Class: Xampl::FromXML

Inherits:
Object
  • Object
show all
Defined in:
lib/xamplr/from-xml.rb

Constant Summary collapse

@@by_tag = 
{}
@@by_ns_tag = 
{}

Instance Attribute Summary collapse

Class Method Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(recovering = false) ⇒ FromXML

Returns a new instance of FromXML.



18
19
20
21
22
23
24
25
26
27
28
 
# File 'lib/xamplr/from-xml.rb', line 18

def initialize(recovering=false)
  @recovering             = recovering

  @attribute_name         = Array.new(32)
  @attribute_namespace    = Array.new(32)
  @attribute_value        = Array.new(32)

  @insert_end_element     = false
  @faking_an_end_element  = false
  @just_opened_an_element = false
end

Instance Attribute Details

#checkWellFormed ⇒ Object (readonly)

1.9.1 , false



9
10
11
 
# File 'lib/xamplr/from-xml.rb', line 9

def checkWellFormed
  @checkWellFormed
end

#is_realising ⇒ Object (readonly)

1.9.1 , false



10
11
12
 
# File 'lib/xamplr/from-xml.rb', line 10

def is_realising
  @is_realising
end

#tokenise_content ⇒ Object (readonly)

1.9.1 , false



11
12
13
 
# File 'lib/xamplr/from-xml.rb', line 11

def tokenise_content
  @tokenise_content
end

Class Method Details

.register(tag, ns_tag, klass) ⇒ Object 



35
36
37
38
39
40
41
42
43
44
45
 
# File 'lib/xamplr/from-xml.rb', line 35

def FromXML.register(tag, ns_tag, klass)
  @@by_ns_tag[ns_tag] = [klass]
  a                   = @@by_tag[tag]
  if (nil == a) then
    @@by_tag[tag] = [klass]
  else
    found = false
    a.each { |thing| found = found | (thing == klass) }
    a << klass unless found
  end
end

.registered(name) ⇒ Object 



47
48
49
50
51
52
 
# File 'lib/xamplr/from-xml.rb', line 47

def FromXML.registered(name)
  klass = @@by_ns_tag[name]
  klass = @@by_tag[name] unless klass
  klass = [] unless klass
  return klass
end

.reset_registry ⇒ Object 



30
31
32
33
 
# File 'lib/xamplr/from-xml.rb', line 30

def FromXML.reset_registry
  @@by_tag    = {}
  @@by_ns_tag = {}
end

.tokenise_string(str, strip = true) ⇒ Object 



295
296
297
298
299
300
 
# File 'lib/xamplr/from-xml.rb', line 295

def FromXML.tokenise_string(str, strip=true)
  return nil unless str
  str.strip! if strip
  str.gsub!(/[ \n\r\t][ \n\r\t]*/, " ")
  return str
end

Instance Method Details

#attributeCount ⇒ Object 



363
364
365
 
# File 'lib/xamplr/from-xml.rb', line 363

def attributeCount
  return @attribute_name.length
end

#attributeName(i) ⇒ Object 



367
368
369
 
# File 'lib/xamplr/from-xml.rb', line 367

def attributeName(i)
  return @attribute_name[i]
end

#attributeNamespace(i) ⇒ Object 



371
372
373
 
# File 'lib/xamplr/from-xml.rb', line 371

def attributeNamespace(i)
  return @attribute_namespace[i]
end

#attributeValue(i) ⇒ Object 



375
376
377
 
# File 'lib/xamplr/from-xml.rb', line 375

def attributeValue(i)
  return @attribute_value[i]
end

#build_attribute_arrays ⇒ Object 



349
350
351
352
353
354
355
356
357
358
359
360
361
 
# File 'lib/xamplr/from-xml.rb', line 349

def build_attribute_arrays
  @attribute_name.clear
  @attribute_namespace.clear
  @attribute_value.clear

  return unless @reader.attributes?

