Redirects to high-res images on gallery sites, skipping past descriptions and comments
As of
// ==UserScript==
// @name Eza's Image Glutton
// @namespace https://inkbunny.net/ezalias
// @homepage https://greasyfork.org/en/users/4876-ezalias
// @author Ezalias
// @description Redirects to high-res images on gallery sites, skipping past descriptions and comments
// @license MIT
// @license Public domain / No rights reserved
// @include /^https*://www\.furaffinity\.net/(view|full)/.*/
// @include https://inkbunny.net/submissionview.php*
// @include https://inkbunny.net/s/*
// @include http://gelbooru.com/*page=post&s=view*
// @include http://youhate.us/*page=post&s=view*
// @include https://youhate.us/*page=post&s=view*
// @include http://www.gelbooru.com/*s=view*
// @include https://www.gelbooru.com/*s=view*
// @include http://gelbooru.com/*s=view*
// @include https://gelbooru.com/*s=view*
// @include http://danbooru.donmai.us/posts/*
// @include https://danbooru.donmai.us/posts/*
// @include http://*.tumblr.com/image/*
// @include https://*.tumblr.com/image/*
// @include /^http(s|)://e(621|926)\.net/post/show*//
// @include http://*.deviantart.com/art/*
// @include https://*.deviantart.com/art/*
// @include /^https?://w*\.*hentai-foundry\.com/pictures/user/.*/[0-9]*/.*/
// @include /^https*://www\.sofurry\.com/view/*//
// @include https://www.weasyl.com/*
// @include http://www.y-gallery.net/view/*
// @include http://rule34.paheal.net/post/view/*
// @include https://rule34.paheal.net/post/view/*
// @include http://rule34.xxx/index.php?page=post*
// @include http://rule34hentai.net/post/view/*
// @include /^https*://derpibooru.org/.*/
// @include http://*.booru.org/*s=view*
// @include http://mspabooru.com/*s=view*
// @include http://safebooru.org/*s=view*
// @include http://www.majhost.com/cgi-bin/gallery.cgi?i=*
// @include http://e-hentai.org/s/*
// @include https://e-hentai.org/s/*
// @include http://nijie.info/view.php?id=*
// @include http://www.pixiv.net/member_illust.php?mode=medium&illust_id=*
// @include https://www.pixiv.net/member_illust.php?mode=medium&illust_id=*
// @include http://*sleepymaid.com/*
// @include https://*.sankakucomplex.com/post/*
// @include http://*.bronibooru.com/posts/*
// @include http://luscious.net/c/*
// @include https://luscious.net/c/*
// @include http://imageboard.neko-sentai.com/post/*
// @include https://uberbooru.com/posts/*
// @include https://www.furiffic.com/*/view/*
// @include https://beta.furrynetwork.com/artwork/*
// @include http://hiccears.com/picture.php?pid=*
// @include https://hiccears.com/picture.php?pid=*
// @include http://www.hiccears.com/picture.php?pid=*
// @include https://www.hiccears.com/picture.php?pid=*
// @include http://www.jabarchives.com/main/post/*
// @include https://www.jabarchives.com/main/post/*
// @exclude http://www.deviantart.com/users/outgoing?*
// @exclude *#dnr
// @version 1.26.11
// ==/UserScript==
// Any single-image submission will redirect to the full-size image. On multi-image submissions, every page except the first will redirect to its full-size image.
// If you go "back" to the normal gallery page (to favorite the image, read its description, leave a comment, etc.) then this script will not send you forward again.
// https://greasyfork.org/scripts/4713-eza-s-image-glutton
// https://sleazyfork.org/scripts/4713-eza-s-image-glutton
// TO DO:
// for modify_tumblr: for photoset pages (but everywhere, to be safe) make unlinked images link to themselves. I want nice, clean, chronological tabs for multi-image comics.
// ugh. test without adblock enabled.
// modify_furaffinity to change prev/next/fav links with pre-appended #dnr. not raw html fiddling: use the DOM and getElementsByType or whatever. thingy.href=url_plus_dnr.
// flickr? maybe separately. that whole site is a mess. also full-size images are sometimes gigantic, like dozens of megabytes.
// inkbunny: move page links above first-page preview on multi-image submissions? - by altering divs or css, if possible. minimal molestation implies better future-proofing.
// weasyl: replace bespoke thumbnails with smallest preview images? eh, do these as separate scripts, once userscripts stops fucking around.
// Consider changing some @includes to @match.
// http://thehentaiworld.com/hentai-doujinshi/theres-something-about-sakura-naruto/ ? I already do rule34; there's no pretending this is just about "art."
// Almost deserves a more Pixiv Fixiv-like fix. Maybe just a link dump like that DeviantArt gallery script?
// Swagster.com? Eh. The name alone rubs me the wrong way. Ugh, and they watermark.
// Greasyfork install page as options page?
// This would work faster if I could delay or prevent the loading of images. E.g., execute script before loading page, define CSS that doesn't download embedded images, wait for page to load, scrape image_url, and then redirect as usual. Since the script wouldn't trigger on #dnr (which I should do as an @exclude, I guess) images would load as usual when you clicked 'back.'
