Class: RbSync

Inherits:
Object
  • Object
show all
Defined in:
lib/rbsync.rb

Overview

progress 表示のために fileutils.copy を 自作する

Instance Attribute Summary collapse

Instance Method Summary collapse

Constructor Details

#initializeRbSync

Returns a new instance of RbSync.



101
102
103
104
105
106
107
108
# File 'lib/rbsync.rb', line 101

def initialize()
  @conf ={}
  @conf[:update] = false
  @conf[:excludes] = []
  @conf[:preserve] = true
  @conf[:overwrite] = true
  @conf[:strict] = true
end

Instance Attribute Details

#confObject

Returns the value of attribute conf.



86
87
88
# File 'lib/rbsync.rb', line 86

def conf
  @conf
end

Instance Method Details

#check_hash=(flag) ⇒ Object

flag true or false



468
469
470
# File 'lib/rbsync.rb', line 468

def check_hash=(flag)
  @conf[:use_md5_digest] = flag
end

#collet_hash(file_names, basedir, options = {}) ⇒ Object



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
230
231
232
233
234
235
236
237
238
239
# File 'lib/rbsync.rb', line 204

def collet_hash(file_names,basedir,options={})
  #prepare
  require 'thread'
  self.patch_digest_base
  threads =[]
  output = Hash.new{|s,key| s[key]=[] }
  q      = Queue.new
  limitsize = options[:hash_limit_size]
  # compute digests
  file_names.each{|e| q.push e }
  3.times{
    threads.push(
      Thread.start{
        while(!q.empty?)
          name = q.pop
          #$stdout.puts "reading #{name}" if options[:verbose]
          #$stdout.flush if options[:verbose]
          hash = compute_digest_file(File.expand_path(name,basedir),limitsize)
          output[hash].push name
        end
      }
    )
  }
  if options[:verbose] then
    t = Thread.start{
        until(q.empty?)
          puts( "#{q.size}/#{file_names.size}");
          $stdout.flush;
          sleep 1;
        end
    } 
    threads.push(t )
  end
  threads.each{|t|t.join}
  return output
end

#compute_digest_file(filename, limitsize = nil) ⇒ Object

compute digest md5

limitsize

If file size is very large,
and a few bytes at head of file is enough to compare.
for speed-up, Set limit size to enable to avoid reading a whole of file.
もしファイルがとても巨大で、かつ、先頭の数キロバイトが比較に十分であれば、limitsize 以降をスキップする


265
266
267
# File 'lib/rbsync.rb', line 265

def compute_digest_file(filename, limitsize=nil)
    Digest::MD5.open(filename,limitsize).hexdigest
end

#copy_r(files) ⇒ Object



354
355
356
357
358
359
360
361
362
363
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
410
411
412
413
414
415
416
417
418
419
# File 'lib/rbsync.rb', line 354

def copy_r(files)
  #
  puts ("copy #{files.size} files") if(@conf[:progress])
  $stdout.flush                     if(@conf[:progress])

  files.each_with_index{|e,i|
    #show
      puts ("start #{i+1}/#{files.size}")if(@conf[:progress])
      $stdout.flush if(@conf[:progress])
    #main
    tmp_name = "#{e[1]}.copy_tmp"
    FileUtils.rm(tmp_name) if File.exists?(tmp_name)
    copy_thread = Thread.start{
      FileUtils.mkdir_p File.dirname(e[1]) unless File.exists?(File.dirname(e[1]))
      ## todo copy file as stream for progress
      begin
        FileUtils.copy( e[0] , tmp_name ,{:preserve=>self.preserve?,:verbose=>self.verbose? } )
        FileUtils.mv(tmp_name,e[1])
        rescue Errno::EACCES => err
        puts e[1];puts err
      end
    }
    
