Files
danbooru/app/logical/sources/strategies/deviant_art.rb
2020-06-16 00:10:37 -05:00

331 lines
13 KiB
Ruby

# Asset URLs:
#
## NORMALIZABLE
## * http://orig12.deviantart.net/9b69/f/2017/023/7/c/illustration___tokyo_encount_oei__by_melisaongmiqin-dawi58s.png
## * http://pre15.deviantart.net/81de/th/pre/f/2015/063/5/f/inha_by_inhaestudios-d8kfzm5.jpg
## * http://th00.deviantart.net/fs71/PRE/f/2014/065/3/b/goruto_by_xyelkiltrox-d797tit.png
##
## * http://fc00.deviantart.net/fs71/f/2013/234/d/8/d84e05f26f0695b1153e9dab3a962f16-d6j8jl9.jpg
## * http://th04.deviantart.net/fs71/PRE/f/2013/337/3/5/35081351f62b432f84eaeddeb4693caf-d6wlrqs.jpg
##
## * https://images-wixmp-ed30a86b8c4ca887773594c2.wixmp.com/intermediary/f/52c4a3ad-d416-42f0-90f6-570983e36797/dczr28f-bd255304-01bf-4765-8cd3-e53983d3f78a.jpg
## * https://images-wixmp-ed30a86b8c4ca887773594c2.wixmp.com/intermediary/f/8b472d70-a0d6-41b5-9a66-c35687090acc/d23jbr4-8a06af02-70cb-46da-8a96-42a6ba73cdb4.jpg/v1/fill/w_786,h_1017,q_70,strp/silverhawks_quicksilver_by_edsfox_d23jbr4-pre.jpg
## * https://images-wixmp-ed30a86b8c4ca887773594c2.wixmp.com/f/76098ac8-04ab-4784-b382-88ca082ba9b1/d9x7lmk-595099de-fe8f-48e5-9841-7254f9b2ab8d.png?token=eyJ0eXAiOiJKV1QiLCJhbGciOiJIUzI1NiJ9.eyJzdWIiOiJ1cm46YXBwOiIsImlzcyI6InVybjphcHA6Iiwib2JqIjpbW3sicGF0aCI6IlwvZlwvNzYwOThhYzgtMDRhYi00Nzg0LWIzODItODhjYTA4MmJhOWIxXC9kOXg3bG1rLTU5NTA5OWRlLWZlOGYtNDhlNS05ODQxLTcyNTRmOWIyYWI4ZC5wbmcifV1dLCJhdWQiOlsidXJuOnNlcnZpY2U6ZmlsZS5kb3dubG9hZCJdfQ.KFOVXAiF8MTlLb3oM-FlD0nnDvODmjqEhFYN5I2X5Bc
## * https://images-wixmp-ed30a86b8c4ca887773594c2.wixmp.com/f/fe7ab27f-7530-4252-99ef-2baaf81b36fd/dddf6pe-1a4a091c-768c-4395-9465-5d33899be1eb.png/v1/fill/w_800,h_1130,q_80,strp/stay_hydrated_and_in_the_shade_by_raikoart_dddf6pe-fullview.jpg?token=eyJ0eXAiOiJKV1QiLCJhbGciOiJIUzI1NiJ9.eyJzdWIiOiJ1cm46YXBwOjdlMGQxODg5ODIyNjQzNzNhNWYwZDQxNWVhMGQyNmUwIiwiaXNzIjoidXJuOmFwcDo3ZTBkMTg4OTgyMjY0MzczYTVmMGQ0MTVlYTBkMjZlMCIsIm9iaiI6W1t7ImhlaWdodCI6Ijw9MTEzMCIsInBhdGgiOiJcL2ZcL2ZlN2FiMjdmLTc1MzAtNDI1Mi05OWVmLTJiYWFmODFiMzZmZFwvZGRkZjZwZS0xYTRhMDkxYy03NjhjLTQzOTUtOTQ2NS01ZDMzODk5YmUxZWIucG5nIiwid2lkdGgiOiI8PTgwMCJ9XV0sImF1ZCI6WyJ1cm46c2VydmljZTppbWFnZS5vcGVyYXRpb25zIl19.J0W4k-iV6Mg8Kt_5Lr_L_JbBq4lyr7aCausWWJ_Fsbw
#
## * http://www.deviantart.com/download/135944599/Touhou___Suwako_Moriya_Colored_by_Turtle_Chibi.png
## * https://www.deviantart.com/download/549677536/countdown_to_midnight_by_kawacy-d939hwg.jpg?token=92090cd3910d52089b566661e8c2f749755ed5f8&ts=1438535525
#
## NOT NORMALIZABLE
## * http://th04.deviantart.net/fs70/300W/f/2009/364/4/d/Alphes_Mimic___Rika_by_Juriesute.png
## * http://fc02.deviantart.net/fs48/f/2009/186/2/c/Animation_by_epe_tohri.swf
