Milk processing companies in Uganda, Brachycerini & You've Got a Lot to Answer For

README.org

@@ -1,6 +1,6 @@
 # -*- mode: org; coding: utf-8; -*-
 
-* S.P.A.C.E (working title)
+* Superject (working title)
 
 #+BEGIN_QUOTE
 A proposition, for Whitehead, is neither true nor false in itself. It is a lure for feeling the world that “might be.” A proposition is what places a concrescing subject (an occasion of experience) in relation to a potential predicative pattern. It is the potential togetherness of subject and predicate that characterizes a proposition and not its actual exemplification. But propositions do not hang in the thin air of abstraction; they always emerge from somewhere and are entertained by an actual subject within a particular milieu. The proposition’s relevance and importance (or its “truth”—in William James’ sense) is contingent upon the world that it finds itself within.
@@ -12,13 +12,14 @@ Various notes can be found in the ~notes/~ folder
 
 * propositions (cards)
 
-Card design can be found in the ~cards/~ folder and rendered cards can be found in the  ~cards/_ouput~ folder. Ways of using the cards can be found in file:Process&Reality.org
+Card design can be found in the ~cards/~ folder and rendered cards can be found in the  ~cards/_ouput~ folder. Ways of using the cards can be found in [[file:Process and Reality.org]]
 
  - testing deck creating using [[https://github.com/andymeneely/squib][squib]]
  - edit card descriptions in the ~cards.txt~ files
  - change layout in the ~layout.yml~ file(s)
  - change design logic in the ~deck.rb~ file
 
+#+ATTR_org: :width 600px
 #+CAPTION: cards (example)
 [[file:cards/img/_example.png]]
 