// This thought is mostly driven by opening a bunch of e.g. Gelbooru links all at once. They spend long enough loading that the full-size images are usually half-done before the redirect happens.
// Escape function in JS is encodeURI. Also use in Tumblr Scraper, where we need 'safe' URLs as tag IDs.
// FurAffinity stories redirect to thumbnail, e.g. http://www.furaffinity.net/view/15903888/ - might need to break out a whole complex function here.
// http://seiga.nicovideo.jp/seiga/im4507046 ?
// Nijie.info support might be missing out on multipage submissions? I don't even have an account.
// My Nijie support is basically nonexistant because I didn't have an account. Turns out they're more like Pixiv now, including multi-image posts. This is problematic. (Animations work, though.)
// Make undersized images link to themselves on imgur.
// Eza's image glutton as described on http://cuddle.horse/post/109728993805/a-few-browser-extensions-that-make-furaffinity-a -
// Eza’s Image Glutton: This affects websites beyond just FA but is a unique tool for powerbrowsing and such. When you open a page with a single image it skips all comments and descriptions and just shows the image in the highest quality possible. If you want to see all the items that are hidden all you have to do is go back a page.
// good rundown from a third party. 'A page with a single image' is clearer than 'gallery submission page.'
// http://www.pixiv.net/member_illust.php?mode=medium&illust_id=52107168 404s.
// goes to http://i1.pixiv.net/img-original/img/2015/08/21/19/14/00/52107168_p0.jpg
// should be http://i1.pixiv.net/img-original/img/2015/08/21/19/14/00/52107168_p0.png
// Yet another filetype mismatch. Great!
// Single images submitted as manga don't work. E.g. http://www.pixiv.net/member_illust.php?mode=medium&illust_id=52465388 and that artist's other works.
// It's impossible to find this script after Greasyfork fucked over "adult" scripts. SleazyFork does not appear in search engines, at all. The empty GreasyFork page doesn't show up. All that people will see if they don't already know where to go is userscripts-mirror.org, which was last updated in twenty-fucking-twelve. This is completely unacceptable.
// On the other hand, Yahoo also won't find 'greasyfork ezalias,' so what the fuck. Google finds it. Yahoo just sucks.
// Apparently Chrome doesn't redirect properly - there's no 'back' functionality. Grand.
// ... there's back functionality if and only if the page finishes first. Fuck Chrome.
// Fixing what ain't broken - I could "//@include *" and keep the switch-case business, if only to skirt GreasyFork's dumb "adult" rules.
// Ugh, I'd probably be marked "adult" just for mentioning that it works on "adult" sites. Like e621... but not tumblr or pixiv. What even.
// Could also generalize all extract_image_url_after guff to a for-loop over an array of search strings.
// https://beta.furrynetwork.com/dog-bone/artwork/ - ugh, another one
// FurryNetwork.com. How do they have the same source on every page? Oh god, there's a mess of minified JSON handling in app.js. It's two megabytes! What the fuck!
// Okay, so these people are lunatics. No big deal. Let's grab the image via the DOM. (3712 stylesheet rules, what the fuck what the fuck what the fuck.)
// <img class="image__img onclick" src="https://furrynetwork-beta.s3.amazonaws.com/bl/black-kitten/submission/2016/05/ddc0c7791f329a958a6b6b09cc2a61b0/2500x1500.jpg" alt="I had to xD" data-reactid=".0.2.0.0.1.$601223.1">
// Oh of course, the DOM doesn't include that when the script loads. Fuck me. I guess it needs some sort of delay?
// Hit F12 and it's easy to getelementsbytypename( 'img' ), but doing that from userscript js yields nothing. wtf.
// view-source:https://inkbunny.net/submissionview.php?id=1325657 - inkbunny 'friends only' page -
// Finally on Pawoo, let's add that. Ugh - Twitter-style presentation. Gimme a gallery grid.
// Might be more like that Tumblr bare-image resizer, which this script should also mimic. Bare images are linked. Are those max size?
// https://curate.mastodon.art/gallery/media_attachments/files/000/088/717/original/525713c353f7db21.jpg
// https://files.mastodon.social/media_attachments/files/000/963/359/original/ccfa625b17d2e1d0.png
// https://files.mastodon.social/media_attachments/files/000/963/345/original/d23781d67022b95f.png
// Compare:
// https://pawoo.net/media/r89DgtVaqQeydHyPb1g
// https://img.pawoo.net/media_attachments/files/001/246/940/original/b4a22a1ea4433a9f.jpg
// Totally arbitrary, no way to figure one from another. Hmm.
// Instagram?
// Oh my god instagram is the worst site. Neverfuckingmind. It deserves its own unfucker script.
// Middle-click needs to work. Images need to be exposed for saving.
// The & thing for SankakuComplex is called an "entity reference." See about decoding that better.
// Reload_if not working?
// Some DeviantArt pages don't have a download button.
