Class: CoreEnumerationAlgo

Inherits:
Object
  • Object
show all
Defined in:
lib/mbt-gen.rb

Class Method Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize ⇒ CoreEnumerationAlgo

Returns a new instance of CoreEnumerationAlgo.



1128
1129
1130
1131
1132
1133
 
# File 'lib/mbt-gen.rb', line 1128

def initialize()
  @num_tries = 0
  @num_docs = 0
  @on_new_doc = nil
  @on_unsat = nil
end

Class Method Details

.parse_assignment(str) ⇒ Object 



1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
 
# File 'lib/mbt-gen.rb', line 1266

def self.parse_assignment(str)
  unless str =~ /assignment\((.*)\)/
    return nil
  end
  split = $1.split(",")
  result = {}
  split.each do |e|
    s2 = e.split(":")
    key = s2[0]
    value = s2[1]
    if value == "nil" then
      result[key] = nil
    elsif value == "true" then
      result[key] = true
    elsif value == "false" then
      result[key] = false
    elsif value =~ /\"(.*)\"/
      result[key] = $1
    end
  end
  result
end

.parse_continuation_point(line) ⇒ Object 



1312
1313
1314
1315
1316
1317
1318
1319
1320
 
# File 'lib/mbt-gen.rb', line 1312

def self.parse_continuation_point(line)
  split = line.split(";")
  doc_number = Integer(split[0])
  num_docs = Integer(split[1])
  state = CoreEnumerationAlgo.parse_state(split[2])
  assignment = CoreEnumerationAlgo.parse_assignment(split[3])
  new_assignment = CoreEnumerationAlgo.parse_assignment(split[4])
  return [doc_number, num_docs, state, assignment, new_assignment]
end

.parse_new_assignment(str) ⇒ Object 



1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
 
# File 'lib/mbt-gen.rb', line 1289

def self.parse_new_assignment(str)
  unless str =~ /new_assignment\((.*)\)/
    return nil
  end
  split = $1.split(",")
  result = {}
  split.each do |e|
    s2 = e.split(":")
    key = s2[0]
    value = s2[1]
    if value == "nil" then
      result[key] = nil
    elsif value == "true" then
      result[key] = true
    elsif value == "false" then
      result[key] = false
    elsif value =~ /\"(.*)\"/
      result[key] = $1
    end
  end
  result
end

.parse_state(state_str) ⇒ Object 



1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
 
# File 'lib/mbt-gen.rb', line 1242

def self.parse_state(state_str)
  unless state_str =~ /state\((.*)\)/
    return nil
  end
  split = $1.split("#")
  split.map do |e|
    if e == "nil" then
      nil
    elsif e =~ /\[(.*)\]/ then
      $1.split(",").map do |i|
        if i.strip == "nil" then
          nil
        elsif i.strip == "true" then
          true
        elsif i.strip == "false" then
          false
        elsif i.strip =~ /\"(.*)\"/ then
          $1
        end
      end
    end
  end
end

Instance Method Details

#determine_progress(state, vars) ⇒ Object

NOTE: can only be called when all key-fields are finite-state



1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
 
# File 'lib/mbt-gen.rb', line 1389

def determine_progress(state, vars)
  combinations_per_value = [nil] * vars.size()
  comb = 1
  vars.size().times.to_a.reverse_each do |i|
    combinations_per_value[i] = comb
    comb = values_for_key_field(vars[i][1]).size() * comb
  end

  blocked_combinations = 0
  state.size().times do |i|
    if state[i] then
      blocked_combinations += state[i].size() * combinations_per_value[i]
    end
  end
  blocked_combinations
end

#display_combination_space(assignment, vars, switched, blocked) ⇒ Object

TODO: logging



1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
 
# File 'lib/mbt-gen.rb', line 1324

def display_combination_space(assignment, vars, switched, blocked)
  # enumerate_fixing_clause(assignment, vars, switched) + enumerate_blocking_clause(vars[switched][0], vars[switched][1], blocked)

  # => all vars left of switched are fixed to their assigned values (when the value is nil, then they are empty)
  # => the switched var is prevented to have any of the blocked values
  res = switched.times.to_a.map do |index|
    var_xpath = vars[index][0]
    "#{var_xpath}=#{assignment[var_xpath].inspect}"
  end + ["#{vars[switched][0]}\{#{blocked.map{|v| v.inspect}.join(",")}}"]
  "{#{res.join(", ")}}"
end

