embulk-filter-google_vision_api 0.1.0 → 0.2.1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 129cfa811e9b9b1a2e94004d2c32fc8c7008cc99
-  data.tar.gz: be9ea20997ac41ee0126b5b137c175defbeb2e7d
+  metadata.gz: 89b4d7b05b737af39f0a639df19f18eed4980666
+  data.tar.gz: d7f16dc85867f7eab0330587d7b54ca550f113c4
 SHA512:
-  metadata.gz: 0177f5ffe7ae5f07012ae1a934ba26b2bcbf87ef4dc3ea24946af7a6f3c33a1458fdde84e6066937ec5dde9d5f49c87e74d1b59aab1c1bd6ebebe17897c469f1
-  data.tar.gz: a11a2816eea10df168d759783f5c9d5b2b66071b178cb540bc153b870ccd777ce94f07d228ea63ad427e72aec3e162623da35b155bf9ad5d8631fdb3e1744b91
+  metadata.gz: 1e3a3df9c004434fe12d976cd8cf32521643e098598bce620c31d2afa370796744463b0e60032bb12386dcc307fb99156b0e8cd1b2be68f2e6349678d28d54e6
+  data.tar.gz: 3ceffe09715df5d3dd26b0bbcc2ea6e3a069080384b5f86a5e741223769894b90771ba954a262526f653f85d62d53fa4d3b6d744896a445de81653273084d22e

data/README.md

@@ -26,7 +26,7 @@ Very easy image recognition.
     - { image_path: 'http://www.embulk.org/docs/_images/embulk-logo.png' }
 ```
 
-* respond localfile path and http URI.
+* respond localfile path and http URI(http://〜) and GCS Image(gs://〜).
 
 ### setting
 ```yaml

data/embulk-filter-google_vision_api.gemspec

@@ -1,7 +1,7 @@
 
 Gem::Specification.new do |spec|
   spec.name          = "embulk-filter-google_vision_api"
-  spec.version       = "0.1.0"
+  spec.version       = "0.2.1"
   spec.authors       = ["toyama0919"]
   spec.summary       = "Google Vision Api filter plugin for Embulk"
   spec.description   = "Google Vision Api filter plugin for Embulk. Very easy image recognition."

data/lib/embulk/filter/google_vision_api.rb

@@ -1,16 +1,9 @@
-require "json"
-require "net/http"
-require "uri"
-require "openssl"
-require "base64"
-require "pp"
+require_relative "google_vision_api/vision_client"
 
 module Embulk
   module Filter
-
     class GoogleVisionApi < FilterPlugin
       Plugin.register_filter("google_vision_api", self)
-      ENDPOINT_PREFIX = "https://vision.googleapis.com/v1/images:annotate"
 
       def self.transaction(config, in_schema, &control)
         task = {
@@ -32,16 +25,10 @@ module Embulk
       end
 
       def init
-        @uri = URI.parse("#{ENDPOINT_PREFIX}?key=#{task['google_api_key']}")
-        @http = Net::HTTP.new(@uri.host, @uri.port)
-        @http.use_ssl = true
-        @http.verify_mode = OpenSSL::SSL::VERIFY_NONE
-        @post = Net::HTTP::Post.new(@uri.request_uri, initheader = {'Content-Type' =>'application/json'})
         @image_path_key_name = task['image_path_key_name']
-        @out_key_name = task['out_key_name']
         @delay = task['delay']
         @image_num_per_request = task['image_num_per_request']
-        @features = task['features']
+        @client = VisionClient.new(features: task['features'], google_api_key: task['google_api_key'])
       end
 
       def close
@@ -54,29 +41,14 @@ module Embulk
 
         record_groups.each do |records|
           requests = []
-          records.each do |record|
-            request = {
-              image: {},
-              features: @features
-            }
-            image_body = get_image_body(record)
-            request[:image][:content] = Base64.encode64(image_body)
-            requests << request
+          images = records.map do |record|
+            record[@image_path_key_name]
           end
-          body = {
-            requests: requests
-          }
-          @post.body = body.to_json
-          Embulk.logger.debug "request body => #{@post.body}"
 