## * http://fc08.deviantart.net/files/f/2007/120/c/9/Cool_Like_Me_by_47ness.jpg
##
## * http://fc08.deviantart.net/images3/i/2004/088/8/f/Blackrose_for_MuzicFreq.jpg
## * http://img04.deviantart.net/720b/i/2003/37/9/6/princess_peach.jpg
##
## * http://prnt00.deviantart.net/9b74/b/2016/101/4/468a9d89f52a835d4f6f1c8caca0dfb2-pnjfbh.jpg
## * http://other00.deviantart.net/8863/o/2009/197/3/7/37ac79eaeef9fb32e6ae998e9a77d8dd.jpg
## * http://fc09.deviantart.net/fs22/o/2009/197/3/7/37ac79eaeef9fb32e6ae998e9a77d8dd.jpg
## * http://pre06.deviantart.net/8497/th/pre/f/2009/173/c/c/cc9686111dcffffffb5fcfaf0cf069fb.jpg
#
########################
#
# Page URLs:
#
# * https://www.deviantart.com/noizave/art/test-post-please-ignore-685436408
# * https://noizave.deviantart.com/art/test-post-please-ignore-685436408
# * https://www.deviantart.com/deviation/685436408
# * https://fav.me/dbc3a48
#
# Profile URLs:
#
# * https://noizave.deviantart.com
# * https://www.deviantart.com/noizave
# * https://deviantart.com/noizave
module Sources
module Strategies
class DeviantArt < Base
ASSET_SUBDOMAINS = /(?:fc|th|pre|img|orig|origin-orig)\d*/i
RESERVED_SUBDOMAINS = %r{\Ahttps?://(?:#{ASSET_SUBDOMAINS}|www)\.}i
MAIN_DOMAIN = %r{\Ahttps?://(?:www\.)?deviantart.com}i
TITLE = /(?<title>[a-z0-9_-]+?)/i
ARTIST = /(?<artist>[a-z0-9_-]+?)/i
DEVIATION_ID = /(?<deviation_id>[0-9]+)/i
DA_FILENAME_1 = /[a-f0-9]{32}-d(?<base36_deviation_id>[a-z0-9]+)\./i
DA_FILENAME_2 = /#{TITLE}(?:_by_#{ARTIST}(?:-d(?<base36_deviation_id>[a-z0-9]+))?)?\./i
DA_FILENAME = Regexp.union(DA_FILENAME_1, DA_FILENAME_2)
WIX_FILENAME = %r{d(?<base36_deviation_id>[a-z0-9]+)[0-9a-f-]+\.\w+(?:/\w+/\w+/[\w,]+/(?<title>[\w-]+)_by_(?<artist>[\w-]+)_d\w+-\w+\.\w+)?.+}i
NOT_NORMALIZABLE_ASSET = %r{\Ahttps?://#{ASSET_SUBDOMAINS}\.deviantart\.net/.+/[0-9a-f]{32}(?:-[^d]\w+)?\.}i
DA_ASSET = %r{\Ahttps?://#{ASSET_SUBDOMAINS}\.deviantart\.net/.+/#{DA_FILENAME}}i
WIX_ASSET = %r{\Ahttps?://images-wixmp-ed30a86b8c4ca887773594c2\.wixmp\.com/(?:intermediary/)?\w/[0-9a-f-]+/#{WIX_FILENAME}}i
ASSET = Regexp.union(DA_ASSET, WIX_ASSET)
DA_DOWNLOAD = %r{#{MAIN_DOMAIN}/download/#{DEVIATION_ID}/#{DA_FILENAME_2}?}i
DEVIATION_ART = %r{#{MAIN_DOMAIN}/deviation/#{DEVIATION_ID}\z}i
PATH_ART = %r{#{MAIN_DOMAIN}/#{ARTIST}/art/#{TITLE}-#{DEVIATION_ID}\z}i
SUBDOMAIN_ART = %r{\Ahttps?://#{ARTIST}\.deviantart\.com/art/#{TITLE}-#{DEVIATION_ID}\z}i
PATH_PROFILE = %r{#{MAIN_DOMAIN}/#{ARTIST}/?\z}i
SUBDOMAIN_PROFILE = %r{\Ahttps?://#{ARTIST}\.deviantart\.com/?\z}i
FAVME = %r{\Ahttps?://(?:www\.)?fav\.me/d(?<base36_deviation_id>[a-z0-9]+)\z}i
def domains
["deviantart.net", "deviantart.com", "fav.me"]
end
def wixmp_hosts
["images-wixmp-ed30a86b8c4ca887773594c2.wixmp.com", "api-da.wixmp.com"]
end
def site_name
"Deviant Art"
end
def match?