#enumerate(progress, switched, vars, assignment, message, key, output_dir, config, options, state = [nil] * vars.size()) ⇒ Object

core algo



1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
 
# File 'lib/mbt-gen.rb', line 1337

def enumerate(progress, switched, vars, assignment, message, key, output_dir, config, options, state = [nil] * vars.size())
  enumerate(progress, switched+1, vars, assignment, message, key, output_dir, config, options, state) if switched < vars.size()-1
  is_finite_state = keys_are_finite_state(vars)
  if state[switched] == nil then
    blocked = [assignment[vars[switched][0]]]
    state[switched] = blocked
  else
    blocked = state[switched] + [assignment[vars[switched][0]]]
    state[switched] = blocked
  end
  found = true
  while found do
    found = false
    # - fix all vars left of switched to their values in assignment
    # - block switched from obtaining the values in blocked
    # progress.print_line "DEBUG: #{summarize_fixing_clause(assignment, vars, switched)}, #{summarize_blocking_clause(vars[switched][0], blocked)}"
    clauses = enumerate_fixing_clause(assignment, vars, switched) + enumerate_blocking_clause(vars[switched][0], vars[switched][1], blocked)
    output_doc_filename = prepare_output_doc_filename(@num_tries, output_dir)
    new_assignment = generate_sparse_doc(progress, message, key, config, options, clauses, output_doc_filename)
    generated_doc_number = @num_tries
    store_continuation_point(generated_doc_number, state, assignment, new_assignment, "#{output_dir}/#{CONTINUATION_FILE}")
    if is_finite_state then
      progress.report_progress(determine_progress(state, vars))
    else
      progress.report_progress(nil)
    end
    @num_tries += 1
    if new_assignment then # sat
      @num_docs += 1
      @on_new_doc.call(generated_doc_number, output_doc_filename, new_assignment) if @on_new_doc
      if switched < vars.size()-1 then
        enumerate(progress, switched+1, vars, new_assignment, message, key, output_dir, config, options, state)
      end
      blocked << new_assignment[vars[switched][0]]
      state[switched] = blocked
      found = true
    else # unsat
      @on_unsat.call(generated_doc_number, log_filename_for_output_filename(output_doc_filename), display_combination_space(assignment, vars, switched, blocked)) if @on_unsat
      found = false
    end
  end
  state[switched] = nil
end

#enumerate_blocking_clause(var_xpath, var_def, blocked) ⇒ Object

prevent the var from taking values that already have been enumerated exhaustively



1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
 
# File 'lib/mbt-gen.rb', line 1164

def enumerate_blocking_clause(var_xpath, var_def, blocked)
  clauses = []
  num = 0
  if var_def[:type] == :field then
    blocked.each do |value|
      if value == nil then
        clauses << "(assert (! (not (= #{filled_var_for_field(var_xpath)} false)) :named blocking_clause_#{num}))"
      else
        clauses << "(assert (! (not (and (= #{filled_var_for_field(var_xpath)} true) (= #{value_var_for_field(var_xpath)} #{value_of_type(value, var_def[:datatype])}))) :named blocking_clause_#{num}))"
      end
      num += 1
    end
  elsif var_def[:type] == :structure then
    blocked.each do |value|
      clauses << "(assert (! (not (= #{exists_var_for_structure(var_xpath)} #{value.inspect})) :named blocking_clause_#{num}))"
      num += 1
    end
  end
  "; blocking clauses:\n#{clauses.join("\n")}" 
end

#enumerate_fixing_clause(assignment, vars, switched) ⇒ Object

fix values of all variables left from switched



1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
 
# File 'lib/mbt-gen.rb', line 1137

def enumerate_fixing_clause(assignment, vars, switched)
  clauses = []
  switched.times do |index|
    var_xpath = vars[index][0]
    var_def = vars[index][1]
    if var_def[:type] == :field then
      if assignment[var_xpath] == nil then
        clauses << "(assert (! (= #{filled_var_for_field(var_xpath)} false) :named fixing_clause_#{index}))"
      else
        clauses << "(assert (! (and (= #{filled_var_for_field(var_xpath)} true) (= #{value_var_for_field(var_xpath)} #{value_of_type(assignment[var_xpath], var_def[:datatype])})) :named fixing_clause_#{index}))"
      end
    elsif var_def[:type] == :structure then
      clauses << "(assert (! (= #{exists_var_for_structure(var_xpath)} #{assignment[var_xpath].inspect}) :named fixing_clause_#{index}))"
    end
  end
  "; fixing clauses:\n#{clauses.join("\n")}" 
end

