File: dataset.rb

package info (click to toggle)
libsequel-core-ruby 1.5.1-1
  • links: PTS
  • area: main
  • in suites: lenny
  • size: 648 kB
  • ctags: 840
  • sloc: ruby: 10,949; makefile: 36
file content (423 lines) | stat: -rw-r--r-- 13,640 bytes parent folder | download
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
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
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
76
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
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
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
224
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
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
301
302
303
304
305
306
307
308
309
310
311
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
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
420
421
422
423
require 'time'
require 'date'
require 'yaml'
require 'base64'

require File.join(File.dirname(__FILE__), 'dataset/sql')
require File.join(File.dirname(__FILE__), 'dataset/sequelizer')
require File.join(File.dirname(__FILE__), 'dataset/convenience')
require File.join(File.dirname(__FILE__), 'dataset/callback')
require File.join(File.dirname(__FILE__), 'dataset/pagination')

module Sequel
  # A Dataset represents a view of a the data in a database, constrained by
  # specific parameters such as filtering conditions, order, etc. Datasets
  # can be used to create, retrieve, update and delete records.
  # 
  # Query results are always retrieved on demand, so a dataset can be kept
  # around and reused indefinitely:
  #   my_posts = DB[:posts].filter(:author => 'david') # no records are retrieved
  #   p my_posts.all # records are now retrieved
  #   ...
  #   p my_posts.all # records are retrieved again
  #
  # In order to provide this functionality, dataset methods such as where, 
  # select, order, etc. return modified copies of the dataset, so you can
  # use different datasets to access data:
  #   posts = DB[:posts]
  #   davids_posts = posts.filter(:author => 'david')
  #   old_posts = posts.filter('stamp < ?', Date.today - 7)
  #
  # Datasets are Enumerable objects, so they can be manipulated using any
  # of the Enumerable methods, such as map, inject, etc.
  #
  # === The Dataset Adapter Interface
  #
  # Each adapter should define its own dataset class as a descendant of
  # Sequel::Dataset. The following methods should be overriden by the adapter
  # Dataset class (each method with the stock implementation):
  #
  #   # Iterate over the results of the SQL query and call the supplied
  #   # block with each record (as a hash).
  #   def fetch_rows(sql, &block)
  #     @db.synchronize do
  #       r = @db.execute(sql)
  #       r.each(&block)
  #     end
  #   end
  #
  #   # Insert records.
  #   def insert(*values)
  #     @db.synchronize do
  #       @db.execute(insert_sql(*values)).last_insert_id
  #     end
  #   end
  #
  #   # Update records.
  #   def update(*args, &block)
  #     @db.synchronize do
  #       @db.execute(update_sql(*args, &block)).affected_rows
  #     end
  #   end
  #
  #   # Delete records.
  #   def delete(opts = nil)
  #     @db.synchronize do
  #       @db.execute(delete_sql(opts)).affected_rows
  #     end
  #   end
  class Dataset
    include Enumerable
    include Sequelizer
    include SQL
    include Convenience
    include Callback
    
    attr_accessor :db, :opts, :row_proc
    
    alias_method :size, :count
    
    # Returns an array with all records in the dataset. If a block is given,
    # the array is iterated over.
    def all(opts = nil, &block)
      a = []
      each(opts) {|r| a << r}
      post_load(a)
      a.each(&block) if block
      a
    end
  
    # Constructs a new instance of a dataset with a database instance, initial
    # options and an optional record class. Datasets are usually constructed by
    # invoking Database methods:
    #   DB[:posts]
    # Or:
    #   DB.dataset # the returned dataset is blank
    #
    # Sequel::Dataset is an abstract class that is not useful by itself. Each
    # database adaptor should provide a descendant class of Sequel::Dataset.
    def initialize(db, opts = nil)
      @db = db
      @opts = opts || {}
      @row_proc = nil
      @transform = nil
    end
    
    # Returns a new clone of the dataset with with the given options merged.
    def clone(opts = {})
      c = super()
      c.opts = @opts.merge(opts)
      c.instance_variable_set(:@columns, nil)
      c
    end
    
    NOTIMPL_MSG = "This method must be overriden in Sequel adapters".freeze
    
    # Executes a select query and fetches records, passing each record to the
    # supplied block. Adapters should override this method.
    def fetch_rows(sql, &block)
      # @db.synchronize do
      #   r = @db.execute(sql)
      #   r.each(&block)
      # end
      raise NotImplementedError, NOTIMPL_MSG
    end
  
