Top Level Namespace

Defined Under Namespace

Modules: AWS, S3CP Classes: ProxyIO

Constant Summary collapse

DEBUG = 
true

Instance Method Summary collapse

Instance Method Details

#copy(from, to, options) ⇒ Object 



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
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
 
# File 'lib/s3cp/s3cp.rb', line 488

def copy(from, to, options)
  bucket_from, key_from = S3CP.bucket_and_key(from)
  bucket_to, key_to = S3CP.bucket_and_key(to)

  case direction(from, to)
  when :s3_to_s3
    if options[:recursive]
      keys = []
      @s3.buckets[bucket_from].objects.with_prefix(key_from).each do |entry|
        keys << entry.key
      end
      keys.each do |key|
        if match(key)
          dest = key_path key_to, relative(key_from, key)
          if !options[:overwrite] && s3_exist?(bucket_to, dest)
            $stderr.puts "Skipping s3://#{bucket_to}/#{dest} - already exists."
          else
            s3_to_s3(bucket_from, key, bucket_to, dest, options)
          end
        end
      end
    else
      key_to += File.basename(key_from) if key_to[-1..-1] == "/"
      if !options[:overwrite] && s3_exist?(bucket_to, key_to)
        $stderr.puts "Skipping s3://#{bucket_to}/#{key_to} - already exists."
      else
        s3_to_s3(bucket_from, key_from, bucket_to, key_to, options)
      end
    end
  when :local_to_s3
    if options[:recursive]
      files = Dir[from + "/**/*"]
      files.each do |f|
        if File.file?(f) && match(f)
          #puts "bucket_to #{bucket_to}"
          #puts "no_slash(key_to) #{no_slash(key_to)}"
          #puts "relative(from, f) #{relative(from, f)}"
          key = key_path key_to, relative(from, f)
          if !options[:overwrite] && s3_exist?(bucket_to, key)
            $stderr.puts "Skipping s3://#{bucket_to}/#{key} - already exists."
          else
            local_to_s3(bucket_to, key, File.expand_path(f), options)
          end
        end
      end
    else
      key_to += File.basename(from) if key_to[-1..-1] == "/"
      if !options[:overwrite] && s3_exist?(bucket_to, key_to)
        $stderr.puts "Skipping s3://#{bucket_to}/#{key_to} - already exists."
      else
        local_to_s3(bucket_to, key_to, File.expand_path(from), options)
      end
    end
  when :s3_to_local
    if options[:recursive]
      keys = []
      @s3.buckets[bucket_from].objects.with_prefix(key_from).each do |entry|
        keys << entry.key
      end
      keys.each do |key|
        if match(key)
          dest = if options[:mkdir]
            suffix = mkdir_relative(options[:mkdir], key) or next
            dest = File.join(File.expand_path(to), suffix)
          else
            dest = File.expand_path(to) + '/' + relative(key_from, key)
            dest = File.join(dest, File.basename(key)) if File.directory?(dest)
            dest
          end
          dir = File.dirname(dest)
          FileUtils.mkdir_p dir unless File.exist? dir
          fail "Destination path is not a directory: #{dir}" unless File.directory?(dir)
          if !options[:overwrite] && File.exist?(dest)
            $stderr.puts "Skipping #{dest} - already exists."
          else
            s3_to_local(bucket_from, key, dest, options)
          end
        end
      end
    else
      dest = if options[:mkdir]
        suffix = mkdir_relative(options[:mkdir], key_from) or
          fail("Error: The key '#{key_from}' does not match the --mkdir regular expression '#{options[:mkdir]}'")
        dest = File.join(File.expand_path(to), suffix)
        dir = File.dirname(dest)
        FileUtils.mkdir_p dir unless File.exist? dir
        dest
      else
        dest = File.expand_path(to)
        dest = File.join(dest, File.basename(from)) if File.directory?(dest)
        dest
      end
      if !options[:overwrite] && File.exist?(dest)
        $stderr.puts "Skipping #{dest} - already exists."
      else
        s3_to_local(bucket_from, key_from, dest, options)
      end
    end
  when :local_to_local
    if options[:include_regex].any? || options[:exclude_regex].any?
      fail "Include/exclude not supported on local-to-local copies"
    end
    if options[:recursive]
      FileUtils.cp_r from, to
      FileUtils.rm_r from if options[:move]
    else
      FileUtils.cp from, to
      FileUtils.rm from if options[:move]
    end
  end
end

