Class: Fluent::GithubActivities::Crawler

Inherits:
Object
  • Object
show all
Defined in:
lib/fluent/plugin/github-activities/crawler.rb

Defined Under Namespace

Classes: EmptyRequestQueue

Constant Summary collapse

NO_INTERVAL = 
0
DEFAULT_INTERVAL = 
1
"$github-activities-related-avatar"
"$github-activities-related-organization-logo"
"$github-activities-related-event"
MERGE_COMMIT_MESSAGE_PATTERN = 
/\AMerge pull request #\d+ from [^\/]+\/[^\/]+\n\n/

Instance Attribute Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(options = {}) ⇒ Crawler

Returns a new instance of Crawler.



43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
 
# File 'lib/fluent/plugin/github-activities/crawler.rb', line 43

def initialize(options={})
  @users_manager = UsersManager.new(:users    => options[:watching_users],
                                    :pos_file => options[:pos_file])

  @access_token = options[:access_token]

  @watching_users = options[:watching_users] || []

  @include_commits_from_pull_request = options[:include_commits_from_pull_request]
  @include_foreign_commits = options[:include_foreign_commits]

  @request_queue = options[:request_queue] || []

  @default_interval = options[:default_interval] || DEFAULT_INTERVAL
end

Instance Attribute Details

#interval_for_next_request ⇒ Object (readonly)

Returns the value of attribute interval_for_next_request.



41
42
43
 
# File 'lib/fluent/plugin/github-activities/crawler.rb', line 41

def interval_for_next_request
  @interval_for_next_request
end

#on_emit=(value) ⇒ Object (writeonly)

Sets the attribute on_emit

Parameters:

  • value —

    the value to set the attribute on_emit to.



40
41
42
 
# File 'lib/fluent/plugin/github-activities/crawler.rb', line 40

def on_emit=(value)
  @on_emit = value
end

#request_queue ⇒ Object (readonly)

Returns the value of attribute request_queue.



41
42
43
 
# File 'lib/fluent/plugin/github-activities/crawler.rb', line 41

def request_queue
  @request_queue
end

Instance Method Details

#extra_request_headers(request) ⇒ Object 



127
128
129
130
131
132
133
134
135
136
137
138
139
 
# File 'lib/fluent/plugin/github-activities/crawler.rb', line 127

def extra_request_headers(request)
  headers = {}
  if request[:previous_entity_tag]
    headers["If-None-Match"] = request[:previous_entity_tag]
  elsif request[:type] == TYPE_EVENTS
    position = @users_manager.position_for(request[:user])
    if position
      entity_tag = position["entity_tag"]
      headers["If-None-Match"] = entity_tag if entity_tag
    end
  end
  headers
end

#process_commit(commit, push_event) ⇒ Object 



208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
 
# File 'lib/fluent/plugin/github-activities/crawler.rb', line 208

def process_commit(commit, push_event)
  $log.debug("GithubActivities::Crawler: processing commit #{commit["sha"]}") if $log
  user = commit["author"]["login"]

  if user and (@include_foreign_commits or watching_user?(user))
    commit[RELATED_USER_IMAGE_KEY] = push_event["actor"]["avatar_url"]
    if push_event["org"]
      commit[RELATED_ORGANIZATION_IMAGE_KEY] = push_event["org"]["avatar_url"]
    end
    commit[RELATED_EVENT] = push_event
    emit("commit", commit)
  end

  commit_refs = push_event["payload"]["commits"]
  target_commit_ref = commit_refs.find do |commit_ref|
    commit_ref["sha"] == commit["sha"]
  end
  target_commit_ref["commit"] = commit if target_commit_ref

  completely_fetched = commit_refs.all? do |commit_ref|
    commit_ref["commit"]
  end
  emit("push", push_event) if completely_fetched
end

