Class: NHKore::App

Inherits:
Object
  • Object
show all
Includes:
CLI::FXCmd, CLI::GetCmd, CLI::NewsCmd, CLI::SearchCmd, CLI::SiftCmd
Defined in:
lib/nhkore/app.rb

Constant Summary collapse

NAME = 
'nhkore'
DEFAULT_SLEEP_TIME =

So that sites don’t ban us (i.e., think we are human)

0.1
COLOR_OPTS = 
i[c color].freeze
NO_COLOR_OPTS = 
i[C no-color].freeze
SPINNER_MSG = 
'[:spinner] :title:detail...'
CLASSIC_SPINNER = 
TTY::Spinner.new(SPINNER_MSG,format: :classic)
DEFAULT_SPINNER = 
TTY::Spinner.new(SPINNER_MSG,interval: 5,
frames: ['〜〜〜','日〜〜','日本〜','日本語'])
NO_SPINNER_MSG = 
'%{title}%{detail}...'

Constants included from CLI::SiftCmd

CLI::SiftCmd::DEFAULT_SIFT_EXT, CLI::SiftCmd::DEFAULT_SIFT_FUTSUU_FILE, CLI::SiftCmd::DEFAULT_SIFT_YASASHII_FILE, CLI::SiftCmd::SIFT_EXTS

Constants included from CLI::NewsCmd

CLI::NewsCmd::DEFAULT_NEWS_SCRAPE

Constants included from CLI::GetCmd

CLI::GetCmd::DEFAULT_GET_CHUNK_SIZE, CLI::GetCmd::DEFAULT_GET_URL_LENGTH, CLI::GetCmd::GET_URL, CLI::GetCmd::GET_URL_FILENAME

Instance Attribute Summary collapse

Attributes included from CLI::SiftCmd

#sift_datetime_text, #sift_search_criteria

Instance Method Summary collapse

Methods included from CLI::SiftCmd

#build_sift_cmd, #build_sift_filename, #run_sift_cmd

Methods included from CLI::SearchCmd

#build_search_cmd, #run_search_cmd, #run_search_help, #show_search_urls

Methods included from CLI::NewsCmd

#build_news_cmd, #run_news_cmd, #scrape_news_article, #scraped_news_article?

Methods included from CLI::GetCmd

#build_get_cmd, #run_get_cmd

Methods included from CLI::FXCmd

#build_fx_cmd, #run_fx_cmd, #test_fx_progress_bar, #test_fx_spinner

Constructor Details

#initialize(args = ARGV) ⇒ App 



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
 
# File 'lib/nhkore/app.rb', line 71

def initialize(args = ARGV)
  super()

  @args = args
  @cmd = nil
  @cmd_args = nil
  @cmd_opts = nil
  @high = HighLine.new
  @rainbow = Rainbow.new
  @progress_bar = :default # [:default, :classic, :no]
  @scraper_kargs = {}
  @sleep_time = DEFAULT_SLEEP_TIME
   = DEFAULT_SPINNER

  autodetect_color

  build_app_cmd

  build_fx_cmd
  build_get_cmd
  build_news_cmd
  build_search_cmd
  build_sift_cmd
  build_version_cmd

  @app_cmd.add_command Cri::Command.new_basic_help
end

Instance Attribute Details

#cmd ⇒ Object (readonly)

Returns the value of attribute cmd.



63
64
65
 
# File 'lib/nhkore/app.rb', line 63

def cmd
  @cmd
end

#cmd_args ⇒ Object (readonly)

Returns the value of attribute cmd_args.



64
65
66
 
# File 'lib/nhkore/app.rb', line 64

def cmd_args
  @cmd_args
end

#cmd_opts ⇒ Object (readonly)

Returns the value of attribute cmd_opts.



65
66
67
 
# File 'lib/nhkore/app.rb', line 65

def cmd_opts
  @cmd_opts
end

#progress_bar ⇒ Object

Returns the value of attribute progress_bar.



66
67
68
 
# File 'lib/nhkore/app.rb', line 66

def progress_bar
  @progress_bar
end

#scraper_kargs ⇒ Object

Returns the value of attribute scraper_kargs.



67
68
69
 
# File 'lib/nhkore/app.rb', line 67

def scraper_kargs
  @scraper_kargs
end

#sleep_time ⇒ Object

Returns the value of attribute sleep_time.