#debug(str) ⇒ Object 



44
45
46
 
# File 'lib/s3cp/completion.rb', line 44

def debug(str)
  @debug.puts(str) if @debug
end

#depth(path) ⇒ Object 



71
72
73
 
# File 'lib/s3cp/s3du.rb', line 71

def depth(path)
  path.count("/")
end

#direction(from, to) ⇒ Object 



211
212
213
214
215
216
217
218
219
220
221
 
# File 'lib/s3cp/s3cp.rb', line 211

def direction(from, to)
  if s3?(from) && s3?(to)
    :s3_to_s3
  elsif !s3?(from) && s3?(to)
    :local_to_s3
  elsif s3?(from) && !s3?(to)
    :s3_to_local
  else
    :local_to_local
  end
end

#key_path(prefix, key) ⇒ Object 



476
477
478
479
480
481
482
 
# File 'lib/s3cp/s3cp.rb', line 476

def key_path(prefix, key)
  if (prefix.nil? || prefix.strip == '')
    key
  else
    no_slash(prefix) + '/' + key
  end
end

#local_to_s3(bucket_to, key, file, options = {}) ⇒ Object 



283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
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
363
364
365
366
367
368
369
370
371
372
373
374
 
# File 'lib/s3cp/s3cp.rb', line 283

def local_to_s3(bucket_to, key, file, options = {})
  log(with_headers("#{operation(options)} #{file} to s3://#{bucket_to}/#{key}"))

  expected_md5 = if options[:checksum] || options[:sync]
     S3CP.md5(file)
  end

  actual_md5 = if options[:sync]
    md5 = s3_checksum(bucket_to, key)
    case md5
    when :not_found
      nil
    when :invalid
      $stderr.puts "Warning: No MD5 checksum available and ETag not suitable due to multi-part upload; file will be force-copied."
      nil
    else
      md5
    end
  end

  if actual_md5.nil? || (options[:sync] && expected_md5 != actual_md5)
    retries = 0
    begin
      if retries == options[:retries]
        fail "Unable to upload to s3://#{bucket_to}/#{key} after #{retries} attempts."
      end
      if retries > 0
        delay = options[:retry_delay] * (options[:retry_backoff] ** retries)
        $stderr.puts "Sleeping #{"%0.2f" % delay} seconds.  Will retry #{options[:retries] - retries} more time(s)."
        sleep delay
      end

      begin
        obj = @s3.buckets[bucket_to].objects[key]

        s3_options = {}
        S3CP.set_header_options(s3_options, @headers)
        s3_options[:acl] = options[:acl] if options[:acl]
        s3_options[:content_length] = File.size(file)

        multipart_threshold = options[:multipart].is_a?(Fixnum) ?  options[:multipart] : AWS.config.s3_multipart_threshold
        if (expected_md5 != nil) && (File.size(file) >= multipart_threshold) && options[:multipart]
          meta = s3_options[:metadata] || {}
          meta[:md5] = expected_md5
          s3_options[:metadata] = meta
        end

        if options[:multipart]
          s3_options[:multipart_threshold] = multipart_threshold
        else
          s3_options[:single_request] = true
        end

        progress_bar = if $stdout.isatty
          ProgressBar.new(File.basename(file), File.size(file)).tap do |p|
            p.file_transfer_mode
          end
        end

        File.open(file) do |io|
          obj.write(ProxyIO.new(io, progress_bar), s3_options)
        end

        progress_bar.finish if progress_bar

        if options[:checksum]
          actual_md5 = s3_checksum(bucket_to, key)
          if actual_md5.is_a? String
            if actual_md5 != expected_md5
              $stderr.puts "Warning: invalid MD5 checksum.  Expected: #{expected_md5} Actual: #{actual_md5}"
            end
          else
            $stderr.puts "Warning: invalid MD5 checksum in metadata: #{actual_md5.inspect}; skipped checksum verification."
            actual_md5 = nil
          end
        end
      rescue => e
        actual_md5 = "bad"
        if progress_bar
          progress_bar.clear
          $stderr.puts "Error copying #{file} to s3://#{bucket_to}/#{key}"
        end
        raise e if !options[:checksum] || e.is_a?(AWS::S3::Errors::AccessDenied)
        $stderr.puts e
      end
      retries += 1
    end until options[:checksum] == false || actual_md5.nil? || expected_md5 == actual_md5
  else
    log "Already synchronized."
  end
  FileUtils.rm file if options[:move]
end

#log(msg) ⇒ Object 



248
249
250
 
# File 'lib/s3cp/s3cp.rb', line 248

def log(msg)
  puts msg if @verbose
end

#match(path) ⇒ Object 



223
224
225
226
227
228
 
# File 'lib/s3cp/s3cp.rb', line 223

def match(path)
  matching = true
  return false if @includes.any? && !@includes.any? { |regex| regex.match(path) }
  return false if @excludes.any? &&  @excludes.any? { |regex| regex.match(path) }
  true
end