// https://tohad.deviantart.com/art/The-Exchange-706969244
// Arg, sleepymaid broke her whole site. (Nevermind, she reverted it.)
// FF57 broke FA. God dangit. Probably a Greasemonkey/WebExtensions problem, but fix it anyway. Broke Pixiv as well. I need an FF57 profile.
// Well, great. GreaseMonkey saved the trouble of juggling profiles by breaking FFDE55 as well.
// Greasemonkey 4 is fucked. They removed the menu items, because they removed the menu.
// https://www.pixiv.net/member_illust.php?mode=medium&illust_id=66547435 seems broken.
// Tumblr's still broken, by the way.
// https://www.hentai-foundry.com/pictures/user/teku/566798/Daisy-Darret-and-Penny#dnr loads an image from the description?
// Hentai-Foundry is broken in a stupid way.
// https://www.hentai-foundry.com/pictures/user/TanoshyHT/586603/Kaede-cosplay-force#dnr - works.
// https://hentai-foundry.com/pictures/user/darkdragoon/586678/EV.Allura - doesn't. ah, www. fuck you.
// Tampermonkey gets stuck in redirect loops on e.g. InkBunny. Going back does go back (thank god) but then goes forward again.
// Using Waterfox for testing TM, because thanks to my insane browser needs, I already had it installed. v38 -> v47 -> v50
// I've got Pale Moon, too. No IceWeasel though.
// #dnr works on Safebooru. Weird.
// https://inkbunny.net/s/1589436#dnr#dnr#dnr didn't... oh, TM has crappy indication of when it's enabled. Clicking the checkmark does disable it, but there's no feedback.
// So I know it's adding #dnr, that's nice. @exclude is totally fucking broken now, isn't it?
// Seems to be wait_for_dnr. I need a more explicit check for "#dnr".
// Fixed.
// Find a better Pixiv-safe redirect. I've been extracting text from the page to inject into the address bar since... forever, really. This is hilariously unsafe.
// Testing simple redirects everywhere, found Youhate.us isn't working. Weird. Also, unrelated.
// DeviantArt is broken in a bunch of pages, again, but that's not related either.
// Rule34.xxx is broken. Unrelated.
// Rule34hentai.net is broken. Unrelated.
// Derpibooru tries to download the file, WTF.
// Nijiee is broken, because it was never well-supported.
// Y!Gallery doesn't really exist anymore.
// Luscious.net doesn't work from album links. Different URL structure - change @includes.
// Neko-Sentai doesn't seem to exist anymore.
// Uberbooru is broken from search? E.g. https://uberbooru.com/posts/1845083?tags=kankri_vantas#dnr
// FurryNetwork.com doesn't work. Unrelated.
// So the simple solution works, even if it's past time to re-fix all the sites that re-broke.
// Since I'm just leafing through HTML (usually), can I jump to the image /before/ trying to load the page? GreaseMonkey has a wonky option for running the script before the page runs, but I don't think we get all the HTML first. Maybe... maybe AJAX the page we're on? Like, @RunAtStart or whatever, then create a little blank page, then grab the URL via XmlHTTPgetObject or whatever, then read the HTML as responseText. The trouble (I expect) would be going back to the normal page when someone hits 'back.' This script shouldn't run... but any browser will probably have cached the fake page.
// Owyn Tyler has a ridiculously replete script with similar goals called Handy Just Image - http://userscripts.org/scripts/show/166494
// The supported-site list is waaay longer than mine, and/but his goals are more complex. Image Glutton exists only to deliver the image.
// He's having trouble with back-trapping, though. His solution sounds absurdly complex even compared to mine. Test the script and recommend help if possible.
// Changes since last upload:
// Fixed HentaiFoundry again. My fault this time.
// Added HTTPS support to Paheal. Surprise.
// Added JabArchives.
// Fixed Tumblr /image pages. Twice.
// Fixed Hentai-Foundry to work with "www" prefix. Why the fuck doesn't @include just work on TLD? HTTP vs HTTPS, www vs not, fuck you, it's the same site.
// Clarified #dnr / wait_for_dnr logic for TamperMonkey's sake.
// Changed all sites to a simpler redirect, as referral issues seem to be fixed.
// global variables, for simplicity
var image_url = ''; // location of the full-size image to redirect to
var wait_for_dnr = false; // some site URLs use "#" liberally, so if this var isn't empty, only "#dnr" will stop a redirect
var simple_redirect = false; // some domains are kicking back my JS redirect (for native referral), so do naive location=url instead
var page_failed = false; // If the page 503s or otherwise forces us to reload, wait a moment, then reload.
// detect site, extract image URL, then decide whether or not to redirect
switch( document.domain.replace( 'www.', '' ) ) { // Remove "www" to avoid cases where both example.com and www.example.com are supported.