    #progress of each file
    puts "#{e[0]}" if self.verbose? || self.debug?
    progress_thread = nil
    if(@conf[:progress])
      progress_thread = Thread.start{
        bar = ProgressBar.new
        bar.size = 30
        src_size = File.size(e[0])
        dst_size = -1
        bar.start("copying #{e[0]} \r\n   to #{e[1]}")
        cnt = 0
        dst_name = tmp_name
        while(src_size!=dst_size)
          dst_name = e[1] if File.exists?(e[1]) and not File.exists?(tmp_name)
          unless File.exists?(dst_name) then
            cnt = cnt + 1
            if cnt > 25 then
              puts "copying #{e[1]} is terminated.\r\n timeout error"
              throw Error
              break
            end
            sleep 0.2
            next
          end
          src_size = File.size(e[0]).to_f
          dst_size = File.size(dst_name).to_f
          break if src_size == 0 # preven zero divide
          # next  if dst_size == 0 # preven zero divide
          percent = dst_size/src_size*100
          bar.progress(percent.to_int)
          sleep 0.6
        end
        src_size = File.size(e[0]).to_f
        dst_size = File.size(e[1]).to_f
        percent = dst_size/src_size*100
        bar.progress(percent.to_int)
        bar.end("done")
      }
    end
    progress_thread.join if progress_thread 
    copy_thread.join
  }
end

#debug_mode=(flag) ⇒ Object Also known as: debug=

flag true or false



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

def debug_mode=(flag)
  self.conf[:debug] = flag
end

#debug_mode?Boolean Also known as: debug?

for setting

Returns:

  • (Boolean)


442
443
444
# File 'lib/rbsync.rb', line 442

def debug_mode?
  self.conf[:debug] ==true
end

#excludesObject



478
479
480
# File 'lib/rbsync.rb', line 478

def excludes
  @conf[:excludes]
end

#excludes=(glob_pattern) ⇒ Object



475
476
477
# File 'lib/rbsync.rb', line 475

def excludes=(glob_pattern)
  @conf[:excludes].push glob_pattern
end

#find_as_relative(dir_name, excludes = []) ⇒ Object

collect file paths. paths are relatetive path.



110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
# File 'lib/rbsync.rb', line 110

def find_as_relative(dir_name,excludes=[])
  files =[]
  excludes = [] unless excludes
  #todo write this two line . exculude initialize test
  excludes = excludes.split(",") if excludes.class == String
  excludes = [excludes]          unless excludes.class == Array
  
  Dir.chdir(dir_name){ 
    files = Dir.glob "./**/*", File::FNM_DOTMATCH
    exclude_files =[]
    exclude_files = excludes.map{|g| Dir.glob "./**/#{g}",File::FNM_DOTMATCH } 
    files = files.reject{|e| File.directory?(e)  }
    files = files - exclude_files.flatten
  }
  files = files.reject{|e| [".",".."].any?{|s| s== File::basename(e)  }}
end

#find_files(src, dest, options) ⇒ Object

compare two directory by name and FileUtis.cmp



127
128
129
130
131
132
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
# File 'lib/rbsync.rb', line 127

def find_files(src,dest,options)
  src_files  = self.find_as_relative(  src, options[:excludes] )
  dest_files = self.find_as_relative( dest, options[:excludes] )

  # output target files
  puts "  元フォルダ:"  +  src_files.size.to_s + "" if self.debug?
  puts "同期先フォルダ:"  + dest_files.size.to_s + "" if self.debug?
  #pp src_files if self.debug?
  sleep 1 if self.debug?

  #両方にあるファイル名で中身が違うもので src の方が古いもの
  same_name_files = (dest_files & src_files)
  same_name_files.reject!{|e|
      #ファイルが同じモノは省く
      next unless File.exists?( File.expand_path(e,dest))
      puts "compare file bin.  #{e}" if self.debug? || self.verbose?
      $stdout.flush if self.debug?
      FileUtils.cmp( File.expand_path(e,src) , File.expand_path(e,dest) ) 
  } if options[:strict]
  same_name_files.reject!{|e|
      #ファイルサイズが同じモノを省く(全部比較する代替手段)
      next unless File.exists?( File.expand_path(e,dest))
      puts "size/mtime compare #{e}" if self.debug? || self.verbose?
      File.size(File.expand_path(e,src)) == File.size( File.expand_path(e,dest))
      #&& File.mtime(File.expand_path(e,src)) == File.mtime( File.expand_path(e,dest) )
  } unless options[:strict]
  if options[:update] then
    same_name_files= same_name_files.select{|e|
        puts "mtime is newer   #{e}" if self.debug? || self.verbose?
        (File.mtime(File.expand_path(e,src)) > File.mtime( File.expand_path(e,dest)))
    }
  end
  if options[:overwrite] == false then
    same_name_files= same_name_files.reject{|e|
        puts "can over write?  #{e}" if self.debug? || self.verbose?
        (File.exists?(File.expand_path(e,src)) && File.exists?( File.expand_path(e,dest)))
    }
  end
  $stdout.flush if self.debug?
  files_not_in_dest = (src_files - dest_files)
  #files
  files =[]
  files = (files_not_in_dest + same_name_files ).flatten
  files
end