return false if parsed_url.nil?
parsed_url.domain.in?(domains) || parsed_url.host.in?(wixmp_hosts)
end
def image_urls
[image_url]
end
def image_url
# work is private, deleted, or the url didn't contain a deviation id; use image url as given by user.
if api_deviation.blank?
url
elsif api_deviation[:is_downloadable]
api_download[:src]
elsif api_deviation[:flash].present?
api_deviation.dig(:flash, :src)
elsif api_deviation[:videos].present?
api_deviation[:videos].max_by { |x| x[:filesize] }[:src]
else
src = api_deviation.dig(:content, :src)
if deviation_id && deviation_id.to_i <= 790_677_560 && src =~ %r{\Ahttps://images-wixmp-} && src !~ /\.gif\?/
src = src.sub(%r{(/f/[a-f0-9-]+/[a-f0-9-]+)}, '/intermediary\1')
src = src.sub(%r{/v1/(fit|fill)/.*\z}i, "")
end
src = src.sub(%r{\Ahttps?://orig\d+\.deviantart\.net}i, "http://origin-orig.deviantart.net")
src = src.gsub(/q_\d+,strp/, "q_100")
src
end
end
def page_url
if api_deviation.present?
api_deviation[:url]
elsif deviation_id.present?
page_url_from_image_url
else
nil
end
end
def page_url_from_image_url
artist, title, id = artist_name_from_url, title_from_url, deviation_id
if artist.present? && title.present? && id.present?
"https://www.deviantart.com/#{artist}/art/#{title}-#{id}"
elsif id.present?
"https://www.deviantart.com/deviation/#{id}"
else
nil
end
end
def normalize_for_source
page_url_from_image_url
end
def profile_url
return nil if artist_name.blank?
"https://www.deviantart.com/#{artist_name.downcase}"
end
# Prefer the name from the url because the api metadata won't be present when
# the input url doesn't contain a deviation id, or the deviation is private or deleted.
def artist_name
if artist_name_from_url.present?
artist_name_from_url
elsif api_metadata.present?
api_metadata.dig(:author, :username)
else
nil
end
end
def artist_commentary_title
api_metadata[:title]
end
def artist_commentary_desc
api_metadata[:description]
end
def tags
if api_metadata.blank?
return []
end
api_metadata[:tags].map do |tag|
[tag[:tag_name], "https://www.deviantart.com/tag/#{tag[:tag_name]}"]
end
end
def dtext_artist_commentary_desc
DText.from_html(artist_commentary_desc) do |element|
# Convert embedded thumbnails of journal posts to 'deviantart #123'
# links. Strip embedded thumbnails of image posts. Example:
# https://sa-dui.deviantart.com/art/Commission-Meinos-Kaen-695905927.
if element.name == "a" && element["data-sigil"] == "thumb"
element.name = "span"
# <a href="https://sa-dui.deviantart.com/journal/About-Commissions-223178193" data-sigil="thumb" class="thumb lit" ...>
if element["class"].split.include?("lit")
deviation_id = element["href"][/-(\d+)\z/, 1].to_i
element.content = "deviantart ##{deviation_id}"
else
element.content = ""
end
end
if element.name == "a" && element["href"].present?
