Module: Persist

Defined in:
lib/rbbt/persist.rb,
lib/rbbt/persist/tsv.rb,
lib/rbbt/persist/tsv/cdb.rb,
lib/rbbt/persist/tsv/lmdb.rb,
lib/rbbt/persist/tsv/adapter.rb,
lib/rbbt/persist/tsv/leveldb.rb,
lib/rbbt/persist/tsv/sharder.rb,
lib/rbbt/persist/tsv/kyotocabinet.rb,
lib/rbbt/persist/tsv/packed_index.rb,
lib/rbbt/persist/tsv/tokyocabinet.rb,
lib/rbbt/persist/tsv/fix_width_table.rb,
lib/rbbt/persist/tsv/tokyocabinet/marshal.rb

Defined Under Namespace

Modules: CDBAdapter, FWTAdapter, KCAdapter, LMDBAdapter, LevelDBAdapter, PKIAdapter, SharderAdapter, TCAdapter, TSVAdapter Classes: Sharder

Constant Summary collapse

MEMORY = 
{}
MAX_FILE_LENGTH = 
150
TRUE_STRINGS = 
Set.new ["true", "True", "TRUE", "t", "T", "1", "yes", "Yes", "YES", "y", "Y", "ON", "on"]
CONNECTIONS = 
{}

Class Attribute Summary collapse

Class Method Summary collapse

Class Attribute Details

.cachedir ⇒ Object

Returns the value of attribute cachedir.



10
11
12
 
# File 'lib/rbbt/persist.rb', line 10

def cachedir
  @cachedir
end

.lock_dir ⇒ Object

Returns the value of attribute lock_dir.



18
19
20
 
# File 'lib/rbbt/persist.rb', line 18

def lock_dir
  @lock_dir
end

Class Method Details

.get_filename(source) ⇒ Object 



39
40
41
42
43
44
45
46
47
48
49
50
 
# File 'lib/rbbt/persist/tsv.rb', line 39

def self.get_filename(source)
  case
  when Path === source
    source
  when (source.respond_to?(:filename) and source.filename)
    source.filename
  when source.respond_to?(:cmd)
    "CMD-#{Misc.digest(source.cmd)}"
  when TSV === source
    "TSV[#{Misc.digest Misc.fingerprint(source)}]"
  end || source.object_id.to_s
end

.get_result(path, type, persist_options, lockfile, &block) ⇒ Object 



268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
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
 
# File 'lib/rbbt/persist.rb', line 268

def self.get_result(path, type, persist_options, lockfile, &block)
  res = yield path
  stream = res if IO === res
  stream = res.stream if res.respond_to? :stream

  if stream
    if persist_options[:no_load] == :stream 
      callback = stream.respond_to?(:callback)? stream.callback : nil
      abort_callback = stream.respond_to?(:abort_callback)? stream.abort_callback : nil

      # This is to avoid calling the callbacks twice, since they have been
      # moved to the new 'res' stream
      #stream.callback = nil
      #stream.abort_callback = nil

      res = tee_stream(stream, path, type, callback, abort_callback, lockfile)

      #res.lockfile = lockfile

      raise KeepLocked.new res 
    else
      stream = res.get_stream if res.respond_to? :get_stream
      begin
        Open.write(path, stream)
        Open.open(path) do |stream|
          case type
          when :array
            stream.read.split "\n"
          when :tsv
            TSV.open(stream)
          else
            stream.read
          end
        end
      rescue
        stream.abort if stream.respond_to? :abort
        raise $!
      end
    end
  else
    res
  end
end

.is_persisted?(path, persist_options = {}) ⇒ Boolean

Returns:

  • (Boolean) 


46
47
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
 
# File 'lib/rbbt/persist.rb', line 46

def self.is_persisted?(path, persist_options = {})
  return true if Open.remote?(path)
  return true if Open.ssh?(path)
  return false if not Open.exists? path
  return false if TrueClass === persist_options[:update]

  expiration = persist_options[:expiration]
  if expiration
    seconds = Misc.timespan(expiration)
    patht = Open.mtime(path)
    return false if Time.now > patht + seconds
  end

  check = persist_options[:check]
  return true if check.nil?

