首页 探索 帮助
注册 登录
J1X
/
huginn
关注 2
点赞 0
派生 0

浏览代码
Allow error codes, such as 404, to be considered successful, potential solution for #1272

George Opritescu 9 年之前
父节点
d6410833b7
当前提交
6232189688
共有 2 个文件被更改，包括 79 次插入 和 1 次删除
显示文件统计
  1. 30 1
      app/models/agents/website_agent.rb
  2. 49 0
      spec/models/agents/website_agent_spec.rb

+ 30 - 1
app/models/agents/website_agent.rb
查看文件 

@@ -103,6 +103,8 @@ module Agents
103 103
104 104 
       Set `unzip` to `gzip` to inflate the resource using gzip.
105 105
106 
+      Set `consider_http_error_success` to an array of ints, ex: `[404]` to consider also 404 as successes, and to scrape it.
107 
+
106 108 
       # Liquid Templating
107 109
108 110 
       In Liquid templating, the following variable is available:
 
@@ -149,6 +151,7 @@ module Agents
149 151 
       errors.add(:base, "either url, url_from_event, or data_from_event are required") unless options['url'].present? || options['url_from_event'].present? || options['data_from_event'].present?
150 152 
       errors.add(:base, "expected_update_period_in_days is required") unless options['expected_update_period_in_days'].present?
151 153 
       validate_extract_options!
154 
+      validate_consider_http_success_option!
152 155
153 156 
       # Check for optional fields
154 157 
       if options['mode'].present?
 
@@ -166,6 +169,27 @@ module Agents
166 169 
       validate_web_request_options!
167 170 
     end
168 171
172 
+    def validate_consider_http_success_option!
173 
+      consider_success = options["consider_http_error_success"]
174 
+      if consider_success != nil
175 
+
176 
+        if consider_success.class != Array
177 
+          errors.add(:base,"Must be an array and specify at least one status code")
178 
+        else
179 
+          if consider_success.uniq.count != consider_success.count
180 
+            errors.add(:base,"Duplicate http code found")
181 
+          else
182 
+            if consider_success.map(&:class).uniq != [Fixnum]
183 
+              errors.add(:base,"Please make sure to use only integer values for code")
184 
+            else
185 
+              @error_codes_considered_success = consider_success
186 
+            end
187 
+          end
188 
+        end
189 
+
190 
+      end
191 
+    end
192 
+
169 193 
     def validate_extract_options!
170 194 
       extraction_type = (extraction_type() rescue extraction_type(options))
171 195 
       case extract = options['extract']
 
@@ -273,7 +297,7 @@ module Agents
273 297 
       uri = Utils.normalize_uri(url)
274 298 
       log "Fetching #{uri}"
275 299 
       response = faraday.get(uri)
276 
-      raise "Failed: #{response.inspect}" unless response.success?
300 
+      raise "Failed: #{response.inspect}" unless consider_response_successful?(response)
277 301
278 302 
       interpolation_context.stack {
279 303 
         interpolation_context['_response_'] = ResponseDrop.new(response)
 
@@ -353,6 +377,11 @@ module Agents
353 377 
     end
354 378
355 379 
     private
380 
+    def consider_response_successful?(response)
381 
+      response.success? || begin
382 
+        @error_codes_considered_success.present? && @error_codes_considered_success.include?(response.status)
383 
+      end
384 
+    end
356 385
357 386 
     def handle_event_data(data, event, existing_payload)
358 387 
       handle_data(data, event.payload['url'], existing_payload)

+ 49 - 0
spec/models/agents/website_agent_spec.rb
查看文件 

@@ -40,6 +40,23 @@ describe Agents::WebsiteAgent do
40 40 
         expect(@checker).not_to be_valid
41 41 
       end
42 42
43 
+      it 'should validate the consider_http_error_success fields' do
44 
+        @checker.options['consider_http_error_success'] = [404]
45 
+        expect(@checker).to be_valid
46 
+
47 
+        @checker.options['consider_http_error_success'] = [404, 404]
48 
+        expect(@checker).not_to be_valid
49 
+
50 
+        @checker.options['consider_http_error_success'] = [404.0]
51 
+        expect(@checker).not_to be_valid
52 
+
53 
+        @checker.options['consider_http_error_success'] = ["not_a_code"]
54 
+        expect(@checker).not_to be_valid
55 
+
56 
+        @checker.options['consider_http_error_success'] = []
57 
+        expect(@checker).not_to be_valid
58 
+      end
59 
+
43 60 
       it "should validate uniqueness_look_back" do
44 61 
         @checker.options['uniqueness_look_back'] = "nonsense"
45 62 
         expect(@checker).not_to be_valid
 
@@ -169,6 +186,38 @@ describe Agents::WebsiteAgent do
169 186 
       end
170 187 
     end
171 188
189 
+    describe 'consider_http_error_success' do
190 
+      it 'should allow scraping from a 404 result' do
191 
+        json = {
192 
+          'response' => {
193 
+            'version' => 2,
194 
+            'title' => "hello!"
195 
+          }
196 
+        }
197 
+        zipped = ActiveSupport::Gzip.compress(json.to_json)
198 
+        stub_request(:any, /gzip/).to_return(body: zipped, headers: { 'Content-Encoding' => 'gzip' }, status: 404)
199 
+        site = {
200 
+          'name' => "Some JSON Response",
201 
+          'expected_update_period_in_days' => "2",
202 
+          'type' => "json",
203 
+          'url' => "http://gzip.com",
204 
+          'mode' => 'on_change',
205 
+          'consider_http_error_success': [404],
206 
+          'extract' => {
207 
+            'version' => { 'path' => 'response.version' },
208 
+          },
209 
+          # no unzip option
210 
+        }
211 
+        checker = Agents::WebsiteAgent.new(:name => "Weather Site", :options => site)
212 
+        checker.user = users(:bob)
213 
+        checker.save!
214 
+
215 
+        checker.check
216 
+        event = Event.last
217 
+        expect(event.payload['version']).to eq(2)
218 
+      end
219 
+    end
220 
+
172 221 
     describe 'unzipping' do
173 222 
       it 'should unzip automatically if the response has Content-Encoding: gzip' do
174 223 
         json = {