Creates a new page showing just the images from any Tumblr
// ==UserScript==
// @name Eza's Tumblr Scrape
// @namespace https://inkbunny.net/ezalias
// @description Creates a new page showing just the images from any Tumblr
// @license MIT
// @license Public domain / No rights reserved
// @include http://*?ezastumblrscrape*
// @include https://*?ezastumblrscrape*
// @include http://*/ezastumblrscrape*
// @include http://*.tumblr.com/
// @include https://*.tumblr.com/
// @include http://*.tumblr.com/page/*
// @include https://*.tumblr.com/page/*
// @include http://*.tumblr.com/tagged/*
// @include https://*.tumblr.com/tagged/*
// @include http://*.tumblr.com/search/*
// @include https://*.tumblr.com/search/*
// @include http://*.tumblr.com/post/*
// @include https://*.tumblr.com/post/*
// @include https://*.media.tumblr.com/*
// @include https://media.tumblr.com/*
// @include http://*/archive
// @include https://*/archive
// @include http://*.co.vu/*
// @exclude */photoset_iframe/*
// @exclude *imageshack.us*
// @exclude *imageshack.com*
// @exclude *//scmplayer.*
// @exclude *//wikplayer.*
// @exclude *//www.wikplayer.*
// @exclude *//www.tumblr.com/search*
// @grant GM_registerMenuCommand
// @version 5.17
// ==/UserScript==
// Create an imaginary page on the relevant Tumblr domain, mostly to avoid the ridiculous same-origin policy for public HTML pages. Populate page with all images from that Tumblr. Add links to this page on normal pages within the blog.
// This script also works on off-site Tumblrs, by the way - just add /archive?ezastumblrscrape?scrapewholesite after the ".com" or whatever. Sorry it's not more concise.
// Make it work, make it fast, make it pretty - in that order.
// TODO:
// I'll have to add filtering as some kind of text input... and could potentially do multi-tag filtering, if I can reliably identify posts and/or reliably match tag definitions to images and image sets.
// This is a good feature for doing /scrapewholesite to get text links and then paging through them with fancy dynamic presentation nonsense. Also: duplicate elision.
// I'd love to do some multi-scrape stuff, e.g. scraping both /tagged/homestuck and /tagged/art, but that requires communication between divs to avoid constant repetition.
// post-level detection would also be great because it'd let me filter out reblogs. fuck all these people with 1000-page tumblrs, shitty animated gifs in their theme, infinite scrolling, and NO FUCKING TAGS. looking at you, http://neuroticnick.tumblr.com/post/16618331343/oh-gamzee#dnr - you prick.
// Look into Tumblr Saviour to see how they handle and filter out text posts.
// Add a convenient interface for changing options? "Change browsing options" to unhide a div that lists every ?key=value pair, with text-entry boxes or radio buttons as appropriate, and a button that pushes a new URL into the address bar and re-hides the div. Would need to be separate from thumbnail toggle so long as anything false is suppressed in get_url or whatever.
// Dropdown menus? Thumbnails yes/no, Pages At Once 1-20. These change the options_map settings immediately, so next/prev links will use them. Link to Apply Changes uses same ?startpage as current.
// Could I generalize that the way I've generalized Image Glutton? E.g., grab all links from a Pixiv gallery page, show all images and all manga pages.
// Possibly @include any ?scrapeeverythingdammit to grab all links and embed all pictures found on them. single-jump recursive web mirroring. (fucking same-domain policy!)
// now that I've got key-value mapping, add a link for 'view original posts only (experimental).' er, 'hide reblogs?' difficult to accurately convey.
// make it an element of the post-scraping function. then it would also work on scrape-whole-tumblr.
// better yet: call it separately, then use the post-scraping function on each post-level chunk of HTML. i.e. call scrape_without_reblogs from scrape_whole_tumblr, split off each post into strings, and call soft_scrape_page( single_post_string ) to get all the same images.
// or would it be better to get all images from any post? doing this by-post means we aren't getting theme nonsense (mostly).
// maybe just exclude images where a link to another tumblr happens before the next image... no, text posts could screw that up.
// general post detection is about recognizing patterns. can we automate it heuristically? bear in mind it'd be done at least once per scrape-page, and possibly once per tumblr-page.
// user b84485 seems to be using the scrape-whole-site option to open image links in tabs, and so is annoyed by the 500/1280 duplicates. maybe a 'remove duplicates' button after the whole site's done?
// It's a legitimately good idea. Lord knows I prefer opening images in tabs under most circumstances.
// Basically I want a "Browse Links" page instead of just "grab everything that isn't nailed down."
// http://mekacrap.tumblr.com/post/82151443664/oh-my-looks-like-theres-some-pussy-under#dnr - lots of 'read more' stuff, for when that's implemented.
// eza's tumblr scrape: "read more" might be tumblr standard.
// e.g. <p class="read_more_container"><a href="http://ladylovelycocks.tumblr.com/post/66964089115/stupid-comic-continued-under-readmore-more" class="read_more">Read More</a></p>
// http://c-enpai.tumblr.com/ - interesting content visible in /archive, but every page is 'themed' to be a blank front page. wtf.
// chokes on multi-thousand-page tumblrs like actual-vriska, at least when listing all pages. it's just link-heavy text. maybe skip having a div for every page and just append to one div. or skip divs and append to the raw document innerHTML. it could be a memory thing, if ajax elements are never destroyed.
// multi-thousand-page tumblrs make "find image links from all pages" choke. massive memory use, massive CPU load. ridiculous. it's just text. (alright, it's links and ajax requests, but it's doggedly linear.)
// maybe skip individual divs and append the raw pile-of-links hypertext into one div. or skip divs entirely and append it straight to the document innerHTML.
// could it be a memory leak thing? are ajax elements getting properly released and destroyed when their scope ends? kind of ridiculous either way, considering we're holding just a few kilobytes of text per page.
// try re-using the same ajax object.
/* Assorted notes from another text file
. eza's tumblr fixiv? de-style everything by simply erasing the <style> block.
. eza's tumblr scrape - test finishing whole page for displaying updates. (maybe only on ?scrapewholesite.) probably not too smart, but an interesting benchmark. only ever one document.body.innerHTML=thing;.
. eza's tumblr scrape: definitely do everything-at-once page write for thumbnail/browse mode. return one list of urls per page, so e.g. ten separate lists. remove duplicates between all lists. then build the page and do a single html write. prior to that, write 'fetching pages...' or something. it should be pretty quick. it's not like it's terribly responsive when loading pages anwyay. scrolling doesn't work right.
. thinking e.g. http://whatdoesitlumpingmean.tumblr.com/archive?startpage=1?pagesatonce=10?find=/tagged/my-art?ezastumblrscrape?thumbnails which has big blue dots on every post.
. see also http://herblesbians.tumblr.com/ with its gigantic tall banners
. alternate solution: check natural resolution, don't downscale tiny / narrow images.
. eza's tumblr scrape: why doesn't the thumbnail page pick up mspadventures.com gifs? e.g. http://kitkaloid.tumblr.com/page/26, with tavros's face being 'dusted.'
*/
// Soft_scrape_page should return non-image links from imgur, deviantart, etc., then collect them at the bottom of the div in thumbnail mode.
// links to embedded videos? linkdump at top, just below the /page link. looks like https://www.tumblr.com/video_file/119027046245/tumblr_nlt061qtgG1u32sbu/480 - e.g. manyakis.tumblr.
// Tumblr has a standard mobile version. Fuck me, how long has that been there? example.tumblr.com/mobile, no CSS, bare image links. Shit on the fuck.
// Hey hey! This might allow trivial recognition of individual posts and reblogs vs. OC. Via, but no source... weak. Good enough for us, though.
// Every post is between a <p> and </p>, but can contain <p></p> blocks inside. Messy.
// Reblogs say "(via <a href='http//example.tumblr.com/123456'>example</a>)" and original posts don't.
// Dates are noted in <h2> blocks, but they're outside any <p> blocks, so who cares.
// Images are linked (all in _500, grr, but we can obviously deal with that) but posts aren't. Shame. That would've been useful.
// Shit, consider the basics... do tags works? Pagination is just /mobile/page/2, etc. with tags: example.tumblr.com/tagged/homestuck/page/2/mobile.
// Are photosets handled correctly? What about read-more links? Uuugh, photosets just appear as "[video]". Literally that text. No link. Fuck! So close, aaand useless.
// I can use /mobile instead of /archive, but there's no point. It breaks favicons and I still have to fetch the fat-ass normal pages.
// I can probably use mobile pages to match normal pages, since they... wait, are they guaranteed to have the same post count? yes. alice grove has one post per page.
// So to find original posts, I have to fetch both normal and mobile pages, and... shit, and consistently separate posts on normal pages. It has to be identical.
// I can also use mobile for page count, since it's guaranteed to have forward / backward links. Ha! We can start testing at 100!
// Adding /mobile even works on individual posts. You can get a via link from any stupid theme. Yay.
// Add "show via/source links" or just "visit mobile page" as a Greasemonkey action / script command?
// Tumblr's own back/forward links are broken in the mobile theme. Goes to: /mobile/tagged/example. Should go to: /tagged/example/mobile. Modify those pages.
// http://thegirlofthebeach.tumblr.com/archive - it's all still there, but the theme shows nothing but a 'fuck you, bye' message.
// Pages still provide an og:image link. Unfortunately, that's a single image, even for photosets. Time to do some reasoning about photoset URLs and their constituent image URLs.
// Oh yeah - mobile. That gives us a page count, at least, but then photosets don't provide even that first image.
// Add a tag ?usemobile for using /mobile when scraping or browsing images.
// To do when /archive works and provides photosets in addition to /mobile images: http://fotophest.tumblr.com
// Archive does allow seeking by year/month, e.g. http://fotophest.tumblr.com/archive/2012/4
// example.tumblr.com/page/1/mobile always points to example.tumblr.com. Have to do example.tumblr.com/mobile. Ugh.
// Archival note: since Tumblr images basically never disappear, it SHOULD suffice to save the full scrape of a blog into a text file. I don't need to temporarily mass-download the whole image set, as I've been doing.
// Add "tagged example" to title for ?find=/tagged/example, to make this easier.
// Make a browser for these text files. Use the image-browser interface to display ten pages at once (by opening the file via a prompt, since file:// would kvetch about same-origin policy.) Maintain page links.
// Filter duplicates globally, in this mode. No reason not to.
// Use ?lastpage or ?end to indicate last detected page. (Indicate that it's approximate.) I keep checking ?scrapewholesite because I forget if I'm blowing through twenty pages or two hundred.
// Given multiple ?key=value definitions on the same URL, the last one takes precedence. I can just tack on whatever multiple options I please. (The URL-generating function remains useful for cleanliness.)
// Some Tumblrs (e.g. http://arinezez.tumblr.com/) have music players in frames. I mean, wow. Tumblr is dedicated to recreating every single design mistake Geocities allowed.
// This wouldn't be a big deal, except it means the address-bar URL doesn't change when you change pages. That's a hassle.
// Images with e.g. _r1_1280 are revisions? See http://actual-vriska.tumblr.com/post/32788651941/ vs. its source http://cancerousaquarium.tumblr.com/post/32784513645/ - an obvious watermark has been added.
// Tested with a few random _r1 images from a large scrape's textfile. Some return XML errors ('no associated stylesheet') and others 404. Mildly interesting at best.
// Aha - now that we have an 'end page,' there can be a drop down for page selection. Maybe also for pages-at-once? Pages 1-10, 11-20, 21-30, etc. Pages at once: 1, 5, 10, 20/25?
// Possibly separate /post/ links, since they'll obviously be posts from that page. (Ehh, maybe not - I think e.g. Promstuck links to a "masterpost" in the sidebar.)
// Maybe hide links behind a button instead of ignoring them entirely? That way compactness is largely irrelevant.
// Stick 'em behind a button? Maybe ID and Class each link, so we can GetElementsByClassName and this.innerText = this.href.
// If multi-split works, just include T1 / O1 links in-order with everything else. It beats scrolling up and guessing, even vs page themes that suck.
// No multi-split. I'd have to split on one term, then foreach that array and split for another term, then combine all resulting arrays in-order.
// Aha: there IS multi-splitting, using regexes as the delimiter. E.g. "Hello awesome, world!".split(/[\s,]+/); for splitting on spaces and commas.
// Split e.g. src=|src="|src=', then split space/singlequote/doublequote and take first element? We don't care what the first terminator is; just terminate.
// How do YouTube videos count? E.g. http://durma.tumblr.com/post/57768318100/ponponpon%E6%AD%8C%E3%81%A3%E3%81%A6%E3%81%BF%E3%81%9F-verdirk-by-etna102
// Another example of off-site video: http://pizza-omelette.tumblr.com/post/44128050736/2hr-watch-this-its-very-important-i-still
// Some themes have EVERY post "under the cut," e.g. http://durma.tumblr.com/. Photosets show up. Replies to posts don't. ?usemobile should get some different stuff.
// Brute-force method: ajax every single /post/ link. Thus - post separation, read-more, possibly other benefits. Obvious downside is massive latency increase.
// Test on e.g. akitsunsfw.tumblr.com with its many read-more links.
// Probably able to identify new posts vs. reblogs, too. Worth pursuing. At the very least, I could consistently exclude posts whose mobile versions include via/source.
// <!-- GOOGLE CAROUSEL --><script type="application/ld+json">{"@type":"ItemList","url":"http:\/\/crystalpei.com\/page\/252","itemListElement":[{"@type":"ListItem","position":1,"url":"http:\/\/crystalpei.com\/post\/30638270842\/actually-this-is-a-self-portrait"}],"@context":"http:\/\/schema.org"}</script>
// <link rel="canonical" href="http://crystalpei.com/page/252" />
// Does this matter? Seems useful, listing /post/ URLs for this page.
// 10,000-page tumblrs are failing with ?showlinks enabled. the text gets doubled-up. is there a height limit for firefox rendering a page? does it just overflow? try a dummy function that does no ajax and just prints many links.
// There IS a height limit! I printed 100 lines of dummy text for 10,000 pages, and at 'page' 8773, the 27th line is halfway offscreen. I cannot scroll past that.
// Saving as text does save all 10,000 pages. So the text is there... Firefox just won't render it.
// Reducing text size (ctrl-minus) resulted in -less- text being visible. Now it ends at 8729.
// http://shegnyanny.tumblr.com/ redirects to https://www.tumblr.com/dashboard/blog/shegnyanny - from every page. But it's all still there! Fuck, that's aggravating!
// DaveJaders needs that dashboard scrape treatment. Ditto lencrypted.
// I need a 'repeatedly click 'show more notes' function' and maybe that should be part of this script.
// 1000+ page examples: http://neuroticnick.tumblr.com/ - http://teufeldiabolos.co.vu/ - http://actual-vriska.tumblr.com/ - http://cullenfuckers.tumblr.com/ - http://soupery.tumblr.com - http://slangwang.tumblr.com - some with 10,000 pages or more.
// JS 2015 has "Fetch" as standard? Say it's true, even if it's not!
// The secret to getting anything done in-order seems to be creating a function (which you can apparently do anywhere, in any scope) and then calling that_function.then.
// So e.g. function{ do_stuff( fetch(a).etc ) }, followed by do_stuff().then{ use results from stuff }.
// Promise.all( iterable ), ahhhh.
// Promise.map isn't real (or isn't standard?) but is just Promise.all( array.map( n => f(n) ). Works inside "then" too.
// setTimeout( 0 ) is a meaningful concept because it pushes things onto the event queue instead of the stack, so they'll run in-order when the stack is clear.
// Functions can be passed as arguments. Just send the name, and url_array=parse( body, look_for_videos ) allows parse(body,func) to do func( body ) and find videos.
// Test whether updating all these divs is what keeps blocking and hammering the CPU. Could be thrashing? It's plainly not main RAM. Might be VRAM. Test acceleration.
// Doing rudimentary benchmarking, the actual downloading of pages is stupid fast. Hundreds per minute. All the hold-up is in the processing.
// Ooh yeah, updating divs hammers and blocks even when it's just 'pagenum - body.length.'
// Does body.length happen all at once? JS in FF seems aggressively single-threaded.
// Like, 'while( array.shift )' obviously matches a for loop with a shift in it, but the 'while' totally locks up the browser.
// Maybe it's the split()? I could do that manually with a for loop... again. Test tall div updating with no concern for response content.
// Speaking of benchmarks, forEach is faster than map. How in the hell? It's like a 25% speed gain for what should be a more linear process.
// ... not that swapping map and forEach makes a damn bit of difference with the single-digit milliseconds involved here.
// Since it gets worse as the page goes on, is it a getElementById thing?
// Editing the DOM doesn't update an actual HTML file in memory (AFAIK), but you're scanning past 1000-odd elements instead of, like, 2.
// Audio files? Similar to grabbing videos. - http://joeckuroda.tumblr.com/post/125455996815 -
// Add a ?chrono option that works better than Tumblr's. Reverse url_array, count pages backwards, etc.
// Maybe just ?reverse for per-page reversal of url_array. Lazy option.
// http://crotah.tumblr.com desperately needs an original-posts filter.
// I can't return a promise to "then." soft_scrape returns promise.all(etc).then(etc). You can soft_scrape(url).then, but not .then( s=>soft_scrape(s) ).then. Because fuck you.
// I'm frothing with rage over this. This is a snap your keyboard in half, grab the machine by the throat, and scream 'this should work' problem.
// I know you can return a promise in a 'then' because that's the fucking point of 'then.' It's what fetch's response.text() does. It's what fetch does!
// Multi-tag scraping requires post-level detection, or at least, post enumeration instead of image enumeration.
// Grab all /post/ links in /tagged/foo, then all /post/ links in /tagged/bar, then remove duplicates.
// Sort by /post/12345 number? Or give preference to which page they appear on? The latter could get weird.
// CrunchCaptain has no /post/ links. All posts are linked with tmblr.co shortened URLs. Tumblr, you never cease to disappoint me.
// What do I want, exactly? What does innovation look like here?
// Not waiting for page scrapes. Scrapewholesite, then store that, and compose output as needed.
// Change window.location to match current settings, so they're persistent in the event of a reload. Fall back to on-demand page scraping if that happens.
// Maybe a 100-page limit instead of 1000, at least for this ?newscrape or ?reduxmode business.
// Centered UI, for clarity. Next and Previous links side-aligned, however clumsily. Touch-friendly because why not.
// Dropdowns for options_map. Expose more choices.
// Post-level splitting? Tags? Maybe Google Carousel support backed up by hacky fakery.
// I want to be able to grab a post URL while I'm looking at the image in browse mode.
// On-hover links for each image? Text directly below images? CSS nightmares, but presumably tractable.
// What I -really- want - what this damn script was half-invented for - is eliminating reblogs. Filtering down to only original posts. (Or at least posts with added content.)
// Remember that mobile mode needs special consideration. (Which could just be 'for each img' if mobile's all it handles.)
// Maybe skip the height-vs-width thumbnail question by having user options a la Pixiv Fixiv. Just slap more options into the CSS and change the body's class.
// GM menu commands for view-mobile, view-embed, and google-alternate-source?
// I'm overthinking this. Soft scrape: grab each in-domain /post/ link, scrape it, display its contents. The duplicate-finder will handle the crud.
// Can't reliably grab tags from post-by-post scraping because some people link their art/cosplay tags in their theme. Bluh.
// http://iloveyournudity.tumblr.com/archive?ezastumblrscrape?scrapewholesite?find=/?grabrange=1001?lastpage=2592 stops at page 1594. rare, these days.
// Could put the "Post" text over images, if CSS allows some flag for 'zero width.' Images will always(ish) be wider than the word "Post."
// Nailed it, more or less, thanks to https://www.designlabthemes.com/text-over-image-with-css/
// Wrapping on images vs. permalink, and the ability to select images sensibly, are highly dependent on the number and placement of spaces in that text block.
// Permalink in bar across bottom of image, or in padded rectangle at bottom corner? I'm leaning toward bar.
// ... I just realized this breaks 'open in new tabs.' You can't select a bunch of images and open them because you open the posts too.
// Oh well. 'Open in new tabs' isn't browser-standard, and downthemall selection works fine.
// InkBunny user Logitech says the button is kind of crap. It is. I think I said I'd fix it later a year ago.
// Holy shit, Tumblr finally updated /mobile to link to posts. Does it include [video] posts?! ... no. Fuck you, Tumblr.
// Y'know, if page_dupe_hash tracked how many times each /tagged/ URL was seen, it'd make tag tracking more or less free.
