transcribable 0.0.1
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- data/MIT-LICENSE +20 -0
- data/Rakefile +32 -0
- data/lib/generators/templates/assets/stylesheets/simple_frame.css +284 -0
- data/lib/generators/templates/config/documentcloud.yml +3 -0
- data/lib/generators/templates/controller.rb +48 -0
- data/lib/generators/templates/migration.rb +15 -0
- data/lib/generators/templates/migration_add_columns.rb +13 -0
- data/lib/generators/templates/model.rb +10 -0
- data/lib/generators/templates/views/_form.html.erb +32 -0
- data/lib/generators/templates/views/layouts/simple_frame.html.erb +50 -0
- data/lib/generators/templates/views/new.html.erb +28 -0
- data/lib/generators/transcribable_generator.rb +59 -0
- data/lib/tasks/harvester.rake +23 -0
- data/lib/transcribable.rb +156 -0
- data/lib/transcribable/railtie.rb +11 -0
- data/lib/transcribable/version.rb +3 -0
- data/test/dummy/README.rdoc +28 -0
- data/test/dummy/Rakefile +6 -0
- data/test/dummy/app/assets/javascripts/application.js +13 -0
- data/test/dummy/app/assets/stylesheets/application.css +13 -0
- data/test/dummy/app/controllers/application_controller.rb +5 -0
- data/test/dummy/app/helpers/application_helper.rb +2 -0
- data/test/dummy/app/models/filing.rb +6 -0
- data/test/dummy/app/models/transcription.rb +10 -0
- data/test/dummy/app/views/layouts/application.html.erb +14 -0
- data/test/dummy/bin/bundle +3 -0
- data/test/dummy/bin/rails +4 -0
- data/test/dummy/bin/rake +4 -0
- data/test/dummy/config.ru +4 -0
- data/test/dummy/config/application.rb +23 -0
- data/test/dummy/config/boot.rb +5 -0
- data/test/dummy/config/database.yml +25 -0
- data/test/dummy/config/environment.rb +5 -0
- data/test/dummy/config/environments/development.rb +29 -0
- data/test/dummy/config/environments/production.rb +80 -0
- data/test/dummy/config/environments/test.rb +36 -0
- data/test/dummy/config/initializers/backtrace_silencers.rb +7 -0
- data/test/dummy/config/initializers/filter_parameter_logging.rb +4 -0
- data/test/dummy/config/initializers/inflections.rb +16 -0
- data/test/dummy/config/initializers/mime_types.rb +5 -0
- data/test/dummy/config/initializers/secret_token.rb +12 -0
- data/test/dummy/config/initializers/session_store.rb +3 -0
- data/test/dummy/config/initializers/wrap_parameters.rb +14 -0
- data/test/dummy/config/locales/en.yml +23 -0
- data/test/dummy/config/routes.rb +56 -0
- data/test/dummy/db/schema.rb +19 -0
- data/test/dummy/db/test.sqlite3 +0 -0
- data/test/dummy/log/test.log +979 -0
- data/test/dummy/public/404.html +58 -0
- data/test/dummy/public/422.html +58 -0
- data/test/dummy/public/500.html +57 -0
- data/test/dummy/public/favicon.ico +0 -0
- data/test/test_helper.rb +15 -0
- data/test/transcribable_test.rb +72 -0
- metadata +202 -0
@@ -0,0 +1,50 @@
|
|
1
|
+
<!DOCTYPE html>
|
2
|
+
<html>
|
3
|
+
<head>
|
4
|
+
