Class: T2Flow::Parser

Inherits:
Object
  • Object
show all
Defined in:
lib/t2flow/parser.rb

Instance Method Summary collapse

Instance Method Details

#add_annotation(dataflow, annotation) ⇒ Object

:nodoc:


234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
# File 'lib/t2flow/parser.rb', line 234

def add_annotation(dataflow, annotation) # :nodoc:
  node = LibXML::XML::Parser.string("#{annotation}").parse
  content_node = node.find_first("//annotationBean")

  case content_node["class"]
    when /freetextdescription/i
      dataflow.annotations.descriptions << content_node.find_first('./text').content
    when /descriptivetitle/i
      dataflow.annotations.titles << content_node.find_first('./text').content
    when /author/i
      dataflow.annotations.authors << content_node.find_first('./text').content
    when /semanticannotation/i
      dataflow.annotations.semantic_annotation = parse_semantic_annotation(dataflow, content_node)
    end # case
end

#add_coordination(dataflow, condition) ⇒ Object

:nodoc:


225
226
227
228
229
230
231
232
# File 'lib/t2flow/parser.rb', line 225

def add_coordination(dataflow, condition) # :nodoc:
  coordination = Coordination.new
  
  coordination.control = condition["control"]
  coordination.target = condition["target"]
  
  dataflow.coordinations << coordination
end

:nodoc:


201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
# File 'lib/t2flow/parser.rb', line 201

def add_link(dataflow, link) # :nodoc:
  datalink = Datalink.new
  
  if sink = link.find_first('./t2:sink')
    if processor = sink.find_first('./t2:processor')
      datalink.sink = "#{processor.content}:"
    else
      datalink.sink = ""
    end
    datalink.sink += "#{sink.find_first('./t2:port').content}"
  end

  if source = link.find_first('./t2:source')
    if processor = source.find_first('./t2:processor')
      datalink.source = "#{processor.content}:"
    else
      datalink.source = ""
    end
    datalink.source += "#{source.find_first('./t2:port').content}"
  end

  dataflow.datalinks << datalink
end

#add_processor(dataflow, element) ⇒ Object

:nodoc:


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
# File 'lib/t2flow/parser.rb', line 110

def add_processor(dataflow, element) # :nodoc:
  processor = Processor.new
  
  temp_inputs = []
  temp_outputs = []
  
  element.each_element do |elt|
    case elt.name
      when "name"
        processor.name = elt.content
      when /inputports/i # ports from services
        elt.each_element { |port| port.each_element { |x| temp_inputs << x.content if x.name=="name" }}
      when /outputports/i # ports from services
        elt.each_element { |port| port.each_element { |x| temp_outputs << x.content if x.name=="name" }}
      when "annotations"
        extract_annotations(processor, elt)
      when "activities" # a processor can only have one kind of activity
        activity = elt.find_first('./t2:activity')
        activity.each_element do |node|
          if node.name == "configBean"
              activity_node = node.child

              if activity_node.name == "net.sf.taverna.t2.component.ComponentActivityConfigurationBean"
                processor.configuration[:component] = {}
                processor.configuration[:component][:registry] = activity_node.find_first('./registryBase').content
                processor.configuration[:component][:family_name] = activity_node.find_first('./familyName').content
                processor.configuration[:component][:name] = activity_node.find_first('./componentName').content
                processor.configuration[:component][:version] = activity_node.find_first('./componentVersion').content.to_i
              end
              
              if node["encoding"] == "dataflow"
                processor.dataflow_id = activity_node["ref"]
                processor.type = "workflow"
              else
                processor.type = (activity_node.name =~ /martquery/i ?
                    "biomart" : activity_node.name.split(".")[-2])
                
                activity_node.each_element do |value_node|
                  case value_node.name
                    when "wsdl"
                      processor.configuration[:wsdl] = value_node.content
                    when "operation"
                      processor.configuration[:wsdl_operation] = value_node.content
                    when /endpoint/i
                      processor.configuration[:endpoint] = value_node.content
                    when /servicename/i
                      processor.configuration[:biomoby_service_name] = value_node.content
                    when /authorityname/i
                      processor.configuration[:biomoby_authority_name] = value_node.content
                    when "category"
                      processor.configuration[:biomoby_category] = value_node.content
                    when "script"
                      processor.configuration[:script] = value_node.content
                    when "value"
                      processor.configuration[:value] = value_node.content
                    when "presentationOrigin"
                      processor.configuration[:interaction_page] = value_node.content
                    when "connectionSettings"
                      if processor.type == 'rshell'
                        processor.configuration[:r_server] = value_node.find_first('./host').content + ":" +
                            value_node.find_first('./port').content
                      end
                    when "xpathExpression"
                      processor.configuration[:xpath_expression] = value_node.content
                    when "inputs" # ALL ports present in beanshell
                      value_node.each_element do |input|
                        input.each_element do |x|
                          processor.inputs = [] if processor.inputs.nil?
                          processor.inputs << x.content if x.name == "name" 
                        end
                      end
                    when "outputs" # ALL ports present in beanshell
                      value_node.each_element do |output|
                        output.each_element do |x|
                          processor.outputs = [] if processor.outputs.nil?
                          processor.outputs << x.content if x.name == "name" 
                        end
                      end
                  end # case value_node.name
                end # activity_node.each
              end # if else node["encoding"] == "dataflow"
          end # if node.name == "configBean"
        end # activity.each
    end # case elt.name
  end # element.each
  