#mkdir_relative(regex, key) ⇒ Object 



242
243
244
245
246
 
# File 'lib/s3cp/s3cp.rb', line 242

def mkdir_relative(regex, key)
  m = regex.match(key) or return
  pos = m.captures.empty? ? m.begin(0) : m.begin(1)
  suffix = key[pos..-1]
end

#no_slash(path) ⇒ Object 



230
231
232
233
 
# File 'lib/s3cp/s3cp.rb', line 230

def no_slash(path)
  path = path.match(/\/$/) ? path[0..-2] : path
  path.match(/^\//) ? path[1..-1] : path
end

#nth_occurrence(str, substr, n) ⇒ Object

Returns the index of the nth occurrence of substr if it exists, otherwise -1



76
77
78
79
80
81
82
83
84
85
86
 
# File 'lib/s3cp/s3du.rb', line 76

def nth_occurrence(str, substr, n)
  pos = -1
  if n > 0 && str.include?(substr)
    i = 0
    while i < n do
      pos = str.index(substr, pos + substr.length) if pos != nil
      i += 1
    end
  end
  pos != nil && pos != -1 ? pos + 1 : -1
end

#operation(options) ⇒ Object 



262
263
264
265
266
267
 
# File 'lib/s3cp/s3cp.rb', line 262

def operation(options)
  operation = "Copy"
  operation = "Move" if options[:move]
  operation = "Sync" if options[:sync]
  operation
end

#print(key, size) ⇒ Object 



93
94
95
96
 
# File 'lib/s3cp/s3du.rb', line 93

def print(key, size)
  size = S3CP.format_filesize(size, :unit => @options[:unit], :precision => @options[:precision])
  puts ("%#{7 + @options[:precision]}s " % size) + key
end

#print_buckets(buckets) ⇒ Object 



88
89
90
91
92
93
94
95
96
97
98
99
100
 
# File 'lib/s3cp/completion.rb', line 88

def print_buckets(buckets)
  buckets = buckets.map { |b| @legacy_format ? b + ":" : b + "/"  }
  buckets << buckets[0] + " " if buckets.size == 1
  buckets.each do |bucket|
    if @legacy_format
      debug bucket
      puts bucket
    else
      debug "//#{bucket}"
      puts "//#{bucket}"
    end
  end
end

#print_keys(bucket, keys) ⇒ Object 



75
76
77
78
79
80
81
82
83
84
85
86
 
# File 'lib/s3cp/completion.rb', line 75

def print_keys(bucket, keys)
  keys << keys[0] + " " if keys.size == 1 && @dirs_only
  keys.each do |key|
    if @legacy_format
      debug key
      puts key
    else
      debug "//#{bucket}/#{key}"
      puts "//#{bucket}/#{key}"
    end
  end
end

#relative(base, path) ⇒ Object

relative(“path/to/”, “path/to/file”) => “file” relative(“path/to”, “path/to/file”) => “to/file”



237
238
239
240
 
# File 'lib/s3cp/s3cp.rb', line 237

def relative(base, path)
  dir = base.rindex("/") ? base[0..base.rindex("/")] : ""
  no_slash(path[dir.length..-1])
end

#rule_to_str(r) ⇒ Object 



82
83
84
85
86
87
88
89
90
 
# File 'lib/s3cp/s3lifecycle.rb', line 82

def rule_to_str(r)
  if r.expiration_time
    [ r.prefix || "[root]", r.id, r.status, time_or_date_str("Expire", r.expiration_time)]
  elsif r.glacier_transition_time
    [ r.prefix || "[root]", r.id, r.status, time_or_date_str("Glacier", r.glacier_transition_time)]
  else
    [ r.prefix || "[root]", r.id, r.status, "???"]
  end
end

#s3?(url) ⇒ Boolean

Returns:

  • (Boolean) 


196
197
198
 
# File 'lib/s3cp/s3cp.rb', line 196

def s3?(url)
  S3CP.bucket_and_key(url)[0]
end

#s3_checksum(bucket, key) ⇒ Object 



457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
 
# File 'lib/s3cp/s3cp.rb', line 457

def s3_checksum(bucket, key)
  begin
    metadata = @s3.buckets[bucket].objects[key].head()
    return :not_found unless metadata
  rescue => e
    return :not_found if e.is_a?(AWS::S3::Errors::NoSuchKey)
    raise e
  end

  case
  when metadata[:meta] && metadata[:meta]["md5"]
    metadata[:meta]["md5"]
  when metadata[:etag] && metadata[:etag] !~ /-/
    metadata[:etag].sub(/^"/, "").sub(/"$/, "") # strip beginning and trailing quotes
  else
    :invalid
  end
end

#s3_exist?(bucket, key) ⇒ Boolean

Returns:

  • (Boolean) 