    # Inserts values into the associated table. Adapters should override this
    # method.
    def insert(*values)
      # @db.synchronize do
      #   @db.execute(insert_sql(*values)).last_insert_id
      # end
      raise NotImplementedError, NOTIMPL_MSG
    end
  
    # Updates values for the dataset. Adapters should override this method.
    def update(values, opts = nil)
      # @db.synchronize do
      #   @db.execute(update_sql(values, opts)).affected_rows
      # end
      raise NotImplementedError, NOTIMPL_MSG
    end
  
    # Deletes the records in the dataset. Adapters should override this method.
    def delete(opts = nil)
      # @db.synchronize do
      #   @db.execute(delete_sql(opts)).affected_rows
      # end
      raise NotImplementedError, NOTIMPL_MSG
    end
    
    # Returns the columns in the result set in their true order. The stock 
    # implementation returns the content of @columns. If @columns is nil,
    # a query is performed. Adapters are expected to fill @columns with the
    # column information when a query is performed.
    def columns
      first unless @columns
      @columns || []
    end
    
    def columns!
      first
      @columns || []
    end
    
    # Inserts the supplied values into the associated table.
    def <<(*args)
      insert(*args)
    end
  
    # Updates the dataset with the given values.
    def set(*args, &block)
      update(*args, &block)
    end
    
    # Iterates over the records in the dataset
    def each(opts = nil, &block)
      if graph = @opts[:graph]
        graph_each(opts, &block)
      else
        row_proc = @row_proc unless opts && opts[:naked]
        transform = @transform
        fetch_rows(select_sql(opts)) do |r|
          r = transform_load(r) if transform
          r = row_proc[r] if row_proc
          yield r
        end
      end
      self
    end

    # Returns the the model classes associated with the dataset as a hash.
    def model_classes
      @opts[:models]
    end
    
    # Returns the column name for the polymorphic key.
    def polymorphic_key
      @opts[:polymorphic_key]
    end
    
    # Returns a naked dataset clone - i.e. a dataset that returns records as
    # hashes rather than model objects.
    def naked
      d = clone(:naked => true, :models => nil, :polymorphic_key => nil)
      d.set_model(nil)
      d
    end
    
    # Associates or disassociates the dataset with a model. If no argument or
    # nil is specified, the dataset is turned into a naked dataset and returns
    # records as hashes. If a model class specified, the dataset is modified
    # to return records as instances of the model class, e.g:
    #
    #   class MyModel
    #     def initialize(values)
    #       @values = values
    #       ...
    #     end
    #   end
    # 
    #   dataset.set_model(MyModel)
    #
    # You can also provide additional arguments to be passed to the model's
    # initialize method:
    #
    #   class MyModel
    #     def initialize(values, options)
    #       @values = values
    #       ...
    #     end
    #   end
    # 
    #   dataset.set_model(MyModel, :allow_delete => false)
    #  
    # The dataset can be made polymorphic by specifying a column name as the
    # polymorphic key and a hash mapping column values to model classes.
    #
    #   dataset.set_model(:kind, {1 => Person, 2 => Business})
    #
    # You can also set a default model class to fall back on by specifying a
    # class corresponding to nil:
    #
    #   dataset.set_model(:kind, {nil => DefaultClass, 1 => Person, 2 => Business})
    # 
    # To disassociate a model from the dataset, you can call the #set_model 
    # and specify nil as the class:
    # 
    #   dataset.set_model(nil)
    #
    def set_model(key, *args)
      # pattern matching
      case key
      when nil # set_model(nil) => no
        # no argument provided, so the dataset is denuded
        @opts.merge!(:naked => true, :models => nil, :polymorphic_key => nil)
        self.row_proc = nil
        # extend_with_stock_each
      when Class
        # isomorphic model
        @opts.merge!(:naked => nil, :models => {nil => key}, :polymorphic_key => nil)
        if key.respond_to?(:load)
          # the class has a values setter method, so we use it
          self.row_proc = proc{|h| key.load(h, *args)}
        else
          # otherwise we just pass the hash to the constructor
          self.row_proc = proc{|h| key.new(h, *args)}
        end
        extend_with_destroy
      when Symbol
        # polymorphic model
        hash = args.shift || raise(ArgumentError, "No class hash supplied for polymorphic model")
        @opts.merge!(:naked => true, :models => hash, :polymorphic_key => key)
        if hash.values.first.respond_to?(:load)
          # the class has a values setter method, so we use it
          self.row_proc = proc do |h|
            c = hash[h[key]] || hash[nil] || \
              raise(Error, "No matching model class for record (#{polymorphic_key} => #{h[polymorphic_key].inspect})")
            c.load(h, *args)
          end
        else
          # otherwise we just pass the hash to the constructor
          self.row_proc = proc do |h|
            c = hash[h[key]] || hash[nil] || \
              raise(Error, "No matching model class for record (#{polymorphic_key} => #{h[polymorphic_key].inspect})")
            c.new(h, *args)
          end
        end
        extend_with_destroy
      else
        raise ArgumentError, "Invalid model specified"
      end
      self
    end
    