// If I used the empty-page fetch to prime page_dupe_hash with negative numbers, I could count up to 0 and notice when a tag is legitimately part of a post.
// Should page_number( x ) be a function? Return example.tumblr.com/guff/page/x, but also handle /mobile and so on.
// Pages without any posts don't get any page breaks. That's almost a feature instead of a bug, but it does stem from an empty variable somewhere.
// Debating making the 'experimental' post-by-post mode the main mode. It's objectively better in most ways... but should be better in all ways before becoming the standard.
// For example: new_embedded_display doesn't handle non-1280 images. (They still exist, annoyingly.)
// I'm loathe to ever switch more than once. But do I want two pairs of '10 at once' / '1 at once' links? Which ones get bolded? AABB or ABAB?
// Maybe put the new links on the other side of the page. Alignment: right.
// 'Browse pages' vs. 'Browse posts?' The fact it's -images- needs to prominent. 'Image viewer (pages)' vs. 'Image viewer (posts)?' '100 posts at once?' (Guesstimated.)
// Maybe 'many posts at once' vs. 'few posts at once.' Ehhh.
// The post browser could theoretically grab from /archive, but I'd need to know how to paginate in /archive.
// The post browser could add right-aligned 'source' links in the permalink bar... and once we're detecting that, even badly, we can filter down to original posts!
// Mobile mode is semi-helpful, offering a standard for post separation and "via" / "source" links. But: still missing photosets. Arg.
// ... and naively you'd miss cases where this tumblr has replied and added artwork. Hmm.
// Not sure if ?usemobile works with ?everypost. I think it use to, a couple days ago.
// In "snap rows" mode, max-width:100% doesn't respect aspect ratio. Wide images get squished.
// They're already in containers, so it's something like .container .fixed-height etc. Fix it next version.
// The concept of 'as big as possible' is annoyingly complicated. I think I need the container to have a max-width... but to fit an image?
// Container height fixed, container with max-width, image inside container with 100% container width and height=auto? Whitespace, but whatever.
// Container max-height 240px, max-width 100%... then image width 100% of container... then container height to match image inside it? is min-height a thing?
// Lazy approach to chrono: make a "reverse post order" button. Just re-sort all the numbered IDs.
// To use the post-by-post scraper for scrapwholesite (or a novel scrapewholesite analog), I guess I'd only have to return 'true' from some Promise.all.
// /Post URL, contents, /post URL, contents, repeat. Hide things with divs/spans and they should save to text only when shown.
// Reverse tag-overview order. I hate redoing a coin-flip decision like that, but it's a serious usability shortcoming. I want to hit End and see the big tags.
// GM command or button for searching for missing pages? E.g. google.com?q=tumblr+username-hey-heres-some-art if "there's nothing here."
// http://cosmic-rumpus.tumblr.com/post/154833060516/give-it-up-for-canon-lesbans11-feel-free - theme doesn't show button - also check if this photoset appears in new mode
// The Scrape button fucks up on www.example.tumblr.com, e.g. http://www.fuckyeahhomestuckkids.tumblr.com/tagged/dave_strider
// Easy filter for page-by-page scraper: ?require and hide any images missing the required tag.
// Same deal, ?exclude to e.g. avoid dupes. Go through /tagged/homestuck-art and then go through /tagged/hs-art with ?exclude=/tagged/homestuck-art.
// Oops, Scrape button link shows up in tag overview. Filter for indexOf ?ezastumblrscrape. Also page links. Sloppy on my part.
// Also opengraph, wtf? - http://falaxy.tumblr.com/archive?ezastumblrscrape?scrapewholesite?find=/tagged/Gabbi-draws - 15 http://falaxy.tumblr.com/tagged/Gabbi+draws?og=1 -
// Son of a fuck over the Scrape button - http://ayanak.tumblr.com/ - 'install theme' garbage
// Consider an isolated tag-overview mode for huge tumblrs. Grabbing pages is pretty quick. Processing and displaying them is dog-slow.
// SCM Player nonsense breaks the Scrape button: http://homestuck-arts.tumblr.com/
// In post-by-post interface: 'just this page' next to page links, e.g., 'Page 27 - Scrape just this page' - instead of 'one at once' / 'ten at once' links at top.
// Tag overview / global duplicate finder - convert everything to lowercase. Tumblr doesn't care.
// Redux scrapewholesite: format should be post link on one line, then image(s) on following lines, maybe with one or the other indented.
// Tags included? This could almost get XML-ish.
// Holy shit, speaking of XML - http://magpizza.tumblr.com/sitemap1.xml
// It's not the first page. That 'breathing-breathless-heaving-breaths' post (9825320766) shows up on page... 647? the fuck?
// Good lord, it might be by original post dates. Oh wow. It is. And each file is huge. 500 posts each!
// I don't need to parse the XML because the only content inside each structure is a URL and a last-modified date code.
// Missing files return a standard "not found" page, like you request /page ten billion. So it's safe.
// Obviously this requires (and encourages!) a post-level scraper.
// Even 1000-page Tumblrs will only have tens of these compact XML files, so it's reasonable to grab all of them, list URLs, and then fetch in reverse-chrono.
// Oh, neverfuckingmind - /sitemap.xml points to all /sitemap#.xml files. Kickass.
// Make sure this is standard before getting ahead of yourself. Tispy has it. 'Kinsie' has it, both HTTP. Catfish.co.vu has it.
// HTTPS? PunGhostHere has it, but it's empty, obviously. "m-arci-a" has it. I think it's fully standard. Hot dang.
// Does it work for tags? Like at all? Useful for ?scrapewholesite regardless.
// http://tumblino.tumblr.com/tagged/dercest/sitemap1.xml - not it.
// http://amigos-americas.tumblr.com/sitemap-pages.xml - is this also standard? side links, custom pages.
// /sitemap.xml will list /sitemap-pages if it exists. Ignore it.
// Might just need to make a single-page Ajax-y "app" viewer for this.
// Get the "Scrape" link off of photoset iframes
// Maybe make tag overview point directly to scrape links? Or just add scrape links.
// Fuck about in the /embed source code, see if there's any kind of /standard post that it's displaying without a theme.
// If that's all in the back-end then there's nothing to be done.
// http://slangwang.tumblr.com/post/61921070822/so-scandalous/embed
// https://embed.tumblr.com/embed/post/_tZgwiPCHYybfPeUR0XC9A/61921070822?width=542&language=en_US&did=d664f54e36ba2c3479c46f7b8c877002f3aa5791
// Okay so the post ID is in that hot mess of a URL, but there's two high-entropy strings that are probably hidden crap.
// Changing the post ID (to e.g. 155649716949, another post on the same blog) does work within the blog, at least for now.
// Changing the post ID to posts from other blogs, or to made-up numbers, does not work. "This post is gone now" is the error. Nuts.
// "&did" value doesn't matter. Might be my account's ID? The only script in /embed pages mostly registers page-views.
// http://sometipsygnostalgic.tumblr.com/post/157273191096/anyway-i-think-that-iris-is-the-hardest-champion/embed
// https://embed.tumblr.com/embed/post/4RwtewsxXp-k1ReCcdAgXg/157273191096?width=542&language=en_US&did=2098cf23085faa58c1b48d717a4a5072f591adb5
// Nope, "&did" is different on this sometipsygnostalgic post.
// http://sometipsygnostalgic.tumblr.com/post/157273080206/the-role-of-champions-in-the-games-is-confusing/embed
// https://embed.tumblr.com/embed/post/4RwtewsxXp-k1ReCcdAgXg/157273080206?width=542&language=en_US&did=86e94fe484d6739f7a26f3660bb907f82d600120
// the _tZg etc. value is probably Tumblr's internal reference to the user.
// So I'd only have to get it once, and then I could generate these no-bullshit unthemed URLs based on /post IDs from sitemap.xml.
// /embed source code refers to the user-unique ID as the "embed_key". It's not present in /archive, which would've been super convenient.
// global.build.js has references to embed_key; dunno if I can just grab it.
// It seems like this script should be placing the key in links - data-post-id, data-embed-key, data-embed-did, etc. Just not seeing it in /archive source.
// Those data-etc values are in a <ul> (the fuck is a ul?) inside a div with classes like "popover".
// One div class="tumblr-post" explicitly links the /embed/post version with the embed_key and post_id. Where is that?
// post_micro_157274781274 in the link div class ID is just the /post number. What's post_tumblelog_4b5d01b9ddd34b704fef28e6c9d83796 about?
// Screw it, it's in every /embed page. Just grab one. (Sadly it does require a real /post ID.)
// Pass ?embed_id along in all links, otherwise modes that need it will have to rediscover it.
// Don't fetch /archive or whatever. Grab some /post from the already-loaded innerHTML before removing it.
// Wait, shit! I can't get embed.tumblr.com because it's on a different subdomain! Bastards!
// Nevermind, anything.tumblr.com/embed works. Anything. That's weirdly accomodating.
// Seems they broke it before I got around to implementing this. Nerts.
// So: grab a post/12345/embed link, get the user id, redirect(?) to some ?embedded=xyz123 url.
// embed_key":"gLSpeeMF7v3RdBncMIHy2w is not hard to find.
// The use of sitemap.xml could be transformative to this script, or even split off into another script entirely. It finally allows a standard theme!
// Just do it. Entries look like:
// <url><loc>http://leopha.tumblr.com/post/57593339717/hello-a-friend-found-my-blog-somehow-so-url</loc><lastmod>2013-08-07T06:42:58Z</lastmod></url></urlset>
// So split on <url><loc> and terminate at </loc>. Or see if JS has native XML parsing, for honest object-orientation instead of banging at text files.
// Grab normally for now, I guess, just to get it implemented. Make ?usemobile work. Worry about /embed stuff later.
// Check if dashboard-only blogs have xml files exposed.
// http://baklavalon.tumblr.com/post/16683271151/theyre-porking-in-that-room needs this treatment, its theme is just the word "poop".
// Holy shit, /archive div classes include tags like not_mine, is_reblog, is_original, with_permalink, has_source - this is a fucking goldmine.
// http://miyuli.tumblr.com/archive?startpage=16?pagesatonce=5?thumbnails?find=/tagged/fanart?ezastumblrscrape?everypost?lastpage=22 breaks on:
// http://miyuli.tumblr.com/tagged/fanart/page/16
// http://miyuli.tumblr.com/archive?startpage=16?pagesatonce=5?thumbnails?find=/tagged/fanart?ezastumblrscrape?lastpage=22 - old method works
// Could just be dupes from nearby pages? Nope.
// Same deal here - http://miyuli.tumblr.com/archive?startpage=36?pagesatonce=5?thumbnails?find=/tagged/original?ezastumblrscrape?everypost?lastpage=39 - breaks:
// http://miyuli.tumblr.com/tagged/original/page/36
// http://miyuli.tumblr.com/archive?startpage=36?pagesatonce=5?thumbnails?find=/tagged/original?ezastumblrscrape?lastpage=39 - works
// Or... seems to be getting everything, but putting it under the wrong div. "Toggle page breaks" just breaks in the wrong places. Huh.
// http://northernvehemence.tumblr.com/archive?startpage=21?pagesatonce=5?thumbnails?ezastumblrscrape?lastpage=42?everypost - everything's there, but misplaced
// a-box.tumblr.com redirects on all pages, but there's still content there. Scrapes normally, though? Good case for embedded links instead of bare /post links.
// Oddball feature idea: add ?source=[url]?via=[url] for tumblrs with zero hint of content in their URLs.
// E.g. if https://milkybee.tumblr.com/post/31692413218 goes missing, I have no fucking idea what's in it. It is a mystery.
// Easy answer, from ?everypost - add image URL after permalink.
// Make ?ezastumblrscrape (no ?scrapewholesite) some kind of idiot-proof landing-page with links to various functions. As always, it's mostly for my sake, but it's good to have.
// /page mode - "Previous/Next (pagesatonce) pages" and also have "Previous/Next 1 page".
// http://sybillineturtle.tumblr.com/post/138070422744/crookedxhill-cosplay-crookedxhill - I need a way for eza's tumblr scrape to handle these bullshit dashboard-only blogs
// Oh my god: studiomoh.com/fun/tumblr_originals/?tumblr=TUMBLRNAME returns original posts for TUMBLRNAME.tumblr.com.
// http / https changes break all the ?key=pair values, but there's not much to be done about it.
// If I used Tumblr the way normal people do, I'd probably want an image browser for the dashboard.
// Is there any sort of sitemap for generic www.tumblr.com/tagged/example pages?
// I do almost nothing on the dashboard, but a way to dump entire tags would be fucking incredible.
// The tag overview is presumably what's causing the huge CPU spike when a long scrapewholesite ends. That or the little "Done" at the top.
// Presumed solution: insert those things in whatever way the scraped pages are inserted. Generate divs at the start, I guess.
// Failed to have any effect. Am I doing something I should be peppering with setTimeout(0)s? JS multitasking is dumb.
// Can I do anything for the identifiability of pages with no text after the number?
// E.g. woofkind-blog.tumblr.com/post/13147883338 += #tumblr_luzrb2F9Hv1qm6tzgo1_1280.jpg so there's something to go on if the blog dies.
// Tag overview still picks up other tumblrs, which is okay-ish, but it borks the "X posts" links that go straight to scraping. Filter them.
// Embed-based scraping could probably identify low-reblog content (low note count) for archival purposes.
// Options_url() seems to add an extra question mark to the end of every URL. Doesn't break anything; just weird.
// sarahfuarchive needs that /embed-based scrape. ?usemobile is a band-aid.
// Okay, so the "Done." does generate a newline when it appears. Added a BR and crossed my fingers.
// Is it just the tag overview? Is it just the huge foreach over page_dupe_hash?
// Lag sometimes happens in the middle of a long scrape. If it's painting-related, maybe it's from long URLs forcing the screen to become wider.
// Consider fetching groups of 10 instead of 25 when grabbing in the high thousands. Speed is not an issue, and Tumblr seems to lag more for 'deeper' pages.
// Consider newline after bottom-div prev/next buttons, because popup url notifiers are fucking garbage. Put it back in my status bar, Firefox. That's what it's for.
// "Tumblr Image Size" is no longer maintained (where'd I even get it?) and should maybe be folded into this script. Ajax instead of fetch because we'd only look for error?
// Maybe as part of Image Glutton. Also include Twitter :large or :orig nonsense.
// Pretty sure I can use that embed.tumblr view in ?scrapemode=everypost just by futzing the permalink URLs. Make it separate there, like ?usemobile.
// Also change embedded tags to search THIS domain instead of linking to general Tumblr searches. What even.
// czechadamick.tumblr.com as an example with no tags anywhere
// I should really test for photosets that might not appear. Maybe some ?verbose mode where it links anything that looks like a URL.
// Is there any way to resize inline images? They're so tiny!
// http://static.tumblr.com/8781b0eeeb05c5d67462fb52fb349b45/zpdyuqm/LQ4ojqrq2/tumblr_static_bcae2zpugw000c0sog404804k.png - this one is huge!
// https://66.media.tumblr.com/a38c7bd5148aded5494d4f8a70fce217/tumblr_inline_mlqzd2xLoZ1qz4rgp.jpg - this one is tiny. Not the same piece, obviously.
// https://media.tumblr.com/0625918cc22c673b4fb6db33ee44188b/tumblr_inline_mm3w29L9xN1qz4rgp.jpg
// https://static.tumblr.com/865f4c82ae6fc0a9dc0631a646475a86/ghtmrj5/8hCop2k79/tumblr_static_9usneomihhc0s0oc0w0s0sw8c_2048_v2.gif
// The _2048_v2 is definitely sizing, because that image works without it.
// http://studentsmut.tumblr.com/post/161643756857/anoneifanocs-mage0fheart-anoneifanocs-hey has an inline image that works with with _raw. It's https://media.tumblr.com/3a5ddf70b08a79c593fb05063ce011ed/tumblr_inline_orb5yhJuoV1tcvfhg_raw.png - and doesn't work with any other sizes? The hell?
// http://68.media.tumblr.com/a04c862afb3930c1250e067c23a887fc/tumblr_inline_owo0nfrLnO1rrmi7y_1280.png
// Oh! Duh! media.tumblr -> 66.media.tumblr! Nope, still breaks a lot.
// ?everypost isn't getting page/1 of http://shojoheart.tumblr.com/archive?startpage=1?pagesatonce=5?thumbnails=fixed-width?ezastumblrscrape?scrapemode=pagebrowser
// Not a general problem; this mode still works on other tumblrs. I think it's their theme. Yeesh.
// Might be more general - e.g. fetching /page/1 redirects to /, so the page fetched might look empty. Arg. Ship it, it's not new.
// http://grimdarkcake.tumblr.com tag overview doesn't work even though the pages have tags - ditto ticklishivories
// Scrape link from /archive is broken; erase ?find if it reads '/archive'.
// http://shittyhorsey.tumblr.com/ and http://venomous-sausage.tumblr.com/ fail before counting pages. WTF. Page counts use /mobile; there no way for themes to interfere.
// "The Same Origin Policy disallows reading the remote resource at https://www.tumblr.com/safe-mode?url=http://shittyhorsey.tumblr.com/page/1000/mobile."
// Okay, looking on the bright side, Tumblr has a "safe-mode" that might be easily abused. Look into that later. (Nope, just redirects. Poop.)
// Raiseshipseerve is fucked now too. What bullshit is Tumblr pulling? Other pages still work, obviously, since I've been using this for days straight.
// NRFW still works! Good, it's not totally fucky for https tumblrs.
// fetch( 'https://www.tumblr.com/safe-mode?url=http://nrfw.tumblr.com/page/2/mobile' ).then(s=>console.log(s)) does work from www.tumblr.com. It's not all crazy.
// -And- it's not a redirect (somehow), so... no it still returns an empty page. Fuck you, Tumblr.
// Yeah, you can set site_and_tags to some safe-mode link, but it still can't count pages. Fuck.
// embed.tumblr.com should still work, once I implement it. Even if you have to go to an /embed page to get the secret ID without fetches.
// Back to roots, at least for my personal use, maybe a forced theme? Just get all images on the page right now and put them in standard format.
// Oh thank god - Greasyfork user Petr Savelyev identified it as a cookies-related issue and gave an easy fix. fetch( url, { credentials: 'include' } ).
// Trust but verify: 'including credentials' does not appear to allow anything skeezy.
// No apparent way to set this for the whole script. Nuts.
// So, uh. How much of Tumblr works with credentials? Can I get any subdomain from any other subdomain? No, apparently not. Unfortunate.
// http://mediarama.tumblr.com/archive?startpage=1?pagesatonce=10?find=/tagged/cosplay?ezastumblrscrape?scrapemode=scrapewholesite?autostart is fucked somehow
// Make tag overview on slightly-gay-pogohammer work.
// Tag overview links choke on sites that end every tag with /chrono.
// The /search method supports boolean operators in weird ways. Foo+bar returns posts with both words. Foo|bar returns different results... somehow. Fewer?
// "Creates a new page showing just the images from any Tumblr" -> "Shows just the images from any Tumblr, using a new page?"
// Memory use and slowdown on 1000-page scrapes: am I just using "var" where I should use "let?" Be more HTML5-y.
// Is it a document.write thing? Just the fact I'm dropping in raw HTML? FF might have to re-parse the whole long-ass page.
// _raw files also work on the data.tumblr subdomain.
// e.g. http://data.tumblr.com/80882d3b1c6103e73b4de89761c804bd/tumblr_oe54v7tKhN1uctdepo1_raw.png
// xekstrin.tumblr.com fails every couple pages? It's been long enough that I forgot what causes this. Video?
// ?scrapewholesite should probably separate posts / links / images below each /page. Just a text indicator on a new line.
// Oh, and indicate there's a tag overview at the bottom. "Scroll do the end for an overview of tags." Bottom, not end. Summary of tags. Anchor link? No, people will manage.
// Add buttons for ?usesmall and ?notraw... and for neither. Genericize it into ?maxres=500 or something.
// Enable this only after adding resize functionality? That might need to be optional. I don't know if GreaseMonkey has, like, script cookies. JSON crap, I guess.
// Ooh, maybe make it a button on bare images. Little hovering button in the corner: 500, 1280, raw. (Still good to have a default.)
// 1280 with a raw button is an acceptable compromise. Raw images can be stupid large.
// Eza's Resize Automator. Eza's Image Embiggener. Eza's Bigass Imager. Eza's Image Magnifier. Eza's Squint Disabler. Eza's Size Queen. Eza's Resolution Glutton.
// Once this is integrated, I should link ?usesmall to _raw and then scale down as needed. Assuming that works. Might fuck up more on XML / 404s than going up from JPG.