  missing = check.reject{|file| Open.exists?(file) }
  return false if missing.any?

  return true unless ENV["RBBT_UPDATE"]

  if Array === check
    newer = check.select{|file| newer? path, file}
    return true if newer.empty?
    Log.medium "Persistence check for #{path} failed in: #{ Misc.fingerprint(newer)}"
    return false 
  else
    ! newer?(path, check)
  end
end

.load_file(path, type) ⇒ Object 



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
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/rbbt/persist.rb', line 114

def self.load_file(path, type)
  begin
    case (type || :marshal).to_sym
    when :path
      Path.setup(Open.read(path).strip)
    when :nil
      nil
    when :boolean
      TRUE_STRINGS.include? Open.read(path).chomp.strip
    when :annotations
      Annotated.load_tsv TSV.open(path)
    when :tsv
      TSV.open(path)
    when :marshal_tsv
      TSV.setup(Marshal.load(Open.open(path)))
    when :fwt
      FixWidthTable.get(path) 
    when :string, :text
      Open.read(path)
    when :binary
      f = Open.open(path, :mode => 'rb')
      res = f.read
      f.close
      res.force_encoding("ASCII-8BIT") if res.respond_to? :force_encoding
      res
    when :array
      res = Open.read(path).split("\n", -1)
      res.pop if res.last and res.last.empty?
      res
    when :marshal
      Open.open(path) do |stream|
        content = stream.read.unpack("m").first
        Marshal.load(content) 
      end
    when :json
      Open.open(path) do |stream|
        JSON.parse(stream.read)
      end
    when :yaml
      Misc.load_yaml(path)
    when :float
      Open.read(path).to_f
    when :integer
      Open.read(path).to_i
    else
      raise "Unknown persistence: #{ type }"
    end
  rescue
    Log.medium "Exception loading #{ type } #{ path }: #{$!.message}"
    raise $!
  end
end

.memory(name, options = {}, &block) ⇒ Object 



505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
 
# File 'lib/rbbt/persist.rb', line 505

def self.memory(name, options = {}, &block)
  case options
  when nil
    persist name, :memory, :file => name, &block
  when String
    persist name, :memory, :file => name + "_" << options, &block
  else
    options = options.dup
    file = name
    repo = options.delete :repo if options and options.any?
    update = options.delete :update if options and options.any?
    file << "_" << (options[:key] ? options[:key] : Misc.hash2md5(options)) if options and options.any?
    persist name, :memory, {:repo => repo, :update => update, :persist => true, :file => file}.merge(options), &block
  end
end

.newer?(path, file, by_link = false) ⇒ Boolean

Is ‘file’ newer than ‘path’? return non-true if path is newer than file

Returns:

  • (Boolean) 


29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
 
# File 'lib/rbbt/persist.rb', line 29

def self.newer?(path, file, by_link = false)
  return true if not Open.exists?(file)
  path = path.find if Path === path
  file = file.find if Path === file
  if by_link
    patht = File.exist?(path) ? File.lstat(path).mtime : nil
    filet = File.exist?(file) ? File.lstat(file).mtime : nil
  else
    patht = Open.mtime(path)
    filet = Open.mtime(file)
  end
  return true if patht.nil? || filet.nil?
  diff = patht - filet
  return diff if diff < 0
  return false
end

.open_cdb(path, write, serializer = nil) ⇒ Object 



128
129
130
131
132
133
134
135
136
137
138
139
140
141
 
# File 'lib/rbbt/persist/tsv/cdb.rb', line 128

def self.open_cdb(path, write, serializer = nil)
  write = true unless File.exist? path

  FileUtils.mkdir_p File.dirname(path) unless File.exist?(File.dirname(path))

  database = Persist::CDBAdapter.open(path, write)

  #unless serializer == :clean
  #  TSV.setup database
  #  database.serializer = serializer if serializer
  #end

  database
end

.open_database(path, write, serializer = nil, type = "HDB", options = {}) ⇒ Object 