////////// Simple extract_image_url_after sites
case 'e621.net': image_url = document.getElementById( 'highres' ).href; break;
case 'e926.net': image_url = document.getElementById( 'highres' ).href; break;
case 'weasyl.com': extract_image_url_after( '<div id="detail-art">', '/' ); break; // also redirects to plaintext/HTML on stories, haha
// case 'hentai-foundry.com': extract_image_url_after( '<center><img', '//' ); reload_if( '<h1>An error occurred.' ); break;
case 'y-gallery.net':
extract_image_url_after( 'id="idPreviewImage"', 'http://' );
break; // Fucked, but they say they're coming back eventually
case 'rule34.xxx': extract_image_url_after( '>Edit</a></li>', '//' ); break;
// case 'derpiboo.ru': extract_image_url_after( '>Download</a>', '//' ); break; // No longer exists
case 'derpibooru.org': extract_image_url_after( 'Download</a>', '//' ); simple_redirect = true; break;
case 'chan.sankakucomplex.com': extract_image_url_after( '<li>Original:', '//' ); image_url = image_url.replace( '&', '&' ); break;
case 'idol.sankakucomplex.com': extract_image_url_after( '<li>Original:', '//' ); image_url = image_url.replace( '&', '&' ); break;
case 'furiffic.com': extract_image_url_after( 'onload="$', '//' ); break; // Not using og:image because different URL causes image to re-load is user hits Back
case 'jabarchives.com': extract_image_url_after( 'class="group1"', '/main' );
////////// Slightly complicated extract_image_url_after sites
case 'rule34hentai.net': extract_image_url_after( 'shm-zoomer', '/_images/' ); wait_for_dnr = true; reload_if( '<h2>Rate limit hit' ); break; // wtf? even 'view image' returns text nonsense. images save fine. I bet the site's lying about the mime type. google's not helping for other answers, and I can't fix that, so 'meh' for now. sorry.
case 'rule34.paheal.net': extract_image_url_after( 'shm-zoomer', 'http://' ); wait_for_dnr = true; break;
case 'majhost.com': image_url = document.getElementsByTagName( "img" )[0].src; break; // first and only <img> tag
case 'luscious.net': image_url = document.getElementsByClassName( 'icon-download' )[0].href; wait_for_dnr = true; break;
case 'gelbooru.com': extract_image_url_after( "og:image", '//' ); simple_redirect = true; break;
////////// Simple custom sites
case 'sofurry.com':
image_url = window.location.href.replace('sofurry.com/view/','sofurryfiles.com/std/content?page=');
if( document.body.outerHTML.indexOf( '<div id="sfContentImage' ) < 0 ) { image_url = ''; } // Do not redirect from stories
if( document.body.outerHTML.indexOf( '<div class="sf-story"' ) > 0 ) { image_url = ''; } // Really do not redirect from stories
break;
case 'danbooru.donmai.us':
extract_image_url_after( '% of original (', '/data/' ); // resized images will say "X% of original (view full" or something like that
if( image_url === '' ) {extract_image_url_after( 'twitter:image', 'http://' ); // otherwise just grab the preview-sized image (this also works on pages claiming you need Gold to see them)
image_url = image_url.replace( '/sample/sample-', '/' ); } // if the preview-sized image is a sample, fix that - this sometimes fails for PNG images with JPG previews
break;
case 'furaffinity.net': // This is a mess because I'm trying not to redirect from stories / music... but FA kindly links the thumbnail images for those.
reload_if( 'center;">Error 503' );
//extract_image_url_after( 'File type</strong>:', 'd.facdn' );
// extract_image_url_after( 'Favorites</a></b>', '//' );
extract_image_url_after( '<div class="alt1 actions', '//' ); // Works even when not signed in
//if( document.getElementsByTagName('html')[0].innerHTML.indexOf( 'Category:</b> Story' ) > 0 ) { image_url = ''; } // Don't redirect from stories.
//if( document.getElementsByTagName('html')[0].innerHTML.indexOf( 'File type</strong>' ) > 0 ) { image_url = ''; } // Don't redirect from non-images
// Arg, still redirecting from music. There's a table with the text in it... but it has no unique ID. Has different width on images? Blarg, also width=1% on music (to fit to content).
// Can't just check 'category: art' because there's a wide variety.
// Maybe just check for embedded mp3 or flash.
// Are music/story image filenames always .txt.gif or similar? Nope.
// I might be an idiot: just use the Download link.
// No, even that's a little screwy: you can submit a big image and categorize it as a Story, so you get an illustration and write the story itself in the description. (Maybe do this anyway.)