#generate_doc_return_key_values(progress, message, key, output_filename, config, options) ⇒ Object 



1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
 
# File 'lib/mbt-gen.rb', line 1201

def generate_doc_return_key_values(progress, message, key, output_filename, config, options)
  log = options[:log] || nil
  log.puts "generating doc #{output_filename}" if log
  options[:solverLog] = log_filename_for_output_filename(output_filename)

  key_fields = collect_key_fields(key, message)

  z3path = config[:z3path] || DEFAULT_Z3_PATH
  z3 = Z3Solver.new(progress, z3path)
  begin
    model = z3.query_model(progress, options) do |solver|
      set_standard_options(progress, solver)
      log.puts "translating structure" if log
      translate_structure_to_SMTLIB(progress, solver, message, :options => options, :config => config)
      solver.to_solver(progress, "")
    end
  ensure
    z3.close()
  end

  if model then
    result_doc = build_document_from_model(progress, message, model)
    dir = File.dirname(output_filename)
    FileUtils.mkdir_p(dir) unless Dir.exist?(dir)
    File.open(output_filename, "w") do |f|
      f.puts result_doc.to_xml(:save_with => Nokogiri::XML::Node::SaveOptions::AS_XML)
    end
    progress.print_line "DONE. Generated XML was written to #{output_filename}, SMTLIB log was written to #{options[:solverLog]}"
    return extract_key_field_values_from_model(key_fields, model)
  else
    progress.print_line "FAILED. No XML document was generated. For more details, please see the SMTLIB log in #{options[:solverLog]}"
    return nil
  end
end

#generate_docs_for_key(message, key, output_dir, config, options) ⇒ Object 



1440
1441
1442
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
1487
1488
1489
1490
1491
1492
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505
1506
1507
1508
1509
1510
1511
1512
1513
1514
1515
1516
1517
1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
1584
 
# File 'lib/mbt-gen.rb', line 1440

def generate_docs_for_key(message, key, output_dir, config, options)

  number_of_combinations = calc_combinations(message, key)
  if number_of_combinations.is_a?(String) then
    progress = ProgressBar.new(100000000000000)
  else
    progress = ProgressBar.new(number_of_combinations)
  end
  begin
  ensure
  end
  progress.report_progress(0.0)
  progress.print_line "generating sparse documents efficiently..."
  progress.print_line "there is a theoretical maximum of #{number_of_combinations} combinations."
  progress.print_line "Note, however, that we are generating only those documents that are valid, which are usually much less."

  progress.print_line "preparing output dir #{output_dir}..."
  if options[:continue] then
    unless Dir.exist?(output_dir) then
      raise FatalError.new("FATAL: directory #{output_dir} not found. There seems to be nothing to continue from...")
    end
    log_mode = "a"
  else
    unless Dir.exist?(output_dir) then
      FileUtils.mkdir_p(output_dir)
    end
    log_mode = "w"
  end
  log_filename = "#{output_dir}/#{COMBINATION_LOG}"

  key_fields = collect_key_fields(key, message)
  key_fields_list = key_fields.each_pair.to_a
  key_fields_list.sort! {|a,b| num_combinations(b[1]) <=> num_combinations(a[1]) }

  if options[:continue] then
    stored_key_fields_str = File.read("#{output_dir}/#{VARS_LOG}")
    unless key_fields_list.to_yaml == stored_key_fields_str then
      raise FatalError.new("vars are differing! please specify the same key in order to use -continue!")
    end
    progress.print_line "continuing enumerating documents for key #{key}!"
  else
    # store vars
    File.open("#{output_dir}/#{VARS_LOG}", "w") do |f|
      f.puts(key_fields_list.to_yaml)
    end
  end