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
77
78
79
 
# File 'lib/rbbt/persist/tsv.rb', line 52

def self.open_database(path, write, serializer = nil, type = "HDB", options = {})
  case type
  when "LevelDB"
    Persist.open_leveldb(path, write, serializer)
  when "CDB"
    Persist.open_cdb(path, write, serializer)
  when "LMDB"
    Persist.open_lmdb(path, write, serializer)
  when 'kch', 'kct'
    Persist.open_kyotocabinet(path, write, serializer, type)
  when 'fwt'
    value_size, range, update, in_memory, pos_function = Misc.process_options options.dup, :value_size, :range, :update, :in_memory, :pos_function
    if pos_function
      Persist.open_fwt(path, value_size, range, serializer, update, in_memory, &pos_function)
    else
      Persist.open_fwt(path, value_size, range, serializer, update, in_memory)
    end
  when 'pki'
    pattern, pos_function = Misc.process_options options.dup, :pattern, :pos_function
    if pos_function
      Persist.open_pki(path, write, pattern, &pos_function)
    else
      Persist.open_pki(path, write, pattern)
    end
  else
    Persist.open_tokyocabinet(path, write, serializer, type)
  end
end

.open_fwt(path, value_size, range = false, serializer = nil, update = false, in_memory = false, &pos_function) ⇒ Object 



100
101
102
103
104
105
106
107
108
109
110
111
 
# File 'lib/rbbt/persist/tsv/fix_width_table.rb', line 100

def self.open_fwt(path, value_size, range = false, serializer = nil, update = false, in_memory = false, &pos_function)
  FileUtils.mkdir_p File.dirname(path) unless File.exist?(File.dirname(path))

  database = Persist::FWTAdapter.open(path, value_size, range, update, in_memory, &pos_function)

  unless serializer == :clean
    TSV.setup database
    database.serializer = serializer || database.serializer
  end

  database
end

.open_kyotocabinet(path, write, serializer = nil, kyotocabinet_class = 'kch') ⇒ Object 



80
81
82
83
84
85
86
87
88
89
90
91
92
93
 
# File 'lib/rbbt/persist/tsv/kyotocabinet.rb', line 80

def self.open_kyotocabinet(path, write, serializer = nil,  kyotocabinet_class= 'kch')
  write = true unless File.exist? path

  FileUtils.mkdir_p File.dirname(path) unless File.exist?(File.dirname(path))

  database = Persist::KCAdapter.open(path, write, kyotocabinet_class)

  unless serializer == :clean
    TSV.setup database
    database.serializer = serializer || database.serializer
  end

  database
end

.open_leveldb(path, write, serializer = nil) ⇒ Object 



106
107
108
109
110
111
112
113
114
115
116
117
118
119
 
# File 'lib/rbbt/persist/tsv/leveldb.rb', line 106

def self.open_leveldb(path, write, serializer = nil)
  write = true unless File.exist? path

  FileUtils.mkdir_p File.dirname(path) unless File.exist?(File.dirname(path))

  database = Persist::LevelDBAdapter.open(path, write)

  unless serializer == :clean
    TSV.setup database
    database.serializer = serializer || database.serializer
  end

  database
end

.open_lmdb(path, write, serializer = nil) ⇒ Object 



47
48
49
50
51
52
53
54
55
56
57
58
59
60
 
# File 'lib/rbbt/persist/tsv/lmdb.rb', line 47

def self.open_lmdb(path, write, serializer = nil)
  write = true unless File.exist? path

  FileUtils.mkdir_p File.dirname(path) unless File.exist?(File.dirname(path))

  database = Persist::LMDBAdapter.open(path, write)

  unless serializer == :clean
    TSV.setup database
    database.serializer = serializer || database.serializer
  end

  database
end

.open_pki(path, write, pattern, &pos_function) ⇒ Object 



98
99
100
101
102
103
104
105
106
107
108
 
# File 'lib/rbbt/persist/tsv/packed_index.rb', line 98

def self.open_pki(path, write, pattern, &pos_function)
  FileUtils.mkdir_p File.dirname(path) unless File.exist?(File.dirname(path))

  database = Persist::PKIAdapter.open(path, write, pattern, &pos_function)

