Module: Advance

Defined in:
lib/advance.rb,
lib/advance/version.rb

Constant Summary collapse

RESET = 
"\e[0m"
BOLD = 
"\e[1m"
ITALIC = 
"\e[3m"
UNDERLINE = 
"\e[4m"
CYAN = 
"\e[36m"
GRAY = 
"\e[37m"
GREEN = 
"\e[32m"
MAGENTA = 
"\e[35m"
RED = 
"\e[31m"
WHITE = 
"\e[1;37m"
YELLOW = 
"\e[33m"
VERSION = 
"0.4.3"

Class Method Summary collapse

Instance Method Summary collapse

Class Method Details

.included(pipeline_module) ⇒ Object 



24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
 
# File 'lib/advance.rb', line 24

def self.included(pipeline_module)
  $pipeline = caller_locations.first.path
  meta =
    if File.exist?(".meta")
      JSON.parse(File.read(".meta"))
    else
      {}
    end
  last_run_number = meta["last_run_number"] ||= -1
  $run_number = last_run_number + 1
  $cores=`nproc`.to_i
  puts "Multi steps will use #{$cores} cores"

  $verbose_logging = case ENV["ADVANCE_VERBOSE_LOGGING"]
                     when "true"; true
                     when "false"; false
                     when nil
                       puts "For detailed logging of multi steps, rerun after 'export ADVANCE_VERBOSE_LOGGING=true'"
                       false
                     else
                       puts "env variable ADVANCE_VERBOSE_LOGGING should be 'true', 'false', or not present (defaults to 'false')"
                       puts "currently set to >#{ENV["ADVANCE_VERBOSE_LOGGING"]}<"
                     end
end

Instance Method Details

#add_dir_to_path(dir) ⇒ Object 



357
358
359
360
361
362
363
 
# File 'lib/advance.rb', line 357

def add_dir_to_path(dir)
  bin_dir = File.expand_path(dir)
  path = ENV["PATH"]

  return if path.include?(bin_dir)
  ENV["PATH"] = [path, bin_dir].join(":")
end

#advance(processing_mode, label, command) ⇒ Object 



109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
 
# File 'lib/advance.rb', line 109

def advance(processing_mode, label, command)
  $redo_mode ||= :checking
  $step ||= 0
  previous_dir_path = get_previous_dir_path

  $step += 1
  dir_prefix = step_dir_prefix($step)
  dir_name = "#{dir_prefix}_#{label}"

  puts "#{CYAN}advance #{$step} #{label}#{WHITE}... #{RESET}"

  if $redo_mode != :checking || !(File.exist?(dir_name) || File.exist?(dir_name + '.tgz'))
    clean_previous_step_dirs(dir_prefix)

    if previous_dir_path =~ /\.tgz$/
      do_command_wo_log "tar xzf #{previous_dir_path}"
    end
    previous_dir_path = previous_dir_path.gsub(/\.tgz$/, "")
    start_time = Time.now
    send(processing_mode, command, previous_dir_path, dir_name)
    file_count = count_files(dir_name)
    duration = Time.now - start_time
    update_meta($step, processing_mode, label, command, start_time, duration, file_count)
  end
  previous_dir_path = previous_dir_path.gsub(/\.tgz$/, "")
  if File.basename(previous_dir_path) =~ /^step_/
    if !File.exist?("#{previous_dir_path}.tgz")
      do_command_wo_log "tar czf #{previous_dir_path}.tgz #{File.basename(previous_dir_path)}"
    end
    do_command_wo_log "rm -rf #{previous_dir_path}"
  end
end

#capture_column_names_from_csv ⇒ Object 



77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
 
# File 'lib/advance.rb', line 77

def capture_column_names_from_csv
  if $step.nil?
    raise "capture_column_names_from_csv cannot be the first step"
  end

  if File.exist?(".meta")
    if read_column_names_from_meta
      return
    end
  end

  previous_dir_path = get_previous_dir_path
  input_file_path = previous_file_path(previous_dir_path)
  CSV.foreach(input_file_path, :headers => true) do |row|
    $column_names = row.headers.map(&:to_sym)
    break
  end
end

#clean_previous_step_dirs(dir_prefix) ⇒ Object 



189
190
191
192
193
194
 
# File 'lib/advance.rb', line 189

def clean_previous_step_dirs(dir_prefix)
  while (step_dir = find_step_dir(dir_prefix))
    puts "## removing #{step_dir}"
    FileUtils.rm_rf step_dir
  end
end

#count_files(dir) ⇒ Object 



146
147
148
149
150
151
152
153
154
155
 
# File 'lib/advance.rb', line 146

def count_files(dir)
  file_count = 0
  Find.find(dir) do |path|
    next if File.directory?(path)
    next if File.basename(path) == "log"
    next if File.basename(path) =~ /^\./
    file_count += 1
  end
  file_count
end