  @reader.attributes.each do |name, value|
    @attribute_name << name
    @attribute_namespace << nil
    @attribute_value << value
  end
end

#chew ⇒ Object 



125
126
127
128
129
130
131
132
133
 
# File 'lib/xamplr/from-xml.rb', line 125

def chew
  xml   = @reader.outer_xml
  depth = @reader.depth
  @reader.read
  while depth != @reader.depth do
    @reader.read
  end
  return xml
end

#column ⇒ Object 



387
388
389
 
# File 'lib/xamplr/from-xml.rb', line 387

def column
  return @reader.column_number
end

#current_node_type ⇒ Object 



302
303
304
305
306
307
308
 
# File 'lib/xamplr/from-xml.rb', line 302

def current_node_type
  if @faking_an_end_element then
    Nokogiri::XML::Node::ELEMENT_DECL
  else
    @reader.node_type
  end
end

#depth ⇒ Object 



379
380
381
 
# File 'lib/xamplr/from-xml.rb', line 379

def depth
  return @reader.depth
end

#find_the_first_element ⇒ Object 



340
341
342
343
344
345
346
347
 
# File 'lib/xamplr/from-xml.rb', line 340

def find_the_first_element
  while true do
    break if start_element?
    break unless next_reader_event
  end
  @just_opened_an_element = start_element?
  @insert_end_element     = (@just_opened_an_element and @reader.empty_element?)
end

#line ⇒ Object 



383
384
385
 
# File 'lib/xamplr/from-xml.rb', line 383

def line
  return @reader.line_number
end

#next_reader_event ⇒ Object 



310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
 
# File 'lib/xamplr/from-xml.rb', line 310

def next_reader_event
  if @insert_end_element then
    @faking_an_end_element = true
    @insert_end_element    = false
    return
  end

  @faking_an_end_element  = false

  begin
    okay = @reader.read
  rescue => e
    raise RuntimeError, "WHAT?? -- #{ e }", e.backtrace
  end

  @just_opened_an_element = self.start_element?
  @insert_end_element     = (@just_opened_an_element and @reader.empty_element?)
  okay
end

#parse(filename, tokenise_content = true, is_realising = false) ⇒ Object 



101
102
103
104
105
106
107
108
109
 
# File 'lib/xamplr/from-xml.rb', line 101

def parse(filename, tokenise_content=true, is_realising=false)
  begin
    setup_parse(filename, tokenise_content, is_realising)
    element, ignore = parse_element
    return element
  rescue => e
    raise RuntimeError, "trouble parsing file: '#{filename}' -- #{ e }", e.backtrace
  end
end

#parse_element(parent = nil, target = nil) ⇒ Object 



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
 
# File 'lib/xamplr/from-xml.rb', line 136

def parse_element(parent=nil, target=nil)
#      puts caller(0)[0..5]

  find_the_first_element
  return unless start_element?

  namespace        = @reader.namespace_uri
  name             = @reader.local_name
  existing_element = nil
  element          = nil

  requires_caching = false

  build_attribute_arrays

  if ((nil != namespace) and (0 < namespace.size)) then
    klass_name = "{#{namespace}}#{name}"
    klasses    = FromXML.registered(klass_name)
    if (0 == klasses.size) then
      # The class has not been registered (either it was never generated, or it was never loaded)
      begin
        #discard this node and all children, but say something
        thing = chew
        puts "#{ ::File.basename __FILE__ }:#{ __LINE__ } [#{__method__}] UNRECOGNISED CHILD ELEMENTS: class: #{ klass_name }\n#{ thing }"
        return nil, true
      rescue => e
        puts "Ohhhh NO! #{ e }"
        puts e.backtrace
        raise e
      end
    end
    if (1 < klasses.size) then
      raise XamplException.new("there is more than one '#{name}' tag in namespace '#{namespace}'\nplease report this error")
    end
  else
    klasses = FromXML.registered(name)
    if (0 == klasses.size) then
      raise XamplException.new("do not recognise tag '#{name}' (no namespace specified)")
    end
    if (1 < klasses.size) then
      raise XamplException.new("there is more than one '#{name}' tag (no namespace specified)")
    end
  end

  unless @is_realising then
    @attribute_value.size.times do |i|
      FromXML.tokenise_string @attribute_value[i]
    end
  end