#process_create_event(event) ⇒ Object 



296
297
298
299
300
301
302
303
304
 
# File 'lib/fluent/plugin/github-activities/crawler.rb', line 296

def process_create_event(event)
  payload = event["payload"]
  case payload["ref_type"]
  when "branch"
    emit("branch", event)
  when "tag"
    emit("tag", event)
  end
end

#process_issue_event(event) ⇒ Object 



237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
 
# File 'lib/fluent/plugin/github-activities/crawler.rb', line 237

def process_issue_event(event)
  payload = event["payload"]
  case payload["action"]
  when "opened"
    emit("issue-open", event)
  when "closed"
    emit("issue-close", event)
  when "reopened"
    emit("issue-reopen", event)
  when "assigned"
    emit("issue-assign", event)
  when "unassigned"
    emit("issue-unassign", event)
  when "labeled"
    emit("issue-label", event)
  when "unlabeled"
    emit("issue-unlabel", event)
  end
end

#process_issue_or_pull_request_comment_event(event) ⇒ Object 



286
287
288
289
290
291
292
293
294
 
# File 'lib/fluent/plugin/github-activities/crawler.rb', line 286

def process_issue_or_pull_request_comment_event(event)
  payload = event["payload"]
  if payload["issue"]["pull_request"]
    emit("pull-request-comment", event)
    # emit("pull-request.cancel", event)
  else
    emit("issue-comment", event)
  end
end

#process_pull_request_event(event) ⇒ Object 



257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
 
# File 'lib/fluent/plugin/github-activities/crawler.rb', line 257

def process_pull_request_event(event)
  payload = event["payload"]
  case payload["action"]
  when "opened"
    emit("pull-request", event)
  when "closed"
    if payload["pull_request"]["merged"]
      emit("pull-request-merged", event)
    else
      emit("pull-request-cancelled", event)
    end
  when "reopened"
    emit("pull-request-reopen", event)
  end
end

#process_push_event(event) ⇒ Object 



192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
 
# File 'lib/fluent/plugin/github-activities/crawler.rb', line 192

def process_push_event(event)
  payload = event["payload"]
  commit_refs = payload["commits"]
  if !@include_commits_from_pull_request and
       push_event_from_merged_pull_request?(event)
    return
  end
  commit_refs.reverse.each do |commit_ref|
    @request_queue.push(:type => TYPE_COMMIT,
                        :uri  => commit_ref["url"],
                        :sha  => commit_ref["sha"],
                        :push => event)
  end
  # emit("push", event)
end

#process_request ⇒ Object 



59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
 
# File 'lib/fluent/plugin/github-activities/crawler.rb', line 59

def process_request
  request = @request_queue.shift
  $log.debug("GithubActivities::Crawler: processing request: #{request.inspect}") if $log
  if request[:process_after] and
       Time.now.to_i < request[:process_after]
    @request_queue.push(request)
    @interval_for_next_request = NO_INTERVAL
    return false
  end

  uri = request_uri(request)
  extra_headers = extra_request_headers(request)

  $log.debug("GithubActivities::Crawler: requesting to #{uri.inspect}") if $log
  response = http_get(uri, extra_headers)
  $log.debug("GithubActivities::Crawler: response: #{response.inspect}") if $log

  case response
  when Net::HTTPSuccess
    $log.trace("GithubActivities::Crawler: Net::HTTPSuccess / request type: #{request[:type]}") if $log
    body = JSON.parse(response.body)
    case request[:type]
    when TYPE_EVENTS
      events = body
      $log.trace("GithubActivities::Crawler: events size: #{events.size}") if $log
      process_user_events(request[:user], events)
      reserve_user_events(request[:user], :previous_response => response)
      @users_manager.save_position_for(request[:user], :entity_tag => response["ETag"])
    when TYPE_COMMIT
      process_commit(body, request[:push])
    end
  when Net::HTTPNotModified
    $log.trace("GithubActivities::Crawler: Net::HTTPNotModified / request type: #{request[:type]}") if $log
    case request[:type]
    when TYPE_EVENTS
      reserve_user_events(request[:user],
                          :previous_response => response,
                          :previous_entity_tag => extra_headers["If-None-Match"])
    end
    @interval_for_next_request = @default_interval
    return true
  else
    $log.trace("GithubActivities::Crawler: UnknownType / request type: #{request[:type]}") if $log
    case request[:type]
    when TYPE_COMMIT
      fake_body = {
        "sha"    => request[:sha],
        "author" => {},
      }
      process_commit(fake_body, request[:push])
    end
  end
  @interval_for_next_request = @default_interval
  return true
rescue StandardError => error
  $log.error(error.inspect)
end