#hash_limit_sizeObject



448
449
450
# File 'lib/rbsync.rb', line 448

def hash_limit_size
  @conf[:hash_limit_size]
end

#hash_limit_size=(int_byte_size) ⇒ Object



471
472
473
# File 'lib/rbsync.rb', line 471

def hash_limit_size=(int_byte_size)
  @conf[:hash_limit_size] = int_byte_size
end

#overwrite=(flag) ⇒ Object



487
488
489
# File 'lib/rbsync.rb', line 487

def overwrite=(flag)
  @conf[:overwrite] = flag
end

#overwrite?Boolean

Returns:

  • (Boolean)


490
491
492
# File 'lib/rbsync.rb', line 490

def overwrite?
  @conf[:overwrite]
end

#patch_digest_baseObject



240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
# File 'lib/rbsync.rb', line 240

def patch_digest_base()
    require 'digest/md5'
    s = %{
    class Digest::Base
      def self.open(path,limitsize=nil)
        obj = new
        File.open(path, 'rb') do |f|
          buf = ""
          while f.read(256, buf)
            obj << buf
            break if f.pos > (limitsize or f.pos+1)
          end
        end
        return obj
      end
    end
    }
    eval s
end

#preserve=(flag) ⇒ Object



481
482
483
# File 'lib/rbsync.rb', line 481

def preserve=(flag)
  @conf[:preserve] = false
end

#preserve?Boolean

Returns:

  • (Boolean)


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

def preserve?
  @conf[:preserve]
end

#sync(src, dest, options = {}) ⇒ Object



420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
# File 'lib/rbsync.rb', line 420

def sync(src,dest,options={})
  options[:excludes]        = self.excludes.push(options[:excludes]).flatten.uniq if options[:excludes]
  options[:update]          = @conf[:update] if options[:update] == nil
  options[:strict]          = @conf[:strict] if options[:strict] == nil
  options[:check_hash]      = options[:check_hash] or @conf[:check_hash]
  options[:hash_limit_size] = @conf[:hash_limit_size]                       if options[:hash_limit_size] == nil
  options[:overwrite]       = @conf[:overwrite]                             if options[:overwrite] == nil
  options[:overwrite]       = false                                         if options[:no_overwrite]
  FileUtils.mkdir_p dest unless File.exists? dest
  if options[:rename]
    return self.sync_by_anothername(src,dest,options)
  elsif options[:backup]
    return self.sync_with_backup(src,dest,options)
  elsif options[:check_hash]
    return self.sync_by_hash(src,dest,options)
  else
    return self.sync_normally(src,dest,options)
  end
end

#sync_by_anothername(src, dest, options) ⇒ Object

別名で名前をつけて転送する



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
# File 'lib/rbsync.rb', line 323

def sync_by_anothername(src,dest,options)
  # 上書き付加の場合
  # 
  #ファイル一覧を取得する
  files = find_as_relative(src,options[:excludes])
  #中身が同じモノを排除
  files = files.reject{|e|
    FileUtils.cmp(File.expand_path(e,src) , File.expand_path(e,dest))
  }
  if options[:update] then
    #更新日が当たらしいモノを排除
    files = files.reject{|e|
      File.mtime(File.expand_path(e,src)) < File.mtime(File.expand_path(e,dest))
    }
  end
  #別名をつける
  files = files.map{|e|
    extname = File.extname(e)
    basename = File.basename(e).gsub(extname,"")
    candidate = ""
    100.times{|i|
      candidate = File.expand_path("#{basename}(#{i+1})#{extname}",dest)
      break unless File.exists?(File.expand_path(candidate,dest))
      raise "upto #{i} files are already exists ." if  i >=1000
      next FileUtils.cmp(File.expand_path(e,src) , File.expand_path(candidate,dest))
    }
    [File.expand_path(e,src) , File.expand_path(candidate,dest)]
  }
  #コピーする
  self.copy_r(files)
end