element["href"] = element["href"].gsub(%r{\Ahttps?://www\.deviantart\.com/users/outgoing\?}i, "")
# href may be missing the `http://` bit (ex: `inprnt.com`, `//inprnt.com`). Add it if missing.
uri = Addressable::URI.heuristic_parse(element["href"]) rescue nil
if uri.present? && uri.path.present?
uri.scheme ||= "http"
element["href"] = uri.to_s
end
end
end.gsub(/\A[[:space:]]+|[[:space:]]+\z/, "")
end
def self.deviation_id_from_url(url)
if url =~ NOT_NORMALIZABLE_ASSET
nil
elsif url =~ ASSET || url =~ FAVME
$~[:base36_deviation_id].try(:to_i, 36)
elsif url =~ PATH_ART || (url !~ RESERVED_SUBDOMAINS && url =~ SUBDOMAIN_ART) || url =~ DA_DOWNLOAD || url =~ DEVIATION_ART
$~[:deviation_id].to_i
else
nil
end
end
def self.artist_name_from_url(url)
if url =~ NOT_NORMALIZABLE_ASSET
nil
elsif url =~ ASSET || url =~ PATH_ART || url =~ PATH_PROFILE || url =~ DA_DOWNLOAD
$~[:artist].try(:dasherize)
elsif url !~ RESERVED_SUBDOMAINS && (url =~ SUBDOMAIN_ART || url =~ SUBDOMAIN_PROFILE)
$~[:artist]
else
nil
end
end
def self.title_from_url(url)
if url =~ NOT_NORMALIZABLE_ASSET
nil
elsif url =~ ASSET || url =~ PATH_ART || url =~ DA_DOWNLOAD
$~[:title].to_s.titleize.strip.squeeze(" ").tr(" ", "-").presence
elsif url !~ RESERVED_SUBDOMAINS && url =~ SUBDOMAIN_ART
$~[:title].to_s.titleize.strip.squeeze(" ").tr(" ", "-").presence
else
nil
end
end
def deviation_id
self.class.deviation_id_from_url(url) || self.class.deviation_id_from_url(referer_url)
end
def artist_name_from_url
self.class.artist_name_from_url(url) || self.class.artist_name_from_url(referer_url)
end
def title_from_url
self.class.title_from_url(url) || self.class.title_from_url(referer_url)
end
def page
return nil if page_url_from_image_url.blank?
resp = Danbooru::Http.cache(1.minute).get(page_url_from_image_url, follow: {max_hops: 1})
if resp.status.success?
Nokogiri::HTML(resp.body.to_s)
# the work was deleted
elsif resp.code == 404
nil
else
raise "failed to fetch page (got code #{resp.code})"
end
end
memoize :page
# Scrape UUID from <meta property="da:appurl" content="DeviantArt://deviation/12F08C5D-A3A4-338C-2F1A-7E4E268C0E8B">
# For hidden or deleted works the UUID will be nil.
def uuid
return nil if page.nil?
meta = page.at_css('meta[property="da:appurl"]')
return nil if meta.nil?
appurl = meta["content"]
uuid = appurl[%r{\ADeviantArt://deviation/(.*)\z}, 1]
uuid
end
memoize :uuid
def api_client
api_client = DeviantArtApiClient.new(
Danbooru.config.deviantart_client_id,
Danbooru.config.deviantart_client_secret
)
api_client.access_token = Cache.get("da-access-token", 11.weeks) do
api_client.access_token.to_hash
end
api_client
end
memoize :api_client
def api_deviation
return {} if uuid.nil?
api_client.deviation(uuid)
end
memoize :api_deviation
def api_metadata
return {} if uuid.nil?
api_client.metadata(uuid)[:metadata].first
end
memoize :api_metadata
def api_download
return {} unless uuid.present? && api_deviation[:is_downloadable]
api_client.download(uuid)
end
memoize :api_download
def api_response
{
deviation: api_deviation,
metadata: api_metadata,
download: api_download
}
end
end
end
end