Module: MultipartUploadUtils

Included in:
FilestackClient
Defined in:
lib/filestack/utils/multipart_upload_utils.rb

Overview

Includes all the utility functions for Filestack multipart uploads

Instance Method Summary collapse

Instance Method Details

#create_upload_jobs(apikey, filename, filepath, filesize, start_response, storage, options) ⇒ Array

Create array of jobs for parallel uploading

Parameters:

  • apikey (String) —

    Filestack API key

  • filename (String) —

    Name of incoming file

  • filepath (String) —

    Local path to file

  • filesize (Int) —

    Size of incoming file

  • start_response (Typhoeus::Response) —

    Response body from multipart_start

  • options (Hash) —

    User-defined options for multipart uploads

Returns:

  • (Array) 


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
116
117
118
119
120
121
122
123
 
# File 'lib/filestack/utils/multipart_upload_utils.rb', line 90

def create_upload_jobs(apikey, filename, filepath, filesize, start_response, storage, options)
  jobs = []
  part = 1
  seek_point = 0
  while seek_point < filesize
    part_info = {
      seek: seek_point,
      filepath: filepath,
      filename: filename,
      apikey: apikey,
      part: part,
      filesize: filesize,
      uri: start_response['uri'],
      region: start_response['region'],
      upload_id: start_response['upload_id'],
      location_url: start_response['location_url'],
      start_response: start_response,
      store_location: storage
    }
    options = multipart_options(options)
    part_info = part_info.merge!(options) if options

    if seek_point + FilestackConfig::DEFAULT_CHUNK_SIZE > filesize
      size = filesize - (seek_point)
    else
      size = FilestackConfig::DEFAULT_CHUNK_SIZE
    end
    part_info[:size] = size
    jobs.push(part_info)
    part += 1
    seek_point += FilestackConfig::DEFAULT_CHUNK_SIZE
  end
  jobs
end

#get_file_info(file) ⇒ Object 



16
17
18
19
20
21
22
23
24
 
# File 'lib/filestack/utils/multipart_upload_utils.rb', line 16

def get_file_info(file)
  filename = File.basename(file)
  filesize = File.size(file)
  mimetype = MimeMagic.by_magic(File.open(file))
  if mimetype.nil?
    mimetype = 'application/octet-stream'
  end
  [filename, filesize, mimetype.to_s]
end

#multipart_complete(apikey, filename, filesize, mimetype, start_response, parts_and_etags, options, storage, intelligent = false) ⇒ Typhoeus::Response

Send complete call to multipart endpoint

Parameters:

  • apikey (String) —

    Filestack API key

  • filename (String) —

    Name of incoming file

  • filesize (Int) —

    Size of incoming file

  • mimetype (String) —

    Mimetype of incoming file

  • start_response (Typhoeus::Response) —

    Response body from multipart_start

  • security (FilestackSecurity) —

    Security object with policy/signature

  • parts_and_etags (Array) —

    Array of strings defining etags and their associated part numbers

  • options (Hash) —

    User-defined options for multipart uploads

Returns:

  • (Typhoeus::Response) 


207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
 
# File 'lib/filestack/utils/multipart_upload_utils.rb', line 207

def multipart_complete(apikey, filename, filesize, mimetype, start_response, parts_and_etags, options, storage, intelligent = false)
  data = {
    apikey: apikey,
    uri: start_response['uri'],
    region: start_response['region'],
    upload_id: start_response['upload_id'],
    filename: filename,
    size: filesize,
    mimetype: mimetype,
    store_location: storage,
    file: Tempfile.new(filename)
  }
  options = multipart_options(options)
  data.merge!(options) if options
  data.merge!(intelligent ? { multipart: intelligent } : { parts: parts_and_etags.join(';') })

  Typhoeus.post(
    FilestackConfig::MULTIPART_COMPLETE_URL, body: data,
                                             headers: FilestackConfig::HEADERS
  )
end

#multipart_options(options) ⇒ Object 



26
27
28
29
30
31
32
33
34
 
# File 'lib/filestack/utils/multipart_upload_utils.rb', line 26

def multipart_options(options)
  [:region, :container, :path, :access].each do |key|
    if options.has_key?(key)
      options[:"store_#{key}"] = options[key]
      options.delete(key)
    end
  end
  return options
end

#multipart_start(apikey, filename, filesize, mimetype, security, storage, options = {}, intelligent) ⇒ Typhoeus::Response

Send start response to multipart endpoint

Parameters:

  • apikey (String) —

    Filestack API key

  • filename (String) —

    Name of incoming file

  • filesize (Int) —

    Size of incoming file

  • mimetype (String) —

    Mimetype of incoming file

  • security (FilestackSecurity) —

    Security object with policy/signature

  • options (Hash) (defaults to: {}) —

    User-defined options for multipart uploads

Returns:

  • (Typhoeus::Response) 