// Ech, link _1280 at most. DownThemAll doesn't do JS redirects.
// Generalize ?usesmall and ?notraw to ?maxres=400 etc.
// Done. Now adjust those deprecated option names to maxres settings in the setup.
// Add options to image modes, probably below immediate/persistent scaling options. "Maximum resolution: Raw, 1280, 500, 400, 250, 100." I think those are the right numbers. They're all persistent links because fuck rejiggering the images on the page.
// Incidentally, _100 and _250 are not resized by the tumblr bare-image resizer I use. So I guess it's time to absorb that functionality. When that happens, add to the max-res options, "Opened links will display in higher resolution" or something like that. Clicked links? Clicked images? Maximum resolution? UI is hard.
// Ah, half-right: "Tumblr Image Size 1.1" doesn't work in //media.tumblr.com URLs. It needs a CDN number, like //66.media.tumblr.com. Might absorb the function anyway.
// Clean up the comments and dead code on this.
// The new method doesn't support ?usemobile, but it kinda can't, since /mobile pages have no post links.
// XML methods would work.
// Finally implemented the XML sitemap mode - and it finds "[video]" image sets with ?usemobile!
// Ironically I have no idea whether it finds actual videos.
// renkasbending doesn't show tags in xml mode, even though they're visible on the page. What.
// emiggax has no sitemap.xml! Whaaat? Ditto muteandthemew.
// So... ?find=/archive does find posts and images. Can I use that?
// Dashboard-only blogs - like https://www.tumblr.com/dashboard/blog/milky-morty - can still have the 'Scrape' link in the right place. (E.g., add /archive.) I can invoke functionality for them. It's just a right bitch to test anything, because the URL never changes and makes no difference. I'm not even sure I can request resources because all addresses resolve to the stupid dashboard thing.
// Oh, or https://milky-morty.tumblr.com/archive shows up fine. That works too.
// As does https://milky-morty.tumblr.com/sitemap.xml? Okay, so I might've accidentally solved this problem.
// Nevermind. At some point that guy took his blog out of dashboard-only mode.
// wwhatevven is dashboard-only. (And far less weird.) Nope, redirects everything to the /dashboard/blog crap. Even /archive and /sitemap.xml.
// Can I check if posts are original? Doesn't seem to be an OpenGraph / 'og:' thing. /embed has is_reblog, but we're not using that... yet. I should.
// /embed pages are desirable because they'll show their goddamn tags.
// Hey dingdong. You can still get thistumblr.tumblr.com/post links from the /page files, under most circumstances.
// Example /post with previous/next links: http://artisticships.tumblr.com/post/71222875632/spencerofspace-more-rose-cosplay-first-set-x
// Dashboard-only blogs: i-just-want-to-die-plz / thedevilandhisfiddle.
// Maybe always link to _1280 versions in thumbnail view? This would make ?maxres more useful, e.g. for DownThemAll.
// Nontrivial, surprisingly. I use a standard function for standardizing images - so the link starts the same as the inline image.
// Any /embed link for a wrong /post number delivers a standard 'not found' page, with the stupid animations.
// _raw still exists in some form but is restricted. I'd need to find examples in the wild.
// You know, I could still get /post numbers from /page... pages... so long as the blog works. This would allow per-post tags and image, but wouldn't fix theme-fucked blogs. It wouldn't work from /mobile, afaik. The main benefit would be 'under the cut' images.
// Note count would be nice. Maybe once I get standardized links through /embed trickery.
// media.tumblr.com just stopped working.
// ... started working again two days later. Tumblr was in a sorry state in the meantime. For once, a change on their end was clearly a bug on their end.
// Test grabbing 1000pgs at once.
// http://rosedai.tumblr.com/archive?ezastumblrscrape?scrapewholesite?find= - NaN pages, wtf?
// But http://rosedai.tumblr.com/archive?startpage=1?pagesatonce=10?ezastumblrscrape?scrapemode=scrapewholesite works. Wat.
// https://nightcigale.tumblr.com/archive?ezastumblrscrape?scrapewholesite?find= fails too. wtf. ?find=/ doesn't help.
// https://nightcigale.tumblr.com/archive?startpage=1?pagesatonce=10?ezastumblrscrape?scrapemode=scrapewholesite doesn't work either. Oh no. Did they break mobile?
// No, https://nightcigale.tumblr.com/page/10/mobile still works. Thank fuck. HTTP redirects to HTTPS as well, but test if it's that. (Shouldn't be; many tumblrs do HTTPS.)
// https://nightcigale.tumblr.com/page/100/mobile also works, for this 40-ish-page tumblr. Ditto https://nightcigale.tumblr.com/page/100000/mobile for the upper bound.
// And now it works suddenly. Fuck me, I guess. Tumblr just randomly wants to not respond to /mobile pages in a way that breaks the bounds check.
// Tumblr is deleting everything NSFW within two weeks. Fucking hell.
// Quick and dirty post-by-post method for text? Maybe extend XML. E.g. http://pinewreaths.tumblr.com/post/164287584755/smutmas-2017-masterpost
// It works, expose it.
// Okay, post-by-post needs to work with tags. Jerry-rig some system that takes /page + /mobile and filters posts matching the current domain.
// e.g. http://tat-buns.tumblr.com/archive?ezastumblrscrape?scrapewholesite?find= is too big
// http://tat-buns.tumblr.com/post/104224472370/sparks-part-1-mabeldipper-gravity-falls - http://tat-buns.tumblr.com/tagged/tat%27s-fanfiction
// https://mrdaxxonford.tumblr.com/post/139839107484/little-things
// https://mrdaxxonford.tumblr.com/archive?startpage=1?pagesatonce=10?find=/tagged/my-stuff?ezastumblrscrape?scrapemode=xml?lastpage=9?sitemap=1?xmlcount=9?story?usemobile?tagscrape - managed to change the page colors, wtf. These are /mobile posts! Whatever, makes no difference to a plain text file.
// Massive slowdown from /tagged/pinecest - possibly re-applying CSS from every single /post across 90 /pages?
// Ah fuck, virulentmalapropisms does it too. For "links only!" What the fuck?
// Tried reversion testing - it's not just tumblr fucking with me, it -is- something in my code. Uuugh.
// Forgot to make "story" stuff conditional.
// mrdaxxonford still has different style settings, what the fuck. Oh - because I'm doing ?story and ?tagscrape, so it's not /mobile. Maybe.
// No: it's already grabbing /mobile. WTF.
// It would be more correct to add a 'post-by-post but just for this tag' link after the page count has been found. That'd allow breaking out in 100-page chunks.
// Images 'under the cut' don't work in /tagscrape. Because fuck me.
// http://slightly-gay-pogohammer.tumblr.com/archive?startpage=1?pagesatonce=10?find=/tagged/nsfw-under-cut?ezastumblrscrape?scrapemode=xml?autostart?lastpage=3?sitemap=1?xmlcount=3?tagscrape
// http://manicpeixesdreamgirl.tumblr.com/archive?startpage=1?pagesatonce=10?ezastumblrscrape?scrapemode=xml?lastpage=478?sitemap=5?story?usemobile?usemobile - breaks at 35% even with ?usemobile. I really need fetches to fail safely. Either fuck with .catch() (where does that even go, for a promise.then chain?) or make absolutely goddamn sure every URL begins with window.location.hostname. Oh, and sitemap=11 gets 0% done.
// Sloppy fix: change every fetch to relative URLs.
// Function for relativizing URLs: replace https:// and http:// both. Replacing window.location.domain or whatever, including ".com/". Add a slash to front of remaining URL.
// Really fuckin' wish I'd figured out dashboard blogs now.
// /amp is a thing, like /embed or /mobile. Quickly use it.
// http://diediedie3344-deactivated-204913.tumblr.com/archive still works what the fuck. That would've saved so much extra shit.
// https://unwrapping.tumblr.com/post/126390533972/handy-tumblr-urls - son of a bitch I would've loved to know this a year ago.
// Archive, filter by... Photo — http://<site>.tumblr.com/archive/filter-by/photo
// Posts (from your blog) — https://<www>.tumblr.com/blog/<site> - oh, just 'dashboard mode.' Or 'side mode' or whatever it's called.
// Day (posts by date) — http://<site>.tumblr.com/day/<year>/<month>/<day> example: http://unwrapping.tumblr.com/day/2015/04/01
// Tumblrs below 10 pages don't find pages correctly anymore.
// Also, seeing a lot of "last page is between 9 and 100" or "99 and 1000." That first check is 1, 10, 100, etc. I didn't fucking ask for 99. What the hell?
// Eats CPU cycles on the landing page. Not my fault, Tumblr just bites now. /archive is a fat bitch.
// Alternate standard landing page? /amp, maybe. /mobile. ?test works, obviously. &test gets eaten.
// /mobile works, but might not be 100.0% reliable. I think I remember some sites not having it? Might be thinking of sitemap.xml.
// Whoops, no longer has "Archive" in the title, so textfile filenames would be different.
// standard_landing_page should probably be a variable. I've already messed up replacing it in the several odd spots where it's relevant.
// /archive/1999/1? Any year before Tumblr existed. "No posts yet."
// Consider checking the /amp version of /post pages. Should be standardized. Might still show lingering NSFW content. Again: fuck this website.
// If anyone from Tumblr is reading this, fuck you personally. I don't care what you do there or when you joined. Your company has destroyed art and culture on a grand scale. Millions of hours of artistic effort, the archives of deceased creators, and countless human interactions, gone. No excuse can matter.
// Would <wbr> break text links? I could insert it all over long-ass strings, but I don't know if it screws up how e.g. DownThemAll interprets those strings. HREF is unchanged.
// The world needs an Eza's Twitter Scrape. I just hope to god someone besides me has done it already.
// Why in the name of god is there no way to decode entity references in standard fucking Javascript? Did they think nobody would want to?!
// I should probably detect rate-limiting and keep retrying. Each page on a full-site scrape gets its own div, right?
// Always fetching /post/12345/words-etc/amp is an option. /page/2/amp doesn't work. So reliable tags are not trivial. Dangit.
// Added first-page special case to /mobile handling, and there's some code duplication going on. Double-check what's going on for both scrapers.
// Should offsite images read (Waiting for offsite image)? I'd like to know at a glance whether a Tumblr image failed or some shit from Minus didn't load.
// "Programming is like this amazing puzzle game, where the puzzles are created by your own stupidity."
// Maybe just... like... ?exclude=/tagged/onepiece... to do a full rip of /tagged/onepiece and load that into page_dupe_hash? Or ?include maybe? Logical and & or?
// Or some kind of text-file combiner, not because that's convenient to anyone else, but because I've been saving a shitload of text-files.
// Re-editing this from whatever's on GreasyFork, in Linux. November 2020.
// Changes since last version:
// In ?scrapemode=www, moved each /post's permalink discovery ahead of de-duplication.
// Added navigation controls and timing information for www.tumblr.com/tagged mode.
// Refactored onclick code for "immediate" image size changes.
// Added navigation to specific dates for www.tumblr.com/tagged mode.
// 2020 deep hibernation return:
// Fixed the goddamn titles. Back to simplicity.
// Included max sizes for new-format image URLs, but de-duplication needs more than standardization.
// ------------------------------------ Global variables ------------------------------------ //
var options_map = new Object(); // Associative array for ?key=value pairs in URL.
// Here's the URL options currently used. Scalars are at their default values; boolean flags are all set to false.
options_map[ "startpage" ] = 1; // Page to start at when browsing images.
options_map[ "pagesatonce" ] = 10; // How many Tumblr pages to browse images from at once.
options_map[ "thumbnails" ] = false; // For browsing mode, 240px-wide images vs. full-size.
options_map[ "find" ] = ""; // What goes after the Tumblr URL. E.g. /tagged/art or /chrono.
// Only two hard problems in computer science...
var page_dupe_hash = {}; // Dump each scraped page into this, then test against each page - to remove inter-page duplicates (i.e., same shit on every page)
// Using a global variable for this (defined later) instead of repeatedly defining it or trying to pass within the page using black magic.
var site_and_tags = ""; // To be filled with e.g. http: + // + example.tumblr.com + /tagged/sherlock once options_map.find is parsed from the URL
// Global list of post_id numbers that have been used, corresponding to span IDs based on them
var posts_placed = new Array;
// Global list of posts to be scraped, for use by sitemap.xml-based scrape function.
var post_urls = new Array;
// "Safe" landing page - Tumblr-standard, no theme, ideally few images.
// This can't be /archive or else immediate resizing doesn't work on blogs. But /mobile doesn't work for the new www mode. /archive/amp is consistent but "Not found."
// Blog CSS still won't work right with /archive/amp are you kidding me.
var standard_landing_page = "/mobile/page/amp";
// ------------------------------------ Script start, general setup ------------------------------------ //
// First, determine regime: Bare image? Scrape page? Normal page with a scrape link?
if( window.location.href.indexOf( 'media.tumblr.com/' ) > -1 ) { maximize_image_size(); } // If it's an image, attempt to resize it
// First, determine if we're loading many pages and listing/embedding them, or if we're just adding a convenient button to that functionality.
else if( window.location.href.indexOf( 'ezastumblrscrape' ) > -1 ) { // If we're scraping pages:
// Replace Tumblr-standard Archive page with our own custom nonsense
let original_title = document.title;
document.head.innerHTML = ""; // Delete CSS. We'll start with a blank page.
// document.title = ''; // Clear title. Script adds stuff 'to the end,' as before. Fetch below eventually fixes the front.
// Since we're avoiding /archive (due to content security policy interfering with image-resize gimmicks) we need to get the page title from elsewhere. (Oh, and favicon.)
/*
fetch( "/archive" ).then( r => r.text() ).then( text => { // This is now the wrong title for www stuff. Mmmeh.
title = text.split( '<title>' )[1].split( '</title>' )[0];
// title = text.split( '<title' )[1].split( '>' )[1].split( '</title' )[0]; // Attempted kludge
document.title = window.location.hostname + " - " + htmlDecode( title ) + " " + document.title; // Beat the race condition by losing. (If we're first, we append a blank.)
favicon = text.split( '<link' ).map( x => x.split( '>' )[0] ).filter( x => x.indexOf( 'shortcut icon' ) > 0 )[0];
document.head.innerHTML += '<link ' + favicon + '>';
// The favicon fails sometimes - even on blogs where it otherwise works - even on pages where reloading makes it work. I do not understand how.
} )
*/
// Fuck it:
document.title = window.location.hostname + " - " + original_title.replace( /[\"\\\/]/g, '' ) + " " + document.title;
document.body.outerHTML = "<div id='maindiv'><div id='fetchdiv'></div></div><div id='bottom_controls_div'></div>"; // This is our page. Top stuff, content, bottom stuff.
let css_block = "<style> "; // Let's break up the CSS for readability and easier editing, here.
css_block += "body { background-color: #DDD; } "; // Light grey BG to make image boundaries more obvious
css_block += "h1,h2,h3{ display: inline; } ";
css_block += ".fixed-width img{ width: 240px; } ";
css_block += ".fit-width img{ max-width: 100%; } ";
css_block += ".fixed-height img{ height: 240px; max-width:100%; } "; // Max-width for weirdly common 1px-tall photoset footers. Ultrawide images get their own row.
css_block += ".fit-height img{ max-height: 100%; } ";
css_block += ".smart-fit img{ max-width: 100%; max-height: 100%; } "; // Might ditch the individual fit modes. This isn't Pixiv Fixiv. (No reason not to have them?)
css_block += ".pagelink{ display:none; } ";
css_block += ".showpagelinks .pagelink{ display:inline; } ";
// Text blocks over images hacked together after https://www.designlabthemes.com/text-over-image-with-css/ - CSS is the devil
css_block += ".container { position: relative; padding: 0; margin: 0; } "; // Holds an image and the on-hover permalink
css_block += ".postlink { position: absolute; width: 100%; left: 0; bottom: 0; opacity: 0; background: rgba(255,255,255,0.7); } "; // Permalink across bottom
css_block += ".container:hover .postlink { opacity: 1; } ";
css_block += ".hidden { display: none; } ";
css_block += "a.interactive{ color:green; } "; // For links that don't go anywhere, e.g. 'toggle image size' - signalling.
css_block += "</style>";
document.body.innerHTML += css_block; // Has to go in all at once or the browser "helpfully" closes the style tag upon evaluation
var mydiv = document.getElementById( "maindiv" ); // I apologize for the generic names. This script used to be a lot simpler.
// Identify options in URL (in the form of ?key=value pairs)
var key_value_array = window.location.href.split( '?' ); // Knowing how to do it the hard way is less impressive than knowing how not to do it the hard way.
key_value_array.shift(); // The first element will be the site URL. Durrrr.
for( dollarsign of key_value_array ) { // forEach( key_value_array ), including clumsy homage to $_
var this_pair = dollarsign.split( '=' ); // Split key=value into [key,value] (or sometimes just [key])
if( this_pair.length < 2 ) { this_pair.push( true ); } // If there's no value for this key, make its value boolean True
if( this_pair[1] == "false " ) { this_pair[1] = false; } // If the value is the string "false" then make it False - note fun with 1-ordinal "length" and 0-ordinal array[element].
else if( !isNaN( parseInt( this_pair[1] ) ) ) { this_pair[1] = parseInt( this_pair[1] ); } // If the value string looks like a number, make it a number
options_map[ this_pair[0] ] = this_pair[1]; // options_map.key = value
}
if( options_map.find[ options_map.find.length - 1 ] == "/" ) { options_map.find = options_map.find.substring( 0, options_map.find.length - 1 ); } // Prevents .com//page/2
// if( options_map.find.indexOf( '/chrono' ) > 0 ) { options_map.chrono = true; } else { options_map.chrono = false; } // False case, to avoid unexpected persistence? Hm.
// Convert old URL options to new key-value pairs
if( options_map[ "scrapewholesite" ] ) { options_map.scrapemode = "scrapewholesite"; options_map.scrapewholesite = false; }
if( options_map[ "everypost" ] ) { options_map.scrapemode = "everypost"; options_map.everypost = false; }
if( options_map[ "thumbnails" ] == true ) { options_map.thumbnails = "fixed-width"; } // Replace the original valueless key with the default value
if( options_map[ "notraw" ] ) { options_map.maxres = "1280"; options_map.notraw = false; }
if( options_map[ "usesmall" ] ) { options_map.maxres = "400"; options_map.usesmall = false; }
document.body.className = options_map.thumbnails; // E.g. fixed-width, fixed-height, as matches the CSS. Persistent thumbnail options. Failsafe = original size.
// Oh yeah, we have to do this -after- options_map.find is defined:
site_and_tags = window.location.protocol + "//" + window.location.hostname + options_map.find; // e.g. http: + // + example.tumblr.com + /tagged/sherlock
// Grab an example page so that duplicate-removal hides whatever junk is on every single page
// This remains buggy due to asynchronicity. It's a race condition where the results are, at worst, mildly annoying.
// Previous notes mention jeffmacanolinsfw.tumblr.com for some reason.
// Can just .then this function?
// Can I create cookies? That'd work fine. On load, grab cookies for this site and for this script, use /page/1 crap.
if( options_map.startpage != 1 && options_map.scrapemode != "scrapewholesite" ) // Not on first page, so on-every-page stuff appears somewhere
{ exclude_content_example( site_and_tags + '/page/1' ); } // Since this doesn't happen on page 1, let's use page 1. Low pages are faster somehow. .
// Add tags to title, for archival and identification purposes
document.title += options_map.find.split('/').join(' '); // E.g. /tagged/example/chrono -> "tagged example chrono"
// In Chrome, /archive pages monkey-patch and overwrite Promise.all and Promise.resolve.
// Clunky solution to clunky problem: grab the default property from a fresh iframe.
// Big thanks to inu-no-policeman for the iframe-based solution. Prototypes were not helpful.
var iframe = document.createElement( 'iframe' );
document.body.appendChild( iframe );
window['Promise'] = iframe.contentWindow['Promise'];
document.body.removeChild( iframe );
mydiv.innerHTML = "Not all images are guaranteed to appear.<br>"; // Thanks to JS's wacky accomodating nature, mydiv is global despite appearing in an if-else block.
// Go to image browser or link scraper according to URL options.
switch( options_map.scrapemode ) {
case "scrapewholesite": scrape_whole_tumblr(); break;
case "xml" : scrape_sitemap(); break;
case "everypost": setTimeout( new_embedded_display, 500 ); break; // Slight delay increases odds of exclude_content_example actually fucking working
case "www": scrape_www_tagged(); break; // www.tumblr.com/tagged, and eventually your dashboard, maybe.
default: scrape_tumblr_pages(); // Sensible delays do not work on the original image browser. Shrug.