<meta http-equiv="Content-Type" content="text/html; charset=utf-8">
|
5
|
+
<meta name="viewport" content="width=device-width">
|
6
|
+
<%%= stylesheet_link_tag :simple_frame %>
|
7
|
+
<%%= yield :css %>
|
8
|
+
<!-- switchy -->
|
9
|
+
<meta name="viewport" content="width=device-width, maximum-scale=1.6">
|
10
|
+
<!-- /switchy -->
|
11
|
+
<script src="//s3.amazonaws.com/s3.documentcloud.org/viewer/loader.js"></script>
|
12
|
+
<%%= yield :javascripts %>
|
13
|
+
<%%= yield :header %>
|
14
|
+
<title><%%= yield :title %></title>
|
15
|
+
</head>
|
16
|
+
<body class="app" id="simple_wrapper">
|
17
|
+
<div id="banner">
|
18
|
+
<div class="wrapper">
|
19
|
+
<div id="banner-identity">
|
20
|
+
<p id="app-tagline">My Transcribable App</p>
|
21
|
+
</div>
|
22
|
+
</div>
|
23
|
+
</div>
|
24
|
+
|
25
|
+
<div id="banner-nav">
|
26
|
+
<div class="wrapper">
|
27
|
+
<div style="clear:both"></div>
|
28
|
+
</div>
|
29
|
+
</div>
|
30
|
+
|
31
|
+
<div id="content">
|
32
|
+
<div class="wrapper">
|
33
|
+
<div id="modal-screen-coverer"></div>
|
34
|
+
<div id="simple_frame">
|
35
|
+
<%%= yield %>
|
36
|
+
</div>
|
37
|
+
</div>
|
38
|
+
</div>
|
39
|
+
|
40
|
+
<div id="fb-root"></div>
|
41
|
+
<div id="footer">
|
42
|
+
<div class="wrapper">
|
43
|
+
<div id="footer-logo">
|
44
|
+
<p id="footer-logo-image">My Transcribable App</p>
|
45
|
+
</div>
|
46
|
+
<div style="clear:both;"></div>
|
47
|
+
</div>
|
48
|
+
</div>
|
49
|
+
</body>
|
50
|
+
</html>
|
@@ -0,0 +1,28 @@
|
|
1
|
+
<%% if flash[:alert] %>
|
2
|
+
<div class="warn"><%%= flash[:alert] %></div>
|
3
|
+
<%% end %>
|
4
|
+
<%% if flash[:notice] %>
|
5
|
+
<div class="notice"><%%= flash[:notice] %></div>
|
6
|
+
<%% end %>
|
7
|
+
|
8
|
+
<h1><%%= "Tell Us About This Document" %></h1>
|
9
|
+
<div id="left-dc-embed">
|
10
|
+
<div id="doc">
|
11
|
+
</div>
|
12
|
+
</div>
|
13
|
+
<div id="right-column">
|
14
|
+
<%%= render 'form' %>
|
15
|
+
</div>
|
16
|
+
|
17
|
+
<script>
|
18
|
+
var CURRENT_DOCUMENT = DV.load("https://documentcloud.org/documents/<%%= @<%= @table.singularize %>.dc_slug %>.js", {
|
19
|
+
container : '#doc',
|
20
|
+
width : 660,
|
21
|
+
height : 900,
|
22
|
+
zoom : 'auto',
|
23
|
+
showSidebar : false,
|
24
|
+
embedded : true,
|
25
|
+
showAnnotations : false
|
26
|
+
//afterLoad : setupViewer
|
27
|
+
});
|
28
|
+
</script>
|
@@ -0,0 +1,59 @@
|
|
1
|
+
require "rails/generators"
|
2
|
+
require "rails/generators/active_record"
|
3
|
+
require "transcribable"
|
4
|
+
|
5
|
+
|
6
|
+
class TranscribableGenerator < ActiveRecord::Generators::Base
|
7
|
+
include Transcribable
|
8
|
+
|
9
|
+
desc "Generates transcriptions table"
|
10
|
+
# to get around AR Generators requiring a NAME param
|
11
|
+
argument :name, type: :string, default: 'random_name'
|
12
|
+
|
13
|
+
source_root File.expand_path('../templates', __FILE__)