  #TSV.setup database

  #database.serializer = :clean

  database
end

.open_sharder(path, write, serializer = nil, type = TokyoCabinet::HDB, options, &shard_function) ⇒ Object 



245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
 
# File 'lib/rbbt/persist/tsv/sharder.rb', line 245

def self.open_sharder(path, write, serializer = nil, type = TokyoCabinet::HDB, options, &shard_function)
  write = true unless File.exist? path

  FileUtils.mkdir_p File.dirname(path) unless File.exist?(File.dirname(path))

  database = Persist::SharderAdapter.open(path, write, type, options, &shard_function)

  if type.to_s == 'pki'
    TSV.setup database
    database.type = :list
    database.serializer = :clean 
  else
    if serializer != :clean 
      TSV.setup database
      database.serializer = serializer if serializer
    end
  end

  database
end

.open_tokyocabinet(path, write, serializer = nil, tokyocabinet_class = TokyoCabinet::HDB) ⇒ Object 



87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
 
# File 'lib/rbbt/persist/tsv/tokyocabinet.rb', line 87

def self.open_tokyocabinet(path, write, serializer = nil, tokyocabinet_class = TokyoCabinet::HDB)
  write = true unless File.exist? path

  FileUtils.mkdir_p File.dirname(path) unless File.exist?(File.dirname(path))

  database = Persist::TCAdapter.open(path, write, tokyocabinet_class)

  unless serializer == :clean
    TSV.setup database
    database.write_and_read do
      database.serializer = serializer
    end if serializer && database.serializer != serializer
  end

  database
end

.persist(name, type = nil, persist_options = {}, &block) ⇒ Object 



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
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
 
# File 'lib/rbbt/persist.rb', line 382

def self.persist(name, type = nil, persist_options = {}, &block)
  type ||= :marshal

  persist_options ||= {}
  if type == :memory && persist_options[:file] && persist_options[:persist] 
    repo = persist_options[:repo] || Persist::MEMORY
    if persist_options[:persist] == :update || persist_options[:update]
      repo.delete persist_options[:file]
    end
    return repo[persist_options[:file]] ||= yield
  end

  if FalseClass === persist_options[:persist]
    yield
  else
    persist_options[:update] = true if persist_options[:persist].to_s == "update"
    other_options = Misc.process_options persist_options, :other
    path = persistence_path(name, persist_options, other_options || {})

    if ENV["RBBT_UPDATE_TSV_PERSIST"] == 'true' and name and Open.exists?(name)
      persist_options[:check] ||= []
      persist_options[:check] << name
    else
      check_options = {}
    end

    case 
    when type.to_sym == :memory
      repo = persist_options[:repo] || Persist::MEMORY
      path = path.find if Path === path
      repo.delete path if persist_options[:update]
      repo[path] ||= yield

    when (type.to_sym == :annotations and (persist_options.include?(:annotation_repo) || persist_options.include?(:repo)))

      repo = persist_options[:annotation_repo] || persist_options[:repo]

      keys = nil
      subkey = name + ":"

      if String === repo
        repo = repo.find if Path === repo
        repo = Persist.open_tokyocabinet(repo, false, :list, "BDB")
        repo.read_and_close do
          keys = repo.range subkey + 0.chr, true, subkey + 254.chr, true
        end
      else
        repo.read_and_close do
          keys = repo.range subkey + 0.chr, true, subkey + 254.chr, true
        end
      end

      repo.read

      case
      when (keys.length == 1 and keys.first == subkey + 'NIL')
        nil
      when (keys.length == 1 and keys.first == subkey + 'EMPTY')
        []
      when (keys.length == 1 and keys.first =~ /:SINGLE$/)
        key = keys.first
        values = repo.with_read do
          repo[key]
        end
        Annotated.load_tsv_values(key, values, "literal", "annotation_types", "JSON")
      when (keys.any? and not keys.first =~ /ANNOTATED_DOUBLE_ARRAY/)
        repo.with_read do
          keys.sort_by{|k| k.split(":").last.to_i}.collect{|key|
            v = repo[key]
            Annotated.load_tsv_values(key, v, "literal", "annotation_types", "JSON")
          }
        end
      when (keys.any? and keys.first =~ /ANNOTATED_DOUBLE_ARRAY/)
        repo.with_read do