}
} else { // If it's just a normal Tumblr page, add a link to the appropriate /ezastumblrscrape URL
// Add link(s) to the standard "+Follow / Dashboard" nonsense. Before +Follow, I think - to avoid messing with users' muscle memory.
// This is currently beyond my ability to dick with JS through a script in a plugin. Let's kludge it for immediate usability.
// kludge by Ivan - http://userscripts-mirror.org/scripts/review/65725.html
// Preserve /tagged/tag/chrono, etc. Also preserve http: vs https: via "location.protocol".
var find = window.location.pathname;
if( find.indexOf( "/page/chrono" ) <= 0 ) { // Basically checking for posts /tagged/page, thanks to Detective-Pony. Don't even ask.
if( find.lastIndexOf( "/page/" ) >= 0 ) { find = find.substring( 0, find.lastIndexOf( "/page/" ) ); } // Don't include e.g. /page/2. We'll add that ourselves.
if( find.lastIndexOf( "/post/" ) >= 0 ) { find = find.substring( 0, find.lastIndexOf( "/post" ) ); }
if( find.lastIndexOf( "/archive" ) >= 0 ) { find = find.substring( 0, find.lastIndexOf( "/archive" ) ); }
// On individual posts (and the /archive page), the link should scrape the whole site.
}
var url = window.location.protocol + "//" + window.location.hostname + standard_landing_page + "?ezastumblrscrape?scrapewholesite?find=" + find;
if( window.location.host == "www.tumblr.com" ) { url += "?scrapemode=www?thumbnails=fixed-width"; url = url.replace( "?scrapewholesite", "" ); }
// "Don't clean this up. It's not permanent."
// Fuck it, it works and it's fragile. Just boost its z-index so it stops getting covered.
var scrape_button = document.createElement("a");
scrape_button.setAttribute( "style", "position: absolute; top: 26px; right: 1px; padding: 2px 0 0; width: 50px; height: 18px; display: block; overflow: hidden; -moz-border-radius: 3px; background: #777; color: #fff; font-size: 8pt; text-decoration: none; font-weight: bold; text-align: center; line-height: 12pt; z-index: 1000; " );
scrape_button.setAttribute("href", url);
scrape_button.innerHTML = "Scrape";
var body_ref = document.getElementsByTagName("body")[0];
body_ref.appendChild(scrape_button);
// Pages where the button gets split (i.e. clicking top half only redirects tiny corner iframe) are probably loading this script separately in the iframe.
// Which means you'd need to redirect the window instead of just linking. Bluh.
// Greasemonkey supports user commands through its add-on menu! Thus: no more manually typing /archive?ezastumblrscrape?scrapewholesite on uncooperative blogs.
GM_registerMenuCommand( "Scrape whole Tumblr blog", go_to_scrapewholesite );
// If a page is missing (post deleted, blog deleted, name changed) a reblog can often be found based on the URL
// Ugh, these should only appear on /post/ pages.
// Naming these commands is hard. Both look for reblogs, but one is for if the post you're on is already a reblog.
// Maybe "because this Tumblr changed names / moved / is missing" versus "because this reblog got deleted?" Newbs might not know either way.
// "As though" this blog is missing / this post is missing?
// "Search for this Tumblr under a different name" versus "search for other blogs that've reblogged this?"
GM_registerMenuCommand( "Google for reblogs of this original Tumblr post", google_for_reblogs );
GM_registerMenuCommand( "Google for other instances of this Tumblr reblog", google_for_reblogs_other ); // two hard problems
// if( window.location.href.indexOf( '?browse' ) > -1 ) { browse_this_page(); } // Experimental single-page scrape mode - DEFINITELY not guaranteed to stay
}
function go_to_scrapewholesite() {
// let redirect = window.location.protocol + "//" + window.location.hostname + "/archive?ezastumblrscrape?scrapewholesite?find=" + window.location.pathname;
let redirect = window.location.protocol + "//" + window.location.hostname + standard_landing_page
+ "?ezastumblrscrape?scrapewholesite?find=" + window.location.pathname;
window.location.href = redirect;
}
function google_for_reblogs() {
let blog_name = window.location.href.split('/')[2].split('.')[0]; // e.g. http//example.tumblr.com -> example
let content = window.location.href.split('/').pop(); // e.g. http//example.tumblr.com/post/12345/hey-i-drew-this -> hey-i-drew-this
let redirect = "https://google.com/search?q=tumblr " + blog_name + " " + content;
window.location.href = redirect;
}
function google_for_reblogs_other() {
let content = window.location.href.split('/').pop().split('-'); // e.g. http//example.tumblr.com/post/12345/hey-i-drew-this -> hey,i,drew,this
let blog_name = content.shift(); // e.g. examplename-hey-i-drew-this -> examplename
content = content.join('-');
let redirect = "https://google.com/search?q=tumblr " + blog_name + " " + content;
window.location.href = redirect;
}
// ------------------------------------ Whole-site scraper for use with DownThemAll ------------------------------------ //
// Monolithic scrape-whole-site function, recreating the original intent (before I added pages and made it a glorified multipage image browser)
function scrape_whole_tumblr() {
// console.log( page_dupe_hash );
var highest_known_page = 0;
options_map.startpage = 1; // Reset to default, because other values do goofy things to the image-browsing links below
// Link to image-viewing version, preserving current tags
mydiv.innerHTML += "<br><h1><a id='browse10' href='" + options_url( {scrapemode:'pagebrowser', thumbnails:true} ) +"'>Browse images (10 pages at once)</a> </h1>";
mydiv.innerHTML += "<h3><a id='browse5' href='" + options_url( {scrapemode:'pagebrowser', thumbnails:true, pagesatonce:5} ) + "'>(5 pages at once)</a> </h3>";
mydiv.innerHTML += "<h3><a id='browse1' href='" + options_url( {scrapemode:'pagebrowser', thumbnails:true, pagesatonce:1} ) + "'>(1 page at once)</a></h3><br><br>";
mydiv.innerHTML += "<a id='exp10' href='" + options_url( {scrapemode:'everypost', thumbnails:true, pagesatonce:10} ) + "'>Experimental fetch-every-post image browser (10 pages at once)</a> ";
mydiv.innerHTML += "<a id='exp5' href='" + options_url( {scrapemode:'everypost', thumbnails:true, pagesatonce: 5} ) + "'>(5 pages at once)</a> ";
mydiv.innerHTML += "<a id='exp1' href='" + options_url( {scrapemode:'everypost', thumbnails:true, pagesatonce: 1} ) + "'>(1 page at once)</a><br><br>";
mydiv.innerHTML += "<a id='xml_all' href='" + options_url( {scrapemode:'xml' } ) + "'>Post-by-post images and text</a> <-- New option for saving stories<br><br>";
// Find out how many pages we need to scrape.
if( isNaN( options_map.lastpage ) ) {
// Find upper bound in a small number of fetches. Ideally we'd skip this - some themes list e.g. "Page 1 of 24." I think that requires back-end cooperation.
mydiv.innerHTML += "Finding out how many pages are in <b>" + site_and_tags.substring( site_and_tags.indexOf( '/' ) + 2 ) + "</b>:<br><br>";
// Returns page number if there's no Next link, or negative page number if there is a Next link.
// Only for use on /mobile pages; relies on Tumblr's shitty standard theme
function test_next_page( body ) {
var link_index = body.indexOf( 'rel="canonical"' ); // <link rel="canonical" href="http://shadygalaxies.tumblr.com/page/100" />
var page_index = body.indexOf( '/page/', link_index );
var terminator_index = body.indexOf( '"', page_index );
var this_page = parseInt( body.substring( page_index+6, terminator_index ) );
if( body.indexOf( '>next<' ) > 0 ) { return -this_page; } else { return this_page }
}
// Generates an array of length "steps" between given boundaries - or near enough, for sanity's sake
function array_between_bounds( lower_bound, upper_bound, steps ) {
if( lower_bound > upper_bound ) { // Swap if out-of-order.
var temp = lower_bound; lower_bound = upper_bound, upper_bound = temp;
}
var bound_range = upper_bound - lower_bound;
if( steps > bound_range ) { steps = bound_range; } // Steps <= bound_range, but steps > 1 to avoid division by zero:
var pages_per_test = parseInt( bound_range / steps ); // Steps-1 here, so first element is lower_bound & last is upper_bound. Off-by-one errors, whee...
var range = Array( steps )
.fill( lower_bound )
.map( (value,index) => value += index * pages_per_test );
range.push( upper_bound );
return range;
}
// DEBUG
// site_and_tags = 'https://www.tumblr.com/safe-mode?url=http://shittyhorsey.tumblr.com';
// Given a (presumably sorted) list of page numbers, find the last that exists and the first that doesn't exist.
function find_reasonable_bound( test_array ) {
return Promise.all( test_array.map( pagenum => fetch( site_and_tags + '/page/' + pagenum + '/mobile', { credentials: 'include' } ) ) )
.then( responses => Promise.all( responses.map( response => response.text() ) ) )
.then( pages => pages.map( page => test_next_page( page ) ) )
.then( numbers => {
var lower_index = -1;
numbers.forEach( (value,index) => { if( value < 0 ) { lower_index++; } } ); // Count the negative numbers (i.e., count the pages with known content)
if( lower_index < 0 ) { lower_index = 0; }
var bounds = [ Math.abs(numbers[lower_index]), numbers[lower_index+1] ]
mydiv.innerHTML += "Last page is between " + bounds[0] + " and " + bounds[1] + ".<br>";
return bounds;
} )
}
// Repeatedly narrow down how many pages we're talking about; find a reasonable "last" page
find_reasonable_bound( [2, 10, 100, 1000, 10000, 100000] ) // Are we talking a couple pages, or a shitload of pages?
.then( pair => find_reasonable_bound( array_between_bounds( pair[0], pair[1], 10 ) ) ) // Narrow it down. Fewer rounds of more fetches works best.
.then( pair => find_reasonable_bound( array_between_bounds( pair[0], pair[1], 10 ) ) ) // Time is round count, fetches add up, selectivity is fetches x fetches.
// Quit fine-tuning numbers and just conditional in some more testing for wide ranges.
.then( pair => { if( pair[1] - pair[0] > 50 ) { return find_reasonable_bound( array_between_bounds( pair[0], pair[1], 10 ) ) } else { return pair; } } )
.then( pair => { if( pair[1] - pair[0] > 50 ) { return find_reasonable_bound( array_between_bounds( pair[0], pair[1], 10 ) ) } else { return pair; } } )
.then( pair => {
options_map.lastpage = pair[1];
document.getElementById( 'browse10' ).href += "?lastpage=" + options_map.lastpage; // Add last-page indicator to Browse Images link
document.getElementById( 'browse5' ).href += "?lastpage=" + options_map.lastpage; // ... and the 5-pages-at-once link.
document.getElementById( 'browse1' ).href += "?lastpage=" + options_map.lastpage; // ... and the 1-page-at-onces link.
document.getElementById( 'exp10' ).href += "?lastpage=" + options_map.lastpage; // ... and this fetch-every-post link.
document.getElementById( 'exp5' ).href += "?lastpage=" + options_map.lastpage; // ... and this fetch-every-post link.
document.getElementById( 'exp1' ).href += "?lastpage=" + options_map.lastpage; // ... and this fetch-every-post link.
document.getElementById( 'xml_all' ).href += "?lastpage=" + options_map.lastpage; // ... and this XML post-by-post link, why not.
start_scraping_button();
} );
}
else { // If we're given the highest page by the URL, just use that
start_scraping_button();
}
// Add "Scrape" button to the page. This will grab images and links from many pages and list them page-by-page.
function start_scraping_button() {
if( options_map.grabrange ) { // If we're only grabbing a 1000-page block from a huge-ass tumblr:
mydiv.innerHTML += "<br>This will grab 1000 pages starting at <b>" + options_map.grabrange + "</b>.<br><br>";
} else { // If we really are describing the last page:
mydiv.innerHTML += "<br>Last page is <b>" + options_map.lastpage + "</b> or lower. ";
mydiv.innerHTML += "<a href=" + options_url( {lastpage: false} ) + ">Find page count again?</a> <br><br>";
}
if( options_map.lastpage > 1500 && !options_map.grabrange ) { // If we need to link to 1000-page blocks, and aren't currently inside one:
for( let x = 1; x < options_map.lastpage; x += 1000 ) { // For every 1000 pages...
let decade_url = window.location.href + "?grabrange=" + x + "?lastpage=" + options_map.lastpage;
mydiv.innerHTML += "<a href='" + decade_url + "'>Pages " + x + "-" + (x+999) + "</a><br>"; // ... link a range of 1000 pages.
}
}
// Add button to scrape every page, one after another.
// Buttons within GreaseMonkey are a huge pain in the ass. I stole this from stackoverflow.com/questions/6480082/ - thanks, Brock Adams.
var button = document.createElement ('div');
button.innerHTML = '<button id="myButton" type="button">Find image links from all pages</button>';
button.setAttribute ( 'id', 'scrape_button' ); // I'm really not sure why this id and the above HTML id aren't the same property.
document.body.appendChild ( button ); // Add button (at the end is fine)
document.getElementById ("myButton").addEventListener ( "click", scrape_all_pages, false ); // Activate button - when clicked, it triggers scrape_all_pages()
if( options_map.autostart ) { document.getElementById ("myButton").click(); } // Getting tired of clicking on every reload - debug-ish
if( options_map.lastpage <= 26 ) { document.getElementById ("myButton").click(); } // Automatic fetch (original behavior!) for a single round
}
}
function scrape_all_pages() { // Example code implies that this function /can/ take a parameter via the event listener, but I'm not sure how.
var button = document.getElementById( "scrape_button" ); // First, remove the button. There's no reason it should be clickable twice.
button.parentNode.removeChild( button ); // The DOM can only remove elements from a higher level. "Elements can't commit suicide, but infanticide is permitted."
mydiv.innerHTML += "Scraping page: <div id='pagecounter'></div><div id='afterpagecounter'><br></div><br>"; // This makes it easier to view progress,
// Create divs for all pages' content, allowing asynchronous AJAX fetches
var x = 1;
var div_end_page = options_map.lastpage;
if( !isNaN( options_map.grabrange ) ) { // If grabbing 1000 pages from the middle of 10,000, don't create 0..10,000 divs
x = options_map.grabrange;
div_end_page = x + 1000; // Should be +999, but whatever, no harm in tiny overshoot
}
for( ; x <= div_end_page; x++ ) {
var siteurl = site_and_tags + "/page/" + x;
if( options_map.usemobile ) { siteurl += "/mobile"; } // If ?usemobile is flagged, scrape the mobile version.
if( x == 1 && options_map.usemobile ) { siteurl = site_and_tags + "/mobile"; } // Hacky fix for redirect from example.tumblr.com/page/1/anything -> example.tumblr.com
var new_div = document.createElement( 'div' );
new_div.id = '' + x;
document.body.appendChild( new_div );
}
// Fetch all pages with content on them
var page_counter_div = document.getElementById( 'pagecounter' ); // Probably minor, but over thousands of laggy page updates, I'll take any optimization.
pagecounter.innerHTML = "" + 1;
var begin_page = 1;
var end_page = options_map.lastpage;
if( !isNaN( options_map.grabrange ) ) { // If a range is defined, grab only 1000 pages starting there
begin_page = options_map.grabrange;
end_page = options_map.grabrange + 999; // NOT plus 1000. Stop making that mistake. First page + 999 = 1000 total.
if( end_page > options_map.lastpage ) { end_page = options_map.lastpage; } // Kludge
document.title += " " + (parseInt( begin_page / 1000 ) + 1); // Change page title to indicate which block of pages we're saving
}
// Generate array of URL/pagenum pair-arrays
url_index_array = new Array;
for( var x = begin_page; x <= end_page; x++ ) {
var siteurl = site_and_tags + "/page/" + x;
if( options_map.usemobile ) { siteurl += "/mobile"; } // If ?usemobile is flagged, scrape the mobile version. No theme shenanigans... but also no photosets. Sigh.
if( x == 1 && options_map.usemobile ) { siteurl = site_and_tags + "/mobile"; } // Hacky fix for redirect from example.tumblr.com/page/1/anything -> example.tumblr.com
url_index_array.push( [siteurl, x] );
}
// Fetch, scrape, and display all URLs. Uses promises to work in parallel and promise.all to limit speed and memory (mostly for reliability's sake).
// Consider privileging first page with single-element fetch, to increase apparent responsiveness. Doherty threshold for frustration is 400ms.
var simultaneous_fetches = 25;
var chain = Promise.resolve(0); // Empty promise so we can use "then"
var order_array = [1]; // We want to show the first page immediately, and this is a callback rat's-nest, so let's make an array of how many pages to take each round
for( var x = 1; x < url_index_array.length; x += simultaneous_fetches ) { // E.g. [1, simultaneous_fetchs, s_f, s_f, s_f, whatever's left]
if( url_index_array.length - x > simultaneous_fetches ) { order_array.push( simultaneous_fetches ); } else { order_array.push( url_index_array.length - x ); }
}
order_array.forEach( (how_many) => {
chain = chain.then( s => {
var subarray = url_index_array.splice( 0, how_many ); // Shift a reasonable number of elements into separate array, for partial array.map
return Promise.all( subarray.map( page =>
Promise.all( [ fetch( page[0], { credentials: 'include' } ).then( s => s.text() ), page[1], page[0] ] ) // Return [ body of page, page number, page URL ]
) )
} )
.then( responses => responses.map( s => { // Scrape URLs for links and images, display on page
var pagenum = s[1];
var page_url = s[2];
var url_array = soft_scrape_page_promise( s[0] ) // Surprise, this is a promise now
.then( urls => {
// Sort #link URLs to appear first, because we don't do that in soft-scrape anymore
urls.sort( (a,b) => -a.indexOf( "#link" ) ); // Strings containing "#link" go before others - return +1 if not found in 'a.' Should be stable.
// Print URLs so DownThemAll (or similar) can grab them
var bulk_string = "<br><a href='" + page_url + "'>" + page_url + "</a><br>"; // A digest, so we can update innerHTML just once per div
// DEBUG-ish - on theory that 1000-page-tall scraping/rendering fucks my VRAM
if( options_map.smalltext ) { bulk_string = "<p style='font-size:1px'>" + bulk_string; } // If ?smalltext flag is set, render text unusably small, for esoteric reasons
urls.forEach( (value,index,array) => {
if( options_map.plaintext ) {
bulk_string += value + '<br>';
} else {
bulk_string += '<a href ="' + value + '">' + value + '</a><br>';
}
} )
document.getElementById( '' + pagenum ).innerHTML = bulk_string;
if( parseInt( pagecounter.innerHTML ) < pagenum ) { pagecounter.innerHTML = "" + pagenum; } // Increment pagecounter (where sensible)
} );
} )
)
} )
chain = chain.then( s => {
document.getElementById( 'afterpagecounter' ).innerHTML = "Done. Use DownThemAll (or a similar plugin) to grab all these links.";
// Divulge contents of page_dupe_hash to check for common tags
// Ugh, I'm going to have to turn this from an associative array into an array-of-arrays if I want to sort it.
let tag_overview = "<br>" + "Tag overview: " + "<br>";
let score_tag_list = new Array; // This will hold an array of arrays so we can sort this associative array by its values. Wheee.
for( let url in page_dupe_hash ) {
if( url.indexOf( '/tagged/' ) > 0 // If it's a tag URL...
&& page_dupe_hash[ url ] > 1 // and non-unique...
&& url.indexOf( '/page/' ) < 0 // and not a page link...
&& url.indexOf( '?og' ) < 0 // and not an opengraph link...
&& url.indexOf( '?ezas' ) < 0 // and not this script, wtf...
) { // So if it's a TAG, in other words...
score_tag_list.push( [ page_dupe_hash[ url ], url ] ); // ... store [ number of times seen, tag URL ] for sorting.