#do_command(command, feedback = true) ⇒ Object 



310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
 
# File 'lib/advance.rb', line 310

def do_command(command, feedback = true)
  puts "#{YELLOW}#{command}#{RESET}  " #if feedback
  start_time = Time.now
  stdout, stderr, status = Open3.capture3(command)
  elapsed_time = Time.now - start_time
  File.open("log", "w") do |f|
    f.puts "%%% command: >#{command}<"
    f.puts "%%% returned status: >#{status}<"
    f.puts "%%% elapsed time: #{elapsed_time} seconds"
    f.puts "%%% stdout:"
    f.puts stdout
    f.puts "%%% stderr:"
    f.puts stderr
  end
  if !status.success?
    raise "step #{$step} failed with #{status}\n#{stderr}"
  end
end

#do_command_wo_log(command, feedback = true) ⇒ Object 



329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
 
# File 'lib/advance.rb', line 329

def do_command_wo_log(command, feedback = true)
  puts "#{YELLOW}#{command}#{RESET}  " if feedback
  stdout, stderr, status = Open3.capture3(command)
  if !status.success?
    error_msg = [
      "step #{$step} failed",
      "%%% command: >#{command}<",
      "%%% returned status: >#{status}<",
      "%%% stdout:",
      stdout,
      "%%% stderr:",
      stderr
    ].join("\n")

    raise error_msg
  end
end

#ensure_bin_on_path ⇒ Object 



347
348
349
350
351
352
353
354
355
 
# File 'lib/advance.rb', line 347

def ensure_bin_on_path
  $LOAD_PATH << File.expand_path(File.join(caller_locations.first.path, "../../lib"))

  advance_bin_path = File.expand_path(File.join(File.dirname(__FILE__), "../bin"))
  add_dir_to_path(advance_bin_path)

  caller_path = File.dirname(caller[0].split(/:/).first)
  add_dir_to_path(caller_path)
end

#file_path_template(dir_path, files) ⇒ Object 



300
301
302
303
304
305
306
307
308
 
# File 'lib/advance.rb', line 300

def file_path_template(dir_path, files)
  file = files.first
  file_path = File.join(dir_path, file)
  if File.directory?(file_path)
    File.join(dir_path, "{file}", "{file}")
  else
    File.join(dir_path, "{file}")
  end
end

#find_step_dir(dir_prefix) ⇒ Object 



196
197
198
199
 
# File 'lib/advance.rb', line 196

def find_step_dir(dir_prefix)
  dirs = Dir.entries(".")
  dirs.find { |d| d =~ /^#{dir_prefix}/ }
end

#get_previous_dir_path ⇒ Object 



171
172
173
174
175
176
177
178
179
 
# File 'lib/advance.rb', line 171

def get_previous_dir_path
  relative_path = case $step
                  when 0
                    "."
                  else
                    File.join(".", Dir.entries(".").find { |d| d =~ /^#{step_dir_prefix($step)}/ })
                  end
  File.expand_path(relative_path)
end

#multi(command, previous_dir_path, dir_name) ⇒ Object 



216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
 
# File 'lib/advance.rb', line 216

def multi(command, previous_dir_path, dir_name)
  work_in_sub_dir(dir_name) do
    file_paths = Find.find(previous_dir_path).reject { |p| File.basename(p) =~ %r(^\.) || FileTest.directory?(p) || File.basename(p) == "log" }

    last_progress = ""
    progress_proc = ->(index, max_index) do
      latest_progress = sprintf("%3i%%", index.to_f / max_index * 100)
      puts latest_progress if last_progress != latest_progress
      last_progress = latest_progress
    end
    TeamEffort.work(file_paths, $cores, progress_proc: progress_proc) do |file_path|
      begin
        path_relative_to_step_dir = File.dirname(file_path.gsub(%r(^#{previous_dir_path}/?), ""))
        path_relative_to_step_dir = path_relative_to_step_dir == "." ? "" : path_relative_to_step_dir
        basename = File.basename(file_path)
        new_dir_name = path_relative_to_step_dir == "" ? basename : File.join(path_relative_to_step_dir, basename)
        root_file_name = basename.gsub(%r(\.[^.]+$), '')

        command.gsub!("{input_dir}", File.dirname(file_path))
        command.gsub!("{input_file}", file_path)
        command.gsub!("{file_name}", basename)
        command.gsub!("{file_name_without_extension}", root_file_name)
        puts "#{YELLOW}#{command}#{RESET}  " if $verbose_logging
        work_in_sub_dir(new_dir_name) do
          do_command command, $verbose_logging
        end
      rescue
        puts "%%%% error while processing >>#{file_path}<<"
        raise
      end
    end
  end
end