  cont_filename = "#{output_dir}/#{CONTINUATION_FILE}"
  if options[:continue] then
    unless File.exist?(cont_filename) then
      raise FatalError.new("FATAL: file #{cont_filename} not found. There seems to be no continuation point to continue from...")
    end
    doc_number, num_docs, state, assignment, new_assignment = CoreEnumerationAlgo.parse_continuation_point(File.read(cont_filename).chomp)
    if is_final_state(state, new_assignment, key_fields_list) then
      progress.report_progress(number_of_combinations)
      raise FatalError.new("FATAL: The enumeration is finished. There is nothing left to do. Resuming the enumeration hence does not make sense.")
    end
    progress.print_line "RESUMING enumeration from the following continuation point:"
    progress.print_line "|doc_number=#{doc_number}"
    progress.print_line "|num_docs=#{num_docs}"
    progress.print_line "|state=#{state.inspect}"
    progress.print_line "|assignment=#{assignment.inspect}"
    progress.print_line "|new_assignment=#{new_assignment.inspect}"
    if keys_are_finite_state(key_fields_list) then
      progress.report_progress(determine_progress(state, key_fields_list))
    else
      progress.report_progress(nil)
    end

    if new_assignment && !new_assignment.empty? then
      assignment = new_assignment
    end

    # restore state at continuation point
    @num_tries = doc_number+1 # we do not want to overwrite the last document written before the STOP
    number_of_combinations += 1 # account for this in the safety mechanism
    @num_docs = num_docs
    doc_counter = 0
    File.open(log_filename, log_mode) do |log|

      @on_new_doc = Proc.new { |num_doc, doc_filename, combination|
        # combination logging
        log.puts "#{doc_filename} <- {#{combination.map{|k,v| "#{k}=#{v.inspect}"}.join(", ")}}"

        doc_counter += 1
        if options[:max_num_docs] && doc_counter >= options[:max_num_docs] then
          raise NormalProgramTermination.new("-max-num-docs exceeded. Stopping!")
        end

        # safety mechanism
        if @num_docs > number_of_combinations then
          raise RuntimeError.new("This should never happen! Number of combinations exceeded. #{@num_docs} documents generated, but there are only #{number_of_combinations} possible combinations of the key vars.")
        end
      }
      @on_unsat = Proc.new { |num_doc, log_filename, combination_str|
            log.puts "CONTRADICTION (see: #{log_filename}) <- #{combination_str}"
            progress.print_line "doc#{num_doc} NOT generated"
            progress.print_line "No more valid combinations of the free vars could be found. -> Switching"
      }

      # core algo - resume from state
      enumerate(progress, 0, key_fields_list, assignment, message, key, output_dir, config, options, state)
    end
  else
    # init safety mechanism
    @num_tries = 0
    @num_docs = 0
    doc_counter = 0
    # first try - a freely generated document
    File.open(log_filename, log_mode) do |log|
      output_doc_filename = prepare_output_doc_filename(@num_tries, output_dir)
      key_values = generate_doc_return_key_values(progress, message, key, output_doc_filename, config, options)
      log.puts "#{output_doc_filename} <- {#{key_values.map{|k,v| "#{k}=#{v.inspect}"}.join(", ")}}"
      doc_counter += 1
      @num_tries += 1
      @on_new_doc = Proc.new { |num_doc, doc_filename, combination|
        # combination logging
        log.puts "#{doc_filename} <- {#{combination.map{|k,v| "#{k}=#{v.inspect}"}.join(", ")}}"

        doc_counter += 1
        if options[:max_num_docs] && doc_counter >= options[:max_num_docs] then
          raise NormalProgramTermination.new("-max-num-docs exceeded. Stopping!")
        end

        # safety mechanism
        unless number_of_combinations.is_a?(String) then
          if @num_docs > number_of_combinations then
            raise RuntimeError.new("This should never happen! Number of combinations exceeded. #{@num_docs} documents generated, but there are only #{number_of_combinations} possible combinations of the key vars.")
          end
        end
      }
      @on_unsat = Proc.new { |num_doc, log_filename, combination_str|
            log.puts "CONTRADICTION (see: #{log_filename}) <- #{combination_str}"
            progress.print_line "doc#{num_doc} NOT generated"
            progress.print_line "No more valid combinations of the free vars could be found. -> Switching"
      }