68
69
70
 
# File 'lib/nhkore/app.rb', line 68

def sleep_time
  @sleep_time
end

#spinner ⇒ Object

Returns the value of attribute spinner.



69
70
71
 
# File 'lib/nhkore/app.rb', line 69

def 
  
end

Instance Method Details

#autodetect_color ⇒ Object 



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
124
125
126
127
128
129
130
131
 
# File 'lib/nhkore/app.rb', line 99

def autodetect_color
  Cri::Platform.singleton_class.prepend(CriColorExt)

  color = nil # Must be nil, not true/false

  if !@args.empty?
    # Kind of hacky, but necessary for Rainbow.

    color_opts = opts_to_set(COLOR_OPTS)
    no_color_opts = opts_to_set(NO_COLOR_OPTS)

    @args.each do |arg|
      if color_opts.include?(arg)
        color = true
        break
      end

      if no_color_opts.include?(arg)
        color = false
        break
      end

      break if arg == '--'
    end
  end

  if color.nil?
    # - https://no-color.org/
    color = ($stdout.tty? && ENV['TERM'] != 'dumb' && !ENV.key?('NO_COLOR'))
  end

  enable_color(color)
end

#build_app_cmd ⇒ Object 



133
134
135
136
137
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
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
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
 
# File 'lib/nhkore/app.rb', line 133

def build_app_cmd
  app = self

  @app_cmd = Cri::Command.define do
    name    NAME
    usage   "#{NAME} [OPTIONS] [COMMAND]..."
    summary 'NHK News Web (Easy) scraper for Japanese language learners.'

    description "      Scrapes NHK News Web (Easy) to create a list of each word and its\n      frequency (how many times it was used) for Japanese language learners.\n\n      This is similar to a core word/vocabulary list.\n    DESC\n\n    flag :s,:'classic-fx',<<-DESC do |_value,_cmd|\n      use classic spinner/progress special effects (in case of no Unicode support) when running long tasks\n    DESC\n      app.progress_bar = :classic\n      app.spinner = CLASSIC_SPINNER\n    end\n    flag COLOR_OPTS[0],COLOR_OPTS[1],\"force color output (for commands like '| less -R')\" do |_value,_cmd|\n      app.enable_color(true)\n    end\n    flag :n,:'dry-run',<<-DESC\n      do a dry run without making changes; do not write to files, create directories, etc.\n    DESC\n    # Big F because dangerous.\n    flag :F,:force,\"force overwriting files, creating directories, etc. (don't prompt); dangerous!\"\n    flag :h,:help,'show this help' do |_value,cmd|\n      puts cmd.help\n      exit\n    end\n    option :m,:'max-retry',<<-DESC,argument: :required,default: 3 do |value,_cmd|\n      maximum number of times to retry URLs (-1 or integer >= 0)\n    DESC\n      value = value.to_i\n      value = nil if value < 0\n\n      app.scraper_kargs[:max_retries] = value\n    end\n    flag NO_COLOR_OPTS[0],NO_COLOR_OPTS[1],'disable color output' do |_value,_cmd|\n      app.enable_color(false)\n    end\n    flag :X,:'no-fx','disable spinner/progress special effects when running long tasks' do |_value,_cmd|\n      app.progress_bar = :no\n      app.spinner = {} # Still outputs status & stores tokens\n    end\n    option :o,:'open-timeout',<<-DESC,argument: :required do |value,_cmd|\n      seconds for URL open timeouts (-1 or decimal >= 0)\n    DESC\n      value = value.to_f\n      value = nil if value < 0.0\n\n      app.scraper_kargs[:open_timeout] = value\n    end\n    option :r,:'read-timeout',<<-DESC,argument: :required do |value,_cmd|\n      seconds for URL read timeouts (-1 or decimal >= 0)\n    DESC\n      value = value.to_f\n      value = nil if value < 0.0\n\n      app.scraper_kargs[:read_timeout] = value\n    end\n    option :z,:sleep,<<-DESC,argument: :required,default: DEFAULT_SLEEP_TIME do |value,_cmd|\n      seconds to sleep per scrape (i.e., per page/article) so don't get banned (i.e., fake being human)\n    DESC\n      app.sleep_time = value.to_f\n      app.sleep_time = 0.0 if app.sleep_time < 0.0\n    end\n    option :t,:timeout,<<-DESC,argument: :required do |value,_cmd|\n      seconds for all URL timeouts: [open, read] (-1 or decimal >= 0)\n    DESC\n      value = value.to_f\n      value = nil if value < 0.0\n\n      app.scraper_kargs[:open_timeout] = value\n      app.scraper_kargs[:read_timeout] = value\n    end\n    option :u,:'user-agent',<<-DESC,argument: :required do |value,_cmd|\n      HTTP header field 'User-Agent' to use instead of a random one\n    DESC\n      value = app.check_empty_opt(:'user-agent',value)\n\n      app.scraper_kargs[:header] ||= {}\n      app.scraper_kargs[:header]['user-agent'] = value\n    end\n    flag :v,:version,'show the version and exit' do |_value,_cmd|\n      app.show_version\n      exit\n    end\n\n    run do |_opts,_args,cmd|\n      puts cmd.help\n    end\n  end\nend\n"