#sync_by_hash(src, dest, options = {}) ⇒ Object



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
# File 'lib/rbsync.rb', line 172

def sync_by_hash(src,dest,options={})
  src_files   = collet_hash(find_as_relative(src, options[:excludes]), src, options)
  dest_files  = collet_hash(find_as_relative(dest,options[:excludes]),dest, options)
  target  = src_files.keys - dest_files.keys
  target = target.reject{|key|
    e = src_files[key].first
    options[:update] && 
    File.exists?( File.expand_path(e,dest)) &&
    (File.mtime(File.expand_path(e,src)) < File.mtime( File.expand_path(e,dest)))
  }
  if options[:overwrite] == false then
    target = target .reject{|key|
        e = src_files[key].first
        (File.exists?(File.expand_path(e,src)) && File.exists?( File.expand_path(e,dest)))
    }
  end
  puts "同期対象ファイル" if self.debug?
  puts target.each{|key|puts src_files[key].first} if self.debug?
  puts "同期対象はありません" if self.debug? and target.size==0

  files= target.map{|key|
      e = src_files[key].first
      [File.expand_path(e,src) , File.expand_path(e,dest)]
  }
  self.copy_r(files)

  ret = files.map{|e|
    FileTest.exist?(e[1])
  }
  puts "同期が終りました"     if ret.select{|e|!e}.size == 0 && self.debug?
  puts "同期に失敗したみたい" if ret.select{|e|!e}.size != 0 && self.debug?
end

#sync_normally(src, dest, options = {}) ⇒ Object

called from sync



270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
# File 'lib/rbsync.rb', line 270

def sync_normally(src,dest,options={})
  Thread.abort_on_exception = true if self.debug?
  files = self.find_files(src,dest,options)
  puts "同期対象のファイルはありません" if self.debug? && files.size==0
  return true if files.size == 0
  puts "次のファイルを同期します" if self.debug?
  pp files                        if self.debug?
  
  #srcファイルをdestに上書き
  #todo options を取り出す
  self.copy_r(files.map{|e|[File.expand_path(e,src) , File.expand_path(e,dest)]})
  
  #checking sync result
  files = self.find_files(src,dest,options)

  puts "同期が終りました"       if files.size == 0 && self.debug?
  puts "同期に失敗がありました" if files.size != 0 && self.debug?
  pp files                      if files.size != 0 && self.debug?
  return files.size == 0
end

#sync_with_backup(src, dest, options) ⇒ Object

同期先に同名ファイルがあったらファイルを別名にバックアップしてから転送します



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
# File 'lib/rbsync.rb', line 291

def sync_with_backup(src,dest,options)
  # 上書き付加の場合
  # 
  #ファイル一覧を取得する
  files = find_as_relative(src,options[:excludes])
  #中身が同じモノを排除
  files = files.reject{|e|
    FileUtils.cmp(File.expand_path(e,src) , File.expand_path(e,dest))
  }
  #更新日が当たらしいモノを排除
  if options[:update] then
    #更新日が当たらしいモノを排除
    files = files.reject{|e|
      File.mtime(File.expand_path(e,src)) < File.mtime(File.expand_path(e,dest))
    }
  end
  #別名をつける
  files = files.map{|e|
    extname = File.extname(e)
    basename = File.basename(e).gsub(extname,"")
      # 同名のファイルがあった場合
      # ファイルをリネームする
      if File.exists?(File.expand_path(e,dest)) then
          candidate = File.expand_path("#{basename}_#{Time.now.strftime('%Y%m%d%H%M%S')}#{extname}",dest)
          File.rename( File.expand_path(e,dest),candidate )
      end
    [File.expand_path(e,src) , File.expand_path(e,dest)]
  }
  #コピーする
  self.copy_r(files)
end

#updated_file_only=(flag) ⇒ Object Also known as: update=, newer=

flag true or false



463
464
465
# File 'lib/rbsync.rb', line 463

def updated_file_only=(flag)
  @conf[:update] = flag
end

#verbose=(flag) ⇒ Object

flag true or false



457
458
459
460
# File 'lib/rbsync.rb', line 457

def verbose=(flag)
  self.conf[:verbose] = flag
  $stdout.sync = flag
end

#verbose?Boolean

Returns:

  • (Boolean)


445
446
447
# File 'lib/rbsync.rb', line 445

def verbose?
  self.conf[:verbose] == true
end