Class: CorrectHorseBatteryStaple::Corpus::Sqlite

Inherits:
Base show all
Defined in:
lib/correct_horse_battery_staple/corpus/sqlite.rb

Constant Summary collapse

MAX_ITERATIONS = 
1000

Instance Attribute Summary

Attributes inherited from Base

#frequency_mean, #frequency_stddev, #original_size, #probability_mean, #probability_stddev, #weighted_size

Class Method Summary collapse

Instance Method Summary collapse

Methods inherited from Base

#candidates, #compose_filters, #count, #count_by_options, #count_candidates, #entropy_per_word, #entropy_per_word_by_filter, #filter, #filter_for_options, #inspect, #load_stats_from_hash, #precache, #recalculate, #reset, #result, #stats, #words

Methods included from Memoize

included

Methods included from CorrectHorseBatteryStaple::Common

#array_sample, #logger, #random_in_range, #random_number, #set_sample

Methods inherited from CorrectHorseBatteryStaple::Corpus

format_for

Constructor Details

#initialize(file) ⇒ Sqlite

Returns a new instance of Sqlite.



7
8
9
10
11
12
 
# File 'lib/correct_horse_battery_staple/corpus/sqlite.rb', line 7

def initialize(file)
  super
  @db = SQLite3::Database.open file
  @statements = []
  load_stats
end

Class Method Details

.read(file) ⇒ Object 



14
15
16
 
# File 'lib/correct_horse_battery_staple/corpus/sqlite.rb', line 14

def self.read(file)
  self.new file
end

Instance Method Details

#_pick_discrete_n(percentile, length, count = 1) ⇒ Object 



194
195
196
197
198
199
 
# File 'lib/correct_horse_battery_staple/corpus/sqlite.rb', line 194

def _pick_discrete_n(percentile, length, count = 1)
  statement = prepare "select #{COLUMNS.join(", ")} from entries where " +
    " percentile = ? and wordlength = ? and randunit < ? limit ?"

  statement.execute!(percentile, length, random_number, count)
end

#_pick_discrete_n_ids(percentile, length, count = 1) ⇒ Object 



232
233
234
235
236
237
 
# File 'lib/correct_horse_battery_staple/corpus/sqlite.rb', line 232

def _pick_discrete_n_ids(percentile, length, count = 1)
  statement = prepare "select id from entries where " +
    " percentile = ? and wordlength = ? and randunit > ? limit ?"

  statement.execute!(percentile, length, random_number, count)
end

#close ⇒ Object 



241
242
243
244
 
# File 'lib/correct_horse_battery_staple/corpus/sqlite.rb', line 241

def close
  @statements.each { |x| x.close }
  super
end

#each(&block) ⇒ Object

some core Enumerable building blocks



20
21
22
 
# File 'lib/correct_horse_battery_staple/corpus/sqlite.rb', line 20

def each(&block)
  entries.each &block
end

#entries ⇒ Object

our own collection operations



32
33
34
 
# File 'lib/correct_horse_battery_staple/corpus/sqlite.rb', line 32

def entries
  @entries ||= table
end

#frequencies ⇒ Object 



40
41
42
 
# File 'lib/correct_horse_battery_staple/corpus/sqlite.rb', line 40

def frequencies
  @frequencies ||= @db.execute("select frequency from entries").map {|x| x.first}
end

#get_words_for_ids(ids) ⇒ Object 



98
99
100
101
102
103
104
105
106
107
 
# File 'lib/correct_horse_battery_staple/corpus/sqlite.rb', line 98

def get_words_for_ids(ids)
  ids = Array(ids)
  rows = @db.execute("select #{COLUMNS.join(", ")} from entries where id in (#{ids.join(',')})")

  words = []
  ids.each do |id|
    words << rows.find {|r| r[0] == id }
  end
  words.map {|row| word_from_row(row)}
end

#pick(count, options = {}) ⇒ Object

optimized pick variants - they do NOT support :filter, though

Raises:

  • (NotImplementedError) 


47
48
49
50
51
52
53
 
# File 'lib/correct_horse_battery_staple/corpus/sqlite.rb', line 47

def pick(count, options = {})
  # incompat check
  raise NotImplementedError, "SQLite does not support :filter option" if options[:filter]

  strategy = options.delete(:strategy) || ENV['pick_strategy'] || "discrete"
  send("pick_#{strategy}", count, options)
end

#pick_discrete(count, options = {}) ⇒ Object

discrete method



168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
 
# File 'lib/correct_horse_battery_staple/corpus/sqlite.rb', line 168

def pick_discrete(count, options = {})
  p_range = options[:percentile] or 0..100
  l_range = options[:word_length] or 4..12

  result = []
  iterations = 0
  while (iterations < 4 || result.length < count) && iterations < MAX_ITERATIONS
    percentile  = random_in_range(p_range)
    length      = random_in_range(l_range)
    result     += _pick_discrete_n(percentile, length, 1)
    iterations += 1
  end