#build_dir(opt_key, default_dir: '.') ⇒ Object 



231
232
233
234
235
236
237
238
239
240
241
242
 
# File 'lib/nhkore/app.rb', line 231

def build_dir(opt_key,default_dir: '.')
  # Protect against fat-fingering.
  default_dir = Util.strip_web_str(default_dir)
  dir = Util.strip_web_str(@cmd_opts[opt_key].to_s)

  dir = default_dir if dir.empty?

  # '~' will expand to home, etc.
  dir = File.expand_path(dir) unless dir.nil?

  return (@cmd_opts[opt_key] = dir)
end

#build_file(opt_key, default_dir: '.', default_filename: '') ⇒ Object 



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
 
# File 'lib/nhkore/app.rb', line 244

def build_file(opt_key,default_dir: '.',default_filename: '')
  # Protect against fat-fingering.
  default_dir = Util.strip_web_str(default_dir)
  default_filename = Util.strip_web_str(default_filename)
  file = Util.strip_web_str(@cmd_opts[opt_key].to_s)

  if file.empty?
    # Do not check `default_dir.empty?()`.
    file = if default_filename.empty?
             nil # NOTE: nil is very important for BingScraper.init()!
           else
             File.join(default_dir,default_filename)
           end
  # Directory?
  elsif File.directory?(file) || Util.dir_str?(file)
    file = File.join(file,default_filename)
  # File name only? (no directory)
  elsif Util.filename_str?(file)
    file = File.join(default_dir,file)
  else
    # Passed in both: 'directory/file'
  end

  # '~' will expand to home, etc.
  file = File.expand_path(file) unless file.nil?

  return (@cmd_opts[opt_key] = file)
end

#build_in_dir(opt_key, **kargs) ⇒ Object 



273
274
275
 
# File 'lib/nhkore/app.rb', line 273

def build_in_dir(opt_key,**kargs)
  return build_dir(opt_key,**kargs)
end

#build_in_file(opt_key, **kargs) ⇒ Object 



277
278
279
 
# File 'lib/nhkore/app.rb', line 277

def build_in_file(opt_key,**kargs)
  return build_file(opt_key,**kargs)
end

#build_out_dir(opt_key, **kargs) ⇒ Object 



281
282
283
 
# File 'lib/nhkore/app.rb', line 281

def build_out_dir(opt_key,**kargs)
  return build_dir(opt_key,**kargs)
end

#build_out_file(opt_key, **kargs) ⇒ Object 



285
286
287
 
# File 'lib/nhkore/app.rb', line 285

def build_out_file(opt_key,**kargs)
  return build_file(opt_key,**kargs)
end

#build_progress_bar(title, download: false, total: 100, type: @progress_bar, width: 33, **kargs) ⇒ Object 



289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
 
# File 'lib/nhkore/app.rb', line 289

def build_progress_bar(title,download: false,total: 100,type: @progress_bar,width: 33,**kargs)
  case type
  when :default,:classic
    require 'tty-progressbar'

    msg = "#{title} [:bar] :percent :eta".dup
    msg << ' :byte_rate/s' if download

    return TTY::ProgressBar.new(msg,total: total,width: width,**kargs) do |config|
      if type == :default
        config.incomplete = '.'
        config.complete   = '/'
        config.head       = 'o'
      end