// "File type" message may only appear on non-images! I stumbled my way into a solution. Oh, but that misses stories-in-descriptions. Double-check Category, I guess. (Nope: fails on music.)
if( document.getElementsByTagName('html')[0].innerHTML.indexOf('/themes/beta') > -1 ) { // Total kludge. If beta theme, use full-url, audio files be damned.
image_url = unsafeWindow.full_url;
}
break;
case 'e-hentai.org': image_url = document.getElementById( 'img' ).src; break;
case 'nijie.info': // Lord, I don't even care about this site.
extract_image_url_after( 'name="twitter:image"', 'http://' ); // some images are behind some sort of barrier, so let's grab the twitter-size image instead...
image_url = image_url.replace( '/sp/', '/' ); // ... and drop the /sp/ to get the full-size URL.
break;
case 'sleepymaid.com':
case 'yay.sleepymaid.com':
image_url = document.getElementById( 'the-image' ).src;
if( document.getElementById( 'next' ) ) { image_url = ''; } // Don't redirect on comic pages
break;
case 'imageboard.neko-sentai.com': image_url = document.getElementById( 'main_image' ).src; break;
case 'uberbooru.com':
//scrape_booru(); simple_redirect = true; break;
//extract_image_url_after( 'id="image-container', 'https://' );
extract_image_url_after( 'Size: <a', '/data' );
break;
case 'hiccears.com': extract_image_url_after( 'href="./upl0ads', './' ); break; // Wow, long garbage names. Can we use Download titles? Apparently not.
case 'hentai-foundry.com':
// extract_image_url_after( 'class="center"', '//' ); // For images displayed full-sized
extract_image_url_after( ' ', '//pictures.' );
if( image_url.indexOf( "';" ) > 0 ) { image_url = image_url.substring( 0, image_url.indexOf( "';" ) ) } // Singlequote terminate, more or less - only on resizable images
reload_if( '<h1>An error occurred.' );
break;
// https://thumbs.hentai-foundry.com/thumb.php?pid=540436&size=1250
// https://pictures.hentai-foundry.com/m/My_Pet_Tentacle_Monster/540436/My_Pet_Tentacle_Monster-540436-Plumerias_Cute_Little_Henchmen.jpg
/*
case 'instagram.com':
// Oh my god, this site is so hostile to tabbing. Middle-click should never open in the same tab.
// And #dnr doesn't work because it plays stupid URL games when you do use a new tab.
extract_image_url_after( 'og:image', '//' );
break;
*/
////////// Sites complex enough to shove into a function down below
// case 'gelbooru.com': scrape_gelbooru(); break;
case 'youhate.us': scrape_gelbooru(); break; // Image links are all to Gelbooru anyway?
case 'inkbunny.net': scrape_inkbunny(); break;
case 'pixiv.net': scrape_pixiv(); break;
case 'mspabooru.com': scrape_booru(); break;
case 'safebooru.org': scrape_booru(); break;
case 'bronibooru.com': scrape_booru(); break;
}
////////// Holdovers from the previous method; domains that don't neatly conform to document.domain switch selection.
if( address_bar_contains( 'tumblr.com' ) ) { extract_image_url_after( '"presentation">', 'http' ); }
if( address_bar_contains( 'deviantart.com' ) ) { scrape_deviantart(); wait_for_dnr = true; }
if( address_bar_contains( '.booru.org' ) ) { scrape_booru(); }
if( address_bar_contains( 'beta.furrynetwork.com' ) ) { setTimeout( scrape_furrynetwork, 500 ); } // With delay. Also, doesn't work. This site's designers are loons.
// If the page didn't load properly, but could be fixed by reloading, then wait a moment and reload
if( page_failed ) { // If we get a 503 or other 'please reload' error
image_url = ''; // do not redirect this time
setTimeout( function inline_reload() { location.reload(); }, Math.floor((Math.random() * 10) + 1) * 1000 ); // 1s-10s pause. Can't believe you have to name inline functions.
}
// Don't redirect if the filetype is obviously not an image. SWF, TXT, MP3, etc.
// Arguably include Webm? Opening many Gelbooru webms in tabs is a cacaphony.
// It's tedious to detect flash, story, and music pages on every website supported, so instead let's just cancel redirection based on those file extensions.
// Possibly implement as an array with a For loop instead of a list of OR operations. This could get silly, with text, music, and video formats galore.
// txt, doc, pdf, swf, mp3, mp4, webm, midi, mid, wav,
var ext = image_url.substring( image_url.lastIndexOf( '.' ) + 1, image_url.length ); // e.g. "png"
//var not_images = [ 'mp3', 'swf', 'txt', 'webm', 'mp4', 'docx', 'pdf', 'doc', 'rtf', 'midi', 'mid', 'wav', 'flv', 'cab' ];
var not_images = [ 'mp3', 'swf', 'txt', 'mp4', 'docx', 'pdf', 'doc', 'rtf', 'midi', 'mid', 'wav', 'flv', 'cab' ];
for( var n in not_images ) { if( ext == not_images[n] ) { image_url = ''; } } // If the extension is in our blacklist, don't redirect.
// Oh right. Doesn't work on FA because FA points to the icon. Yaaayfuck.
simple_redirect = true; // Originally a debug - testing other redirect methods for referral. Works on Pixiv. Either they relaxed or FF fixed metadata in simple redirects.
// having defined image_url by scraping the page's HTML, modify the current URL to prevent back-traps, then redirect to that full image
// do nothing if image_url is empty. ignore pages with a "#", unless wait_for_dnr makes you wait for a full "#dnr".