}
}
score_tag_list.sort( (a,b) => a[0] > b[0] ); // Ascending order, now - most common tags at the very bottom, for easier access
score_tag_list.map( pair => {
pair[1] = pair[1].replace( '/chrono', '' ); // Remove /chrono from sites that append it automatically, since it breaks the 'N pages' autostart links. (/chrono/ might fail.)
var this_tag = pair[1].split('/').pop(); // e.g. example.tumblr.com/tagged/my-art -> my-art
//if( this_tag == '' ) { let this_tag = pair[1].split('/').pop(); } // Trailing slash screws this up, so get the second-to-last thing instead
var scrape_link = options_url( {find: '/tagged/'+this_tag, lastpage: false, grabrange: false, autostart: true} ); // Direct link to ?scrapewholesite for e.g. /tagged/my-art
tag_overview += "<br><a href='" + scrape_link + "'>" + pair[0] + " posts</a>:\t" + "<a href='" + pair[1] + "'>" + pair[1] + "</a>";
} )
document.body.innerHTML += tag_overview;
} )
}
// ------------------------------------ Multi-page scraper with embedded images ------------------------------------ //
function scrape_tumblr_pages() {
// Grab an empty page so that duplicate-removal hides whatever junk is on every single page
// This is DEBUG-ish. It might be slow, barring caching. It might not work due to asynchrony. It could block actual content thanks to 'my best posts' sidebars.
if( isNaN( parseInt( options_map.startpage ) ) || options_map.startpage <= 1 ) { options_map.startpage = 1; } // Sanity check
mydiv.innerHTML += "<br>" + html_previous_next_navigation() + "<br>";
document.getElementById("bottom_controls_div").innerHTML += "<br>" + html_page_count_navigation() + "<br>" + html_previous_next_navigation();
mydiv.innerHTML += "<br>" + html_ezastumblrscrape_options() + "<br>";
mydiv.innerHTML += "<br>" +image_size_options();
mydiv.innerHTML += "<br><br>" + image_resolution_options();
// Fill an array with the page URLs to be scraped (and create per-page divs while we're at it)
var pages = new Array( parseInt( options_map.pagesatonce ) )
.fill( parseInt( options_map.startpage ) )
.map( (value,index) => value+index );
pages.forEach( pagenum => {
mydiv.innerHTML += "<hr><div id='" + pagenum + "'><b>Page " + pagenum + " </b></div>";
} )
pages.map( pagenum => {
var siteurl = site_and_tags + "/page/" + pagenum; // example.tumblr.com/page/startpage, startpage+1, startpage+2, etc.
if( options_map.usemobile ) { siteurl += "/mobile"; } // If ?usemobile is flagged, scrape mobile version. No theme shenanigans... but also no photosets. Sigh.
if( pagenum == 1 && options_map.usemobile ) { siteurl = site_and_tags + "/mobile"; } // Hacky fix for redirect from example.tumblr.com/page/1/anything -> example.tumblr.com
fetch( siteurl, { credentials: 'include' } ).then( response => response.text() ).then( text => {
document.getElementById( pagenum ).innerHTML += "<b>fetched</b><br>" // Immediately indicate the fetch happened.
+ "<a href='" + siteurl + "'>" + siteurl + "</a><br>"; // Link to page. Useful for viewing things in-situ... and debugging.
// For some asinine reason, 'return url_array' causes 'Permission denied to access property "then".' So fake it with ugly nesting.
soft_scrape_page_promise( text )
.then( url_array => {
var div_digest = ""; // Instead of updating each div's HTML for every image, we'll lump it into one string and update the page once per div.
var video_array = new Array;
var outlink_array = new Array;
var inlink_array = new Array;
url_array.forEach( (value,index,array) => { // Shift videos and links to separate arrays, blank out those URLs in url_array
if( value.indexOf( '#video' ) > 0 ) { video_array.push( value ); array[index] = '' }
if( value.indexOf( '#offsite' ) > 0 ) { outlink_array.push( value ); array[index] = '' }
if( value.indexOf( '#local' ) > 0 ) { inlink_array.push( value ); array[index] = '' }
} );
url_array = url_array.filter( url => url === "" ? false : true ); // Remove empty elements from url_array
// Display video links, if there are any
video_array.forEach( value => {div_digest += "Video: <a href='" + value + "'>" + value + "</a><br> "; } )
// Display page links if the ?showlinks flag is enabled
if( options_map.showlinks ) {
div_digest += "Outgoing links: ";
outlink_array.forEach( (value,index) => { div_digest += "<a href='" + value.replace('#offsite#link', '') + "'>O" + (index+1) + "</a> " } );
div_digest += "<br>" + "Same-Tumblr links: ";
inlink_array.forEach( (value,index) => { div_digest += "<a href='" + value.replace('#local#link', '') + "'>T" + (index+1) + "</a> " } );
div_digest += "<br>";
}
// Embed high-res images to be seen, clicked, and saved
url_array.forEach( image_url => {
// Embed images (linked to themselves) and link to photosets
if( image_url.indexOf( "#photoset#" ) > 0 ) { // Before the first image in a photoset, print the photoset link.
var photoset_url = image_url.split( "#" ).pop();
// URL is like tumblr.com/image#photoset#http://tumblr.com/photoset_iframe - separate past last hash... t.
div_digest += " <a href='" + photoset_url + "'>Set:</a>";
}
div_digest += "<a id='" + encodeURI( image_url ) + "' target='_blank' href='" + image_url + "'>" + "<img alt='(Waiting for image)' onerror='" + error_function( image_url ) + "' src='" + image_url + "'></a> ";
// div_digest += "<a id='" + encodeURI( image_url ) + "' target='_blank' href='" + image_url + "'>" + "<img alt='(Waiting for image)' src='" + image_url + "'></a> ";
} )
div_digest += "<br><a href='" + siteurl + "'>(End of " + siteurl + ")</a>"; // Another link to the page, because I'm tired of scrolling back up.
document.getElementById( pagenum ).innerHTML += div_digest;
} ) // End of 'then( url_array => { } )'
} ) // End of 'then( text => { } )'
} ) // End of 'pages.map( pagenum => { } )'
}
// ------------------------------------ Whole-site scraper based on post-by-post method ------------------------------------ //
// The use of sitemap.xml could be transformative to this script, or even split off into another script entirely. It finally allows a standard theme!
// Just do it. Entries look like:
// <url><loc>http://leopha.tumblr.com/post/57593339717/hello-a-friend-found-my-blog-somehow-so-url</loc><lastmod>2013-08-07T06:42:58Z</lastmod></url></urlset>
// So split on <url><loc> and terminate at </loc>. Or see if JS has native XML parsing, for honest object-orientation instead of banging at text files.
// Grab normally for now, I guess, just to get it implemented. Make ?usemobile work. Worry about /embed stuff later.
// Check if dashboard-only blogs have xml files exposed.
// While we're at it, check that the very first <loc> for sitemap2 doesn't point to a /page/n URL. Not useful, just interesting.
// Simplicate for now. Have a link to grab an individual XML file, get everything inside it, then add images (and tags?) from each page.
// For an automatic 'get the whole damn site' mode, maybe just call each XML file sequentially. Not a for() loop - have each page maybe call the next page when finished.
// Oh right, these XML files are like 500 posts each. Not much different from grabbing 50 pages at a time - and we currently grab 25 at once - but significant.
// If I have to pass the /post URL list to a serial function in order to rate-limit this, I might as well grab all XML files and fill the list -once.-
// Count down? So e.g. scrape(x) calls scrape(x-1). It's up to the root function to call the high value initially.
// Single-sitemap pages (e.g. scraping and displaying sitemap2) will need prev/next controls (e.g. to sitemap1 & sitemap3).
// These will be chronological. That's fine, just mention it somewhere.
// I'm almost loathe to publicize this. It's not faster or more reliable than the monolithic scraper. It's not a better image-browsing method, in my opinion.
// It's only really useful for people who think a blank theme is the same as erasing their blog... and the less those jerks know, the better. Don't delete art.
// I guess it'll be better when I list tags with each post, but at present they're strongly filtered out.
// Wait, wtf? I ripped sometipsygnostalgic as a test cast (98 sitemaps, ~50K posts) and the last page scraped is from years ago.
// None of the sitemaps past sitemap51 exist. So the first 51 are present - the rest 404. (To her theme, not as a generic tumblr error with animated background.)
// It's not a URL-generation error; the same thing happens copy-pasting the XML URLs straight from sitemap.xml.
// Just... fuck. Throw a warning for now, see if there's anything to be done later.
// If /mobile versions of /post URLs unexpectedly had proper [video] references... do they contain previous/next post info? Doesn't look like it. Nuts.
// In addition to tags, this should maybe list sources and 'via' links. It's partially archival, after all.
// Sort of have to conditionally unfilter /tagged links from both duplicate-remover and whichever function removes tumblr guff. 'if !xml remove tagged.' bluh.
// Better idea: send the list to a 'return only tags' filter first, non-destructively. Then filter them out.
// I guess this is what I'd modify to get multi-tag searches - like homestuck+hs+hamsteak. Scrape by /tagged and /page, populate post_urls, de-dupe and sort.
// Of course then I'd want it to 'just work' with the existing post-by-post image browser, which... probably won't happen. I'm loathe to recreate it exactly or inexactly.
// This needs some method to find /post numbers from /page and /tagged pages.
// Maybe trigger a later post-by-post function from a modified scrapewholesite approach? Code quality is a non-issue right now.
// Basically have to fill some array post_urls with /post URLs (strings) and then call scrape_post_urls(). Oh, that array is global. This was already a janky first pass.
// Instead of XML files, use /page or /page + /mobile... pages.
// Like 'if /tagged/something then grab all links and filter for indexof /post.'
// This is the landing function for this mode - it grabs sitemap.xml, parses options, sets any necessary variables, and invokes scrape_sitemap_x for sitemap1 and so on.
function scrape_sitemap() {
document.title += ' sitemap';
// Flow control.
// Two hard problems. Structure now, names later.
if( options_map.sitemap ) {
mydiv.innerHTML += "<br>Completion: <div id='pagecounter'>0%</div><div id='afterpagecounter'></div><br>"; // This makes it easier to view progress,
let final_sitemap = 0; // Default: no recursion. 'Else considered harmful.'
if( options_map.xmlcount ) { final_sitemap = options_map.xmlcount; } // If we're doing the whole site, indicate the highest sitemap and recurse until you get there.
scrape_sitemap_x( options_map.sitemap, final_sitemap );
} else {
// Could create a div for this enumeration of sitemapX files, and always have it at the top. It'd get silly for tumblrs with like a hundred of them.
// Maybe at the bottom?
fetch( window.location.protocol + "//" + window.location.hostname + '/sitemap.xml', { credentials: 'include' } ) // Grab text-like file
.then( r => r.text() )
.then( t => { // Process the text of sitemap.xml
let sitemap_list = t.split( '<loc>' );
sitemap_list.shift(); // Get rid of data before first location
sitemap_list = sitemap_list.map( e => e.split('</loc>')[0] ); // Terminate each entry
sitemap_list = sitemap_list.filter( e => { return e.indexOf( '/sitemap' ) > 0 && e.indexOf( '.xml' ) > 0; } ); // Remove everything but 'sitemapX.xml' links
mydiv.innerHTML += '<br>' + window.location.hostname + ' has ' + sitemap_list.length + ' sitemap XML file';
if( sitemap_list.length > 1 ) { mydiv.innerHTML += 's'; } // Pluralization! Not sexy, just functional.
// mydiv.innerHTML += ' (Up to ' + sitemap_list.length * 500 + ' posts.) <br>'
mydiv.innerHTML += '. (' + Math.ceil(sitemap_list.length / 2) + ',000 posts or fewer.)<br><br>' // Kludge math, but I pefer this presentation.
if( sitemap_list.length > 50 ) { mydiv.innerHTML += 'Sitemaps past 50 may not work.<br><br>'; } // Pluralization! Not sexy, just functional.
// List everything: <Links only>
// List sitemap1: <Links only> or <Links & thumbnails>
// List sitemap2: <Links only> or <Links & thumbnails> etc.
mydiv.innerHTML += 'List everything: ' + '<a href=' + options_url( { sitemap:1, xmlcount:sitemap_list.length } ) + '>Links only</a> - <a href=' + options_url( { sitemap:1, xmlcount:sitemap_list.length, story:true, usemobile:true } ) + '>Links and text (stories)</a><br><br>';
if( options_map.find && options_map.lastpage ) {
mydiv.innerHTML += 'List just this tag (' + options_map.find + '): ' + '<a href=' + options_url( { sitemap:1, xmlcount:options_map.lastpage, tagscrape:true } ) + '>Links only</a> - <a href=' + options_url( { sitemap:1, xmlcount:options_map.lastpage, story:true, usemobile:true, tagscrape:true } ) + '>Links and text (stories)</a><br><br>';
}
// mydiv.innerHTML += "<br><h1><a id='browse10' href='" + options_url( {scrapemode:'pagebrowser', thumbnails:true} ) +"'>Browse images (10 pages at once)</a> </h1>";
for( n = 1; n <= sitemap_list.length; n++ ) {
let text_link = options_url( { sitemap:n } );
let images_link = options_url( { sitemap:n, thumbnails:'xml' } );
let story_link = options_url( { sitemap:n, story:true, usemobile:true } );
mydiv.innerHTML += 'List sitemap' + n + ': ' + '<a href=' + text_link + '>Links only</a> <a href=' + images_link + '>Links & thumbnails</a> <a href=' + story_link + '>Links & text</a><br>'
}
} )
}
}
// Text-based scrape mode for sitemap1.xml, sitemap2.xml, etc.
function scrape_sitemap_x( sitemap_x, final_sitemap ) {
document.title = window.location.hostname + ' - sitemap ' + sitemap_x; // We lose the original tumblr's title, but whatever.
if( sitemap_x == final_sitemap ) { document.title = window.location.hostname + ' - sitemap complete'; }
if( options_map.story ) { document.title += ' with stories'; }
// Last-minute kludge: if options_map.tagscrape, use /page instead of any xml, get /post URLs matching this domain.
// Only whole-hog for now - sitemap_x might allow like ten or a hundred pages at once, later.
var base_site = window.location.protocol + "//" + window.location.hostname; // e.g. http://example.tumblr.com
var sitemap_url = base_site + '/sitemap' + sitemap_x + '.xml';
if( options_map.tagscrape ) {
sitemap_url = base_site + options_map.find + '/page/' + sitemap_x;
document.title += ' ' + options_map.find.replace( /\//g, ' '); // E.g. 'tagged my-stuff'.
}
mydiv.innerHTML += "Finding posts from " + sitemap_url + ".<br>";
fetch( sitemap_url, { credentials: 'include' } ) // Grab text-like file
.then( r => r.text() ) // Get txt from HTTP response
.then( t => { // Process text to extract links
let location_list = t.split( '<loc>' ); // Break at each location declaration
location_list.shift(); location_list.shift(); // Ditch first two elements. First is preamble, second is base URL (e.g. example.tumblr.com).
location_list = location_list.map( e => e.split( '</loc>' )[0] ); // Terminate each entry at location close-tag
if( options_map.tagscrape ) {
// Fill location_list with URLs on this page, matching this domain, containing e.g. /post/12345.
// Trailing slash not guaranteed. Quote terminator unknown. Fuck it, just assume it's all doublequotes.
location_list = t.split( 'href="' );
location_list.shift();
location_list = location_list.map( e => e.split( '"' )[0] ); // Terminate each entry at location close-tag
location_list = location_list.filter( u => u.indexOf( window.location.hostname + '/post' ) > -1 );
location_list = location_list.filter( u => u.indexOf( '/embed' ) < 0 );
location_list = location_list.filter( u => u.indexOf( '#notes' ) < 0 ); // I fucking hate this website.
// https://www.facebook.com/sharer/sharer.php?u=https://shamserg.tumblr.com/post/55985088311/knight-of-vengence-by-shamserg
// I fucking hate the web.
location_list = location_list.map( e => window.location.protocol + '//' + window.location.hostname + e.split( window.location.hostname )[1] );
// location_list = location_list.filter( u => u.indexOf( '?ezastumblrscrape' ) <= 0 ); // Exclude links to this script. (Nope, apparently that's introduced elsewhere. WTF.)
// Christ, https versus http bullshit AGAIN. Just get it done.
// location_list = location_list.map( e => window.location.protocol + '//' + e.split( '//' )[1] ); // http -> https, or https -> http, as needed.
// I don't think I ever remove duplicates. On pages with lots of silly "sharing" bullshit, this might grab pages multiple times. Not sure I care. Extreme go horse.
}
// location_list should now contain a bunch of /post/12345 URLs.
if( options_map.usemobile ) { location_list = location_list.map( (url) => { return url + '/mobile' } ); }
console.log( location_list );
// return location_list; // Promises are still weird. (JS's null-by-default is also weird.)
post_urls = post_urls.concat( location_list ); // Append these URLs to the global array of /post URLs. (It's not like Array.push() because fuck you.)
// console.log( post_urls );
// It seems like bad practice to recurse from here, but for my weird needs it seems to make sense.
if( sitemap_x < final_sitemap ) {
scrape_sitemap_x( sitemap_x + 1, final_sitemap ); // If there's more to be done, recurse.
} else {
scrape_post_urls(); // If not, fetch & display all these /post URLs.
}
} )
}
// Fetch all the URLs in the global array of post_urls, then display contents as relevant
function scrape_post_urls() {
// Had to re-check how I did rate-limited fetches for the monolithic main scraper. It involved safe levels of 'wtf.'
// Basically I put n URLs in an array, or an array of fetch promises I guess, then use promise.all to wait until they're all fetched. Repeat as necessary.
// Can I use web workers? E.g. instantiate 25 parallel scripts that each shift from post_urls and spit back HTML. Ech. External JS is encouraged and DOM access is limited.
// Screw it, batch parallelism will suffice.
// Can I define a variable as a function? I mean, can I go var foobar = function_name, so I can later invoke foobar() and get function_name()?
// If so: I can pick a callback before all of this, fill a global array with e.g. [url, contents] stuff, and jump to a text- or image-based display when that's finished.
// Otherwise I think I'm gonna be copying myself a lot in order to display text xor images.
// Create divs for each post? This is going to be a lot of divs.
// 1000 is fine for the traditional monolithic scrape mode, but that covers ten or twenty times as many posts as a single sitemapX.xml.
// And it probably doesn't belong right here.
// Fuck it, nobody's using this unless they read the source code. Gotta build something twice to build it right.
// Create divs for each post, so they can be fetched and inserted asynchronously
post_urls.forEach( (url) => {
let new_div = document.createElement( 'div' );
new_div.id = '' + url;
document.body.appendChild( new_div );
} )
// Something in here is throwing "Promise rejection value is a non-unwrappable cross-compartment wrapper." I don't even know what the fuck.
// Apparently Mozilla doesn't know what the fuck either, because all discussion of this seems to be 'well it ought to throw a better error.'
// Okay somehow this happens even with everything past 'console.log( order_array );' commented out, so it's not actually a wrong thing involving the Promise nest. Huh.
// It's the for loop.
// Is it because we try using location_list.length? Even though it's a global array and not a Promise? And console.log() has no problem... oh, console.log quietly chokes.
// I'm trying to grab a sitemap2 that doesn't exist because I was being clever in opposite directions.
// Nope, still shits the bed, same incomprehensible error. Console.log shows location_list has... oh goddammit. location_list was the local variable; post_urls is the global.
// Also array.concat doesn't work for some reason. post_urls still has zero elements.
// Oh for fuck's sake it has no side effects! You have to do thing=thing.concat()! God damn the inconsistent behavior versus array.push!
// We now get as far as chain=chain.then and the first console.log('foobar'), but it happens exactly once.
// Splicing appears to happen correctly.
// We enter responses.map, and the [body of post, post url] array is passed correctly and contains right-looking data.
// Ah. links_from_text is not a function. It's links_from_page. Firefox, that's a dead simple error - say something.
// Thumbnails? True thumbnails, and only for tumblr images with _100 available. Still inadvisable for the whole damn tumblr at once.
// So limit that to single-sitemap modes (like linking out to 1000-pages-at-once monolithic scrapes) and call it good. 500 thumbnails is what /archive does anyway.
// Terrible fix for unique tags: can I edit CSS from within the page? Can I count inside that?
// I'd be trying to do something like <tag id='my_art>my_art</tag> and increment some tags['my_art'] value - then only show <tag> with a corresponding value over one.
// Utter hack: I need to get the text of each /post to go above or below it. The key is options_map( 'story' ) == true.
// Still pass it as a "link" string. Prepend with \n or something, a non-URL-like control character, which we're okay to print accidentally. Don't link that "link."
// /mobile might work? Photosets show up as '[video]', but some text is in the HTML. Dunno if it's everything.
// Oh god they're suddenly fucking with mobile.
