHSJ+BkYG9ibiHtgkuvpZWOod3uFESqpOF/ElT/q958h5OtQx/iexY/evY9ACJjBewUtkXHEPyBYUNQQuCETuggNrF2DHF3A4kfUMmLB+BoUYAg4nIX0TTNnbfCjg8HDBGuR4z4Ij+813giAC8rcrkXusjdQEpMpcYt5rCzrMaea7tqD9JLhpCyrMZeadpmApssPaUOszXQALGHz67De0/2gpMGPP014VFpizfgJ5zWXnAF8MryW1rj3x5l8LJANQF1lZQH5f8AAWpNcUs6HAEAAAAABJRU5ErkJggg=="
- ){
-
- if(
- preg_match(
- '/^\/\//',
- $return
- )
- ){
-
- return 'https:' . $return;
- }
-
- return $return;
- }
-
- return null;
- }
- }
-
- private function parsecarousels(&$item_to_remove = false){
-
- $carousels =
- $this->fuckhtml
- ->getElementsByClassName(
- $this->findstyles(
- [
- "padding" => "16px",
- "position" => "relative"
- ],
- self::is_class
- )
- );
-
- $return = [];
-
- for($i=0; $i<count($carousels); $i++){
-
- if(!isset($carousels[$i]["outerHTML"])){
-
- continue;
- }
-
- $this->fuckhtml->load($carousels[$i]);
-
- if($item_to_remove !== false){
-
- $item_to_remove =
- str_replace(
- $carousels[$i]["outerHTML"],
- "",
- $item_to_remove
- );
- }
-
- $pcitems =
- $this->fuckhtml
- ->getElementsByClassName(
- "pcitem",
- "div"
- );
-
- foreach($pcitems as $pcitem){
-
- $this->fuckhtml->load($pcitem);
-
- $out = [
- "url" => null,
- "ref" => null,
- "image" => null,
- "thumb_width" => null,
- "thumb_height" => null,
- "image_width" => null,
- "image_height" => null,
- "title" => null,
- "description" => null,
- "subtext" => null,
- "date" => null
- ];
-
- $url =
- $this->unshiturl(
- $this->fuckhtml
- ->getElementsByTagName("a")
- [0]
- ["attributes"]
- ["href"],
- true
- );
-
- // set ref
- $out["ref"] = $url["ref"];
-
- // set url
- $out["url"] = $url["url"];
-
- // set sizes
- $out["thumb_width"] = $url["thumb_width"];
- $out["thumb_height"] = $url["thumb_height"];
- $out["image_width"] = $url["image_width"];
- $out["image_height"] = $url["image_height"];
-
- // get image
- $out["image"] =
- $this->fuckhtml
- ->getElementsByTagName(
- "img"
- );
-
- if(count($out["image"]) !== 0){
-
- // get title from image
- if(isset($out["image"][0]["attributes"]["alt"])){
-
- $out["title"] =
- $this->titledots(
- $this->fuckhtml
- ->getTextContent(
- $out["image"][0]["attributes"]["alt"]
- )
- );
- }
-
- // get image url
- if(isset($out["image"][0]["attributes"]["id"])){
-
- $out["image"] = $this->getimage($out["image"][0]["attributes"]["id"]);
- }
-
- elseif(isset($out["image"][0]["attributes"]["data-ll"])){
-
- $out["image"] =
- $this->fuckhtml
- ->getTextContent(
- $out["image"][0]["attributes"]["data-ll"]
- );
- }else{
-
- // failed to get image information
- $out["image"] = null;
- }
-
- if($out["image"] == 'data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAABgAAAAYAgMAAACdGdVrAAAADFBMVEVMaXFChfRChfRChfT0tCPZAAAAA3RSTlMAgFJEkGxNAAAAL0lEQVR4AWPADxgdwBT3BTDF9AUiuhdC6WNK/v///y+UggrClSA07EWVglmEFwAA5eYSExeCwigAAAAASUVORK5CYII='){
-
- // found arrow image base64, skip entry
- continue;
- }
- }else{
-
- // Could not find any image in node
- $out["image"] = null;
- }
-
- // get title from spans
- $title =
- $this->fuckhtml
- ->getElementsByClassName(
- $this->findstyles(
- [
- "color" => "#1967d2"
- ],
- self::is_class
- ),
- "span"
- );
-
- if(count($title) !== 0){
-
- $out["title"] =
- $this->fuckhtml
- ->getTextContent(
- $title[0]
- );
- }
-
- // get textnodes
- $textnodes =
- $this->fuckhtml
- ->getElementsByClassName(
- $this->findstyles(
- [
- "white-space" => "pre-line",
- "word-wrap" => "break-word"
- ],
- self::is_class
- )
- );
-
- $subtext = null;
-
- if(count($textnodes) !== 0){
-
- // get date
- $date =
- $this->fuckhtml
- ->getTextContent(
- $textnodes[count($textnodes) - 1],
- true
- );
-
- if(str_replace("\n", " ", $date) == $title){
-
- $date = null;
- }else{
-
- if(strpos($date, "\n") !== false){
-
- $date = explode("\n", $date);
- $date = $date[count($date) - 1];
- }
- elseif(strpos($date, "•") !== false){
-
- $date = explode("•", $date);
- $date = ltrim($date[count($date) - 1]);
- }else{