//if( image_url !== '' && (!address_bar_contains('#') || wait_for_dnr) )
var do_we_redirect = true; // If we've come this far we'll probably go to an image.
if( image_url == '' ) { do_we_redirect = false; } // Don't redirect to an empty string. (Emptying this string is how some functions fail safe.)
if( !wait_for_dnr && address_bar_contains( '#' ) ) { do_we_redirect = false; } // Don't redirect if the wait_for_dnr flag is false and there's a hash. (E.g. FA comments.)
if( address_bar_contains( '#dnr' ) ) { do_we_redirect = false; } // Don't redirect if there's a #dnr in the URL. Obviously.
// I just realized I'm editing this in the FF ESR folder to test it in Waterfox. That would've been some curse words for the rubber duck.
if( do_we_redirect == true ) // Fuck boolean logic, be clear.
{
// some images don't redirect properly, even if you manually "view image" - so we append ".jpg" to URLs without extensions, forcing the browser to consider them images
// even if this doesn't work, the new URL should just 404, which is better than the semi-modal "octet stream" dialog seen otherwise.
if( image_url.lastIndexOf( '/' ) > image_url.lastIndexOf( '.' ) ) { image_url = image_url + '.jpg'; } // if there's not a "." after the last "/" then slap a file extension on there
if( image_url[ image_url.length - 1 ] == '.' ) { image_url = image_url + 'jpg'; } // if the URL ends with a dot, slap a file extension on there
// modify current location, so that when the user clicks "back," they aren't immediately sent forward again
modified_url = window.location.href + '#dnr'; // add do-not-redirect tag to current URL
history.replaceState( {foo:'bar'}, 'Do-not-redirect version', modified_url ); // modify URL without redirecting. the {foo:'bar'} thing is a state object that I don't care about, but the function needs one.
// window.location.href = image_url; // redirect to full image
// alert( image_url ); // DEBUG - holy shit, Gelbooru's blocking redirects!
if( simple_redirect ) { window.location.href = image_url; } // This has different referral properties than clicking a link or displaying an image, so some sites 403
location.assign("javascript:window.location.href=\""+image_url+"\";"); // pixiv-friendly redirect to full image: maintains referral, happens within document's scope instead of within greasemonkey's
// Note: this has always been a mildly horrifying attack surface, and should be changed while I'm dicking with do-not-redirect stuff.
} // end of main execution
// ----- // Functions for readability
function extract_image_url_after( string_before_url, url_begins_with ) { // extract the first quote-delimited string that appears after unique first var and begins with second var
var html_elements = document.getElementsByTagName('html'); // this way we avoiding doing getElementsEtc every time, and we still access the whole page's HTML by reference
var string_index = html_elements[0].innerHTML.indexOf( string_before_url ); // find a unique string somewhere before the image URL
if( string_index > -1 ) {
var image_index = html_elements[0].innerHTML.indexOf( url_begins_with, string_index ); // find where the image URL starts after the unique string
var delimiter_index = html_elements[0].innerHTML.indexOf( '"', image_index ); // find first doublequote after the image URL starts
image_url = html_elements[0].innerHTML.substring( image_index, delimiter_index ); // grab the image URL up to the next doublequote
}
}
function address_bar_contains( string_to_look_for ) { // I'm so tired of typing out window.location.etc == -1. It's stupidly verbose and it looks terrible.
return (window.location.href.indexOf( string_to_look_for ) !== -1); // this makes code more concise and readable. if( address_bar_contains( 'tld.com' ) ) { do tld.com stuff; }
}
function reload_if( error_string ) {
var html_elements = document.getElementsByTagName('html'); // this way we avoiding doing getElementsEtc every time, and we still access the whole page's HTML by reference
var string_index = html_elements[0].innerHTML.indexOf( error_string ); // look for a string indicating the page failed to load
if( string_index > -1 ) { page_failed = true; }
}
// ----- // Functions for individual websites (separated for being especially long)
// DeviantArt sometimes doesn't redirect until you F5. I suspect it's their fancy-pants not-actually-redirecting nonsense. Websites - stop acting stupid and just /be documents./ You are not an app.
function scrape_deviantart() { // this doesn't use ditch_html_before because data-super-full-img's appear for random links - we need to avoid grabbing one from the ass-end of small-image pages
/*
var image_index = document.body.outerHTML.indexOf( 'class="dev-view-deviation"' ); // jump to unique and hopefully universal dev-view-deviation div
if( image_index > 0 ) { // Don't redirect on pages without a deviation (e.g. "oops not found" faux-404s).
var image_index = document.body.outerHTML.indexOf( 'src=', image_index+1 ); // jump to first src (preview size)
var image_index = document.body.outerHTML.indexOf( 'http://', image_index+1 ); // jump to the URL defined in src
image_url = document.body.outerHTML.substring( image_index, document.body.outerHTML.indexOf( '"', image_index ) ); // grab URL, delimited by doublequote
if( image_url.indexOf( "/PRE/" ) > 0 ) { image_url = image_url.replace( "/PRE/", "/" ); } // fix preview-size image to be full-size
}
if( document.body.outerHTML.indexOf( '<div id="flashed-in"' ) > 0 ) { image_url = ''; } // Do not redirect on flash pages
*/
// This goddamn well ought to be an image, but is sometimes a page with an image on it, if you click on it. Because DeviantArt is the devil.
image_url = document.getElementsByClassName( "dev-page-download" )[0].href;
if( document.body.outerHTML.indexOf( '<div id="flashed-in"' ) > 0 ) { image_url = ''; } // Do not redirect on flash pages
}
// Haha, I fixed it for the 2/22/2016 redesign, and it doesn't work on my images. Only the user's own images? Ahh, probably the 'edit' options at top.