// In /mobile, get everything from '</h1>' to '<div id="navigation">'.
var simultaneous_fetches = 25;
var order_array = new Array; // No reason to show immediate results, so fill with n, n, n until the sum matches the number of URLs.
for( x = post_urls.length; x > simultaneous_fetches; x -= simultaneous_fetches ) {
order_array.push( simultaneous_fetches );
}
if( x != 0 ) { order_array.push( x ); } // Any remainder from counting-down for() loop becomes the last element
// console.log( order_array );
// order_array = [5,5]; // Debug - limits fetches to just a few posts
var chain = Promise.resolve(0); // Empty promise so we can use "then"
order_array.forEach( (how_many, which_entry) => {
chain = chain.then( s => {
// console.log( 'foobar' );
var subarray = post_urls.splice( 0, how_many ); // Shift some number of elements into a separate array, for partial array.map
// console.log( 'foobar2', subarray );
return Promise.all( subarray.map( post_url =>
Promise.all( [ fetch( post_url, { credentials: 'include' } ).then( s => s.text() ), post_url ] ) // Return [body of post, post URL]
) )
} )
. then( responses => responses.map( s => {
// console.log( 'foobar3', s );
var post_url = s[1];
// var url_array = soft_scrape_page_promise( s[0] ) // Surprise, this is a promise now
// Wait, do I need it to be a promise? Because otherwise I might prefer to use links_from_text().
var sublinks = links_from_page( s[0] );
let tag_links = sublinks.filter( return_tags ); // Copy tags into some new array. No changes to other filters required! (Test on eixn, ~2000 pages.)
tag_links = tag_links.filter( u => u.indexOf( '?ezastumblrscrape' ) < 0 );
// Do we want to filter down to images for this scrape function? It's not necessarily displaying them.
// Arg, you almost have to, to filter out the links to tumblrs that reblogged each post.
sublinks = sublinks.filter( s => { return s.indexOf( '.jpg' ) > 0 || s.indexOf( '.jpeg' ) > 0 || s.indexOf( '.png' ) > 0 || s.indexOf( '.gif' ) > 0; } );
sublinks = sublinks.filter( tumblr_blacklist_filter ); // Remove avatars and crap
sublinks = sublinks.map( image_standardizer ); // Clean up semi-dupes (e.g. same image in different sizes -> same URL)
sublinks = sublinks.filter( novelty_filter ); // Global duplicate remover
// sublinks = sublinks.concat( tag_links ); // Add tags back in. Not ideal, but should be functional.
if( options_map.story ) {
// Janky last-ditch text ripping, because Tumblr is killing NSFW sites:
// Assume we're using /mobile. Grab everything between the end of the blog name and the standard navigation div.
// let story_start = s[0].indexOf( '</h1>' ) + 5;
let story_start = s[0].indexOf( '<p>' ) + 3; // Better to skip ahead a bit instead of showing the <h2> date
let story_end = s[0].indexOf( '<div id="navigation">' );
let story = s[0].substring( story_start, story_end );
// Problem: images are still showing up. We sort of don't want that. Clunk.
story = story.replace( /<img/g, '<notimg' ); // Clunk clunk clunk get it done. It is extreme go horse time.
// I should really convert inline images to links.
// Embedded videos what the fuck?
story = story.replace( /<iframe/g, '<notiframe' ); // Clunk clunk clunk!
story = story.replace( /<style/g, '<notstyle' ); // CLUNK CLUNK.
//story = story.replace( /<span/g, '<notspan' ); // Clunk? // Oh, no. It's the /embed crap.
sublinks.push( story ); // This is just an array of strings, right?
// console.log( story ); // DEBUG
}
// Oh right, this doesn't really return anything. We arrange and insert HTML from here.
var bulk_string = "<br><a href='" + post_url + "'>" + post_url + "</a><br>"; // A digest, so we can update innerHTML just once per div
sublinks.forEach( (link) => {
let contents = link;
if( options_map.thumbnails == 'xml' && link.indexOf( '_1280' ) > -1 ) { // If we're showing thumbnails and this image can be resized, do, then show it
let img = link.replace( '_1280', '_100' );
contents = '<img src=' + img + '>' + link; // <img> deserves a class, for consistent scale.
}
this_link = '<a href ="' + link + '">' + contents + '</a><br>';
// How is this not what's causing the CSS bleed?
// if( link.substring(0) == '\n' ) { // If this is text
if( link.indexOf( '\n' ) > -1 ) { // If this is text
this_link = link;
}
bulk_string += this_link;
} )
var tag_string = "";
tag_links.forEach( (link) => {
// let tag = link.split( '/tagged/' )[1]
tag_string += '#' + link.split( '/tagged/' )[1] + ' ';
} )
bulk_string += tag_string + "<br>";
// Tags should be added here. If we have them.
// console.log( bulk_string );
document.getElementById( '' + post_url ).innerHTML = bulk_string; // Yeeeah, I should probably create these div IDs before this happens.
// And here's where I'd increase the page counter, if we had one.
// Let's use order_array to judge how done we are - and make it a percent, not a page count. Use order_array.length and pretend the last element's the same size.
// document.getElementById( 'pagecounter' ).innerHTML = '%';
// No wait, we can't do it here. This is per-post, not per-page. Or... we could do it real half-assed.
} )
)
.then( s => { // I don't think we take any actual data here. This just fires once per 'responses' group, so we can indicate page count etc.
let completion = Math.ceil( 100 * (which_entry+1) / order_array.length ); // Zero-ordinal index to percentage. Blugh.
document.getElementById( 'pagecounter' ).innerHTML = '' + completion + '%';
} )
} )
// "Promises allow a flat execution pattern!" Fuck you, you liars. Look at that rat's nest of alternating braces.
// If you've done all the work in spaghetti functions somewhere else, maybe it's fine, but if you want code to happen where it fucking starts, anonymous functions SUCK.
}
// ------------------------------------ Post-by-post scraper with embedded images ------------------------------------ //
// Scrape each page for /post/ links, scrape each /post/ for content, display in-order with less callback hell
// New layout & new scrape method - not required to be compatible with previous functions
function new_embedded_display() {
if( isNaN( parseInt( options_map.startpage ) ) || options_map.startpage <= 1 ) { options_map.startpage = 1; }
mydiv.innerHTML += "<center>" + html_previous_next_navigation() + "<br><br>" + html_page_count_navigation() + "</center><br>";
document.getElementById("bottom_controls_div").innerHTML += "<center>" + html_page_count_navigation() + "<br><br>" + html_previous_next_navigation() + "</center>";
// Links out from this mode - scrapewholesite, original mode, maybe other crap
//mydiv.innerHTML += "This mode is under development and subject to change."; // No longer true. It's basically feature-complete.
// mydiv.innerHTML += " - <a href=" + options_url( { everypost:false } ) + ">Return to original image browser</a>" + "<br>" + "<br>";
mydiv.innerHTML += "<br>" + html_ezastumblrscrape_options() + "<br><br>";
// "Pages 1 to 10 (of 100) from http://example.tumblr.com"
mydiv.innerHTML += "Pages " + options_map.startpage + " to " + (options_map.startpage + options_map.pagesatonce - 1);
if( !isNaN(options_map.lastpage) ) { mydiv.innerHTML += " (of " + options_map.lastpage + ")"; }
mydiv.innerHTML += " from " + site_and_tags + "<br>";
mydiv.innerHTML += image_size_options() + "<br><br>";
mydiv.innerHTML += image_resolution_options() + "<br><br>";
// Messy inline function for toggling page breaks - they're optional because we have post permalinks now
mydiv.innerHTML += "<a href='javascript: void(0);' class='interactive'; onclick=\"(function(o){ \
let mydiv = document.getElementById('maindiv'); \
if( mydiv.className == '' ) { mydiv.className = 'showpagelinks'; } \
else { mydiv.className = ''; } \
})(this)\">Toggle page breaks</a><br><br>";
mydiv.innerHTML += "<span id='0'></span>"; // Empty span for things to be placed after.
posts_placed.push( 0 ); // Because fuck special cases.
// Scrape some pages
for( let x = options_map.startpage; x < options_map.startpage + options_map.pagesatonce; x++ ) {
fetch( site_and_tags + "/page/" + x, { credentials: 'include' } ).then( r => r.text() ).then( text => {
scrape_by_posts( text, x );
} )
}
}
// Take the HTML from a /page, fetch the /post links, display images
// Probably ought to be despaghettified and combined with the above function, but I was fighting callback hell -hard- after the last major version
// Alternately, split it even further and do some .then( do_this ).then( do_that ) kinda stuff above.
function scrape_by_posts( html_copy, page_number ) {
// console.log( page_dupe_hash ); // DEBUG
let posts = links_from_page( html_copy ); // Get links on page
posts = posts.filter( link => { return link.indexOf( '/post/' ) > 0 && link.indexOf( '/photoset' ) < 0; } ); // Keep /post links but not photoset iframes
posts = posts.map( link => { return link.replace( '#notes', '' ); } ); // post/1234 is the same as /post/1234#notes
posts = posts.filter( link => link.indexOf( window.location.host ) > 0 ); // Same-origin filter. Not necessary, but it unclutters the console. Fuckin' CORS.
if( page_number != 1 ) { posts = posts.filter( novelty_filter ); } // Attempt to remove posts linked on every page, e.g. commission info. Suffers a race condition.
posts = remove_duplicates( posts ); // De-dupe
// 'posts' now contains an array of /post URLs
// Display link and linebreak before first post on this page
let first_id = posts.map( u => parseInt( u.split( '/' )[4] ) ).sort( ).pop(); // Grab ID from its place in each URL, sort accordingly, take the top one
let page_link = "<span class='pagelink'><hr><a target='_blank' href='" + site_and_tags + "/page/" + page_number + "'> Page " + page_number + "</a>";
if( posts.length == 0 ) { first_id = 1; page_link += " - No images found."; } // Handle empty pages with dummy content. Out of order, but whatever.
page_link += "<br><br></span>";
display_post( page_link, first_id + 0.5 ); // +/- on the ID will change with /chrono, once that matters
posts.map( link => {
fetch( link, { credentials: 'include' } ).then( r => r.text() ).then( text => {
let sublinks = links_from_page( text );
sublinks = sublinks.filter( s => { return s.indexOf( '.jpg' ) > 0 || s.indexOf( '.jpeg' ) > 0 || s.indexOf( '.png' ) > 0 || s.indexOf( '.gif' ) > 0; } );
sublinks = sublinks.filter( tumblr_blacklist_filter ); // Remove avatars and crap
sublinks = sublinks.map( image_standardizer ); // Clean up semi-dupes (e.g. same image in different sizes -> same URL)
sublinks = sublinks.filter( novelty_filter ); // Global duplicate remover
// Oh. Photosets sort of just... work? That might not be reliable; DownThemAll acts like it can't see the iframes on some themes.
// Yep, they're there. Gonna be hard to notice if/when they fail. Oh well, "not all images are guaranteed to appear."
// Videos will still be weird. (But it does grab their preview thumbnails.)
// Wait, can I filter reblogs here? E.g. with a ?noreblogs flag, and then checking if any given post has via/source links. Hmm. Might be easier in /mobile pages.
// Seem to get a lot of duplicate images? e.g. both
// https://media.tumblr.com/tumblr_m2gktkD7u31qdcy3io1_640.jpg and
// https://media.tumblr.com/tumblr_m2gktkD7u31qdcy3io1_1280.jpg
// Oh! Do I just not handle _640?
// Get ID from post URL, e.g. http//example.tumblr.com/post/12345/title => 12345
let post_id = parseInt( link.split( '/' )[4] ); // 12345 as a NUMBER, not a string, doofus
if( sublinks.length > 0 ) { // If this post has images we're displaying -
let this_post = new String;
sublinks.map( url => {
this_post += '<span class="container">';
this_post += '<a target="_blank" id="' + encodeURI( url ) + '" href="' + url + '">';
this_post += '<img onerror=\'' + error_function( url ) + '\' id="'+url+'" src="' + url + '"></a>';
this_post += '<span class="postlink"><a target="_blank" href="' + link + '">Permalink</a></span></span> ';
} )
display_post( this_post, post_id );
}
} )
} )
}
// Place content on page in descending order according to post ID number
// Consider rejiggering the old scrape method to use this. Move to 'universal' section if so. Alter or spin off to link posts instead?
// Turns out I never implemented ?chrono or ?reverse, so nevermind that for now.
// Remember to set options_map.chrono if ?find contains /chrono or whatever.
function display_post( content, post_id ) {
let this_node = document.createElement( "span" );
this_node.innerHTML = content;
this_node.id = post_id
// Find lower-numbered node than post_id
let target_id = posts_placed.filter( n => n <= post_id ).sort( ).pop(); // Take the highest number less than (or equal to) post_id
if( options_map.find.indexOf( '/chrono' ) > 0 ) {
target_id = posts_placed.filter( n => n <= post_id ).sort( ).shift(); // Take the... fuck... lowest? What am I doing again?
// Fuuuck, this is really inconsistent. Nevermind the looney-toons syntax I used here, =>n<=.
// Screw it, use the old scraper for now.
}
let target_node = document.getElementById( target_id );
// http://stackoverflow.com/questions/4793604/how-to-do-insert-after-in-javascript-without-using-a-library
target_node.parentNode.insertBefore( this_node, target_node ); // Insert our span after the lower-ID node
posts_placed.push( post_id ); // Remember that we added this ID
// No return value
}
// Return ascending or descending order depending on "chrono" setting
// function post_order_sort( a, b )
// ------------------------------------ Specific handling for www.tumblr.com (tag search, possibly dashboard) ------------------------------------ //
// URLs like https://www.tumblr.com/tagged/wooloo?before=1560014505 don't follow easy sequential pagination, so we have to (a) be linear or (b) guess. First whack is (a).
// Tumblr dashboard is obvious standardized, so we can make assumptions about /post links in relation to images.
// We do not fetch any individual /posts. We can't. They're on different subdomains, and Tumblr CORS remains tight-assed. But we can link them like in the post scrape mode.
// Ooh, I could maybe get "behind the jump" content via /embed URLs. Posts here should contain the blog UUID and a post number.
// Copied notes from above:
// https://www.tumblr.com/tagged/homestuck sort of works. Trouble is, pages go https://www.tumblr.com/tagged/homestuck?before=1558723097 - ugh.
// Next page is https://www.tumblr.com/tagged/homestuck?before=1558720051 - yeah this is a Unix timestamp. It's epoch time.
// https://www.tumblr.com/dashboard/2/185117399018 - but this one -is- based on /post numbers. Tumblr.com: given two choices, take all three.
// Being able to scrape and archive site-wide tags or your own dashboard would be useful. Dammit.
// Just roll another mode into this script. It's already a hot mess. The new code just won't run on individual blogs.
// Okay, so dashboard and site-wide tag modes.
// Dashboard post numbers don't have to be real post numbers. Tag-search timestamps obviously don't have to relate to real posts.
// We de-dupe, so overkill is fine... ish. Tumblr's touchy about "rate limit exceeded" these days.
// Tag scrape would be suuuper useful if we can grab blog posts from www.tumblr.com. Like "hiveswapcomicscontest."
// Still no luck on scraping dashboard-only blogs. Bluh.
// This is so aggravating. I can see the content, obviously. "View page source" just returns the dashboard source. But document.body.outerHTML contains the blog proper.
// Consider /embed again:
// https://embed.tumblr.com/embed/post/4RwtewsxXp-k1ReCcdAgXg/185288559546?width=542&language=en_US&did=a5c973d33a43ace664986204d72d7739de31b614
// This works but provides no previous/next link. (We need the ID, but we can get it from www.tumblr.com, then redirect.)
// Using DaveJaders for testing. https://davejaders.tumblr.com/archive does not redirect, so we can use that for same-origin fetches. Does /page stuff work?
// fetch( '/' ).then( r => r.text() ).then( t => document.body.outerHTML = t ) - CORS failure. "The Same Origin Policy disallows reading the remote resource at https://www.tumblr.com/login_required/davejaders . (Reason: CORS header ‘Access-Control-Allow-Origin’ missing)." Fuck me again, apparently.
// Circle back to dashboard-only blogs by showing the actual content. Avoid clearing body.innerHTML, let Tumblr do its thing, interact with the sidebar deal.
// I could still estimate page count, using a binary search. Assuming constant post rates.
// Or, get fancy, and estimate total posts from time between posts on sparse samples. Each fetch is ten posts in-order.
// It does say "No posts found." See https://www.tumblr.com/tagged/starlight-brigade?before=1503710342
// Between the ?before we fetch and the Next link, we have the span of time it took for ten posts to be made. I would completely exclude the "last" page, if found.
// I guess... integrate in blocks? E.g., given a data point, assume that rate continues to the next data point. Loose approximations are fine.
// E.g. push timestamp/rate pairs into an array, sort by timestamp, find delta between each timestamp, and multiply each rate by each period. One period is bidirectional.
// Lerping is not meaningfully harder. It's the above, plus each period times half the difference between the rates at either end of that period.
// Specific date entry? input type = "date".
// Specific date entry is useful and simple enough to finish before posting a new version. Content estimation can wait.
// Oh, idiot: make the "on or before" display also the navigation mechanic.
// We now go right here for any ?ezastumblrscrape URL on www.tumblr.com. Sloppy but functional.
function scrape_www_tagged( ) {
// ?find=/tagged/whatever is already populated by existing scrape links, but any ?key=value stuff gets lost.
// (If no ?before=timestamp, fetch first page. ?before=0 works. Some max_int would be preferable for sorting. No exceptions needed, then.)
is_first_page = false; // Implicit global, eat me. Clunk.
if( isNaN( options_map.before ) ) {
options_map.before = parseInt( Date.now() / 1000 ); // Current Unix timestamp in seconds, not milliseconds. The year is not 51413.
is_first_page = true;
}
if( options_map.before > parseInt( Date.now() / 1000 ) ) { is_first_page = true; } // Also handle when going days or years "into the future."
// Standard-ish intial steps: clear page (handled), add controls, maybe change window title.
// We can't use truly standard prev/next navigation. Even officially, you only get "next" links. (Page count options should still work.)
let www_tagged_next = "<a class='www_next' href='" + options_url() + "'>Next >>></a>";
let pages_www_tagged = "";
pages_www_tagged += "<a class='www_next' href='" + options_url( {"pagesatonce": 10} ) + "'> 10</a>, ";
pages_www_tagged += "<a class='www_next' href='" + options_url( {"pagesatonce": 5} ) + "'> 5</a>, ";
pages_www_tagged += "<a class='www_next' href='" + options_url( {"pagesatonce": 1} ) + "'> 1</a> - >>>";
// Periods of time, in seconds, because Unix epoch timestamps.
let one_day = 24 * 60 * 60;
let one_week = one_day * 7;
let one_year = one_day * 365.24; // There's no reason not to account for leap years.
let approximate_place = "Posts <a href='" + options_map.find + "'>" + options_map.find.split('/').join(' ') + "</a>"; // options_map.find as relative link. Convenient.
let date_string = new Date( options_map.before * 1000 ).toISOString().slice( 0, 10 );
approximate_place += " on or before <input type='date' id='on_or_before' value='" + date_string + "'>";
// How do I associate this entry with the button so that pressing Enter triggers the button? Eh, minor detail.
let coverage = "<span id='coverage'></span>"; // Started writing this string, then realized I can't know its value until all pages are fetched.
let time_www_tagged = ""; // Browse forward / backward by day / week / year.
time_www_tagged += "<a href='" + options_url( {"before": options_map.before + one_day} ) + "'><<< </a> One day";
time_www_tagged += "<a href='" + options_url( {"before": options_map.before - one_day} ) + "'> >>></a> - ";
time_www_tagged += "<a href='" + options_url( {"before": options_map.before + one_week} ) + "'><<< </a> One week";
time_www_tagged += "<a href='" + options_url( {"before": options_map.before - one_week} ) + "'> >>></a> - ";
time_www_tagged += "<a href='" + options_url( {"before": options_map.before + one_year} ) + "'><<< </a> One year";
time_www_tagged += "<a href='" + options_url( {"before": options_map.before - one_year} ) + "'> >>></a>";
// Change the displayed date and it'll go there. Play stupid games, win stupid prizes.
let jump_button_code = 'window.location += "?before=" + parseInt( new Date( document.getElementById( "on_or_before" ).value ).getTime() / 1000 );';
let date_jump = "<button onclick='" + jump_button_code + "'>Go</button>";
mydiv.innerHTML += "<center>" + www_tagged_next + "<br><br>" + pages_www_tagged + "</center><br>";
mydiv.innerHTML += "<center>" + approximate_place + date_jump + "<br>" + coverage + "<br>" + time_www_tagged + "</center>";
mydiv.innerHTML += "This mode is under development and subject to change.<br><br>";
document.getElementById("bottom_controls_div").innerHTML += "<center>" + pages_www_tagged + "<br><br>" + www_tagged_next + "</center>";
mydiv.innerHTML += image_size_options() + "<br><br>";
mydiv.innerHTML += image_resolution_options() + "<br><br>";
// Fetch first page specified by ?before=timestamp.
let tagged_url = "" + options_map.find + "?before=" + options_map.before; // Relative URLs are guaranteed to be same-domain, even if they're garbage.
fetch( tagged_url, { credentials: 'include' } ).then( r => r.text() ).then( text => {
display_www_tagged( text, options_map.before, options_map.pagesatonce ); // ... pagesatonce gets set in the pre-amble, right? It should have a default.