  processor.inputs = temp_inputs if processor.inputs.nil? && !temp_inputs.empty?
  processor.outputs = temp_outputs if processor.outputs.nil? && !temp_outputs.empty?
  dataflow.processors << processor
end

#add_sink(dataflow, port) ⇒ Object

:nodoc:


102
103
104
105
106
107
108
# File 'lib/t2flow/parser.rb', line 102

def add_sink(dataflow, port) # :nodoc:
  return if port.nil? || port.content.chomp.strip.empty?
  
  sink = Sink.new      
  (sink, port)      
  dataflow.sinks << sink
end

#add_source(dataflow, port) ⇒ Object

:nodoc:


94
95
96
97
98
99
100
# File 'lib/t2flow/parser.rb', line 94

def add_source(dataflow, port) # :nodoc:
  return if port.nil? || port.content.chomp.strip.empty?

  source = Source.new
  (source, port)            
  dataflow.sources << source
end

#create_model(element, version) ⇒ Object

:nodoc:


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
# File 'lib/t2flow/parser.rb', line 42

def create_model(element, version) # :nodoc:
  model = Model.new
  
  local_depends = element.find("//localDependencies")
  if local_depends
    local_depends.each do |dependency|
      dependency.each do |dep|
        model.dependencies = [] if model.dependencies.nil?
        model.dependencies << dep.content unless dep.content =~ /^\s*$/
      end
    end
    model.dependencies.uniq! if model.dependencies
  end

  element.each_element do |dataflow|
    next if dataflow["id"].nil? || dataflow["id"].chomp.strip.empty?

    dataflow_obj = Dataflow.new
    dataflow_obj.dataflow_id = dataflow["id"]
    dataflow_obj.role = dataflow["role"]
    
    dataflow.each_element do |elt|
      case elt.name
        when "name"
          dataflow_obj.annotations.name = elt.content
        when "inputPorts"
          elt.each_element { |port| add_source(dataflow_obj, port) }
        when "outputPorts"
          elt.each_element { |port| add_sink(dataflow_obj, port) }
        when "processors"
          elt.each_element { |proc| add_processor(dataflow_obj, proc) }
        when "datalinks"
          elt.each_element { |link| add_link(dataflow_obj, link) }
        when "conditions"
          elt.each_element { |coord| add_coordination(dataflow_obj, coord) }
        when "annotations"
          elt.each_element { |ann| add_annotation(dataflow_obj, ann) }
      end # case elt.name
    end # dataflow.each
    
    model.dataflows << dataflow_obj
  end # element.each
  
  temp = model.processors.select { |x| x.type == "workflow" }
  temp.each do |proc|
    df = model.dataflow(proc.dataflow_id)
    df.annotations.name = proc.name
  end
  
  return model   
end

#parse(t2flow) ⇒ Object

Returns the model for the given t2flow_file. The method accepts objects of classes File, StringIO and String only.

Usage

foo = ... # stuff to initialize foo here
bar = T2Flow::Parser.new.parse(foo)

22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
# File 'lib/t2flow/parser.rb', line 22

def parse(t2flow)
  case t2flow.class.to_s
    when /^string$/i
      document = LibXML::XML::Parser.string(t2flow, :options => LibXML::XML::Parser::Options::NOBLANKS).parse
    when /^stringio|file$/i
      t2flow.rewind
      document = LibXML::XML::Parser.string(t2flow.read, :options => LibXML::XML::Parser::Options::NOBLANKS).parse
    else 
      raise "Error parsing file."
  end

  root = document.root
  root.namespaces.default_prefix = "t2"

  raise "Doesn't appear to be a workflow!" if root.name != "workflow"
  version = root["version"]
  
  create_model(root, version)
end