cards/deck-neural.rb

@@ -0,0 +1,200 @@
+#!/usr/bin/env ruby -W0
+
+# aprox. running time
+# ./deck-neural.rb  144.08s user 10.49s system 28% cpu 8:55.15 total
+# ./deck-neural.rb  155.95s user 10.38s system 30% cpu 9:00.29 total
+# cards-36.csv
+# ./deck-neural.rb  30.74s user 2.63s system 105% cpu 31.507 total
+# 50 cards
+# ./deck-neural.rb  76.23s user 6.31s system 132% cpu 1:02.50 total
+
+require 'squib'
+require 'rest-client'
+require 'open-uri'
+require 'json'
+require 'digest'
+require 'mini_magick'
+require 'unsplash'
+require 'pexels'
+
+# cards from local csv and images vis text2image GAN
+# https://deepai.org/machine-learning-model/text2img
+$deepai_api_key = "59b70a3d-d42a-42be-979d-d09ec08ec7e0"
+
+$pexels_api_key = "563492ad6f917000010000016e741b30a565426c90e5b65b6b34063b"
+
+# prod the GAN in a potentially relevant direction with a keyword or phrase prefix
+$keyword = ""
+
+# where to get images? try "t2i" or "unsplash" or "pexels"
+$img_source = "unsplash"
+
+# cards descriptions
+#$cards_csv = 'cards-p50-1.csv'    # 23:02
+#$cards_csv = 'cards-p50-2.csv'    # 21:24
+#$cards_csv = 'cards-p50-3.csv'    # 22:52
+
+# unsplash config
+Unsplash.configure do |config|
+  config.application_access_key = "FWIsA-zAhsj3biPbVH3-H6qoeB30EVazjQpIjlrnaoE"
+  config.application_secret = "Qcr0ETekpsPVMsGBRSmHSsF7NvAnqWkiZmKw9KP9Ros"
+  config.application_redirect_uri = "urn:ietf:wg:oauth:2.0:oob"
+  config.utm_source = "pandemic"
+end
+
+# cards descriptions input from csv 
+data = Squib.csv file: $cards_csv
+
+# text2image slurp
+
+def get_img_t2i (phrase)
+  puts "test2img> #{$keyword} #{phrase}"
+  d0 = Digest::RMD160.hexdigest phrase
+  r0 = RestClient::Request.execute(method: :post, url: 'https://api.deepai.org/api/text2img', timeout: 600,
+                                  headers: {'api-key' => $deepai_api_key },
+                                  payload: { 'text' => "#{$keyword} #{$phrase}" } )
+  f0 = "img/#{d0}.png"
+  t0 = MiniMagick::Image.open(JSON.parse(r0)['output_url'])
+  t0.format('png').write(f0)
+  puts "wrote: " + f0
+  return f0
+end
+
+# unsplash
+# - https://github.com/unsplash/unsplash_rb
+# - https://www.rubydoc.info/github/unsplash/unsplash_rb/
+
+def get_img_unsplash (phrase)
+  puts "unsplash> #{$keyword} #{phrase}"
+  d0 = Digest::RMD160.hexdigest phrase
+  r0 = Unsplash::Photo.random(query: "#{$keyword} #{phrase}", orientation: "squarish")
+
+  if true
+      f0 = "img/#{d0}.png"
+      t0 = MiniMagick::Image.open(r0.urls.regular)
+      t0.format('png').write(f0)
+      puts "wrote: " + f0
+      return f0
+  else
+    p "couldn't find an image for #{$keyword} #{phrase}..."
+    return "img/blank.png"
+  end
+  rescue
+    p "wonk? err..."
+    return "img/blank.png"
+end
+
+# pexels
+$pxl = Pexels::Client.new($pexels_api_key)
+
+def get_img_pexels (phrase)
+  puts "pexels> #{$keyword} #{phrase}"
+  d0 = Digest::RMD160.hexdigest phrase
+  r0 = $pxl.photos.search("#{$keyword} #{phrase}", per_page: 1)
+  f0 = "img/#{d0}.png"
+  if !r0.photos.empty?
+    r1 = r0.photos[0].src["portrait"] 
+    t0 = MiniMagick::Image.open(r1)
+    t0.format('png').write(f0)
+    p "wrote: " + f0
+    return f0
+  else
+    p "couldn't find an image for #{$keyword} #{phrase}..."
+    return "img/blank.png"
+  end
+end
+
+# other options for automated image collation
+# - https://pixabay.com/api/docs/#api_search_images
+# - https://github.com/public-apis/public-apis#photography
+# - https://en.wikipedia.org/wiki/Wikipedia:Public_domain_image_resources
+# - https://commons.wikimedia.org/wiki/Category:Images
+# - https://rapidapi.com/collection/image-apis
+
+#layouts = ['economy.yml', 'space-layout-1.yml']
+layouts = ['space-layout-3.yml']
+
+c1 = '#9f68ed'
+c2 = '#ed9f68'
+c3 = '#b5ed68'
+c4 = '#eeeeee'
+
+# conditional coloring
+#     https://squib.readthedocs.io/en/v0.15.0/colors.html?highlight=color#samples
+
+color = c1
+
+Squib::Deck.new cards: data['name'].size, width: '38mm', height: '60mm', layout: layouts do
+
+  # set background colour per card type
+  bg = data['type'].map do |t|
+    if (t.eql? 'O')
+      c1
+    elsif (t.eql? 'Q')
+      c2
+    elsif (t.eql? 'Q')
+      c3      
+    else
+      c4
+    end
+  end
+  
+  # set fill colour per card type
+  fg = data['type'].map do |t|
+   if (t.eql? 'O')
+      c2
+    elsif (t.eql? 'Q')
+      c1
+    elsif (t.eql? 'Q')
+      c4  
+    else
+      c4
+   end
+  end
+
+  # associative image search and/or generation
+  #  https://experiments.runwayml.com/generative_engine/
+  #  https://deepai.org/machine-learning-model/text2img
+
+  # generate & collect images
+  case $img_source
+  when "t2i"
+    dlimages = data['name'].map { |i| get_img_t2i i}
+  when "unsplash"
+    dlimages = data['name'].map { |i| get_img_unsplash i}
+  when "pexels"
+    dlimages = data['name'].map { |i| get_img_pexels i}
+  else
+    puts "unknown image source " + $img_source
+  end
+  
+  # sharing & colours  
+  background color: bg 
+  rect fill_color: fg , x: '3mm', y: '3mm', width: '32mm', height: '54mm', radius: 32
+    
+  #rect layout: 'cut'   # cut line as defined by TheGameCrafter in economy.yml
+  #rect layout: 'safe' # safe zone as defined by TheGameCrafter in economy.yml
+
+  # in z-order low->high 
+  text str: data['text'], layout: 'description'
+  png file: dlimages, layout: 'illustration'
+  png file: dlimages, layout: 'overlay'
+  text str: data['type'], layout: 'type'
+  text str: data['name'], layout: 'title'
+  
+  # output a png of each card 
+  tag = Time.now.strftime('%Y%m%d_%H%M')
+
+  save format: :png, prefix: "cards_#{$keyword}_#{tag}_#{$img_source}_", count_format: '%d'
+
+  #.. and pdf of the deck on A4
+  # see also https://squib.readthedocs.io/en/v0.15.0/sprues.html#list-of-sprues
+  save format: :pdf, file: "deck_#{$keyword}_#{tag}_#{$img_source}.pdf",
+       width: '297mm', height: '210mm', margin: '20mm',
+       crop_marks: true,
+       crop_stroke_width: 0.5
+
+  #save_png
+  #save_pdf
+  
+end