// Optionally we could check here if options_map.before == 0 and instead send max_safe_integer.
} )
}
// Either we need a global variable for how many more pages per... page... or else I should pass a how_many_more value to this recursive function.
function display_www_tagged( content, timestamp, pages_left ) {
// First, grab the Next link - i.e. its ?before=timestamp value.
// let next_timestamp_index = content.lastIndexOf( '?before=' );
// let next_timestamp = content.substring( next_timestamp_index + 8, content.indexOf( next_timestamp_index, '"' ) ); // Untested
let next_timestamp = content.split( '?before=' ).pop().split( '"' ).shift(); // The last "?before=12345'" string on the page. Clunky but tolerable.
next_timestamp = "" + parseInt( next_timestamp ); // Guarantee this is a string of a number. (NaN "works.") Pages past the end may return nonsense.
if( pages_left > 1 ) { // If we're displaying more pages then fetch that and recurse.
let tagged_url = "" + options_map.find + "?before=" + next_timestamp; // Relative URLs are guaranteed to be same-domain, even if they're garbage.
// console.log( tagged_url );
fetch( tagged_url, { credentials: 'include' } ).then( r => r.text() ).then( text => {
display_www_tagged( text, next_timestamp, pages_left - 1 );
} )
} else { // Otherwise put that timestamp in our constructed Next link(s).
// I guess... get HTMLcollection of elements for "next" links, and change each one.
// Downside: links will only change once the last page is fetched. We could tack on a ?before for every fetch, but it would get silly. Right?
let next_links = Array.from( document.getElementsByClassName( 'www_next' ) ); // I'm not dealing with a live object unless I have to.
for( link of next_links ) { link.href += "?before=" + next_timestamp; }
// Oh right, and update the header to guesstimate what span of time we're looking at.
let coverage = document.getElementById( 'coverage' );
coverage.innerHTML += "Displaying ";
if( is_first_page ) { coverage.innerHTML += "the most recent "; } // Condition is no longer sensible, but it's a good placeholder.
// We can safely treat ?before as the initial timestamp. Perfect accuracy is not important.
let time_covered = Math.abs( options_map.before - next_timestamp ); // Absolute so I don't care if I have it backwards.
if( time_covered > 48 * 60 * 60 ) { coverage.innerHTML += parseInt( time_covered / (24*60*60) ) + " days of posts"; } // Over two days? Display days.
else if( time_covered > 2 * 60 * 60 ) { coverage.innerHTML += parseInt( time_covered / (60*60) ) + " hours of posts"; } // Over two hours? Display hours.
else if( time_covered > 2 * 60 ) { coverage.innerHTML += parseInt( time_covered / 60 ) + " minutes of posts"; } // Over two minutes? Display minutes.
else { coverage.innerHTML += " several seconds of posts"; } // Otherwise just say it's a damn short time.
if( time_covered == 0 ) { coverage.innerHTML = "Displaying the first available posts"; } // Last page, earlier posts. No "next" page.
}
// Insert div for this timestamp's page.
let new_div = document.createElement( 'span' ); // Span, because divs cause line breaks. Whoops.
new_div.id = "" + timestamp;
let target_node = document.getElementById( 'bottom_controls_div' );
target_node.parentNode.insertBefore( new_div, target_node ); // Insert each page before the footer.
let div_html = "";
// Separate page HTML by posts.
// At least the "li" elements aren't nested, so I can terminate the last one on "</li>". Or... all of them.
let posts = content.split( '<li class="post_container"' );
posts.shift();
posts.push( posts.pop().split( '</li>' )[0] ); // Terminate last element at </li>. Again, not great code, but clunk clunk clunk get it done.
// For each post:
for( post of posts ) {
// Extract images from each post.
let links = links_from_page( post );
links = links.map( image_standardizer ); // This goes before grabbing the permalink because /post URLs do get standardized. No &media guff.
let permalink = links.filter( s => s.indexOf( '.tumblr.com/post' ) > 0 )[0]; // This has to go before de-duping, or posts linking to posts can leave permalinks blank.
links = links.filter( novelty_filter );
links = links.filter( tumblr_blacklist_filter );
// document.body.innerHTML += links.join( "<br>" ) + "<br><br>"; // Debug
// Separate the images.
let images = links.filter( s => s.indexOf( 'media.tumblr.com' ) > 0 ); // Note: this will exclude external images, e.g. embedded Twitter stuff.
// If this post has images:
if( images.length > 0 ) { // Build HTML xor insert div for each post, to display images.
// Get /post URL, including blog name etc.
//let permalink = links.filter( s => s.indexOf( '.tumblr.com/post' ) > 0 )[0];
let post_html = "";
for( image of images ) {
post_html += '<span class="container">';
post_html += '<a target="_blank" id="' + encodeURI( image ) + '" href="' + image + '">';
post_html += '<img onerror=\'' + error_function( image ) + '\' id="' + image + '" src="' + image + '"></a>';
post_html += '<span class="postlink"><a target="_blank" href="' + permalink + '">Permalink</a></span></span> ';
}
div_html += post_html;
}
}
// Insert accumulated HTML into this div.
new_div.innerHTML = div_html;
}
// ------------------------------------ HTML-returning functions for duplication prevention ------------------------------------ //
// Return HTML for standard Previous / Next controls (<<< Previous - Next >>>)
function html_previous_next_navigation() {
let prev_next_controls = "";
if( options_map.startpage > 1 ) {
prev_next_controls += "<a href='" + options_url( "startpage", options_map.startpage - options_map.pagesatonce ) + "'><<< Previous</a> - ";
}
prev_next_controls += "<a href='" + options_url( "startpage", options_map.startpage + options_map.pagesatonce ) + "'>Next >>></a>";
return prev_next_controls;
}
// Return HTML for pages-at-once versions of previous/next page navigation controls (<<< 10, 5, 1 - 1, 5, 10 >>>)
function html_page_count_navigation() {
let prev_next_controls = "";
if( options_map.startpage > 1 ) { // <<< 10, 5, 1 -
prev_next_controls += "<<< ";
prev_next_controls += "<a href='" + options_url( {"startpage": options_map.startpage - 1, "pagesatonce": 1} ) + "'> 1</a>, ";
prev_next_controls += "<a href='" + options_url( {"startpage": options_map.startpage - 5, "pagesatonce": 5} ) + "'> 5</a>, ";
prev_next_controls += "<a href='" + options_url( {"startpage": options_map.startpage - 10, "pagesatonce": 10} ) + "'> 10</a> - ";
}
prev_next_controls += "<a href='" + options_url( {"startpage": options_map.startpage + options_map.pagesatonce, "pagesatonce": 10} ) + "'> 10</a>, ";
prev_next_controls += "<a href='" + options_url( {"startpage": options_map.startpage + options_map.pagesatonce, "pagesatonce": 5} ) + "'> 5</a>, ";
prev_next_controls += "<a href='" + options_url( {"startpage": options_map.startpage + options_map.pagesatonce, "pagesatonce": 1} ) + "'> 1</a> - ";
prev_next_controls += ">>>";
return prev_next_controls;
}
// Return HTML for image-size options (changes via CSS or via URL parameters)
// This used to work. It still works, in the new www mode I just wrote. What the fuck do I have to do for some goddamn onclick behavior?
// "Content Security Policy: The page’s settings blocked the loading of a resource at self (“script-src https://lalilalup.tumblr.com https://assets.tumblr.com/pop/ 'nonce-OTA4NjViZmE2MzZkYTFjMjM1OGZkZGM1MzkwYWU4NTA='”)." What the fuck.
// Jesus Christ, it might be yet again because of Tumblr's tightass settings:
// https://stackoverflow.com/questions/37298608/content-security-policy-the-pages-settings-blocked-the-loading-of-a-resource
// The function this still works in is on a "not found" page. The places it will not work are /archive pages.
// Yeah, on /mobile instead of /archive it works fine. Fuck you, Tumblr.
// Jesus, that means even onError can't work.
function image_size_options() {
var html_string = "Immediate: \t"; // Change <body> class to instantly resize images, temporarily
html_string += "<a class='interactive' href='javascript: void(0);' onclick=\"(function(o){ document.body.className = o; })('')\">Original image sizes</a> - ";
html_string += "<a class='interactive' href='javascript: void(0);' onclick=\"(function(o){ document.body.className = o; })('fixed-width')\">Snap columns</a> - ";
html_string += "<a class='interactive' href='javascript: void(0);' onclick=\"(function(o){ document.body.className = o; })('fixed-height')\">Snap rows</a> - ";
html_string += "<a class='interactive' href='javascript: void(0);' onclick=\"(function(o){ document.body.className = o; })('fit-width')\">Fit width</a> - ";
html_string += "<a class='interactive' href='javascript: void(0);' onclick=\"(function(o){ document.body.className = o; })('fit-height')\">Fit height</a> - ";
html_string += "<a class='interactive' href='javascript: void(0);' onclick=\"(function(o){ document.body.className = o; })('smart-fit')\">Fit both</a><br><br>";
html_string += "Persistent: \t"; // Reload page with different image mode that will stick for previous/next pages
html_string += "<a href=" + options_url( { thumbnails: "original" } ) + ">Original image sizes</a> - "; // This is the CSS default, so any other value works
html_string += "<a href=" + options_url( { thumbnails: "fixed-width" } ) + ">Snap columns</a> - ";
html_string += "<a href=" + options_url( { thumbnails: "fixed-height" } ) + ">Snap rows</a> - ";
html_string += "<a href=" + options_url( { thumbnails: "fit-width" } ) + ">Fit width</a> - ";
html_string += "<a href=" + options_url( { thumbnails: "fit-height" } ) + ">Fit height</a> - ";
html_string += "<a href=" + options_url( { thumbnails: "smart-fit" } ) + ">Fit both</a>";
return html_string;
}
// Return HTML for links to ?maxres versions of the same page, e.g. "_raw" versus "_1280"
function image_resolution_options() {
var html_string = "Maximum resolution: \t";
html_string += "<a href=" + options_url( { maxres: "raw" } ) + ">Raw</a> - "; // I'm not 100% sure "_raw" works anymore, but the error function handles it, so whatever.
html_string += "<a href=" + options_url( { maxres: "1280" } ) + ">1280</a> - ";
html_string += "<a href=" + options_url( { maxres: "500" } ) + ">500</a> - ";
html_string += "<a href=" + options_url( { maxres: "400" } ) + ">400</a> - ";
html_string += "<a href=" + options_url( { maxres: "250" } ) + ">250</a> - ";
html_string += "<a href=" + options_url( { maxres: "100" } ) + ">100</a>";
return html_string;
}
// Return links to other parts of Eza's Tumblr Scrape functionality, possibly excluding whatever you're currently doing
// Switch to full-size images - Toggle image size - Show one page at once - Scrape whole Tumblr - (Experimental fetch-every-post image browser)
// This mode is under development and subject to change. - Return to original image browser
function html_ezastumblrscrape_options() {
let html_string = "";
// "You are browsing" text? Tell people where they are and what they're looking at.
html_string += "<a href='" + options_url( { scrapemode: "scrapewholesite" } ) + "'>Scrape whole Tumblr</a> - ";
html_string += "<a href='" + options_url( { scrapemode: "pagebrowser" } ) + "'>Browse images</a> - "; // Default mode; so any value works
html_string += "<a href='" + options_url( { scrapemode: "everypost" } ) + "'>(Experimental fetch-every-post image browser)</a> ";
return html_string;
}
function error_function( url ) {
// This clunky <img onError> function looks for a lower-res image if the high-res version doesn't exist.
// Surprisingly, this does still matter. E.g. http://66.media.tumblr.com/ba99a55896a14a2e083cec076f159956/tumblr_inline_nyuc77wUR01ryfvr9_500.gif
// This might mismatch _100 images and _250 links because of that self-erasing clause... but it's super rare, so meh.
let on_error = 'if(this.src.indexOf("_raw")>0){this.src=this.src.replace("_raw","_1280").replace("//media","//66.media");}'; // Swap _raw for 1280, add CDN number
on_error += 'else if(this.src.indexOf("_1280")>0){this.src=this.src.replace("_1280","_500");}'; // Swap 1280 for 500
on_error += 'else if(this.src.indexOf("_500")>0){this.src=this.src.replace("_500","_400");}'; // Or swap 500 for 400
on_error += 'else if(this.src.indexOf("_400")>0){this.src=this.src.replace("_400","_250");}'; // Or swap 400 for 250
on_error += 'else{this.src=this.src.replace("_250","_100");this.onerror=null;}'; // Or swap 250 for 100, then give up
on_error += 'document.getElementById("' + encodeURI( url ) + '").href=this.src;'; // Link the image to itself, regardless of size
// 2020: This has to get more complicated, or at least more verbose, to support shitty new image URLs.
// https://66.media.tumblr.com/eb7d40a8683e623f173f81fc253056dc/e4277131a4c174c7-a5/s1280x1920/5b3c54b53a09f91b08f9af7fb88e30a253f3c5db.jpg
//'s400x600'
//'s500x750'
//'s640x960'
//'s1280x1900'
// 's2048x3072'
// This has to go in Image Glutton, because Tumblr only ever gets worse
// Start over.
// on_error = 'this.src = this.src.replace( "_250", "_100" ); '; // Unconditional, reverse ladder pattern. Don't Google that. I just made it up.
on_error = "let oldres = [ '_100', '_250', '_400', '_500', '_1280' ]; "; // Inverted quotes, Eh.
on_error += "let newres = [ 's400x600', 's500x750', 's640x960', 's1280x1920' ];"; // Should probably include slashes. Rare collisions.
// on_error += "for( let i = old.length-1; i > 1; i-- ) { this.src.replace( old[i], old[i-1] ); } "; // 1280 -> 500, 500 -> 400... shit.
on_error += "for( let i = 1; i < oldres.length; i++ ) { this.src = this.src.replace( oldres[i+1], oldres[i] ); } "; // 250 -> 100, 400 -> 250, etc. One step per error.
on_error += "for( let i = 1; i < newres.length; i++ ) { this.src = this.src.replace( newres[i+1], newres[i] ); } ";
on_error += 'document.getElementById("' + encodeURI( url ) + '").href=this.src;'; // Original quotes - be cautious when editing
// God dammit. The links work, but the embedded images don't.
// https://66.media.tumblr.com/eb7d40a8683e623f173f81fc253056dc/e4277131a4c174c7-a5/s1280x1920/ba0f0dd5d5e76726639782f57c4b732f156a5219.jpg
// https://66.media.tumblr.com/eb7d40a8683e623f173f81fc253056dc/e4277131a4c174c7-a5/s1280x1920/5b3c54b53a09f91b08f9af7fb88e30a253f3c5db.jpg
/*
x = 'https://66.media.tumblr.com/eb7d40a8683e623f173f81fc253056dc/e4277131a4c174c7-a5/s1280x1920/44b457df0d9826135b062563cd802afdfe496888.jpg'
newres = [ 's400x600', 's500x750', 's640x960', 's1280x1920' ];
for( let i = 1; i < newres.length; i++ ) { this.src.replace( newres[i+1], newres[i] ); }
*/
return on_error;
}
// ------------------------------------ Universal page-scraping function (and other helper functions) ------------------------------------ //
// Add URLs from a 'blank' page to page_dupe_hash (without just calling soft_scrape_page_promise and ignoring its results)
function exclude_content_example( url ) {
fetch( url, { credentials: 'include' } ).then( r => r.text() ).then( text => {
let links = links_from_page( text );
links = links.filter( novelty_filter ); // Novelty filter twice, because image_standardizer munges some /post URLs
links = links.map( image_standardizer );
links = links.filter( novelty_filter );
} )
// No return value
}
// Spaghetti to reduce redundancy: given a page's text, return a list of URLs.
function links_from_page( html_copy ) {
// Cut off the page at the "More you might like" / "Related posts" footer, on themes that have one
html_copy = html_copy.split( '="related-posts' ).shift();
let http_array = html_copy.split( /['="']http/ ); // Regex split on anything that looks like a source or href declaration
http_array.shift(); // Ditch first element, which is just <html><body> etc.
http_array = http_array.map( s => { // Theoretically parallel .map instead of maybe-linear .forEach or low-level for() loop
if( s.indexOf( "&" ) > -1 ) { s = htmlDecode( s ); } // Yes a fucking " should match a goddamn regex for terminating on quotes!
s = s.split( /['<>"']/ )[0]; // Terminate each element (split on any terminator, take first subelement)
s = s.replace( /\\/g, '' ); // Remove escaping backslashes (e.g. http\/\/ -> http//)
if( s.indexOf( "%3A%2F%2F" ) > -1 ) { s = decodeURIComponent( s ); } // What is with all the http%3A%2F%2F URLs?
// s = s.split( '"' )[0]; // Yes these count as doublequotes you stupid broken scripting language.
return "http" + s; // Oh yeah, add http back in (regex eats it)
} ) // http_array now contains an array of strings that should be URLs
let post_array = html_copy.split( /['="']\/post/ ); // Regex split on anything that defines looks similar to a src="/post" link
post_array.shift(); // Ditch first element, which is just <html><body> etc.
post_array = post_array.map( s => { // Theoretically parallel .map instead of maybe-linear .forEach or low-level for() loop
s = s.split( /['<>"']/ )[0]; // Terminate each element (split on any terminator, take first subelement)
return window.location.protocol + "//" + window.location.hostname + "/post" + s; // Oh yeah, add /post back in (regex eats it)
} ) // post_array now contains an array of strings that should be photoset URLs
http_array = http_array.concat( post_array ); // Photosets are out of order again. Blar.
return http_array;
}
// Filter: Return false for typical Tumblr nonsense (JS, avatars, RSS, etc.)
function tumblr_blacklist_filter( url ) {
if( url.indexOf( "/reblog/" ) > 0 ||
url.indexOf( "/tagged/" ) > 0 || // Might get removed so the script can track and report tag use. Stupid art tags like 'my-draws' or 'art-poop' are a pain to find.
url.indexOf( ".tumblr.com/avatar_" ) > 0 ||
url.indexOf( ".tumblr.com/image/" ) > 0 ||
url.indexOf( ".tumblr.com/rss" ) > 0 ||
url.indexOf( "srvcs.tumblr.com" ) > 0 ||
url.indexOf( "assets.tumblr.com" ) > 0 ||
url.indexOf( "schema.org" ) > 0 ||
url.indexOf( ".js" ) > 0 ||
url.indexOf( ".css" ) > 0 ||
url.indexOf( "twitter.com/intent" ) > 0 || // Weirdly common now
url.indexOf( "tmblr.co/" ) > 0 ||
//https://66.media.tumblr.com/dea691ec31c9f719dd9b057d0c2be8c3/a533a91e352b4ac7-29/s16x16u_c1/f1781cbe6a6eb7417aa6c97e4286d46372e5ba37.jpg
url.indexOf( "s16x16u" ) > 0 || // Avatars
url.indexOf( "s64x64u" ) > 0 || // Avatars
//https://66.media.tumblr.com/49fac1010ada367bcaee721ea49d0de5/3ddc8ebad67ae55a-ca/s64x64u_c1/592c805c7ce2c44cd389421697e0360b83d89f37.jpg
url.indexOf( "u_c1/" ) > 0 || // Avatars
// https://66.media.tumblr.com/a4fe5011f9e07f95588c789128b60dca/603e412a60dc5227-4a/s64x64u_c1_f1/7b50fe887c756759973d442a89ef86ab1359f6e5.gif
url.indexOf( "ezastumblrscrape" ) > 0 ) // Somehow this script is running on pages being fetched, inserting a link. Okay. Sure.
{ return false } else { return true }
}
// Return standard canonical URL for various resizes of Tumblr images - size of _1280, single CDN
// 10/14 - ?usesmall seems to miss the CDN sometimes?