      # core algo - start enumeration
      enumerate(progress, 0, key_fields_list, key_values, message, key, output_dir, config, options)
    end
    
    progress.print_line "No more valid documents could be found. -> Exiting"
    progress.report_progress(number_of_combinations)
  end
end

#generate_sparse_doc(progress, message, key, config, options, code, output_doc_filename) ⇒ Object 



1189
1190
1191
1192
1193
1194
1195
 
# File 'lib/mbt-gen.rb', line 1189

def generate_sparse_doc(progress, message, key, config, options, code, output_doc_filename)
  additional_options = {}
  if code then
      additional_options = { :add_SMTLIB => code }
  end
  return generate_doc_return_key_values(progress, message, key, output_doc_filename, config, options.merge(additional_options))
end

#is_final_state(state, new_assignment, key_fields_list) ⇒ Object

NOTE: can only be called when all key-fields are finite-state



1382
1383
1384
1385
1386
 
# File 'lib/mbt-gen.rb', line 1382

def is_final_state(state, new_assignment, key_fields_list)
  return state[1,state.size()] == [nil] * (state.size()-1) &&
     state[0].to_set == values_for_key_field(key_fields_list[0][1]).to_set &&
     new_assignment.empty?
end

#keys_are_finite_state(list) ⇒ Object 



1424
1425
1426
1427
1428
1429
1430
1431
1432
1433
1434
1435
1436
1437
1438
 
# File 'lib/mbt-gen.rb', line 1424

def keys_are_finite_state(list)
  list.each do |field|
    name = field[0]
    if field[1][:type] == :field then
      if field[1][:datatype] == :enum then
        # enums are finite-state
      elsif field[1][:datatype] == :int and field[1].has_key?(:min) and field[1].has_key?(:max) and field[1][:min] <= field[1][:max] then
        # ints are finite-state when both bounds are given and :min <= :max
      else
        return false
      end
    end
  end
  return true
end

#log_filename_for_output_filename(output_filename) ⇒ Object 



1197
1198
1199
 
# File 'lib/mbt-gen.rb', line 1197

def log_filename_for_output_filename(output_filename)
  "#{output_filename}.smt2"
end

#num_combinations(key_def) ⇒ Object 



1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
 
# File 'lib/mbt-gen.rb', line 1406

def num_combinations(key_def)
  if key_def[:type] == :field then
    if key_def[:datatype] == :enum then
      key_def[:values].size() + (key_def[:optional] ? 1 : 0)
    elsif key_def[:datatype] == :int then
      if key_def.has_key?(:min) and key_def.has_key?(:max) and key_def[:min] <= key_def[:max] then
        1 + key_def[:max] - key_def[:min] + (key_def[:optional] ? 1 : 0)
      else
        nil
      end
    else
      nil
    end
  elsif key_def[:type] == :structure then
    2
  end
end

#store_continuation_point(doc_number, state, assignment, new_assignment, filename) ⇒ Object 



1236
1237
1238
1239
1240
 
# File 'lib/mbt-gen.rb', line 1236

def store_continuation_point(doc_number, state, assignment, new_assignment, filename)
  File.open(filename, "w") do |f|
    f.puts("#{doc_number};#{@num_docs};state(#{state.map{|s| if s then "[#{s.map{|v| v.inspect}.join(",")}]" else "nil" end}.join("#")});assignment(#{assignment.to_a.map{|k,v| "#{k}:#{v.inspect}"}.join(",")});new_assignment(#{new_assignment.to_a.map{|k,v| "#{k}:#{v.inspect}"}.join(",")})")
  end
end

#summarize_blocking_clause(var_xpath, blocked) ⇒ Object 



1185
1186
1187
 
# File 'lib/mbt-gen.rb', line 1185

def summarize_blocking_clause(var_xpath, blocked)
  "blocked(#{var_xpath})=[#{blocked.map{|v| v.inspect}.join(", ")}]"
end

#summarize_fixing_clause(assignment, vars, switched) ⇒ Object 



1155
1156
1157
1158
1159
1160
1161
 
# File 'lib/mbt-gen.rb', line 1155

def summarize_fixing_clause(assignment, vars, switched)
  fixed = switched.times.to_a.map do |index|
    var_xpath = vars[index][0]
    "#{var_xpath}=#{assignment[var_xpath].inspect}"
  end
  "fixed(#{fixed.join(", ")})"
end