-          response_hash = {}
-          @http.start do |h|
-            response = h.request(@post)
-            response_hash = JSON.parse(response.body)
-          end
+          response = @client.request(images)
           records.each_with_index do |record, i|
-            recognized = response_hash['responses'][i]
-            Embulk.logger.warn "Error image => [#{record[@image_path_key_name]}] #{recognized}" if recognized.key?("error")
+            recognized = response.key?("error") ? response : response['responses'][i]
+            Embulk.logger.warn "Error image => [#{record[@image_path_key_name]}] #{recognized}" if response.key?("error")
             page_builder.add(record.values + [recognized])
           end
 
@@ -87,16 +59,6 @@ module Embulk
       def finish
         page_builder.finish
       end
-
-      private
-      def get_image_body(record)
-        image_path = record[@image_path_key_name]
-        if image_path =~ /https?\:\/\//
-          Net::HTTP.get_response(URI.parse(image_path)).body rescue ""
-        else
-          File.read(image_path) rescue ""
-        end
-      end
     end
   end
 end

data/lib/embulk/filter/google_vision_api/vision_client.rb

@@ -0,0 +1,68 @@
+require "json"
+require "net/http"
+require "uri"
+require "openssl"
+require "base64"
+
+module Embulk
+  module Filter
+    class GoogleVisionApi < FilterPlugin
+      class VisionClient
+        ENDPOINT = "https://vision.googleapis.com/v1/images:annotate"
+        def initialize(features:, google_api_key:)
+          uri = URI.parse("#{ENDPOINT}?key=#{google_api_key}")
+          @http = Net::HTTP.new(uri.host, uri.port)
+          @http.use_ssl = true
+          @http.verify_mode = OpenSSL::SSL::VERIFY_NONE
+          @post = Net::HTTP::Post.new(uri.request_uri, initheader = {'Content-Type' =>'application/json'})
+          @features = features
+        end
+
+        def request(images)
+          @post.body = get_body(images).to_json
+          Embulk.logger.debug "request body => #{@post.body}"
+
+          @http.start do |h|
+            response = h.request(@post)
+            JSON.parse(response.body)
+          end
+        end
+
+        private
+        def get_body(images)
+          {
+            requests: get_requests(images)
+          }
+        end
+
+        def get_requests(images)
+          images.map do |image_path|
+            get_request(image_path)
+          end
+        end
+
+        def get_request(image_path)
+          request = {
+            image: Hash.new{|h,k| h[k] = {}},
+            features: @features
+          }
+          if image_path =~ /gs\:\/\//
+            request[:image][:source][:gcs_image_uri] = image_path
+          else
+            image_body = get_image_body(image_path)
+            request[:image][:content] = Base64.encode64(image_body)
+          end
+          request
+        end
+
+        def get_image_body(image_path)
+          if image_path =~ /https?\:\/\//
+            Net::HTTP.get_response(URI.parse(image_path)).body rescue ""
+          else
+            File.read(image_path) rescue ""
+          end
+        end
+      end
+    end
+  end
+end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: embulk-filter-google_vision_api
 version: !ruby/object:Gem::Version
-  version: 0.1.0
+  version: 0.2.1
 platform: ruby
 authors:
 - toyama0919
 autorequire:
 bindir: bin
 cert_chain: []
-date: 2017-01-16 00:00:00.000000000 Z
+date: 2017-01-26 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   requirement: !ruby/object:Gem::Requirement
@@ -81,6 +81,7 @@ files:
 - Rakefile
 - embulk-filter-google_vision_api.gemspec
 - lib/embulk/filter/google_vision_api.rb
+- lib/embulk/filter/google_vision_api/vision_client.rb
 homepage: https://github.com/toyama0919/embulk-filter-google_vision_api
 licenses:
 - MIT