// Well... it's a shippable bug. Leave it for now, fix it with the next serious update.
// Nope, fails on https://inkbunny.net/submissionview.php?id=999328 - which isn't mine. Ditto https://inkbunny.net/submissionview.php?id=1019701 by Aogami.
// ... The old way works. Sure, whatever.
// view-source:https://inkbunny.net/submissionview.php?id=1325657 - private page / friends page
// Haha, /private_files/ vs. /files/ - literally a one-character fix. Removed the leading slash.
function scrape_inkbunny() {
// Old way, pre-2016
// var image_index = document.body.outerHTML.indexOf( 'https://us.ib.metapix.net/files/screen/' ); // look for screen-size image URL
var image_index = document.body.outerHTML.indexOf( 'files/screen/' ); // Find the middle of a screen-sized image URL
image_index = document.body.outerHTML.lastIndexOf( 'https://', image_index ); // ... then back up to the start of it
if( image_index !== -1 ) // if that URL is found
{
var delimiter_index = document.body.outerHTML.indexOf( '"', image_index ); // find first doublequote delimiter after URL
image_url = document.body.outerHTML.substring( image_index, delimiter_index ); // grab delimited URL
image_url = image_url.replace( '/screen/', '/full/' ); // turn screen URL into full URL - we don't care if /screen/ is already full-size, because /full/ will kindly redirect anyway
}
// New way, 2/22/2016 site redesign
// extract_image_url_after( "class='widget_imageFromSubmission", "http" );
// if( unsafeWindow.highdefURL ) { image_url = unsafeWindow.highdefURL; } // sloppy JS - 'if var' as in 'if this exists,' 'if this isn't undefined.'
wait_for_dnr = true;
// if this page is the landing page for a multi-image submission, do not redirect
//if ( document.body.outerHTML.indexOf( '<form id="changethumboriginal_form"' ) !== -1 && !address_bar_contains( '&page=' ) ) {
// Look for 'show custom thumbnails' button (indicating multi-page submission) or #pictop (which doesn't appear on landing pages for multi-page submissions)
if ( document.body.outerHTML.indexOf( '<form id="changethumboriginal_form"' ) !== -1 && !address_bar_contains( '#pictop' ) ) {
image_url = ''; // note: we do redirect on URLs for individual pages, including the first.
}
}
function scrape_furrynetwork() {
var images = document.getElementsByClassName( 'image__img' );
console.log( images.length );
console.log( document.getElementById( 'lightbox-img' ) ); // wtf, null?
image_url = document.getElementById( 'lightbox-img' ).src;
// <img class="image__img" src="https://d3gz42uwgl1r1y.cloudfront.net/ch/cheetahpaws/submission/2016/03/54643764f60aa5db3cef8341dca52d31/2500x1500.jpg" alt="Zootopia Wolves" id="lightbox-img" data-reactid=".0.2.0.0.0.$451553.1">
}
function scrape_pixiv() {
// http://www.pixiv.net/member_illust.php?mode=medium&illust_id=52107168 404s.
// goes to http://i1.pixiv.net/img-original/img/2015/08/21/19/14/00/52107168_p0.jpg
// should be http://i1.pixiv.net/img-original/img/2015/08/21/19/14/00/52107168_p0.png
// Yet another filetype mismatch. Great! No clear idea how to handle this, aside from fetching ?mode=big.
// http://i3.pixiv.net/img-original/img/2015/08/04/22/56/24/51785706_p0.jpg too
// Possible solution: redirect from '.jpg' with 404 text to '.png.' not a loop because not finding the 404 message means we don't redirect.
// https://www.pixiv.net/member_illust.php?mode=medium&illust_id=53192902#dnr 404's to
// https://i.pximg.net/img-original/img/2015/10/24/23/35/39/53192902_p0.jpg
extract_image_url_after( 'class="_illust_modal', '//' ); // Oh, what now? Code below doesn't work for some pages, so do this instead. (This goes first because only the last successful 'extraction' matters.)
extract_image_url_after( 'class="big"', '//' ); // New Pixiv pages (Dec '14) provide the big URL rather directly.
if( image_url === '') { // try this old nonsense first, because god forbid these sites update all their code to be remotely fucking consistent
extract_image_url_after( 'bookmark_modal_thumbnail', '//' ); // grab bookmark-thumbnail image from "medium" landing page
}
if( image_url.indexOf( '/c/' ) > -1 ) { // Convert thumbnail URL to full-size. Used to only happen for bookmark_modal if(), but Pixiv changed something.