453
454
455
 
# File 'lib/s3cp/s3cp.rb', line 453

def s3_exist?(bucket, key)
  @s3.buckets[bucket].objects[key].exists?
end

#s3_size(bucket, key) ⇒ Object 



484
485
486
 
# File 'lib/s3cp/s3cp.rb', line 484

def s3_size(bucket, key)
  @s3.buckets[bucket].objects[key].content_length
end

#s3_to_local(bucket_from, key_from, dest, options = {}) ⇒ Object

Raises:

  • (ArgumentError) 


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
410
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
 
# File 'lib/s3cp/s3cp.rb', line 376

def s3_to_local(bucket_from, key_from, dest, options = {})
  log("#{operation(options)} s3://#{bucket_from}/#{key_from} to #{dest}")
  raise ArgumentError, "source key may not be blank" if key_from.to_s.empty?

  retries = 0
  begin
    if retries == options[:retries]
      File.delete(dest) if File.exist?(dest)
      fail "Unable to download s3://#{bucket_from}/#{key_from} after #{retries} attempts."
    end
    if retries > 0
      delay = options[:retry_delay] * (options[:retry_backoff] ** retries)
      delay = delay.to_i
      $stderr.puts "Sleeping #{"%0.2f" % delay} seconds.  Will retry #{options[:retries] - retries} more time(s)."
      sleep delay
    end
    begin
      expected_md5 = if options[:checksum] || options[:sync]
        md5 = s3_checksum(bucket_from, key_from)
        if options[:sync] && !md5.is_a?(String)
          $stderr.puts "Warning: invalid MD5 checksum in metadata; file will be force-copied."
          nil
        elsif !md5.is_a? String
          $stderr.puts "Warning: invalid MD5 checksum in metadata; skipped checksum verification."
          nil
        else
          md5
        end
      end

      actual_md5 = if options[:sync] && File.exist?(dest)
         S3CP.md5(dest)
      end

      if !options[:sync] || expected_md5.nil? || (expected_md5 != actual_md5)
        f = File.new(dest, "wb")
        begin
          progress_bar = if $stdout.isatty
            size = s3_size(bucket_from, key_from)
            ProgressBar.new(File.basename(key_from), size).tap do |p|
              p.file_transfer_mode
            end
          end
          @s3.buckets[bucket_from].objects[key_from].read_as_stream do |chunk|
            f.write(chunk)
            progress_bar.inc chunk.size if progress_bar
          end
          progress_bar.finish if progress_bar
        rescue => e
          progress_bar.halt if progress_bar
          raise e
        ensure
          f.close()
        end
      else
        log("Already synchronized")
        return
      end
    rescue => e
      raise e if e.is_a?(AWS::S3::Errors::NoSuchKey)
      raise e unless options[:checksum]
      $stderr.puts e
    end

    if options[:checksum] && expected_md5 != nil
      actual_md5 = S3CP.md5(dest)
      if actual_md5 != expected_md5
        $stderr.puts "Warning: invalid MD5 checksum.  Expected: #{expected_md5} Actual: #{actual_md5}"
      end
    end

    retries += 1
  end until options[:checksum] == false || expected_md5.nil? || S3CP.md5(dest) == expected_md5

  @s3.buckets[bucket_from].objects[key_from].delete() if options[:move]
end

#s3_to_s3(bucket_from, key, bucket_to, dest, options = {}) ⇒ Object 



269
270
271
272
273
274
275
276
277
278
279
280
281
 
# File 'lib/s3cp/s3cp.rb', line 269

def s3_to_s3(bucket_from, key, bucket_to, dest, options = {})
  log(with_headers("#{operation(options)} s3://#{bucket_from}/#{key} to s3://#{bucket_to}/#{dest}"))
  s3_source = @s3.buckets[bucket_from].objects[key]
  s3_dest = @s3.buckets[bucket_to].objects[dest]
  s3_options = {}
  S3CP.set_header_options(s3_options, @headers)
  s3_options[:acl] = options[:acl] if options[:acl]
  unless options[:move]
    s3_source.copy_to(s3_dest, s3_options)
  else
    s3_source.move_to(s3_dest, s3_options)
  end
end

#time_or_date_str(msg, t) ⇒ Object 



74
75
76
77
78
79
80
 
# File 'lib/s3cp/s3lifecycle.rb', line 74

def time_or_date_str(msg, t)
  if t.is_a?(Fixnum) || t.to_s =~ /^\d+$/
    "%s after %d days" % [msg, t.to_i]
  else
    "%s on %s" % [msg, t]
  end
end

#with_headers(msg) ⇒ Object 



254
255
256
257
258
259
260
 
# File 'lib/s3cp/s3cp.rb', line 254

def with_headers(msg)
  unless @headers.empty?
    msg += " with headers:"
    msg += @headers.collect{|k,v| "'#{k}: #{v}'"}.join(", ")
  end
  msg
end