// e.g. http://mooseman-draws.tumblr.com/archive?startpage=1?pagesatonce=5?thumbnails?ezastumblrscrape?scrapemode=everypost?lastpage=37?usesmall
// https://66.media.tumblr.com/d970fff86185d6a51904e0047de6e764/tumblr_ookdvk7foy1tf83r7o1_400.png sometimes redirects to 78.media and _raw. What?
// Oh, probably not my script. I fucked with the Tumblr redirect script I use, but didn't handle the lack of CDN in _raw sizes.
// 2020: tumblr scrape:
// https://66.media.tumblr.com/b1515c45637955e1e52ec213944db662/08f2b2e47bbce703-70/s400x600/9077891dda98f127c040bd77581014ce4019fe75.gifv
// obviously can go to .gif, but that .gif saves as tumblr_b1515c45637955e1e52ec213944db662_9077891d_400.gif.
// Also, bumping up to /s500x750 works, so I can probably still slam everything to maximum size via 'canonical' urls.
function image_standardizer( url ) {
// Some lower-size images are automatically resized. We'll change the URL to the maximum size just in case, and Tumblr will provide the highest resolution.
// Replace all resizes with _1280 versions. Nearly all _1280 URLs resolve to highest-resolution versions now, so we don't need to e.g. handle GIFs separately.
// Oh hey, Tumblr now has _raw for a no-bullshit as-large-as-possible setting.
// _raw only works without the CDN - so //media.tumblr yes, but //66.media.tumblr no. This complicates things.
// Does //media and _raw always work? No, of course not. So we still need on_error.
// url = url.replace( "_540.", "_1280." ).replace( "_500.", "_1280." ).replace( "_400.", "_1280." ).replace( "_250.", "_1280." ).replace( "_100.", "_1280." );
let maxres = "1280"; // It is increasingly unlikely that _raw still works. Reconsider CDN handling if that's the case.
if( options_map.maxres ) { maxres = options_map.maxres } // If it's set, use it. Should be _100, _250, whatever. ?usesmall should set it to _400. ?notraw, _1280.
maxres = "_" + maxres + "."; // Keep the URL options clean: "400" instead of "_400." etc.
url = url.replace( "_raw", maxres ).replace( "_1280.", maxres ).replace( "_640.", maxres ).replace( "_540.", maxres )
.replace( "_500.", maxres ).replace( "_400.", maxres ).replace( "_250.", maxres ).replace( "_100.", maxres );
// henrythehangman.tumblr.com has doubled images from /image posts in ?scrapemode=everypost. Lots of _1280.jpg?.jpg nonsense.
// Is that typical for tumblrs with this theme? It's one of those annoying magnifying-glass-on-hover deals. If it's just that one weird fetish site, remove this later.
url = url.split('?')[0]; // Ditch anything past the first question mark, if one exists
url = url.split('&')[0]; // Ditch anything past the first ampersand, if one exists - e.g. speikobrarote.tumblr.com
if( url.indexOf( 'tumblr.com' ) > 0 ) { url = url.split( ' ' )[0]; } // Ditch anything past a trailing space, if one exists - e.g. cinnasmut.tumblr.com
// Standardize media subdomain / CDN subsubdomain, to prevent duplicates and fix _1280 vs _raw complications.
if( url.indexOf( '.media.tumblr.com/' ) > 0 ) {
let url_parts = url.split( '/' )
url_parts[2] = '66.media.tumblr.com'; // This came first. Then //media.tumblr.com worked, even for _raw. Then _raw went away. Now it needs a CDN# again. Bluh.
// url_parts[2] = 'media.tumblr.com'; // 2014: write a thing. 2016: comment out old thing, write new thing. 2018: uncomment old thing, comment new thing. This script.
url = url_parts.join( '/' ).replace( 'http:', 'https:' );
/* // Doesn't work - URLs resolve to HTML when clicked, but don't embed as images. Fuck this website.
// I need some other method for recognizing duplicates - the initial part of the URL is the same.
// Can HTML5 / CSS suppress certain elements if another element is present?
// E.g. #abcd1234#640 is display:none if #abcd1234#1280 exists.
// https://support.awesome-table.com/hc/en-us/articles/115001399529-Use-CSS-to-change-the-style-of-each-row-depending-on-the-content
// The :empty pseudoselector works like :hover. And you can condition it on attributes, like .picture[src=""]:empty.
// The :empty + otherSelector{} syntax is weird, but it's CSS, so of course it's weird.
// Here we'd... generate a style element? Probably just insert a <style> block in innerHTML, honestly. Is inline an option?
// http://rafaelservantez.com/writings_tutorials_web_css_cond.html
// Ech, + requires immediate adjacency. ~ selects within siblings, but only siblings "after" the first element.
// Late 2020: shitty new URLs don't contain _tumblr.
if( url.indexOf( '_tumblr' ) < 0 ) { // Should be !String.match, but Christ, one problem at a time.
let othermax = 's1280x1920';
url = url.replace( 's400x600', othermax ).replace( 's500x750', othermax ).replace( 's640x960', othermax );
//'s400x600'
//'s500x750'
//'s640x960'
//'s1280x1920'
}
*/
}
/* // ?notraw and ?usesmall are deprecated. Use ?maxres=1280 or ?maxres=400 instead.
// Change back to _1280 & CDN for ?scrapewholesite (which does no testing). _raw is unreliable.
if( options_map.scrapemode == 'scrapewholesite' || options_map.notraw ) {
url = url.replace( "_raw.", "_1280." ).replace( '//media', '//66.media' );
}
// Change to something smaller for quick browsing, like _500 or _250
// https://78.media.tumblr.com/166565b897f228352069b290067215c0/tumblr_oozoucFu0O1v1bsoxo2_raw.jpg etc don't work. What?
if( options_map.scrapemode != 'scrapewholesite' && options_map.usesmall ) { // Ignore this on scrapewholesite; that'd be a dumb side effect.
url = url.replace( "_raw.", "_400." ).replace( "_1280.", "_400." ).replace( '//media', '//66.media' );
}
*/
return url;
}
// Remove duplicates from an array (from an iterable?) - returns de-duped array
// Credit to http://stackoverflow.com/questions/9229645/remove-duplicates-from-javascript-array for hash-based string method
function remove_duplicates( list ) {
let seen = {};
list = list.filter( function( item ) {
return seen.hasOwnProperty( item ) ? false : ( seen[ item ] = true );
} );
return list;
}
// Filter: Return true ONCE for any given string.
// Global duplicate remover - return false for items found in page_dupe_hash, otherwise add new items to it and return true
// Now also counts instances of each non-unique argument
function novelty_filter( url ) {
// return page_dupe_hash.hasOwnProperty( url ) ? false : ( page_dupe_hash[ url ] = true );
// console.log( page_dupe_hash ); // DEBUG
url = url.toLowerCase(); // Debug-ish, mostly for "tag overview." URLs can be case-sensitive but collisions will be rare. "Not all images are guaranteed to appear."
if( page_dupe_hash.hasOwnProperty( url ) ) {
page_dupe_hash[ url ] += 1;
return false;
} else {
page_dupe_hash[ url ] = 1;
return true;
}
}
// Filter: Return true for any Tumblr /tagged URL.
// This is used separately from the other filters. Those go X=X.filter( remove stuff ), X=X.filter( remove other stuff ). This should run first, like Y=X.filter( return_tags ).
function return_tags( url ) {
// return url.indexOf( '/tagged' ) > 0; // 'if( condition ) { return true } else { return false }' === return condition.
if( url.indexOf( '/tagged' ) > 0 ) { return true; } else { return false; }
}
// Viewing a bare image, redirect to largest available image size.
// // e.g. https://66.media.tumblr.com/d020ac15b0e04ff9381f246ed08c9f05/tumblr_o2lok9yf8O1ub76b5o1_1280.jpg
// This is a sloppy knockoff of Tumblr Image Size 1.1, because that script doesn't affect //media.tumblr.com URLs.
// Also, this is a distinct mode, not a helper function. Maybe just move it up to the start?
function maximize_image_size() {
if( window.location.href.indexOf( "_1280." ) > -1 ) { return false; } // If it's already max-size, die.
if( window.location.href.indexOf( "_raw." ) > -1 ) { return false; } // If it's already max-size, die.
// Nope, still broken. E.g. https://78.media.tumblr.com/tumblr_mequmpAxgp1r9tb2u.gif
// Needs a positive test for all the replacements... or a check if this URL is different from the changed URL.
// This should probably use image_standardizer to avoid duplicate code.
let replacement = image_standardizer( window.location.href );
if( window.location.href != replacement ) { window.location.href = replacement; }
/*
let maxres = "_1280.";
// maxres = "_" + maxres + "."; // Keep the URL options clean: "400" instead of "_400." etc.
url = window.location.href;
url = url.replace( "_1280.", maxres ).replace( "_540.", maxres ).replace( "_500.", maxres ).replace( "_400.", maxres ).replace( "_250.", maxres ).replace( "_100.", maxres );
if( window.location.href != url ) { window.location.href = url; }
*/
}
// Decode entity references.
// This is officially the stupidest polyfill I've ever used. How is there an entire class of standard escaped text with NO standard decoder in Javascript?
// Copied straight from https://stackoverflow.com/questions/1912501/unescape-html-entities-in-javascript because decency was not an option.
function htmlDecode(input) {
return new DOMParser().parseFromString(input, "text/html").documentElement.textContent;
}
// Convert fetch()'d URLs into local URLs. Offsite URLs will 404 instead of triggering CORS as an unhandled error. This language.
// Hmm. This works, but the test case I thought I had worked anyway (with plain fetches) after several refreshes.
// Currently not used anywhere - needs further testing.
function safe_fetch( ){
// args[0] = args[0]; // Editing TBD
// var video_post = window.location.protocol + "//" + subdomain[4] + ".tumblr.com/post/" + subdomain[5] + "/";
// let local_root = window.location.protocol + "//" + window.location.host + "/";
// arguments[0] = arguments[0]; // Editing TBD - arguments is the argv for JS.
arguments[0] = '/' + arguments[0].split( '/' ).slice( 3 ).join( '/' );
return fetch( ... arguments ); // Straightforward. If this returns a thing that returns a promise, safe_fetch.then() -should- "just work."
}
// Given the bare HTML of a Tumblr page, return an array of Promises for image/video/link URLs
function soft_scrape_page_promise( html_copy ) {
// Linear portion:
let http_array = links_from_page( html_copy ); // Split bare HTML into link and image sources
http_array.filter( url => url.indexOf( '/tagged/' ) > 0 ).filter( novelty_filter ); // Track tags for statistics, before the blacklist removes them
http_array = http_array.filter( tumblr_blacklist_filter ); // Blacklist filter for URLs - typical garbage
function is_an_image( url ) {
// Whitelist URLs with image file extensions or Tumblr iframe indicators
var image_link = false;
if( url.indexOf( ".gif" ) > 0 ) { image_link = true; }
if( url.indexOf( ".jpg" ) > 0 ) { image_link = true; }
if( url.indexOf( ".jpeg" ) > 0 ) { image_link = true; }
if( url.indexOf( ".png" ) > 0 ) { image_link = true; }
if( url.indexOf( "/photoset_iframe" ) > 0 ) { image_link = true; }
if( url.indexOf( ".tumblr.com/video/" ) > 0 ) { image_link = true; }
if( url.indexOf( "/audio_player_iframe/" ) > 0 ) { image_link = true; }
return image_link;
}
// Separate the images
http_array = http_array.map( url => {
if( is_an_image( url ) ) { // If it's an image, get rid of any Tumblr variability about resolution or CDNs, to avoid duplicates with nonmatching URLs
return image_standardizer( url );
} else { // Else if not an image
if( url.indexOf( window.location.host ) > 0 ) { url += "#local" } else { url += "#offsite" } // Mark in-domain vs. out-of-domain URLs.
if( options_map.imagesonly ) { return ""; } // ?imagesonly to skip links on ?scrapewholesite
return url + "#link";
}
} )
.filter( n => { // Remove all empty strings, where "empty" can involve a lot of #gratuitous #tags.
if( n.split("#")[0] === "" ) { return false } else { return true }
} );
http_array = remove_duplicates( http_array ); // Remove duplicates within the list
http_array = http_array.filter( novelty_filter ); // Remove duplicates throughout the page
// Should this be skipped on scrapewholesite? Might be slowing things down.
// Async portion:
// Return promise that resolves to list of URLs, including fetched videos and photoset sub-images
return Promise.all( http_array.map( s => {
if( s.indexOf( '/photoset_iframe' ) > 0 ) { // If this URL is a photoset, return a promise for an array of URLs
return fetch( s, { credentials: 'include' } ).then( r => r.text() ).then( text => { // Fetch URL, get body text from response
var photos = text.split( 'href="' ); // Isolate photoset elements from href= declarations
photos.shift(); // Get rid of first element because it's everything before the first "href"
photos = photos.map( p => p.split( '"' )[0] + "#photoset" ); // Tag all photoset images as such, just because
photos[0] += "#" + s; // Tag first image in set with photoset URL so browse mode can link to it
return photos;
} )
}
else if ( s.indexOf( '.tumblr.com/video/' ) > 0 ) { // Else if this URL is an embedded video, return a Tumblr-standard URL for the bare video file
var subdomain = s.split( '/' ); // E.g. https://www.tumblr.com/video/examplename/123456/500/ -> https,,www.tumblr.com,video,examplename,123456,500
var video_post = window.location.protocol + "//" + subdomain[4] + ".tumblr.com/post/" + subdomain[5] + "/";
// e.g. http://examplename.tumblr.com/post/123456/ - note window.location.protocol vs. subdomain[0], maintaining http/https locally
return fetch( video_post, { credentials: 'include' } ).then( r => r.text() ).then( text => {
if( text.indexOf( 'og:image' ) > 0 ) { // property="og:image" content="http://67.media.tumblr.com/tumblr_123456_frame1.jpg" --> tumblr_123456_frame1.jpg
var video_name = text.split( 'og:image' )[1].split( 'media.tumblr.com' )[1].split( '"' )[0].split( '/' ).pop();
} else if( text.indexOf( 'poster=' ) > 0 ) { // poster='https://31.media.tumblr.com/tumblr_nuzyxqeJNh1rjoppl_frame1.jpg'
var video_name = text.split( "poster='" )[1].split( 'media.tumblr.com' )[1].split( "'" )[0].split( '/' ).pop(); // Bandaid solution. Tumblr just sucks.
} else {
return video_post + '#video'; // Current methods miss the whole page if these splits miss, so fuck it, just return -something.-
}
// tumblr_abcdef12345_frame1.jpg -> tumblr_abcdef12345.mp4
video_name = "tumblr_" + video_name.split( '_' )[1] + ".mp4#video";
video_name = "https://vt.tumblr.com/" + video_name; // Standard Tumblr-wide video server
return video_name; // Should be e.g. https://vt.tumblr.com/tumblr_abcdef12345.mp4
} )
}
else if ( s.indexOf( "/audio_player_iframe/" ) > 0 ) { // Else if this URL is an embedded audio file, return... well not a standard URL perhaps, but an URL.
// How the fuck do I download audio? Video works-ish. Audio is not well-supported.
// http://articulatelydecomposed.tumblr.com/post/176171225450/you-know-i-had-to-do-it-the-new-friendsim-had-a
// Ctrl+F "plays". Points to:
// http://articulatelydecomposed.tumblr.com/post/176171225450/audio_player_iframe/articulatelydecomposed/tumblr_pcaffm6o6H1rc8keu?audio_file=https%3A%2F%2Fwww.tumblr.com%2Faudio_file%2Farticulatelydecomposed%2F176171225450%2Ftumblr_pcaffm6o6H1rc8keu&color=white&simple=1
// Still no bare .mp3 link in that iframe.
// data-stream-url="https://www.tumblr.com/audio_file/articulatelydecomposed/176171225450/tumblr_pcaffm6o6H1rc8keu" ?
// Indeed. Actually that might be in the iframe link. What's that function for decoding URIs?
// Wait, the actual file resolves to https://a.tumblr.com/tumblr_pcaffm6o6H1rc8keuo1.mp3 - this is trivial.
// let audio_name = s.split( '/' ).pop(); // Ignore text before last slash.
// audio_name = audio_name.split( '?' ).shift(); // Ignore text after a question mark, if there is one.
// audio_name should now look something like 'tumblr_12345abdce'.
// return 'https://a.tumblr.com/' + audio_name + '1.mp3#FUCK'; // Standard tumblr-wide video server. Hopefully.
// Inconsistent.
// e.g. http://articulatelydecomposed.tumblr.com/post/176307067285/mintchocolatechimp-written-by-reddit-user shows
// https://a.tumblr.com/tumblr_pce3snfthB1ufch4go1.mp3#offsite#link which works, but also
// https://a.tumblr.com/tumblr_pce3snfthB1ufch4go1.mp3&color=white&simple=1#offsite#link which doesn't and
// https://a.tumblr.com/tumblr_pce3snfthB1ufch4g1.mp3 which also doesn't.
// ... both of the 'go1' links are present even without this iframe-based code, because that audio has a 'download' link.
// Yeah, so much for the simple answer. Fetch. No URL processing seems necessary - 's' is already on this domain.
return fetch( s, { credentials: 'include' } ).then( r => r.text() ).then( text => {
// data-stream-url="https://a.tumblr.com/tumblr_pce3snfthB1ufch4go1.mp3"
let data_url = text.split( 'data-stream-url="' )[1]; // Drop everything before the file declaration.
data_url = data_url.split( '"' )[0]; // Drop everything after the doublequote. Probably not efficient, but fuck regexes.
return data_url + "#audio";
} )
// Alright, well now it sort of works, but sometimes it returns e.g.
// https://www.tumblr.com/audio_file/articulatelydecomposed/176010754365/tumblr_pc1pq1TsD31rc8keu#FRIG which resolves to
// https://a.tumblr.com/tumblr_pc1pq1TsD31rc8keuo1.mp3#FRIG
// Huh. That -is- the correctly-identified "data-stream-url" in some iframes. Shit.
// This is set up to handle multiple return values, right? For photosets? So I -could- return the correct URL and some guesses.
// But at that point - why not just guess from the iframe URL? +1.mp3 and also +o1.mp3.
// Can't just make up a username or post number for the /audio_file sort of URL.
}
return Promise.resolve( [s] ); // Else if this URL is singular, return a single element... resolved as a promise for Promise.all, in an array for Array.concat. Whee.
} ) )
.then( nested_array => { // Given the Promise.all'd array of resolved URLs and URL-arrays
return [].concat.apply( [], nested_array ); // Concatenate array of arrays - apply turns array into comma-separated list, concat turns CSL of arrays into a single array
} )
}
// Returns a URL with all the options_map options in ?key=value format - optionally allowing changes to options in the returned URL
// Valid uses:
// options_url() -> all current settings, no changes
// options_url( "name", number ) -> ?name=number
// options_url( "name", true ) -> ?name
// options_url( {name:number} ) -> ?name=number
// options_url( {name:number, other:true} ) -> ?name=number?other
// Note that simply passing "name" will remove ?name, not add it, because the value will evaluate false. I should probably change this? Eh, { key } without :value causes errors.
function options_url( key, value ) {
var copy_map = new Object();
for( var i in options_map ) { copy_map[ i ] = options_map[ i ]; }
// In any sensible language, this would read "copy_map = object_map." Javascript genuinely does not know how to copy objects. Fuck's sake.
if( typeof key === 'string' ) { // the parameters are optional. just calling options_url() will return e.g. example.tumblr.com/archive?ezastumblrscrape?startpage=1
if( !value ) { value = false; } // if there's no value then use false
copy_map[ key ] = value; // change this key, so we can e.g. link to example.tumblr.com/archive?ezastumblrscrape?startpage=2
}
else if( typeof key === 'object' ) { // If we're passed a hashmap
for( var i in key ) {
if( ! key[ i ] ) { key[ i ] = false; } // Turn any false evaluation into an explicit boolean - this might not be necessary
copy_map[ i ] = key[ i ]; // Press key-object values onto copy_map-object values
}
}
// Construct URL from options
var base_site = window.location.href.substring( 0, window.location.href.indexOf( "?" ) ); // should include /archive, but if not, it still works on most pages
for( var k in copy_map ) { // JS maps are weird. We're actually setting attributes of a generic object. So map[ "thumbnails" ] is the same as map.thumbnails.
if( copy_map[ k ] ) { // Unless the value is False, print a ?key=value pair.
base_site += "?" + k;
if( copy_map[ k ] !== true ) { base_site += "=" + copy_map[ k ]; } // If the value is boolean True, just print the value as a flag.
}
}
return base_site;
}