|
14
|
+
|
15
|
+
def table
|
16
|
+
@table = Transcribable.table
|
17
|
+
end
|
18
|
+
|
19
|
+
def transcribable_attrs
|
20
|
+
Transcribable.transcribable_attrs
|
21
|
+
end
|
22
|
+
|
23
|
+
def new_columns
|
24
|
+
Transcribable.new_columns
|
25
|
+
end
|
26
|
+
|
27
|
+
def copy_files
|
28
|
+
# Copies the migration template to db/migrate.
|
29
|
+
if Transcribable.new_columns.length > 0
|
30
|
+
@migration_name = "add_#{Transcribable.new_columns.keys.join("_and_")}_to_transcriptions"
|
31
|
+
migration_template 'migration_add_columns.rb', "db/migrate/#{@migration_name}.rb"
|
32
|
+
elsif !ActiveRecord::Base.connection.tables.include?("transcriptions")
|
33
|
+
migration_template 'migration.rb', 'db/migrate/create_transcriptions_table.rb'
|
34
|
+
end
|
35
|
+
|
36
|
+
# controller
|
37
|
+
template 'controller.rb', 'app/controllers/transcriptions_controller.rb'
|
38
|
+
|
39
|
+
# model
|
40
|
+
template 'model.rb', 'app/models/transcription.rb'
|
41
|
+
|
42
|
+
# views
|
43
|
+
template 'views/layouts/simple_frame.html.erb', 'app/views/layouts/simple_frame.html.erb'
|
44
|
+
template 'views/_form.html.erb', 'app/views/transcriptions/_form.html.erb'
|
45
|
+
template 'views/edit.html.erb', 'app/views/transcriptions/edit.html.erb'
|
46
|
+
template 'views/new.html.erb', 'app/views/transcriptions/new.html.erb'
|
47
|
+
|
48
|
+
# assets
|
49
|
+
template 'assets/stylesheets/simple_frame.css', 'app/assets/stylesheets/simple_frame.css'
|
50
|
+
|
51
|
+
# config
|
52
|
+
template 'config/documentcloud.yml', 'config/documentcloud.yml'
|
53
|
+
|
54
|
+
# routes
|
55
|
+
route "resources :transcriptions, :only => [:new, :create]"
|
56
|
+
route "resources :#{@table.to_sym}, :only => [:index, :show]"
|
57
|
+
route "root :to => \"#{@table}#index\""
|
58
|
+
end
|
59
|
+
end
|
@@ -0,0 +1,23 @@
|
|
1
|
+
namespace :transcribable do
|
2
|
+
desc "Harvest documents to transcribe from DocumentCloud"
|
3
|
+
task :harvest => :environment do
|
4
|
+
require 'rest-client'
|
5
|
+
klass = Kernel.const_get(Transcribable.table.classify)
|
6
|
+
dc = YAML.load(File.read("#{Rails.root.to_s}/config/documentcloud.yml"))
|
7
|
+
dc_project = JSON.parse(RestClient.get("https://#{CGI::escape(dc['email'])}:#{CGI::escape(dc['password'])}@www.documentcloud.org/api/projects.json"))
|
8
|
+
dc_project = dc_project['projects'].select {|q| q['id'] == dc['project'].scan(/^\d+/)[0].to_i }[0]
|
9
|
+
dc_project['document_ids'].each do |doc_id|
|
10
|
+
begin
|
11
|
+
dc_doc = JSON.parse(RestClient.get("https://www.documentcloud.org/api/documents/#{doc_id}.json"))['document']
|
12
|
+
# this will skip non-public documents
|
13
|
+
rescue RestClient::ResourceNotFound
|
14
|
+
next
|
15
|
+
end
|
16
|
+
obj = klass.find_or_initialize_by_url("https://www.documentcloud.org/documents/#{dc_doc['id']}")
|
17
|
+
# don't plow over verified docs if rerunning the script
|
18
|
+
obj.verified = false if obj.new_record?