          res = keys.sort_by{|k| k.split(":").last.to_i}.collect{|key|
            v = repo[key]
            Annotated.load_tsv_values(key, v, "literal", "annotation_types", "JSON")
          }

          res.first.annotate res
          res.extend AnnotatedArray

          res
        end
      else
        entities = yield

        repo.write_and_read do 
          case
          when entities.nil?
            repo[subkey + "NIL"] = nil
          when entities.empty?
            repo[subkey + "EMPTY"] = nil
          when (not Array === entities or (AnnotatedArray === entities and not Array === entities.first))
            tsv_values = entities.tsv_values("literal", "annotation_types", "JSON") 
            repo[subkey + entities.id << ":" << "SINGLE"] = tsv_values
          when (not Array === entities or (AnnotatedArray === entities and AnnotatedArray === entities.first))
            entities.each_with_index do |e,i|
              next if e.nil?
              tsv_values = e.tsv_values("literal", "annotation_types", "JSON") 
              repo[subkey + "ANNOTATED_DOUBLE_ARRAY:" << i.to_s] = tsv_values
            end
          else
            entities.each_with_index do |e,i|
              next if e.nil?
              tsv_values = e.tsv_values("literal", "annotation_types", "JSON") 
              repo[subkey + i.to_s] = tsv_values
            end
          end
        end

        entities
      end

    else
      path = path.find if Path === path
      persist_file(path, type, persist_options, &block)
    end

  end
end

.persist_file(path, type, persist_options, &block) ⇒ Object 



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
375
376
377
378
379
380
 
# File 'lib/rbbt/persist.rb', line 312

def self.persist_file(path, type, persist_options, &block)
  Misc.insist do
    begin
      if is_persisted?(path, persist_options)
        Log.low "Persist up-to-date: #{ path } - #{Misc.fingerprint persist_options}"
        return path if persist_options[:no_load]
        return load_file(path, type) 
      else
        Open.rm path if Open.exists? path
      end
    rescue Aborted, Interrupt
      Log.warn "Aborted loading persistence (#{ type }) #{ path }: #{$!.message}. Not erasing."
      raise $!
    rescue Exception
      Log.warn "Exception loading persistence (#{ type }) #{ path }: #{$!.message}. Erase and retry."
      Open.rm path if Open.exists? path
      raise $!
    end
  end

  lock_filename = Persist.persistence_path(path + '.persist', {:dir => Persist.lock_dir})
  begin
    lock_options = Misc.pull_keys persist_options, :lock
    lock_options = lock_options[:lock] if Hash === lock_options[:lock]
    Misc.lock lock_filename, lock_options do |lockfile|
      Misc.insist do
        if is_persisted?(path, persist_options)
          Log.low "Persist up-to-date (suddenly): #{ path } - #{Misc.fingerprint persist_options}"
          lockfile.unlock if lockfile.locked?
          return path if persist_options[:no_load]
          return load_file(path, type) 
        end
      end

      Log.medium "Persist create: #{ path } - #{type} #{Misc.fingerprint persist_options}"

      res = get_result(path, type, persist_options, lockfile, &block)

      save_file(path, type, res, lockfile)

      Open.notify_write(path)

      return path if persist_options[:no_load] || type == :path

      res
    end

  rescue Lockfile::StolenLockError
    Log.medium "Lockfile stolen: #{path} - #{lock_filename}"
    Log.exception $!
    sleep 1 + rand(2)
    retry
  rescue TryAgain
    begin
      Open.rm path 
    rescue
    end if Open.exists? path 
    raise $!
  rescue Exception
    Log.medium "Error in persist: #{path}#{Open.exists?(path) ? Log.color(:red, " Erasing") : ""}"

    begin
      Open.rm path 
    rescue
    end if Open.exists? path 

    raise $!
  end
end

.persist_tsv(source, filename = nil, options = {}, persist_options = {}, &block) ⇒ Object 



81
82
83
84
85
86
87
88
89
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
124
125
126
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
172
173
174
175
176
177
178
179
180
 
