indico 0.10.3 → 0.10.4

data/lib/indico.rb

@@ -1,6 +1,7 @@
 require 'indico/version'
 require 'indico/helper'
 require 'indico/image'
+require 'indico/pdf'
 require 'indico/multi'
 require 'indico/settings'
 require 'indico/errors'
@@ -127,6 +128,10 @@ module Indico
     api_handler(text, "summarization", config)
   end
 
+  def self.pdf_extraction(pdf, config = {})
+    api_handler(preprocess_pdf(pdf), "pdfextraction", config)
+  end
+
   def self.fer(image, config = nil)
     size = (config != nil and config["detect"] == true) ? false : 48
     api_handler(preprocess(image, size, false), 'fer', config)

data/lib/indico/pdf.rb

@@ -0,0 +1,26 @@
+require 'base64'
+
+module Indico
+    def self.preprocess_pdf(pdf)
+        if pdf.class == Array
+            # Batch Request
+            pdf_array = Array.new
+
+            # process each image
+            pdf.each do |_pdf|
+                pdf_array.push(preprocess_pdf(_pdf))
+            end
+
+            return pdf_array
+        elsif pdf.class != String
+            raise Exception.new("PDF input must be filename, url or base64 string")
+        end
+
+        begin
+            return Base64.encode64(File.read(pdf))
+        rescue
+            # likely a url or a base64 encoded string already
+            return pdf
+        end
+    end
+end

data/lib/indico/version.rb

@@ -1,3 +1,3 @@
 module Indico
-  VERSION = '0.10.3'
+  VERSION = '0.10.4'
 end

data/spec/pdf_extraction_spec.rb

@@ -0,0 +1,43 @@
+require 'spec_helper'
+
+describe Indico do
+    before do
+        api_key = ENV['INDICO_API_KEY']
+        private_cloud = 'indico-test'
+        @config = { api_key: api_key, cloud: private_cloud}
+    end
+
+    it 'should properly handle pdf urls' do
+        pdf_url = "https://s3-us-west-2.amazonaws.com/indico-test-data/test.pdf"
+
+        response = Indico.pdf_extraction(pdf_url, @config)
+        expected_keys = Set.new(%w(metadata text))
+
+        expect(Set.new(response.keys)).to eql(expected_keys)
+    end
+
+    it 'should properly handle local pdf files' do
+        pdf_path = File.expand_path(
+            File.join(File.dirname(__FILE__), "data", "test.pdf")
+        )
+
+        response = Indico.pdf_extraction(pdf_path, @config)
+        expected_keys = Set.new(%w(metadata text))
+
+        expect(Set.new(response.keys)).to eql(expected_keys)
+    end
+
+    it 'should properly handle array of local pdf files' do
+        pdf_path = File.expand_path(
+            File.join(File.dirname(__FILE__), "data", "test.pdf")
+        )
+        arr = Array.new()
+        arr.push(pdf_path)
+        arr.push(pdf_path)
+
+        response = Indico.pdf_extraction(arr, @config)
+        expected_keys = Set.new(%w(metadata text))
+
+        expect(Set.new(response[0].keys)).to eql(expected_keys)
+    end
+end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: indico
 version: !ruby/object:Gem::Version
-  version: 0.10.3
+  version: 0.10.4
   prerelease: 
 platform: ruby
 authors:
@@ -12,7 +12,7 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2016-10-19 00:00:00.000000000 Z
+date: 2016-12-16 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: inifile
@@ -117,6 +117,7 @@ files:
 - lib/indico/helper.rb
 - lib/indico/image.rb
 - lib/indico/multi.rb
+- lib/indico/pdf.rb
 - lib/indico/settings.rb
 - lib/indico/version.rb
 - spec/config/.indicorc.test
@@ -125,10 +126,12 @@ files:
 - spec/data/dog.jpg
 - spec/data/happy.png
 - spec/data/happy64.txt
+- spec/data/test.pdf
 - spec/imagerecognition_spec.rb
 - spec/indico_batch_spec.rb
 - spec/indico_spec.rb
 - spec/keywords_v2_spec.rb
+- spec/pdf_extraction_spec.rb
 - spec/political_spec.rb
 - spec/settings_spec.rb
 - spec/spec_helper.rb