|
19
|
+
obj.save
|
20
|
+
puts "== added #{obj.url}"
|
21
|
+
end
|
22
|
+
end
|
23
|
+
end
|
@@ -0,0 +1,156 @@
|
|
1
|
+
# adds to Filing
|
2
|
+
require 'uuid'
|
3
|
+
|
4
|
+
module Transcribable
|
5
|
+
require 'transcribable/railtie'
|
6
|
+
extend ActiveSupport::Concern
|
7
|
+
|
8
|
+
mattr_accessor :table
|
9
|
+
mattr_accessor :transcribable_attrs
|
10
|
+
|
11
|
+
included do
|
12
|
+
set_verification_threshhold
|
13
|
+
end
|
14
|
+
|
15
|
+
def self.table
|
16
|
+
# transcribable_attrs() "discovers" the columns
|
17
|
+
# and table. run it if we don't have a table yet.
|
18
|
+
transcribable_attrs if @@table.nil?
|
19
|
+
@@table
|
20
|
+
end
|
21
|
+
|
22
|
+
# creates a hash like...
|
23
|
+
# {'buyer' => :string, 'amount' => :integer}
|
24
|
+
def self.transcribable_attrs
|
25
|
+
return @@transcribable_attrs if @@transcribable_attrs
|
26
|
+
|
27
|
+
@@transcribable_attrs = {}
|
28
|
+
ActiveRecord::Base.connection.tables.reject {|t| t == "schema_migrations" }.each do |table|
|
29
|
+
klass = Kernel.const_get(table.classify)
|
30
|
+
klass.column_names.each do |col|
|
31
|
+
if klass.transcribable?(col)
|
32
|
+
@@table = table
|
33
|
+
@@transcribable_attrs[col] = Filing.columns_hash[col].type
|
34
|
+
end
|
35
|
+
end
|
36
|
+
end
|
37
|
+
|
38
|
+
@@transcribable_attrs
|
39
|
+
end
|
40
|
+
|
41
|
+
# If we've migrated our master table,
|
42
|
+
# and need transcriptions to catch up.
|
43
|
+
# Returns a hash like transcribable_attrs
|
44
|
+
def self.new_columns
|
45
|
+
cols = Transcribable.transcribable_attrs.keys - Transcription.columns_hash.keys
|
46
|
+
cols.reduce(Hash.new(0)) do |memo, it|
|
47
|
+
memo[it] = Kernel.const_get(@@table.classify).columns_hash[it].type
|
48
|
+
memo
|
49
|
+
end
|
50
|
+
end
|
51
|
+
|
52
|
+
module ClassMethods
|
53
|
+
def transcribable(*args)
|
54
|
+
args.each do |k|
|
55
|
+
self.columns_hash[k.to_s].instance_variable_set("@transcribable", true)
|
56
|
+
end
|
57
|
+
include Transcribable::LocalInstanceMethods
|
58
|
+
end
|
59
|
+
|
60
|
+
def transcribable?(_attr)
|
61
|
+
self.columns_hash[_attr].instance_variable_get("@transcribable")
|
62
|
+
end
|
63
|
+
|
64
|
+
def skip_verification(*args)
|
65
|
+
@@skip_verification = args
|
66
|
+
end
|
67
|
+
|
68
|
+
def skipped_attrs
|
69
|
+
@@skip_verification
|
70
|
+
end
|
71
|
+
|
72
|
+
# The number over which people must agree
|
73
|
+
# on every attribute to verify a transcription
|
74
|
+
def set_verification_threshhold(lvl = 1)