-
- $date = null;
- }
- }
-
- if($date !== null){
-
- $date = strtotime($date);
- }
-
- // get description
- $description =
- $this->fuckhtml
- ->getTextContent(
- $textnodes[0]
- );
-
- if($out["title"] === null){
-
- if($date === null){
-
- $out["title"] = $description;
- $description = null;
- }else{
-
- $out["title"] = parse_url($out["url"], PHP_URL_HOST);
- }
- }
-
- if(isset($textnodes[1])){
-
- $out["subtext"] =
- $this->fuckhtml
- ->getTextContent(
- $textnodes[1]
- );
- }
-
- }else{
-
- $date = null;
- $description = null;
- }
-
- $out["date"] = $date;
- $out["description"] = $this->titledots($description);
-
- if($out["url"] === null){
-
- $out["url"] = $out["title"];
- }
-
- if($out["title"] == $out["description"]){
-
- $out["description"] = null;
- }
-
- $return[$i][] = $out;
- }
- }
-
- return $return;
- }
-
- private function unshiturl($url, $return_size = false){
-
- // get parameters from URL
- $url =
- $this->fuckhtml
- ->getTextContent($url);
-
- $newurl = parse_url($url, PHP_URL_QUERY);
-
- if($newurl == ""){
-
- // probably telephone number
- return $url;
- }
-
- $url = $newurl;
- unset($newurl);
-
- parse_str($url, $query);
-
- if(isset($query["imgurl"])){
-
- $url = $query["imgurl"];
- }
- elseif(isset($query["q"])){
-
- $url = $query["q"];
- }
-
- // rewrite URLs to remove extra tracking parameters
- $domain = parse_url($url, PHP_URL_HOST);
-
- if(
- preg_match(
- '/wikipedia.org$/',
- $domain
- )
- ){
-
- // rewrite wikipedia mobile URLs to desktop
- $url =
- $this->replacedomain(
- $url,
- preg_replace(
- '/([a-z0-9]+)(\.m\.)/',
- '$1.',
- $domain
- )
- );
- }
-
- elseif(
- preg_match(
- '/imdb\.com$|youtube\.[^.]+$/',
- $domain
- )
- ){
-
- // rewrite imdb and youtube mobile URLs too
- $url =
- $this->replacedomain(
- $url,
- preg_replace(
- '/^m\./',
- "",
- $domain
- )
- );
-
- }
-
- elseif(
- preg_match(
- '/play\.google\.[^.]+$/',
- $domain
- )
- ){
-
- // remove referrers from play.google.com
- $oldquery = parse_url($url, PHP_URL_QUERY);
- if($oldquery !== null){
-
- parse_str($oldquery, $query);
- if(isset($query["referrer"])){ unset($query["referrer"]); }
- if(isset($query["hl"])){ unset($query["hl"]); }
- if(isset($query["gl"])){ unset($query["gl"]); }
-
- $query = http_build_query($query);
-
- $url =
- str_replace(
- $oldquery,
- $query,
- $url
- );
- }
- }
-
- elseif(
- preg_match(
- '/twitter\.com$/',
- $domain
- )
- ){
-
- // remove more referrers from twitter.com
- $oldquery = parse_url($url, PHP_URL_QUERY);
- if($oldquery !== null){
-
- parse_str($oldquery, $query);
- if(isset($query["ref_src"])){ unset($query["ref_src"]); }
-
- $query = http_build_query($query);
-
- if($query != ""){
-
- $query .= "?" . $query;
- }
-
- $url =
- str_replace(
- '?' . $oldquery,
- $query,
- $url
- );
- }
- }
-
- elseif(
- preg_match(
- '/maps\.google\.[^.]+/',
- $domain
- )
- ){
-
- if(stripos($url, "maps?") !== false){
-
- //https://maps.google.com/maps?daddr=Johnny,+603+Rue+St+Georges,+Saint-J%C3%A9r%C3%B4me,+Quebec+J7Z+5B7
- $query = parse_url($url, PHP_URL_QUERY);
- if($query !== null){
-
- parse_str($query, $query);
-
- if(isset($query["daddr"])){
-
- $url =
- "https://maps.google.com/maps?daddr=" .
- urlencode($query["daddr"]);
- }
- }
- }
- }
-
- if($return_size){
-
- return [
- "url" => $url,
- "ref" => isset($query["imgrefurl"]) ? $query["imgrefurl"] : null,
- "thumb_width" => isset($query["tbnw"]) ? (int)$query["tbnw"] : null,
- "thumb_height" => isset($query["tbnh"]) ? (int)$query["tbnh"] : null,
- "image_width" => isset($query["w"]) ? (int)$query["w"] : null,
- "image_height" => isset($query["h"]) ? (int)$query["h"] : null
- ];
- }
-
- return $url;
- }
-
- private function replacedomain($url, $domain){
-
- return
- preg_replace(
- '/(https?:\/\/)([^\/]+)/',
- '$1' . $domain,
- $url
- );
- }
-
- private function titledots($title){
-
- return rtrim($title, ". \t\n\r\0\x0B");
- }
- }
|