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.



16
17
18
19
20
21
22
23
24
25
26
 
# File 'lib/xamplr/from-xml.rb', line 16

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)

Returns the value of attribute checkWellFormed.



7
8
9
 
# File 'lib/xamplr/from-xml.rb', line 7

def checkWellFormed
  @checkWellFormed
end

#is_realising ⇒ Object (readonly)

Returns the value of attribute is_realising.



8
9
10
 
# File 'lib/xamplr/from-xml.rb', line 8

def is_realising
  @is_realising
end

#tokenise_content ⇒ Object (readonly)

Returns the value of attribute tokenise_content.



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

def tokenise_content
  @tokenise_content
end

Class Method Details

.register(tag, ns_tag, klass) ⇒ Object 



33
34
35
36
37
38
39
40
41
42
43
 
# File 'lib/xamplr/from-xml.rb', line 33

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 



45
46
47
48
49
50
 
# File 'lib/xamplr/from-xml.rb', line 45

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

.reset_registry ⇒ Object 



28
29
30
31
 
# File 'lib/xamplr/from-xml.rb', line 28

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

.tokenise_string(str, strip = true) ⇒ Object 



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

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 



410
411
412
 
# File 'lib/xamplr/from-xml.rb', line 410

def attributeCount
  return @attribute_name.length
end

#attributeName(i) ⇒ Object 



414
415
416
 
# File 'lib/xamplr/from-xml.rb', line 414

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

#attributeNamespace(i) ⇒ Object 



418
419
420
 
# File 'lib/xamplr/from-xml.rb', line 418

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

#attributeValue(i) ⇒ Object 



422
423
424
 
# File 'lib/xamplr/from-xml.rb', line 422

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

#build_attribute_arrays ⇒ Object 



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

def build_attribute_arrays

  @attribute_name.clear
  @attribute_namespace.clear
  @attribute_value.clear

  return unless LibXML::XML::Reader::TYPE_ELEMENT == current_node_type

  if @reader.has_attributes? then
    attribute_count = @reader.attribute_count
    @reader.move_to_first_attribute
    attribute_count.times do | i |
      if @reader.namespace_declaration? then
        @reader.move_to_next_attribute
        next
      end

      @attribute_name << @reader.local_name
      @attribute_namespace << @reader.namespace_uri
      @attribute_value << @reader.value

      @reader.move_to_next_attribute
    end
  end
end

#column ⇒ Object 



434
435
436
 
# File 'lib/xamplr/from-xml.rb', line 434

def column
  return @reader.column_number
end

#current_node_type ⇒ Object 



291
292
293
294
295
296
297
 
# File 'lib/xamplr/from-xml.rb', line 291

def current_node_type
  if @faking_an_end_element then
    LibXML::XML::Reader::TYPE_END_ELEMENT
  else
    @reader.node_type
  end
end

#depth ⇒ Object 



426
427
428
 
# File 'lib/xamplr/from-xml.rb', line 426

def depth
  return @reader.depth
end

#find_the_first_element ⇒ Object 



376
377
378
379
380
381
382
 
# File 'lib/xamplr/from-xml.rb', line 376

def find_the_first_element
  while true do
    break if start_element?
    break unless next_reader_event
  end
  @just_opened_an_element = start_element?
end

#line ⇒ Object 



430
431
432
 
# File 'lib/xamplr/from-xml.rb', line 430

def line
  return @reader.line_number
end

#next_reader_event ⇒ Object

def describe_current_element_type()

  case @reader.node_type
    when LibXML::XML::Reader::TYPE_ATTRIBUTE
      puts "ATTRIBUTE"
    when LibXML::XML::Reader::TYPE_DOCUMENT
      puts "DOCUMENT"
    when LibXML::XML::Reader::TYPE_ELEMENT
      attribute_count = @reader.attribute_count
      puts "ELEMENT #{ @reader.local_name }, ns: #{ @reader.namespace_uri }, #attributes: #{ attribute_count }, depth: #{ @reader.depth }"
      puts "        FAKING END ELEMENT" if @faking_an_end_element
    when LibXML::XML::Reader::TYPE_END_ELEMENT
      puts "END ELEMENT"
    when LibXML::XML::Reader::TYPE_TEXT
      puts "TEXT [[#{ @reader.read_string }]]"
    when LibXML::XML::Reader::TYPE_CDATA
      puts "CDATA [[#{ @reader.read_string }]]"
    when LibXML::XML::Reader::TYPE_SIGNIFICANT_WHITESPACE
      puts "SIGNIFICANT white space [[#{ @reader.read_string }]]"
    when LibXML::XML::Reader::TYPE_ENTITY_REFERENCE
      puts "entity ref"
    when LibXML::XML::Reader::TYPE_WHITESPACE
      puts "whitespace"
    when LibXML::XML::Reader::TYPE_PROCESSING_INSTRUCTION
      puts "processing instruction"
    when LibXML::XML::Reader::TYPE_COMMENT
      puts "comment"
    when LibXML::XML::Reader::TYPE_DOCUMENT_TYPE
      puts "doc type"

    when LibXML::XML::Reader::TYPE_XML_DECLARATION
      puts "xml decl"
    when LibXML::XML::Reader::TYPE_NONE
      puts "NONE!!"
    when LibXML::XML::Reader::TYPE_NOTATION
      puts "notifiation"
    when LibXML::XML::Reader::TYPE_DOCUMENT_FRAGMENT
      puts "doc fragment"
    when LibXML::XML::Reader::TYPE_ENTITY
      puts "entity"
    when LibXML::XML::Reader::TYPE_END_ENTITY
      puts "end entity"
    else
      puts "UNKNOWN: #{@reader.node_type}"
  end
end