|
75
|
+
@@verification_threshhold = lvl
|
76
|
+
end
|
77
|
+
|
78
|
+
def verification_threshhold
|
79
|
+
@@verification_threshhold
|
80
|
+
end
|
81
|
+
|
82
|
+
# Attributes that are potential reasons
|
83
|
+
# to skip a transcription. If enough people
|
84
|
+
# agree to skip, the filing will be marked transcribed.
|
85
|
+
def skip_transcription(*args)
|
86
|
+
@@skippable = args
|
87
|
+
end
|
88
|
+
|
89
|
+
# Override this to write your own assigner
|
90
|
+
# By default, it picks a random filing that
|
91
|
+
# a user has not transcribed. If there's nothing left
|
92
|
+
# for that user to do, it returns nil. This will get slower
|
93
|
+
# the more transcriptions you have, so it'll be a good idea
|
94
|
+
# to index the filing_id (or whatever master table foreign key)
|
95
|
+
# column in your transcriptions table.
|
96
|
+
def assign!(user_id)
|
97
|
+
user_transcribed_filings = Transcription.where(:user_id => user_id).map {|q| q["#{self.table_name.downcase.singularize}_id".to_sym] }.uniq
|
98
|
+
filings = self.where(:verified => [nil, false])
|
99
|
+
if user_transcribed_filings.length > 0
|
100
|
+
filings = filings.where("id NOT IN (?)", user_transcribed_filings)
|
101
|
+
end
|
102
|
+
pick = rand(filings.length - 1)
|
103
|
+
filings.length > 0 ? filings[pick] : nil
|
104
|
+
end
|
105
|
+
end
|
106
|
+
|
107
|
+
module LocalInstanceMethods
|
108
|
+
# Override this to create your own verifier.
|
109
|
+
# By default, all "transcribable" attributes
|
110
|
+
# need to be agreed on by @@verification_threshhold people.
|
111
|
+
def verify!
|
112
|
+
chosen = {}
|
113
|
+
|
114
|
+
attributes = Transcribable.transcribable_attrs.keys.reject do |k|
|
115
|
+
self.class.skipped_attrs.include?(k.to_sym)
|
116
|
+
end
|
117
|
+
|
118
|
+
Rails.logger.info("== Verifying #{attributes.join(", ")}")
|
119
|
+
|
120
|
+
aggregate = transcriptions.reduce({}) do |memo, it|
|
121
|
+
attributes.each do |attribute|
|
122
|
+
memo[attribute] = memo[attribute] ? memo[attribute] : {}
|
123
|
+
memo[attribute][it.instance_values['attributes'][attribute].to_s.upcase] = memo[attribute][it.instance_values['attributes'][attribute].to_s.upcase] ?
|
124
|
+
memo[attribute][it.instance_values['attributes'][attribute].to_s.upcase] + 1 : 1
|
125
|
+
end
|
126
|
+
|
127
|
+
memo
|
128
|
+
end
|
129
|
+
|
130
|
+
aggregate.each do |attribute, answers|
|
131
|
+
answers.each do |answer, answer_ct|
|
132
|
+
if answer_ct > self.class.verification_threshhold
|
133
|
+
chosen[attribute] = answers.each.max_by {|k,v| v}.first
|
134
|
+
end
|
135
|
+
end
|
136
|
+
end
|
137
|
+
|
138
|
+
if chosen.keys.length == attributes.length
|
139
|
+
attributes.each do |a|
|
140
|
+
self[a] = chosen[a]
|
141
|
+
end
|
142
|
+
self.verified = true
|
143
|
+
self.save
|
144
|
+
Rails.logger.info("== Verified #{self.url}")
|
145
|
+
else
|
146
|
+
Rails.logger.info("== Not verified: #{self.url}")
|
147
|
+
end
|
148
|
+
end
|
149
|
+
|
150
|
+
def dc_slug
|
151
|
+
url.split(/\//)[-1]
|
152
|
+
end
|
153
|
+
end
|
154
|
+
end
|
155
|
+
|
156
|
+
ActiveRecord::Base.send :include, Transcribable
|
@@ -0,0 +1,28 @@
|
|
1
|
+
== README
|
2
|
+
|
3
|
+
This README would normally document whatever steps are necessary to get the
|
4
|
+
application up and running.