  # validate that we succeeded
  raise "Cannot find #{count} words matching criteria" if result.length < count

  array_sample(result, count).map {|row| word_from_row(row)}
end

#pick_discrete2(count, options = {}) ⇒ Object

discrete method



203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
 
# File 'lib/correct_horse_battery_staple/corpus/sqlite.rb', line 203

def pick_discrete2(count, options = {})

  p_range = options[:percentile] or 0..100
  l_range = options[:word_length] or 4..12

  ids = []
  iterations = 0
  while (iterations < 3 || ids.length < count) && iterations < MAX_ITERATIONS
    percentile  = random_in_range(p_range)
    length      = random_in_range(l_range)
    ids         = ids.concat(_pick_discrete_n_ids(percentile, length, 25)).uniq
    iterations += 1
  end

  ids = array_sample(ids, count).map {|r| r[0] }
  result = get_words_for_ids(ids)

  # validate that we succeeded
  raise "Cannot find #{count} words matching criteria" if result.length < count
  result
end

#pick_rtree(count, options = {}) ⇒ Object 



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
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
 
# File 'lib/correct_horse_battery_staple/corpus/sqlite.rb', line 55

def pick_rtree(count, options = {})
  base = "select id from index3d "
  wheres = []
  params = []

  wheres << "minR >= ? and maxR <= ?"
  rnd = random_number
  offset = 0.0
  if rnd > 0.8
    offset = 0.8-rnd
  elsif rnd < 0.2
    offset = 0.2-rnd
  end
  params += [rnd - 0.20 + offset, rnd + 0.20 + offset]

  if options[:word_length]
    wheres << "  minL >= ? and maxL <= ? "
    params += [options[:word_length].first, options[:word_length].last]
  end
  if options[:percentile]
    wheres << "  minP >= ? and maxP <= ? "
    params += [options[:percentile].first, options[:percentile].last]
  end
  statement = [base,
               (wheres.empty? ? "" : " WHERE " + wheres.join(" AND ")),
               "limit ?"].join(" ")
  params += [[count,250].max]

  query = prepare(statement)
  ids = array_sample(query.execute!(*params), count).map {|r| r[0]}

  if ids and !ids.empty?
    result = get_words_for_ids(ids)
  else
    result = []
  end

  # validate that we succeeded
  raise "Cannot find #{count} words matching criteria" if result.length < count

  result
end

#pick_standard(count, options = {}) ⇒ Object 



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
 
# File 'lib/correct_horse_battery_staple/corpus/sqlite.rb', line 110

def pick_standard(count, options = {})
  statement = "select #{COLUMNS.join(", ")} from entries "
  params = []
  wheres = []
  if options[:word_length]
    wheres << "  wordlength >= ? and wordlength <= ? "
    params += [options[:word_length].first, options[:word_length].last]
  end
  if options[:percentile]
    wheres << "  percentile >= ? and percentile <= ? "
    params += [options[:percentile].first, options[:percentile].last]
  end
  statement = [statement,
               (wheres.empty? ? "" : " WHERE " + wheres.join(" AND ")),
               "order by RANDOM()",
               "limit ?"].join(" ")
  params << [count, 20].max
  query = prepare(statement)
  result = array_sample(query.execute!(*params), count).
    map { |row| word_from_row(row) }

  # validate that we succeeded
  raise "Cannot find #{count} words matching criteria" if result.length < count

  result
end

#pick_standard2(count, options = {}) ⇒ Object 



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
 
# File 'lib/correct_horse_battery_staple/corpus/sqlite.rb', line 137

def pick_standard2(count, options = {})
  statement = "select id from entries "
  params = []
  wheres = []
  if options[:word_length]
    wheres << "  wordlength >= ? and wordlength <= ? "
    params += [options[:word_length].first, options[:word_length].last]
  end
  if options[:percentile]
    wheres << "  percentile >= ? and percentile <= ? "
    params += [options[:percentile].first, options[:percentile].last]
  end
  statement = [statement,
               (wheres.empty? ? "" : " WHERE " + wheres.join(" AND ")),
               # "order by RANDOM()",
               "limit ?"].join(" ")
  params << [count, 1000].max
  query = prepare(statement)
  ids = array_sample(query.execute!(*params), count).
    map {|r| r[0]}

  result = get_words_for_ids(ids)

  # validate that we succeeded
  raise "Cannot find #{count} words matching criteria" if result.length < count

  result
end

#prepare(statement) ⇒ Object 



187
188
189
190
191
 
# File 'lib/correct_horse_battery_staple/corpus/sqlite.rb', line 187

def prepare(statement)
  res = @db.prepare(statement)
  @statements << res
  res
end

#size ⇒ Object 



24
25
26
 
# File 'lib/correct_horse_battery_staple/corpus/sqlite.rb', line 24

def size
  @size ||= @db.execute("select count(*) from entries").first.first
end

#sorted_entries ⇒ Object 



36
37
38
 
# File 'lib/correct_horse_battery_staple/corpus/sqlite.rb', line 36

def sorted_entries
  entries
end