#pipeline(pipeline_path) ⇒ Object 



142
143
144
 
# File 'lib/advance.rb', line 142

def pipeline(pipeline_path)
  load pipeline_path
end

#previous_file_path(previous_dir_path) ⇒ Object 



296
297
298
 
# File 'lib/advance.rb', line 296

def previous_file_path(previous_dir_path)
  Find.find(previous_dir_path).reject {|p| File.basename(p) =~ %r(^\.) || FileTest.directory?(p) || File.basename(p) == "log"}.first
end

#read_column_names_from_meta ⇒ Object 



96
97
98
99
100
101
102
103
104
105
106
107
 
# File 'lib/advance.rb', line 96

def read_column_names_from_meta
  meta = JSON.parse(File.read(".meta"))
  meta["runs"].each do |run|
    run.each do |step|
      if step["columns"]
        $column_names = step["columns"].map(&:to_sym)
        return true
      end
    end
  end
  false
end

#single(command, previous_dir_path, dir_name) ⇒ Object 



201
202
203
204
205
206
207
208
209
210
211
212
213
214
 
# File 'lib/advance.rb', line 201

def single(command, previous_dir_path, dir_name)
  work_in_sub_dir(dir_name) do
    command.gsub!("{input_dir}", previous_dir_path)
    input_file_path = previous_file_path(previous_dir_path)
    if input_file_path
      basename = File.basename(input_file_path)
      root_file_name = basename.gsub(%r(\.[^.]+$), '')
      command.gsub!("{input_file}", input_file_path)
      command.gsub!("{file_name}", basename)
      command.gsub!("{file_name_without_extension}", root_file_name)
    end
    do_command command
  end
end

#static(processing_mode, label, command) ⇒ Object 



157
158
159
160
161
162
163
164
165
166
167
168
169
 
# File 'lib/advance.rb', line 157

def static(processing_mode, label, command)
  $redo_mode ||= :checking
  $step ||= 0
  previous_dir_path = get_previous_dir_path
  dir_prefix = static_dir_prefix($step)
  dir_name = "#{dir_prefix}_#{label}"
  puts "#{CYAN}static #{$step} #{label}#{WHITE}... #{RESET}"
  return if $redo_mode == :checking && Dir.exist?(dir_name)

  FileUtils.rm_rf dir_name

  send(processing_mode, command, previous_dir_path, dir_name)
end

#static_dir_prefix(step_no) ⇒ Object 



185
186
187
 
# File 'lib/advance.rb', line 185

def static_dir_prefix(step_no)
  "static_%03d" % [step_no]
end

#step_dir_prefix(step_no) ⇒ Object 



181
182
183
 
# File 'lib/advance.rb', line 181

def step_dir_prefix(step_no)
  "step_%03d" % [step_no]
end

#strip_extensions(dir_name) ⇒ Object 



272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
 
# File 'lib/advance.rb', line 272

def strip_extensions(dir_name)
  extensions = %w(
    csv
    csv_nh
    geo_json
    geojson
    gz
    json
    tar
    tgz
    zip
  )

  changed_dir_name = dir_name
  last_dir_name = nil
  until last_dir_name == changed_dir_name do
    last_dir_name = changed_dir_name
    extensions.each do |extension|
      changed_dir_name = changed_dir_name.gsub(%r(\.#{extension}$), "")
    end
  end
  changed_dir_name
end

#update_meta(step_number, processing_mode, label, command, start_time, duration, file_count) ⇒ Object 



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
 
# File 'lib/advance.rb', line 49

def update_meta(step_number, processing_mode, label, command, start_time, duration, file_count)
  meta =
    if File.exist?(".meta")
      JSON.parse(File.read(".meta"))
    else
      {}
    end

  meta["pipeline"] ||= $pipeline
  meta["last_run_number"] = $run_number
  meta["runs"] ||= []

  step_data = {
    "step_number" => step_number,
    "start_time" => start_time,
    "duration" => duration,
    "file_count" => file_count,
    "processing_mode" => processing_mode,
    "label" => label,
    "command" => command,
    "columns" => $column_names
  }
  meta["runs"][$run_number] ||= []
  meta["runs"][$run_number] << step_data

  File.write(".meta", JSON.pretty_generate(meta))
end

#work_in_sub_dir(dir_name) ⇒ Object 



250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
 
# File 'lib/advance.rb', line 250

def work_in_sub_dir(dir_name)
  starting_dir = FileUtils.pwd
  stripped_dir_name = File.join(*(strip_extensions(dir_name).split("/").uniq))
  if $redo_mode == :checking && Dir.exist?(stripped_dir_name)
    return
  end

  $redo_mode = :replacing

  dirs = stripped_dir_name.split("/")
  dirs[-1] = "tmp_#{dirs[-1]}"
  tmp_dir = File.join(dirs)
  FileUtils.rm_rf tmp_dir
  FileUtils.mkdir_p tmp_dir
  FileUtils.cd tmp_dir

  yield

  FileUtils.cd starting_dir
  FileUtils.mv tmp_dir, stripped_dir_name
end