|
5
|
+
|
6
|
+
Things you may want to cover:
|
7
|
+
|
8
|
+
* Ruby version
|
9
|
+
|
10
|
+
* System dependencies
|
11
|
+
|
12
|
+
* Configuration
|
13
|
+
|
14
|
+
* Database creation
|
15
|
+
|
16
|
+
* Database initialization
|
17
|
+
|
18
|
+
* How to run the test suite
|
19
|
+
|
20
|
+
* Services (job queues, cache servers, search engines, etc.)
|
21
|
+
|
22
|
+
* Deployment instructions
|
23
|
+
|
24
|
+
* ...
|
25
|
+
|
26
|
+
|
27
|
+
Please feel free to use a different markup language if you do not plan to run
|
28
|
+
<tt>rake doc:app</tt>.
|
data/test/dummy/Rakefile
ADDED
@@ -0,0 +1,13 @@
|
|
1
|
+
// This is a manifest file that'll be compiled into application.js, which will include all the files
|
2
|
+
// listed below.
|
3
|
+
//
|
4
|
+
// Any JavaScript/Coffee file within this directory, lib/assets/javascripts, vendor/assets/javascripts,
|
5
|
+
// or vendor/assets/javascripts of plugins, if any, can be referenced here using a relative path.
|
6
|
+
//
|
7
|
+
// It's not advisable to add code directly here, but if you do, it'll appear at the bottom of the
|
8
|
+
// compiled file.
|
9
|
+
//
|
10
|
+
// Read Sprockets README (https://github.com/sstephenson/sprockets#sprockets-directives) for details
|
11
|
+
// about supported directives.
|
12
|
+
//
|
13
|
+
//= require_tree .
|
@@ -0,0 +1,13 @@
|
|
1
|
+
/*
|
2
|
+
* This is a manifest file that'll be compiled into application.css, which will include all the files
|
3
|
+
* listed below.
|
4
|
+
*
|
5
|
+
* Any CSS and SCSS file within this directory, lib/assets/stylesheets, vendor/assets/stylesheets,
|
6
|
+
* or vendor/assets/stylesheets of plugins, if any, can be referenced here using a relative path.
|
7
|
+
*
|
8
|
+
* You're free to add application-wide styles to this file and they'll appear at the top of the
|
9
|
+
* compiled file, but it's generally better to create a new file per style scope.
|
10
|
+
*
|
11
|
+
*= require_self
|
12
|
+
*= require_tree .
|
13
|
+
*/
|
@@ -0,0 +1,10 @@
|
|
1
|
+
class Transcription < ActiveRecord::Base
|
2
|
+
belongs_to :user
|
3
|
+
belongs_to :filing
|
4
|
+
|
5
|
+
Transcribable.transcribable_attrs.select {|k,v| [:integer, :float, :decimal].include?(v) }.keys.each do |k|
|
6
|
+
define_method("#{k}=") do |val|
|
7
|
+
write_attribute k.to_sym, val.to_s.gsub(/[^0-9\.]/,"").to_f
|
8
|
+
end
|
9
|
+
end
|
10
|
+
end
|
@@ -0,0 +1,14 @@
|
|
1
|
+
<!DOCTYPE html>
|
2
|
+
<html>
|
3
|
+
<head>
|
4
|
+
<title>Dummy</title>
|
5
|
+
<%= stylesheet_link_tag "application", media: "all", "data-turbolinks-track" => true %>
|
6
|
+
<%= javascript_include_tag "application", "data-turbolinks-track" => true %>
|
7
|
+
<%= csrf_meta_tags %>
|
8
|
+
</head>
|
9
|
+
<body>
|
10
|
+
|
11
|
+
<%= yield %>
|
12
|
+
|
13
|
+
</body>
|
14
|
+
</html>
|