347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
 
# File 'lib/xamplr/from-xml.rb', line 347

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

  #describe_current_element_type

  okay = @reader.read

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

  #describe_current_element_type

  okay
end

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



88
89
90
91
92
93
94
95
96
97
98
 
# File 'lib/xamplr/from-xml.rb', line 88

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

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



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
 
# File 'lib/xamplr/from-xml.rb', line 123

def parse_element(parent=nil, target=nil)
  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
      xml_text = XMLText.new
      xml_text.build(self)
      xml_text = parent.note_adding_text_content(xml_text, @is_realising)
      parent.add_content(xml_text, @tokenise_content) if xml_text
      return xml_text, false
    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].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].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)

    if requires_caching and element and element.persist_required then
      Xampl.cache(element)
    end

    #element = element.note_add_to_parent(parent, @is_realising)
    #element.append_to(parent) if parent
  end

  while next_reader_event
    case current_node_type

=begin
TODO -- can these ever happen?
      when START_DOCUMENT
        return element if @recovering
        return existing_element || element
      when END_DOCUMENT
        return element if @recovering
        return existing_element || element

=end

      when LibXML::XML::Reader::TYPE_ELEMENT
        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 and child
            when XMLText then
              puts "UNRECOGNISED Well-formed XML: #{child.to_s[0..25]}..."
            else
              puts "WHAT IS THIS??? #{child.class.name}"
          end
        end
      when LibXML::XML::Reader::TYPE_END_ELEMENT
        element = element.note_closed(@is_realising)
        return element if @recovering
        return existing_element || element
      when LibXML::XML::Reader::TYPE_TEXT, LibXML::XML::Reader::TYPE_CDATA, LibXML::XML::Reader::TYPE_SIGNIFICANT_WHITESPACE, LibXML::XML::Reader::TYPE_ENTITY_REFERENCE
        if element.has_mixed_content then
          text = @reader.read_string
          the_text = element.note_adding_text_content(text, @is_realising)
          element << the_text
        else
          text = @reader.read_string
          the_text = element.note_adding_text_content(text, @is_realising)
          element.add_content(the_text, false)
        end
      else
    end
  end

  return element if @recovering
  return existing_element || element
end

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



104
105
106
107
108
109
110
111
112
113
114
 
# File 'lib/xamplr/from-xml.rb', line 104

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 Exception => e
    puts "trouble parsing string: '#{string}'"
    puts "Exception: #{e}"
    raise
  end
end

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



100
101
102
 
# File 'lib/xamplr/from-xml.rb', line 100

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

#resolve(name) ⇒ Object 



52
53
54
55
 
# File 'lib/xamplr/from-xml.rb', line 52

def resolve(name)
  #TODO -- ???
  return name
end

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



57
58
59
60
61
62
63
64
65
66
67
68
69
70
 
# File 'lib/xamplr/from-xml.rb', line 57

def setup_parse(filename, tokenise_content=true, is_realising=false)
  @resolver = self

  @is_realising = is_realising
  @tokenise_content = tokenise_content

  @reader = LibXML::XML::Reader.file(filename,
                                     :options => LibXML::XML::Parser::Options::NOENT |
                                             LibXML::XML::Parser::Options::NONET |
                                             LibXML::XML::Parser::Options::NOCDATA |
                                             LibXML::XML::Parser::Options::DTDATTR |
                                             LibXML::XML::Parser::Options::COMPACT)
  #TODO CLOSE THIS THING!!
end

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



72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
 
# File 'lib/xamplr/from-xml.rb', line 72

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

  @is_realising = is_realising
  @tokenise_content = tokenise_content

  #      setInput(string)
  @reader = LibXML::XML::Reader.string(string,
                                       :options => LibXML::XML::Parser::Options::NOENT |
                                               LibXML::XML::Parser::Options::NONET |
                                               LibXML::XML::Parser::Options::NOCDATA |
                                               LibXML::XML::Parser::Options::DTDATTR |
                                               LibXML::XML::Parser::Options::COMPACT)
  #TODO CLOSE THIS THING!!
end

#start_element? ⇒ Boolean

Returns:

  • (Boolean) 


368
369
370
 
# File 'lib/xamplr/from-xml.rb', line 368

def start_element?
  current_node_type == LibXML::XML::Reader::TYPE_ELEMENT
end

#whitespace? ⇒ Boolean

Returns:

  • (Boolean) 


372
373
374
 
# File 'lib/xamplr/from-xml.rb', line 372

def whitespace?
  current_note_type == LibXML::XML::Reader::TYPE_WHITESPACE
end