// convert URL to full-size.
image_url = image_url.replace( '_m.', '.' ); // old style: remove _m for full-size URL
image_url = image_url.replace( '/c/600x600', '' ); // new style: remove /c/600x600, swap image-master for image-original, remove _master1200.
image_url = image_url.replace( '/c/150x150', '' ); // new style: remove /c/150x150, swap image-master for image-original, remove _master1200.
image_url = image_url.replace( '/img-master/', '/img-original/' );
image_url = image_url.replace( '_master1200', '' );
}
// Through sheer accident, the old manga code still works after Pixiv's latest change.
if( document.getElementsByTagName('html')[0].innerHTML.indexOf( '<a href="member_illust.php?mode=manga' ) > 0 ) { // If the works_display preview links to the manga, go there instead
image_url = window.location.href.replace( 'mode=medium', 'mode=manga' ); // manga pages deserve their own HTML, so just go to that page
// Users: please consider Eza's Pixiv Fixiv, which replaces the default manga HTML with full images and none of that scroll-to-load nonsense.
}
// Don't redirect to "Ugoira" animations (ZIP full of JPGs, played as HTML slideshow)
if( document.getElementsByTagName('html')[0].innerHTML.indexOf( 'class="_ugoku' ) > 0 ) { // A little messy since we ditched html_copy, isn't it?
image_url = ''; // prevent redirect by blanking image_url
// add link to ZIP for Ugoira, purely for archival purposes
var ugoku_link = pixiv.context.ugokuIllustData.src;
document.getElementsByClassName( '_ugoku-illust-player-container' )[0].innerHTML += '<br><a href="' + ugoku_link + '">Download Ugoku frames as ZIP file</a>';
}
}
// Maybe clean this up now that Gelbooru's stupid shit gets its own function.
function scrape_booru() { // this works on a wide variety of booru-style imageboards.
extract_image_url_after( '>Resize image</a>', 'http://' ); // for booru's which have automatic resizing and images which require it
// Gelbooru's anti-adblock shit might make the script fail the FIRST time you load a page, but not subsequent times. God dammit.
// Might be time for Gelbooru to get its own scrape function, because god damn.
extract_image_url_after( "$('edit_form')", '//' ); // For booru's with automatic resizing on, use the Original Image link, which appears after the Edit button
if( image_url === '' ) { extract_image_url_after( "$('resized_notice')", '//' ); } // Hey guess what! Gelbooru now serves different sidebars for adblock. Fuck you!
if( image_url === '' ) { extract_image_url_after( 'class="showEditBox">', '//' ); } // Hey guess what!!! Gelbooru now just tells you not to adblock! Fuuuck youuu!
// if( image_url === '//gelbooru.com' ) { image_url = ''; } // Kludge
if( image_url === '' ) { // otherwise, use the image that's being displayed
var container = document.getElementById( 'image' ); // Instead of lurching through raw HTML, let's just grab the display image via the DOM.
image_url = container.src; // "You think it's cool that things don't always have to be a federal fucking issue."
}
}
/*
Test suite of random URLs from the relevant sites:
http://www.hentai-foundry.com/pictures/user/Bottlesoldier/133840/Akibabuse
http://www.hentai-foundry.com/pictures/user/Bottlesoldier/214533/Lil-Gwendolyn
https://inkbunny.net/submissionview.php?id=483550
https://inkbunny.net/submissionview.php?id=374519
http://rule34.xxx/index.php?page=post&s=view&id=1399731
http://rule34.xxx/index.php?page=post&s=view&id=1415193
http://equi.booru.org/index.php?page=post&s=view&id=56940
http://furry.booru.org/index.php?page=post&s=view&id=340299
http://derpibooru.org/470074?scope=scpe80a78d33e96a29ea172a0d93e6e90b47c6a431ea
http://mspabooru.com/index.php?page=post&s=view&id=131809
http://mspabooru.com/index.php?page=post&s=view&id=131804
http://shiniez.deviantart.com/art/thanx-for-5-m-alan-in-some-heavy-makeup-XD-413414430
http://danbooru.donmai.us/posts/1250724?tags=dennou_coil
http://danbooru.donmai.us/posts/1162284?tags=dennou_coildata:text/html,<img src='http://example.com/image.jpg'>
http://www.furaffinity.net/view/12077223/
http://gamesbynick.tumblr.com/post/67039820534/the-secrets-out-guys-the-secret-is-out
http://honeyclop.tumblr.com/post/67122645946/stallion-foursome-commission-for-ciderbarrel-d
http://shubbabang.tumblr.com/post/20990300285/new-headcanon-karkat-is-ridiculously-good-at
http://www.furaffinity.net/view/12092394/
https://e621.net/post/show?md5=25385d2349ae11f2057874f0479422ad
http://sandralvv.tumblr.com/post/64933897836/how-did-varrick-get-that-film-cuz-i-want-a-copy
*/