Class: ArrayIO

Inherits:
Object
  • Object
show all
Includes:
Enumerable
Defined in:
lib/array_io.rb

Instance Attribute Summary collapse

Class Method Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(fd, mode = 'r', io_index = nil) ⇒ ArrayIO

Returns a new instance of ArrayIO.



22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
 
# File 'lib/array_io.rb', line 22

def initialize(fd, mode='r', io_index=nil)
  if mode =~ /u/i
    mode = mode.delete('u')
    @cached = false
  else
    @cached = true
  end
  
  if mode =~ /s/i
    mode = mode.delete('s')
    @strio = true
  else
    @strio = false
  end
  
  if strio?
    mode = mode.delete('s')
    @io = StringIO.open(fd, mode)
    io_index = '' if io_index.nil?
  else
    @io = File.open(fd, mode)
    set_binmode(io)
    io_index = default_index_file if io_index.nil?
  end

  self.io_index = io_index
end

Instance Attribute Details

#io ⇒ Object (readonly)

Returns the value of attribute io.



4
5
6
 
# File 'lib/array_io.rb', line 4

def io
  @io
end

#io_index ⇒ Object

Returns the value of attribute io_index.



4
5
6
 
# File 'lib/array_io.rb', line 4

def io_index
  @io_index
end

Class Method Details

.open(fd, mode = 'r', index_file = nil, auto_index = true, &block) ⇒ Object 



9
10
11
12
13
14
15
16
17
18
19
 
# File 'lib/array_io.rb', line 9

def open(fd, mode='r', index_file=nil, auto_index=true, &block)
  aio = self.new(fd, mode, index_file)
  
  if block_given?
    aio.reindex if auto_index && aio.empty?
    yield(aio)
    aio.close
  else
    aio
  end
end

Instance Method Details

#[](input, length = nil) ⇒ Object 



74
75
76
77
78
79
80
81
82
 
# File 'lib/array_io.rb', line 74

def [](input, length=nil)
  range = range(input, length)
  case range
  when Array
    range.collect { |r| read_entry(r) }
  else
    read_entry(range)
  end
end

#[]=(input, entry) ⇒ Object 



84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
 
# File 'lib/array_io.rb', line 84

def []=(input, entry)
  string = entry_to_str(entry)
  
  range_begin = strio? ? io.string.length :  io.stat.size 
  range_end = range_begin + string.length
  range = format_range(range_begin, range_end)
  
  # do this first in case io is not open for writing.

  set_pos(io, range_begin)
  io << string
  
  if cached?
    io_index[input] = range
  else
    pos = index_pos(input)
    if strio?
      io_index.string[pos...pos+RANGE_SIZE] = range
    else
      #io_index.close

    end
  end
end

#at(index) ⇒ Object 



70
71
72
 
# File 'lib/array_io.rb', line 70

def at(index)
  self[index]
end

#cached? ⇒ Boolean

arrayio functionality

Returns:

  • (Boolean) 


159
160
161
 
# File 'lib/array_io.rb', line 159

def cached?
  @cached
end

#close ⇒ Object 



191
192
193
194
 
# File 'lib/array_io.rb', line 191

def close
  io.close
  io_index.close unless cached?
end

#default_index_file ⇒ Object 



50
51
52
 
# File 'lib/array_io.rb', line 50

def default_index_file
  io.path.chomp(File.extname(io.path)) + '.index'
end

#dump_index(index_file = default_index_file) ⇒ Object 



172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
 
# File 'lib/array_io.rb', line 172

def dump_index(index_file=default_index_file)
  dumping_to_io = io_index.respond_to?(:path) ? 
      File.expand_path(io_index.path) != File.expand_path(index_file) :
      false

  File.open(index_file, 'w') do |file|
    set_binmode(file)
    if cached?
      io_index.each do |range|
        file << [range.begin, range.end].pack(FORMAT)
      end
    else
      file << io_index.read
    end
  end unless dumping_to_io
  
  index_file
end

#each(&block) ⇒ Object 



123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
 
# File 'lib/array_io.rb', line 123

def each(&block)
  raise LocalJumpError("no block given") unless block_given?
  
  if cached?
    io_index.each do |range|
      yield( read_entry(range) )
    end
  else
    io_index.pos = 0
    while !io_index.eof?
      begin_index, end_index = io_index.read(RANGE_SIZE).unpack(FORMAT)
      yield( read_entry(begin_index..end_index) )
    end
  end
end

#each_index(&block) ⇒ Object 



117
118
119
120
121
 
# File 'lib/array_io.rb', line 117

def each_index(&block)
  raise LocalJumpError("no block given") unless block_given?
  