# File 'lib/rbbt/persist/tsv.rb', line 81

def self.persist_tsv(source, filename = nil, options = {}, persist_options = {}, &block)
  persist_options[:prefix] ||= "TSV"

  if data = persist_options[:data]
    Log.debug "TSV persistence creating with data: #{ Misc.fingerprint(data) }"
    yield data
    return data 
  end

  filename ||= get_filename(source)

  if not persist_options[:persist]
    data = {}

    yield(data) 

    return data 
  end

  path = persistence_path(filename, persist_options, options)

  if ENV["RBBT_UPDATE_TSV_PERSIST"] == 'true' and filename
    check_options = {:check => [filename]}
  else
    check_options = {}
  end

  if is_persisted?(path, check_options) and not persist_options[:update]
    path = path.find if Path === path
    Log.debug "TSV persistence up-to-date: #{ path }"
    if persist_options[:shard_function]
      return open_sharder(path, false, nil, persist_options[:engine], persist_options, &persist_options[:shard_function]) 
    else
      return open_database(path, false, nil, persist_options[:engine] || TokyoCabinet::HDB, persist_options) 
    end
  end

  lock_filename = Persist.persistence_path(path, {:dir => TSV.lock_dir})
  Misc.lock lock_filename do
    begin
      if is_persisted?(path, check_options) and not persist_options[:update]
        path = path.find if Path === path
        Log.debug "TSV persistence (suddenly) up-to-date: #{ path }"

        if persist_options[:shard_function]
          return open_sharder(path, false, nil, persist_options[:engine], persist_options, &persist_options[:shard_function]) 
        else
          return open_database(path, false, nil, persist_options[:engine] || TokyoCabinet::HDB, persist_options) 
        end
      end
      path = path.find if Path === path

      FileUtils.rm_rf path if File.exist? path

      Log.medium "TSV persistence creating: #{ path }"

      tmp_path = path + '.persist'

      data = if persist_options[:shard_function]
               open_sharder(tmp_path, true, persist_options[:serializer], persist_options[:engine], persist_options, &persist_options[:shard_function]) 
             else
               open_database(tmp_path, true, persist_options[:serializer], persist_options[:engine] || TokyoCabinet::HDB, persist_options) 
             end

      if TSV === data and data.serializer.nil?
        data.serializer = :type 
      end

      if persist_options[:persist] == :preload
        tmp_tsv = yield({})
        tmp_tsv.annotate data
        data.serializer = tmp_tsv.type
        data.write_and_read do
          tmp_tsv.each do |k,v|
            data[k] = v
          end
        end
      else
        data.write_and_read do
          yield data
        end
      end

      data.write_and_read do
        FileUtils.mv data.persistence_path, path if File.exist? data.persistence_path and not File.exist? path
        tsv = CONNECTIONS[path] = CONNECTIONS.delete tmp_path
        tsv.persistence_path = path

        tsv.fix_io if tsv.respond_to? :fix_io
      end

      data
    rescue Exception
      Log.error "Captured error during persist_tsv. Erasing: #{path}"
      FileUtils.rm_rf tmp_path if tmp_path and File.exist? tmp_path
      FileUtils.rm_rf path if path and File.exist? path
      raise $!
    end
  end
end

.persistence_path(file, persist_options = {}, options = {}) ⇒ Object 



77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
 
# File 'lib/rbbt/persist.rb', line 77

def self.persistence_path(file, persist_options = {}, options = {})
  persistence_file = Misc.process_options persist_options, :file
  return persistence_file unless persistence_file.nil?

  prefix = Misc.process_options persist_options, :prefix

  if prefix.nil?
    perfile = file.to_s.gsub(/\//, '>') 
  else
    perfile = prefix.to_s + ":" + file.to_s.gsub(/\//, '>') 
  end

  perfile.sub!(/\.b?gz$/,'')

  if options.include? :filters
    options[:filters].each do |match,value|
      perfile = perfile + "&F[#{match}=#{Misc.digest(value.inspect)}]"
    end
  end

  persistence_dir = Misc.process_options(persist_options, :dir) || Persist.cachedir 
  Path.setup(persistence_dir) unless Path === persistence_dir

  filename = perfile.gsub(/\s/,'_').gsub(/\//,'>')
  clean_options = options.dup
  clean_options.delete :unnamed
  clean_options.delete "unnamed"

  filename = filename[0..MAX_FILE_LENGTH] << Misc.digest(filename[MAX_FILE_LENGTH+1..-1]) if filename.length > MAX_FILE_LENGTH + 10

  options_md5 = Misc.hash2md5 clean_options
  filename  << ":" << options_md5 unless options_md5.empty?