  if target then
    element            = target
    target.load_needed = false
    target             = nil
    element.init_attributes(@attribute_name, @attribute_namespace, @attribute_value)
    element.note_attributes_initialised(@is_realising)
  else
    if klasses[0].persisted? then
      @attribute_name.each_index do |i|
        if @attribute_name[i] == klasses[0].persisted?.to_s then
          existing_element = Xampl.find_known(klasses[0], @attribute_value[i])
          if existing_element then
            # so we've found the element. Now what??? We can do several
            # reasonable things:
            #
            #    1) continue parsing into the found element
            #    2) simply return the found element
            #    3) replace the found element with the new element
            #
            # The first one is dubious, so we won't.
            # The second and third option both make complete sense
            #
            # We are going to do the second
            #
            # BTW, 'existing element' means a representation of this element already in memory
            # puts "!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!"
            # puts "!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!"
            # puts "!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!"
            # puts "FOUND AN EXISTING THING... #{ klasses[0] } #{ @attribute_value[i] }"
            # puts "!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!"
            # puts "!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!"
            # puts "!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!"
            # caller(0).each { | trace | puts "  #{trace}"}
            #                   existing_element.reset_contents
            #                   element = existing_element
            #                   existing_element = nil
            #                  puts "#{File.basename(__FILE__)} #{__LINE__} EXISTING ELEMENT: #{ existing_element }"
            #                  puts "#{File.basename(__FILE__)} #{__LINE__} WOW, must handle the existing element correctly"
            element = existing_element #TODO -- IS THIS RIGHT????????????????????????
          end
          unless element then
            element          = klasses[0].discriminate(@attribute_name, @attribute_namespace, @attribute_value).new
            requires_caching = @recovering
#                  puts "#{File.basename(__FILE__)} #{__LINE__} WOW, what about recovering????"
            #TODO -- IS THIS RIGHT????????????????????????
            requires_caching = true #!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
            unless @recovering then
              element.force_load if parent
            end
            element.note_created(@is_realising)
          end

          break
        end
      end
    end

    unless element then
      element = klasses[0].discriminate(@attribute_name, @attribute_namespace, @attribute_value).new
      element.note_created(@is_realising)
    end

    element.note_initialise_attributes_with(@attribute_name, @attribute_namespace, @attribute_value, @is_realising)
    element.init_attributes(@attribute_name, @attribute_namespace, @attribute_value)
    element.note_attributes_initialised(@is_realising)

    Xampl.cache(element) if requires_caching && element && element.persist_required
  end

  while next_reader_event do
    if @reader.value? then
      text = @reader.value
      text = text.force_encoding('utf-8') unless 'UTF-8' == text.encoding
      the_text = element.note_adding_text_content(text, @is_realising)
      if element.has_mixed_content then
        element << the_text
      else
        element.add_content(the_text, false)
      end
    elsif Nokogiri::XML::Node::ELEMENT_NODE == @reader.node_type then
      child, ignore_child = parse_element(element)

      unless ignore_child then
        case child
          when XamplObject then
            child = child.note_add_to_parent(element, @is_realising) if child
            child = element.note_add_child(child, @is_realising) if element
            child.append_to(element) if element && child
          when XMLText then
            #TODO -- get rid of this puts
            puts "UNRECOGNISED Well-formed XML: #{child.to_s[0..25]}..."
          else
            #TODO -- get rid of this puts
            puts "WHAT IS THIS??? #{child.class.name}"
        end
      end
    elsif Nokogiri::XML::Node::ELEMENT_DECL == @reader.node_type then
      element = element.note_closed(@is_realising)
      return element if @recovering
      return existing_element || element
    else
      puts "WTF??(#{ @reader.depth }) name: #{ @reader.name }, #{ say_node_type(@reader.node_type)}/#{ @reader.node_type }\n#{ @reader.outer_xml }"
    end
  end

  return element if @recovering
  return existing_element || element
end

#parse_string(string, tokenise_content = true, is_realising = false, target = nil) ⇒ Object 



