Fix rake

README.md

@@ -31,6 +31,4 @@ The rake task ```store_annotations``` has more functions:
   - these functions are idempotent: clipping images will only be fetched if they are not already present
     (otherwise it will output "No fetching [image]"),
     and the clippings.csv will always be regenerated from all the annotations, regardless of whether they are old or new.
-  - note that the rake task needs to look at the annotations as rendered in the site, so you must run ```jekyll build```
-    before running the rake task. Otherwise you will see errors like ```error: pathspec 'annotations/heresies_01' did not match any file(s) known to git.```
-    If this happens, just run ```jekyll build``` and then run the rake task again. (This requirement will be removed in a future release.)
+

Rakefile

@@ -1,12 +1,15 @@
 require 'fileutils'
 require 'json'
+require 'yaml'
 require 'csv'
 require 'slugify'
 require 'sanitize'
 require 'net/http'
 
 require 'byebug'
 
+require 'jekyll'
+
 task :default => [:store_annotations]
 
 task :store_annotations do
@@ -38,9 +41,14 @@ task :store_annotations do
       puts "Updating #{manifest}"
       update_manifest_copy(manifest)
     end
+  end
+  # build jekyll site to get annotations for clippings
+  site = Jekyll::Site.new(Jekyll.configuration({
+    "source" => ".",
+    "destination" => "_site"})).process
 
-    make_clippings(manifest)
-
+  manifests.each do | manifest |
+    make_clippings(manifest, site)
   end
 end
 
@@ -94,21 +102,27 @@ def update_manifest_copy(manifest)
   File.open("iiif/" + manifest + "/manifest.json", 'w+') { |f| f.write("---\nlayout: null\n---\n"+JSON.pretty_generate(manifest_json)) }
 end
 
-def make_clippings(manifest)
-
-  manifest_json = JSON.parse(File.read("iiif/" + manifest + "/manifest.json").gsub(/\A---(.|\n)*?---/, "").to_s)
+def make_clippings(manifest, site)
+
+  manifest_file = File.read("iiif/" + manifest + "/manifest.json").gsub(/\A---(.|\n)*?---/, "").to_s
+  manifest_json = JSON.parse(manifest_file)
 
+  # select canvases with annotations from manifest
   canvasesWithAnnos = manifest_json['sequences'][0]['canvases']
     .select { |canvas| canvas['otherContent'] }
     .select { |canvas| canvas['otherContent'][0]['@type'] == 'sc:AnnotationList' }
-  
+
   clippings = []
-  
+
   canvasesWithAnnos.each do |canvas|
     canvasID = canvas['@id']
     listpath = canvas['otherContent'][0]['@id'].gsub('{{ site.url }}{{ site.baseurl }}/', '')
-    list_json = JSON.parse(File.read('_site/' + listpath).to_s) #TODO remove dependence on generated _site
-
+    puts listpath
+    list_file = File.read('_site/'+listpath).to_s
+
+    # parse list file as JSON
+    list_json = JSON.parse(list_file)
+
     list_json['resources'].each do |resource|
       canvasOn = resource['on'][0]['full']
       next 'WTF canvas ID doesn\'t match' unless canvasID == canvasOn
@@ -120,42 +134,42 @@ def make_clippings(manifest)
       # build label and csv from specified data elements
       labelElements = []
       csvElements = {id: resource['@id'], item: manifest, canvas: canvasID}
-    
+
       canvasNum = canvasID.gsub(/.+\$([0-9]+)\/canvas.*/, '\1')
       labelElements << canvasNum
       csvElements[:canvasNum] = canvasNum
-    
+
       tagElements = []
       tags.each do |tag|
         labelElements << tag['chars']
         tagElements << tag['chars']
       end
       csvElements[:tags] = tagElements.join('|')
-    
+
       textElements = []
       texts.each do |text|
         # strip html markup
         longfilename = Sanitize.clean(text['chars']).strip
-        filename = longfilename.length > 180 ? longfilename[0..179] : longfilename
+        filename = longfilename.length > 100 ? longfilename[0..99] : longfilename # Edited to shorten further for working in deep paths
         labelElements << filename
         textElements << Sanitize.clean(text['chars']).strip
       end
       csvElements[:texts] = textElements.join('|')
-    
+
       labelElements << xywh
       csvElements[:xywh] = xywh
-    
+
       # label ends up like 1-photo-woman-with-film-camera-1235-134-1126-637
       label = labelElements.join(' ').slugify
-    
+
       imageRoot = canvas['images'][0]['resource']['service']['@id']
       clippingURL = imageRoot + '/' + xywh + '/full/0/default.jpg'
       csvElements[:clippingURL] = clippingURL
-    
+
       clippingsPath = 'clippings/' + manifest + '/' + canvasNum
       clippingImage = clippingsPath + '/' + label + '.jpg'
       csvElements[:clippingImage] = clippingImage
-    
+
       FileUtils.mkdir_p clippingsPath
       # fetch clipping image, if not already fetched
       if File.exist?(clippingImage)