  0.upto(length-1, &block)
end

#empty? ⇒ Boolean

Returns:

  • (Boolean) 


139
140
141
 
# File 'lib/array_io.rb', line 139

def empty?
  length == 0
end

#entry_to_str(entry) ⇒ Object 



200
201
202
 
# File 'lib/array_io.rb', line 200

def entry_to_str(entry)
  entry.to_s
end

#fetch(index, default = nil, &block) ⇒ Object 



107
108
109
110
111
 
# File 'lib/array_io.rb', line 107

def fetch(index, default=nil, &block)
  index += index_length if index < 0 
  val = (index >= length ? default : self[index])
  block_given? ? yield(val) : val
end

#first(n = nil) ⇒ Object 



113
114
115
 
# File 'lib/array_io.rb', line 113

def first(n=nil)
  n.nil? ? self[0] : self[0,n]
end

#last(n = nil) ⇒ Object 



143
144
145
146
147
148
149
 
# File 'lib/array_io.rb', line 143

def last(n=nil)
  return self[-1] if n.nil?
  
  start = length-n
  start = 0 if start < 0
  self[start, n]
end

#length ⇒ Object 



62
63
64
65
66
67
68
 
# File 'lib/array_io.rb', line 62

def length 
  if cached?
    io_index.length
  else
    index_length/RANGE_SIZE
  end
end

#load_index(index_file) ⇒ Object 



167
168
169
170
 
# File 'lib/array_io.rb', line 167

def load_index(index_file)
  input = strio? ? index_file : (File.exists?(index_file) ? File.read(index_file) : '')
  @io_index = parse_index(input)
end

#range(input, length = nil) ⇒ 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
 
# File 'lib/array_io.rb', line 204

def range(input, length=nil)
  if cached?
    if length.nil?
      return io_index[input]
    else
      return io_index[input, length] 
    end
  end
    
  unless length.nil?
    return nil if length < 0
    input = input...(input + length)
  end

  if input.kind_of?(Range)
    begin_pos = index_pos(input.begin)
    return nil if begin_pos < 0 || begin_pos >= index_length 
    
    end_pos = index_pos(input.end) + (input.exclude_end? ? 0 : RANGE_SIZE)
    read_length = end_pos-begin_pos
    return [] if read_length <= 0

    set_pos(io_index, begin_pos)
    parse_index(io_index.read(read_length))
  else
    begin_pos = index_pos(input)
    return nil if begin_pos < 0 || begin_pos >= index_length

    set_pos(io_index, begin_pos)
    array = io_index.read(RANGE_SIZE).unpack(FORMAT)
    array.first...array.last
  end
end

#reindex(options = {}, &block) ⇒ Object 



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
263
264
265
266
267
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
 
# File 'lib/array_io.rb', line 238

def reindex(options={}, &block)
  io.rewind
  
  options = {
    :sep_string => $/,
    :break_before => false,
    :exclude_break => false,
    :limit => nil
  }.merge(options)
  
  sep_string = options[:sep_string]
  break_before = options[:break_before]
  exclude_break = options[:exclude_break]
  
  limit = options[:limit]
  total_count = 0
  
  if cached?
    self.io_index = []
  else
    io_index.close 
    if strio?
      self.io_index = ''
    else
      self.io_index = io_index.path
    end
  end
  
  last_pos = 0
  current_pos = 0
  range_begin = 0
  io.each_line(sep_string) do |line|
    # Note positions MUST be built up using line.length

    # io.pos cannot return positions greater than ~2.1e9

    last_pos = current_pos
    current_pos += line.length
    
    if (block_given? ? yield(line) : true)
      range_end = (break_before || exclude_break) ? last_pos : current_pos
      unless range_end == range_begin
        io_index << format_range(range_begin, range_end)
        total_count += 1
      end
      range_begin = (break_before && !exclude_break) ? last_pos : current_pos
  
      break unless limit.nil? || total_count < limit
    end
  end
  if limit.nil? || total_count < limit
    range_end = current_pos
    unless range_end == range_begin
      io_index << format_range(range_begin, range_end)
    end
  end

  # this must be done to re-stat Files, assuring index_length is the length of the io_index file.

  unless cached? || strio?
    io_index.close 
    self.io_index = io_index.path 
  end

  self
end

#size ⇒ Object

array functionality



58
59
60
 
# File 'lib/array_io.rb', line 58

def size
  length
end

#str_to_entry(string) ⇒ Object 



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

def str_to_entry(string)
  string
end

#strio? ⇒ Boolean

Returns:

  • (Boolean) 


163
164
165
 
# File 'lib/array_io.rb', line 163

def strio?
  @strio
end

#values_at(*selectors) ⇒ Object 



151
152
153
 
# File 'lib/array_io.rb', line 151

def values_at(*selectors)
  selectors.collect {|s| self[s]}.flatten
end