  persistence_dir[filename]
end

.save_file(path, type, content, lockfile = nil) ⇒ Object 



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
 
# File 'lib/rbbt/persist.rb', line 167

def self.save_file(path, type, content, lockfile = nil)
  if content.nil?
    lockfile.unlock if lockfile and lockfile.locked?
    return
  end

  case (type || :marshal).to_sym
  when :path
    Open.write(path, content)
  when :nil
    nil
  when :boolean
    Misc.sensiblewrite(path, content ? "true" : "false", :lock => lockfile)
  when :fwt
    content.file.seek 0
    Misc.sensiblewrite(path, content.file.read, :lock => lockfile)
  when :tsv
    content = content.to_s if TSV === content
    Misc.sensiblewrite(path, content, :lock => lockfile)
  when :annotations
    Misc.sensiblewrite(path, Annotated.tsv(content, :all).to_s, :lock => lockfile)
  when :string, :text
    Misc.sensiblewrite(path, content, :lock => lockfile)
  when :binary
    content.force_encoding("ASCII-8BIT") if content.respond_to? :force_encoding
    f = Open.open(path, :mode => 'wb')
    f.puts content
    f.close
    content
  when :array
    case content
    when Array
      if content.empty?
        Misc.sensiblewrite(path, "", :lock => lockfile)
      else
        Misc.sensiblewrite(path, content * "\n" + "\n", :lock => lockfile)
      end
    when IO
      Misc.sensiblewrite(path, content, :lock => lockfile)
    else
      Misc.sensiblewrite(path, content.to_s, :lock => lockfile)
    end
  when :marshal_tsv
    Misc.sensiblewrite(path, Marshal.dump(content.dup), :lock => lockfile)
  when :marshal
    dump = Marshal.dump(content)
    Misc.sensiblewrite(path, [dump].pack("m"), :lock => lockfile)
  when :json
    Misc.sensiblewrite(path, JSON.dump(content), :lock => lockfile)
  when :yaml
    Misc.sensiblewrite(path, YAML.dump(content), :lock => lockfile)
  when :float, :integer, :tsv
    Misc.sensiblewrite(path, content.to_s, :lock => lockfile)
  else
    raise "Unknown persistence: #{ type }"
  end
end

.tee_stream_thread(stream, path, type, callback = nil, abort_callback = nil, lockfile = nil) ⇒ Object Also known as: tee_stream 



225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
 
# File 'lib/rbbt/persist.rb', line 225

def self.tee_stream_thread(stream, path, type, callback = nil, abort_callback = nil, lockfile = nil)
  file, out = Misc.tee_stream(stream)

  out.pair = file
  file.pair = out

  saver_thread = Thread.new do
    begin
      file.threads = []
      Thread.current["name"] = "file saver: " + path
      save_file(path, type, file, lockfile)
    rescue Aborted
      Log.medium "Persist stream thread aborted: #{ Log.color :blue, path }"
      file.abort if file.respond_to? :abort
      raise $!
    rescue Exception
      Log.medium "Persist stream thread exception: #{ Log.color :blue, path }"
      file.abort if file.respond_to? :abort
      raise $!
    rescue Exception
      Log.exception $!
      raise $!
    end
  end

  threads = [saver_thread]
  threads += stream.threads if stream.respond_to?(:threads) && stream.threads
  ConcurrentStream.setup(out, :threads => threads, :filename => path)

  #out.callback = callback
  out.abort_callback = abort_callback
  out.lockfile = stream.lockfile if stream.respond_to? :lockfile and stream.lockfile

  #stream.callback = callback
  #stream.abort_callback = abort_callback

  out
end