48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
 
# File 'lib/filestack/utils/multipart_upload_utils.rb', line 48

def multipart_start(apikey, filename, filesize, mimetype, security, storage, options = {}, intelligent)
  params = {
    apikey: apikey,
    filename: filename,
    mimetype: mimetype,
    size: filesize,
    store_location: storage,
    file: Tempfile.new(filename),
    multipart: intelligent
  }

  options = multipart_options(options)
  params = params.merge!(options) if options

  unless security.nil?
    params[:policy] = security.policy
    params[:signature] = security.signature
  end

  response = Typhoeus.post(
    FilestackConfig::MULTIPART_START_URL, body: params,
                                          headers: FilestackConfig::HEADERS
  )
  if response.code == 200
    JSON.parse(response.body)
  else
    raise RuntimeError.new(response.body)
  end
end

#multipart_upload(apikey, filepath, security, options, timeout, storage, intelligent: false) ⇒ Hash

Run entire multipart process through with file and options

Parameters:

  • apikey (String) —

    Filestack API key

  • filename (String) —

    Name of incoming file

  • security (FilestackSecurity) —

    Security object with policy/signature

  • options (Hash) —

    User-defined options for multipart uploads

Returns:



239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
 
# File 'lib/filestack/utils/multipart_upload_utils.rb', line 239

def multipart_upload(apikey, filepath, security, options, timeout, storage, intelligent: false)
  filename, filesize, mimetype = get_file_info(filepath)
  start_response = multipart_start(
    apikey, filename, filesize, mimetype, security, storage, options, intelligent
  )

  unless start_response['upload_type'].nil?
    intelligent_enabled = ((start_response['upload_type'].include? 'intelligent_ingestion')) && intelligent
  end

  jobs = create_upload_jobs(
    apikey, filename, filepath, filesize, start_response, storage, options
  )

  if intelligent_enabled
    state = IntelligentState.new
    run_intelligent_upload_flow(jobs, state)
    response_complete = multipart_complete(
      apikey, filename, filesize, mimetype,
      start_response, nil, options, storage, intelligent
    )
  else
    parts_and_etags = run_uploads(jobs, apikey, filepath, options)
    response_complete = multipart_complete(
      apikey, filename, filesize, mimetype,
      start_response, parts_and_etags, options, storage
    )
  end
  begin
    Timeout::timeout(timeout) {
      while response_complete.code == 202
        response_complete = multipart_complete(
          apikey, filename, filesize, mimetype,
          start_response, nil, options, storage, intelligent
        )
      end
    }
  rescue
    raise "Upload timed out upon completion. Please try again later"
  end
  JSON.parse(response_complete.body)
end

#run_uploads(jobs, apikey, filepath, options) ⇒ Array

Runs all jobs in parallel

Parameters:

  • jobs (Array) —

    Array of jobs to be run

  • apikey (String) —

    Filestack API key

  • filepath (String) —

    Local path to file

  • options (Hash) —

    User-defined options for multipart uploads

Returns:

  • (Array) —

    Array of parts/etags strings



175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
 
# File 'lib/filestack/utils/multipart_upload_utils.rb', line 175

def run_uploads(jobs, apikey, filepath, options)
  bar = ProgressBar.new(jobs.length)
  results = Parallel.map(jobs, in_threads: 4) do |job|
    response = upload_chunk(
      job, apikey, filepath, options
    )
    if response.code == 200
      bar.increment!
      part = job[:part]
      etag = response.headers[:etag]
      "#{part}:#{etag}"
    end
  end
  results
end

#upload_chunk(job, apikey, filepath, options) ⇒ Typhoeus::Response

Uploads one chunk of the file

Parameters:

  • job (Hash) —

    Hash of options needed to upload a chunk

  • apikey (String) —

    Filestack API key

  • location_url (String) —

    Location url given back from endpoint

  • filepath (String) —

    Local path to file

  • options (Hash) —

    User-defined options for multipart uploads

Returns:

  • (Typhoeus::Response) 


138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
 
# File 'lib/filestack/utils/multipart_upload_utils.rb', line 138

def upload_chunk(job, apikey, filepath, options)
  file = File.open(filepath)
  file.seek(job[:seek])
  chunk = file.read(FilestackConfig::DEFAULT_CHUNK_SIZE)

  md5 = Digest::MD5.new
  md5 << chunk
  data = {
    apikey: apikey,
    part: job[:part],
    size: chunk.length,
    md5: md5.base64digest,
    uri: job[:uri],
    region: job[:region],
    upload_id: job[:upload_id],
    store_location: job[:store_location],
    file: Tempfile.new(job[:filename])
  }
  data = data.merge!(options) if options
  fs_response = Typhoeus.post(
    FilestackConfig::MULTIPART_UPLOAD_URL, body: data,
                                           headers: FilestackConfig::HEADERS
  ).body
  fs_response = JSON.parse(fs_response)
  Typhoeus.put(
    fs_response['url'], headers: fs_response['headers'], body: chunk
  )
end