#process_user_event(user, event) ⇒ Object 



164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
 
# File 'lib/fluent/plugin/github-activities/crawler.rb', line 164

def process_user_event(user, event)
  # see also: https://developer.github.com/v3/activity/events/types/
  event[RELATED_USER_IMAGE_KEY] = event["actor"]["avatar_url"]
  if event["org"]
    event[RELATED_ORGANIZATION_IMAGE_KEY] = event["org"]["avatar_url"]
  end
  case event["type"]
  when "PushEvent"
    process_push_event(event)
  when "CommitCommentEvent"
    emit("commit-comment", event)
  when "IssuesEvent"
    process_issue_event(event)
  when "IssueCommentEvent"
    process_issue_or_pull_request_comment_event(event)
  when "ForkEvent"
    emit("fork", event)
  when "PullRequestEvent"
    process_pull_request_event(event)
  when "CreateEvent"
    process_create_event(event)
  else
    emit(event["type"], event)
  end
rescue StandardError => error
  $log.exception(error)
end

#process_user_events(user, events) ⇒ Object 



146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
 
# File 'lib/fluent/plugin/github-activities/crawler.rb', line 146

def process_user_events(user, events)
  last_event_timestamp = UsersManager::DEFAULT_LAST_EVENT_TIMESTAMP
  position = @users_manager.position_for(user)
  if position and position["last_event_timestamp"]
    last_event_timestamp = position["last_event_timestamp"]
  end

  events = events.sort do |a, b|
    b["created_at"] <=> a["created_at"]
  end
  events.each do |event|
    timestamp = Time.parse(event["created_at"]).to_i
    next if timestamp <= last_event_timestamp
    process_user_event(user, event)
    @users_manager.save_position_for(user, :last_event_timestamp => timestamp)
  end
end

#push_event_from_merged_pull_request?(event) ⇒ Boolean

Returns:

  • (Boolean) 


275
276
277
278
279
280
281
282
283
284
 
# File 'lib/fluent/plugin/github-activities/crawler.rb', line 275

def push_event_from_merged_pull_request?(event)
  payload = event["payload"]
  inserted_requests = []
  commit_refs = payload["commits"]
  if MERGE_COMMIT_MESSAGE_PATTERN =~ commit_refs.last["message"]
    true
  else
    false
  end
end

#request_uri(request) ⇒ Object 



117
118
119
120
121
122
123
124
125
 
# File 'lib/fluent/plugin/github-activities/crawler.rb', line 117

def request_uri(request)
  uri = nil
  case request[:type]
  when TYPE_EVENTS
    uri = user_activities(request[:user])
  else
    uri = request[:uri]
  end
end

#reserve_user_events(user, options = {}) ⇒ Object 



141
142
143
144
 
# File 'lib/fluent/plugin/github-activities/crawler.rb', line 141

def reserve_user_events(user, options={})
  request = @users_manager.new_events_request(user, options)
  @request_queue.push(request)
end

#watching_user?(user) ⇒ Boolean

Returns:

  • (Boolean) 


233
234
235
 
# File 'lib/fluent/plugin/github-activities/crawler.rb', line 233

def watching_user?(user)
  @watching_users.include?(user)
end