    STOCK_TRANSFORMS = {
      :marshal => [
        # for backwards-compatibility we support also non-base64-encoded values.
        proc {|v| Marshal.load(Base64.decode64(v)) rescue Marshal.load(v)}, 
        proc {|v| Base64.encode64(Marshal.dump(v))}
      ],
      :yaml => [
        proc {|v| YAML.load v if v}, 
        proc {|v| v.to_yaml}
      ]
    }
    
    # Sets a value transform which is used to convert values loaded and saved
    # to/from the database. The transform should be supplied as a hash. Each
    # value in the hash should be an array containing two proc objects - one
    # for transforming loaded values, and one for transforming saved values.
    # The following example demonstrates how to store Ruby objects in a dataset
    # using Marshal serialization:
    #
    #   dataset.transform(:obj => [
    #     proc {|v| Marshal.load(v)},
    #     proc {|v| Marshal.dump(v)}
    #   ])
    #
    #   dataset.insert_sql(:obj => 1234) #=>
    #   "INSERT INTO items (obj) VALUES ('\004\bi\002\322\004')"
    #
    # Another form of using transform is by specifying stock transforms:
    # 
    #   dataset.transform(:obj => :marshal)
    #
    # The currently supported stock transforms are :marshal and :yaml.
    def transform(t)
      @transform = t
      t.each do |k, v|
        case v
        when Array
          if (v.size != 2) || !v.first.is_a?(Proc) && !v.last.is_a?(Proc)
            raise Error::InvalidTransform, "Invalid transform specified"
          end
        else
          unless v = STOCK_TRANSFORMS[v]
            raise Error::InvalidTransform, "Invalid transform specified"
          else
            t[k] = v
          end
        end
      end
      self
    end
    
    # Applies the value transform for data loaded from the database.
    def transform_load(r)
      r.inject({}) do |m, kv|
        k, v = *kv
        m[k] = (tt = @transform[k]) ? tt[0][v] : v
        m
      end
    end
    
    # Applies the value transform for data saved to the database.
    def transform_save(r)
      r.inject({}) do |m, kv|
        k, v = *kv
        m[k] = (tt = @transform[k]) ? tt[1][v] : v
        m
      end
    end
    
    # Extends the dataset with a destroy method, that calls destroy for each
    # record in the dataset.
    def extend_with_destroy
      unless respond_to?(:destroy)
        meta_def(:destroy) do
          unless @opts[:models]
            raise Error, "No model associated with this dataset"
          end
          count = 0
          @db.transaction {each {|r| count += 1; r.destroy}}
          count
        end
      end
    end

    @@dataset_classes = []

    def self.dataset_classes #:nodoc:
      @@dataset_classes
    end

    def self.inherited(c) #:nodoc:
      @@dataset_classes << c
    end
    
    # Returns a string representation of the dataset including the class name 
    # and the corresponding SQL select statement.
    def inspect
      '#<%s: %s>' % [self.class.to_s, sql.inspect]
    end

    # Setup mutation (e.g. filter!) methods
    def self.def_mutation_method(*meths)
      meths.each do |meth|
        class_eval("def #{meth}!(*args, &block); mutation_method(:#{meth}, *args, &block) end")
      end
    end
    def def_mutation_method(*meths)
      meths.each do |meth|
        instance_eval("def #{meth}!(*args, &block); mutation_method(:#{meth}, *args, &block) end")
      end
    end

    MUTATION_METHODS = %w'and distinct exclude exists filter from from_self full_outer_join graph
    group group_and_count group_by having inner_join intersect invert_order join
    left_outer_join limit naked or order order_by order_more paginate query reject
    reverse reverse_order right_outer_join select select_all select_more
    set_graph_aliases set_model sort sort_by union unordered where'.collect{|x| x.to_sym}

    def_mutation_method(*MUTATION_METHODS)

    private
      def mutation_method(meth, *args, &block)
        copy = send(meth, *args, &block)
        @opts.merge!(copy.opts)
        self
      end
  end
end