      # config.frequency = 5 # For a big download, set this
      config.interval = 1 if download
    end
  end

  # :no
  return NoProgressBar.new(title,total: total,**kargs)
end

#build_version_cmd ⇒ Object 



313
314
315
316
317
318
319
320
321
322
323
324
325
326
 
# File 'lib/nhkore/app.rb', line 313

def build_version_cmd
  app = self

  @version_cmd = @app_cmd.define_command do
    name    'version'
    usage   'version [OPTIONS] [COMMAND]...'
    aliases :v
    summary "Show the version and exit (aliases: #{app.color_alias('v')})"

    run do |_opts,_args,_cmd|
      app.show_version
    end
  end
end

#check_empty_opt(key, value) ⇒ Object 



328
329
330
331
332
333
334
335
336
 
# File 'lib/nhkore/app.rb', line 328

def check_empty_opt(key,value)
  value = Util.strip_web_str(value) unless value.nil?

  if value.nil? || value.empty?
    raise CLIError,"option[#{key}] cannot be empty[#{value}]"
  end

  return value
end

#check_in_file(opt_key, empty_ok: false) ⇒ Object 



338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
 
# File 'lib/nhkore/app.rb', line 338

def check_in_file(opt_key,empty_ok: false)
  in_file = @cmd_opts[opt_key]

  if Util.empty_web_str?(in_file)
    if !empty_ok
      raise CLIError,"empty input path name[#{in_file}] in option[#{opt_key}]"
    end

    @cmd_opts[opt_key] = nil # nil is very important for BingScraper.init()!

    return true
  end

  in_file = Util.strip_web_str(in_file)

  if !File.exist?(in_file)
    raise CLIError,"input file[#{in_file}] does not exist for option[#{opt_key}]"
  end

  if File.directory?(in_file)
    raise CLIError,"input file[#{in_file}] cannot be a directory for option[#{opt_key}]"
  end

  return true
end

#check_out_dir(opt_key) ⇒ Object 



364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
 
# File 'lib/nhkore/app.rb', line 364

def check_out_dir(opt_key)
  out_dir = @cmd_opts[opt_key]

  if Util.empty_web_str?(out_dir)
    raise CLIError,"empty output directory[#{out_dir}] in option[#{opt_key}]"
  end

  out_dir = Util.strip_web_str(out_dir)

  if File.file?(out_dir)
    raise CLIError,"output directory[#{out_dir}] cannot be a file for option[#{opt_key}]"
  end

  if @cmd_opts[:dry_run]
    puts 'No changes written (dry run).'
    puts "> #{out_dir}"
    puts

    return true
  end

  force = @cmd_opts[:force]

  if !force && Dir.exist?(out_dir) && !Dir.empty?(out_dir)
    puts 'Warning: output directory already exists with files!'
    puts '       : Files inside of this directory may be overwritten!'
    puts "> '#{out_dir}'"

    return false unless @high.agree('Is this okay (yes/no)? ')
    puts
  end

  if !Dir.exist?(out_dir)
    if !force
      puts 'Output directory does not exist.'
      puts "> '#{out_dir}'"

      return false unless @high.agree('Create this directory (yes/no)? ')
    end

    FileUtils.mkdir_p(out_dir,verbose: true)
    puts
  end

  return true
end

#check_out_file(opt_key) ⇒ Object 



411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
 
# File 'lib/nhkore/app.rb', line 411

def check_out_file(opt_key)
  out_file = @cmd_opts[opt_key]

  if Util.empty_web_str?(out_file)
    raise CLIError,"empty output path name[#{out_file}] in option[#{opt_key}]"
  end

  out_file = Util.strip_web_str(out_file)

  if File.directory?(out_file)
    raise CLIError,"output file[#{out_file}] cannot be a directory for option[#{opt_key}]"
  end

  if @cmd_opts[:dry_run]
    puts 'No changes written (dry run).'
    puts "> #{out_file}"
    puts

    return true
  end

  force = @cmd_opts[:force]
  out_dir = File.dirname(out_file)

  if !force && File.exist?(out_file)
    puts 'Warning: output file already exists!'
    puts "> '#{out_file}'"