115
116
117
118
119
120
121
122
123
 
# File 'lib/xamplr/from-xml.rb', line 115

def parse_string(string, tokenise_content=true, is_realising=false, target=nil)
  begin
    setup_parse_string(string, tokenise_content, is_realising)
    element, ignore = parse_element(nil, target)
    return element
  rescue => e
    raise RuntimeError, "trouble parsing string: '#{string}' -- #{ e }", e.backtrace
  end
end

#realise_string(string, tokenise_content = true, target = nil) ⇒ Object 



111
112
113
 
# File 'lib/xamplr/from-xml.rb', line 111

def realise_string(string, tokenise_content=true, target=nil)
  return parse_string(string, tokenise_content, true, target)
end

#resolve(name) ⇒ Object 



54
55
56
57
 
# File 'lib/xamplr/from-xml.rb', line 54

def resolve(name)
  #TODO -- ??? don't seem to need it, this is for specific named entities
  return name
end

#setup_parse(filename, tokenise_content = true, is_realising = false) ⇒ Object 



59
60
61
62
 
# File 'lib/xamplr/from-xml.rb', line 59

def setup_parse(filename, tokenise_content=true, is_realising=false)
  xml = File.read(filename)
  setup_parse_string(xml, tokenise_content, is_realising)
end

#setup_parse_string(string, tokenise_content = true, is_realising = false) ⇒ Object 



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
 
# File 'lib/xamplr/from-xml.rb', line 64

def setup_parse_string(string, tokenise_content=true, is_realising=false)
  @resolver         = self

  @is_realising     = is_realising
  @tokenise_content = tokenise_content

=begin
  STRICT	=	0	 	 Strict parsing
  RECOVER	=	1 << 0	 	 Recover from errors
  NOENT	=	1 << 1	 	 Substitute entities
  DTDLOAD	=	1 << 2	 	 Load external subsets
  DTDATTR	=	1 << 3	 	 Default DTD attributes
  DTDVALID	=	1 << 4	 	 validate with the DTD
  NOERROR	=	1 << 5	 	 suppress error reports
  NOWARNING	=	1 << 6	 	 suppress warning reports
  PEDANTIC	=	1 << 7	 	 pedantic error reporting
  NOBLANKS	=	1 << 8	 	 remove blank nodes
  SAX1	=	1 << 9	 	 use the SAX1 interface internally
  XINCLUDE	=	1 << 10	 	 Implement XInclude substitition
  NONET	=	1 << 11	 	 Forbid network access
  NODICT	=	1 << 12	 	 Do not reuse the context dictionnary
  NSCLEAN	=	1 << 13	 	 remove redundant namespaces declarations
  NOCDATA	=	1 << 14	 	 merge CDATA as text nodes
  NOXINCNODE	=	1 << 15	 	 do not generate XINCLUDE START/END nodes
  DEFAULT_XML	=	RECOVER	 	 the default options used for parsing XML documents
  DEFAULT_HTML	=	RECOVER | NOERROR | NOWARNING | NONET	 	 the default options used for parsing HTML documents
=end

  options           = Nokogiri::XML::ParseOptions::RECOVER | Nokogiri::XML::ParseOptions::NOENT | Nokogiri::XML::ParseOptions::NONET | Nokogiri::XML::ParseOptions::NOCDATA | Nokogiri::XML::ParseOptions::DTDATTR

  utf8_string       = string.force_encoding('utf-8')
  url               = nil
  encoding          = nil

  @reader           = Nokogiri::XML::Reader.from_memory(utf8_string, url, encoding, options)
end

#start_element? ⇒ Boolean

Returns:

  • (Boolean) 


330
331
332
 
# File 'lib/xamplr/from-xml.rb', line 330

def start_element?
  current_node_type == Nokogiri::XML::Node::ELEMENT_NODE
end

#whitespace? ⇒ Boolean

Returns:

  • (Boolean) 


334
335
336
337
338
 
# File 'lib/xamplr/from-xml.rb', line 334

def whitespace?
  #there is no whitespace type with nokogiri
  #TODO -- this is not actually called, so...
  @reader.value? && @reader.value.match(/\S/).nil?
end