    return false unless @high.agree('Overwrite this file (yes/no)? ')
    puts
  end

  if !Dir.exist?(out_dir)
    if !force
      puts 'Output directory does not exist.'
      puts "> '#{out_dir}'"

      return false unless @high.agree('Create this directory (yes/no)? ')
    end

    FileUtils.mkdir_p(out_dir,verbose: true)
    puts
  end

  return true
end

#color(str) ⇒ Object 



458
459
460
 
# File 'lib/nhkore/app.rb', line 458

def color(str)
  return @rainbow.wrap(str)
end

#color_alias(str) ⇒ Object 



462
463
464
 
# File 'lib/nhkore/app.rb', line 462

def color_alias(str)
  return color(str).green
end

#enable_color(enabled) ⇒ Object 



466
467
468
469
 
# File 'lib/nhkore/app.rb', line 466

def enable_color(enabled)
  Cri::Platform.color = enabled
  @rainbow.enabled = enabled
end

#opts_to_set(ary) ⇒ Object 



471
472
473
474
475
476
477
478
 
# File 'lib/nhkore/app.rb', line 471

def opts_to_set(ary)
  set = Set.new

  set.add("-#{ary[0]}") unless ary[0].nil?
  set.add("--#{ary[1]}") unless ary[1].nil?

  return set
end

#refresh_cmd(opts, args, cmd) ⇒ Object 



480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
 
# File 'lib/nhkore/app.rb', line 480

def refresh_cmd(opts,args,cmd)
  new_opts = {}

  # Change symbols with dashes to underscores,
  #   so don't have to type @cmd_opts[:'dry-run'] all the time.
  opts.each do |key,value|
    # %s(max-retry) => :max_retry
    key = key.to_s.gsub('-','_').to_sym

    new_opts[key] = value
  end

  # For now don't set the default proc, as the original code
  #   did not have this in mind.
  # Specifically, SiftCmd.build_sift_filename() is affected by
  #   this due to relying on @cmd_opts[:ext] to be nil.
  #   It's easy to change this one instance, but I'm not sure
  #   at the moment where else might be affected
  #
  # # Cri has a default proc for default values
  # #   that doesn't store the keys.
  # new_opts.default_proc = proc do |hash,key|
  #   # :max_retry => %s(max-retry)
  #   key = key.to_s.gsub('_','-').to_sym
  #
  #   opts.default_proc.call(hash,key)
  # end

  @cmd = cmd
  @cmd_args = args
  @cmd_opts = new_opts

  return self
end

#run ⇒ Object 



515
516
517
 
# File 'lib/nhkore/app.rb', line 515

def run
  @app_cmd.run(@args)
end

#show_version ⇒ Object 



519
520
521
 
# File 'lib/nhkore/app.rb', line 519

def show_version
  puts "#{NAME} v#{VERSION}"
end

#sleep_scraper ⇒ Object 



523
524
525
526
527
528
529
 
# File 'lib/nhkore/app.rb', line 523

def sleep_scraper
  # Do a range to better emulate being a human.
  r = rand(@sleep_time..(@sleep_time + 0.1111))
  s = r.round(3) # Within 1000ms (0.000 - 0.999).

  sleep(s)
end

#start_spin(title, detail: '') ⇒ Object 



531
532
533
534
535
536
537
538
539
540
541
 
# File 'lib/nhkore/app.rb', line 531

def start_spin(title,detail: '')
  if .is_a?(Hash)
    [:detail] = detail
    [:title] = title

    puts(NO_SPINNER_MSG % )
  else
    .update(title: title,detail: detail)
    .auto_spin
  end
end

#stop_spin ⇒ Object 



543
544
545
546
547
548
549
550
 
# File 'lib/nhkore/app.rb', line 543

def stop_spin
  if .is_a?(Hash)
    puts "#{NO_SPINNER_MSG % @spinner} done!"
  else
    .reset
    .stop('done!')
  end
end

#update_spin_detail(detail) ⇒ Object 



552
553
554
555
556
557
558
559
560
 
# File 'lib/nhkore/app.rb', line 552

def update_spin_detail(detail)
  if .is_a?(Hash)
    [:detail] = detail

    puts(NO_SPINNER_MSG % )
  else
    .tokens[:detail] = detail
  end
end