imzml 0.0.2

data/lib/imzml/parser.rb

@@ -0,0 +1,332 @@
+require "ox"
+
+module ImzML
+
+  class Parser
+
+    attr_reader :metadata
+
+    def initialize(filepath)
+
+      sax = ImzML::Sax.new
+      Ox.sax_parse(sax, File.open(filepath))
+      @metadata = sax.metadata
+
+    end
+
+  end
+
+  class Sax < ::Ox::Sax
+
+    attr_reader :metadata
+
+    def initialize()
+      @metadata = Metadata.new
+      @stack = Array.new
+      @elements = Array.new
+
+      # temporary values useful just for data parsing
+      @reference_groups = Hash.new
+      @obo = Hash.new
+    end
+
+    def start_element(name)
+      # p "#{@stack.last} #{@elements.last}"
+      @stack.push(name)
+      @elements.push(Hash.new)
+      # p "#{name} started"
+      # p @stack
+
+      case name
+      when :cvList
+        @cv_list = []
+      when :cv
+        @cv_list << @elements.last
+      end
+    end
+
+    def attr(name, str)
+      element = @elements.last
+      return if element.nil? # skip attributes without correct elements (like <?xml ...)
+      element[name] = str
+    end
+
+    def end_element(name)
+      @stack.pop
+      element = @elements.pop
+      # p @stack
+
+      # open OBO file for each from CV list for further validation
+      if name == :cvList
+        @cv_list.each do |cv|
+          filename = case cv[:id]
+          when "MS"
+            "psi-ms.obo"
+          when "UO"
+            "unit.obo"
+          when "IMS"
+            "imagingMS.obo"
+          end
+          filepath = File.join(File.dirname(__FILE__), "..", "..", "data", filename)
+          @obo[cv[:id].to_s] = Obo::Parser.new(filepath)
+        end
+      end
+
+      # save file content
+      if name == :cvParam && @stack.last == :fileContent
+
+        cv = @obo[element[:cvRef]]
+        stanza = cv.stanza(element[:accession])
+        parent_id = stanza.parent_id
+
+        # init basic structures
+        @metadata.file_description ||= FileDescription.new
+        file_content = (@metadata.file_description.file_content ||= FileContent.new)
+
+        case element[:cvRef]
+        when "MS"
+          # save data file content
+          if parent_id == FileContent::DATA_FILE_CONTENT
+            file_content.data_file_contents ||= Hash.new
+            (file_content.data_file_contents[parent_id] ||= Array.new) << element
+          end
+
+          # save spectrum representation
+          if parent_id == FileContent::SPECTRUM_REPRESENTATION
+            file_content.spectrum_representation = element
+          end
+
+        when "IMS"
+          # save binary type (cannot look by parent because the OBO file is different and
+          # the parser doesn't hadle it well, need to first improve the OBO parser)
+          if stanza.id == FileContent::CONTINUOUS
+            file_content.binary_type = :continuous
+          elsif stanza.id == FileContent::PROCESSED
+            file_content.binary_type = :processed
+          end
+
+          # save checksum type
+          if stanza.id == FileContent::MD5
+            file_content.checksum = element[:value]
+          elsif stanza.id == FileContent::SHA1
+            file_content.checksum = element[:value]
+          end
+
+          # save identifier
+          if stanza.id == FileContent::UNIVERSALLY_UNIQUE_IDENTIFIER
+            file_content.uuid = element[:value]
+          end
+        end
+
+      end
+
+      # save reference group for further usage
+      if name == :cvParam && @stack.last == :referenceableParamGroup
+        (@reference_groups[@elements.last[:id].to_sym] ||= Array.new) << element
+      end
+
+      # save sample list
+      if name == :cvParam && @stack.last == :sample
+        samples = (@metadata.samples ||= Hash.new)
+        samples[@elements.last[:id].to_sym] = element
+      end
+
+      # save software list (raw, without detailed parsing)
+      if name == :software && @stack.last == :softwareList
+        (@metadata.software ||= Array.new) << element
+      end
+
+      # save scan settings
+      if name == :cvParam && @stack.last == :scanSettings
+        scan_settings = (@metadata.scan_settings ||= Hash.new)
+        setting = (scan_settings[@elements.last[:id].to_sym] ||= ScanSettings.new)
+
+        cv = @obo[element[:cvRef]]
+        stanza = cv.stanza(element[:accession])
+        parent_id = stanza.parent_id
+
+        case element[:cvRef]
+        when "IMS"
+
+          # detect correct line scan direction
+          setting.line_scan_direction = case stanza.id
+          when ScanSettings::LINE_SCAN_BOTTOM_UP
+            :bottom_up
+          when ScanSettings::LINE_SCAN_LEFT_RIGHT
+            :left_right
+          when ScanSettings::LINE_SCAN_RIGHT_LEFT
+            :right_left
+          when ScanSettings::LINE_SCAN_TOP_DOWN
+            :top_down
+          else
+            setting.line_scan_direction
+          end
+
+          # detect scan direction
+          setting.scan_direction = case stanza.id
+          when ScanSettings::BOTTOM_UP
+            :bottom_up
+          when ScanSettings::LEFT_RIGHT
+            :left_right
+          when ScanSettings::RIGHT_LEFT
+            :right_left
+          when ScanSettings::TOP_DOWN
+            :top_down
+          else
+            setting.scan_direction
+          end
+
+          # detect scan pattern
+          setting.scan_pattern = case stanza.id
+          when ScanSettings::MEANDERING
+            :meandering
+          when ScanSettings::ONE_WAY
+            :one_way
+          when ScanSettings::RANDOM_ACCESS
+            :random_access
+          when ScanSettings::FLY_BACK
+            :fly_back
+          else
+            setting.scan_pattern
+          end
+
+          # detect scan type
+          setting.scan_type = case stanza.id
+          when ScanSettings::HORIZONTAL_LINE_SCAN
+            :horizontal
+          when ScanSettings::VERTICAL_LINE_SCAN
+            :vertical
+          else
+            setting.scan_type
+          end
+
+          # detect image properties
+          image = (setting.image ||= ImzML::Image.new)
+
+          case stanza.id
+          when ScanSettings::MAX_DIMENSION_X
+            point = (image.max_dimension ||= ImzML::Point.new)
+            point.x = element[:value].to_i
+          when ScanSettings::MAX_DIMENSION_Y
+            point = (image.max_dimension ||= ImzML::Point.new)
+            point.y = element[:value].to_i
+          when ScanSettings::MAX_COUNT_OF_PIXEL_X
+            point = (image.max_pixel_count ||= ImzML::Point.new)
+            point.x = element[:value].to_i
+          when ScanSettings::MAX_COUNT_OF_PIXEL_Y
+            point = (image.max_pixel_count ||= ImzML::Point.new)
+            point.y = element[:value].to_i
+          when ScanSettings::PIXEL_SIZE_X
+            point = (image.pixel_size ||= ImzML::Point.new)
+            point.x = element[:value].to_i
+          when ScanSettings::PIXEL_SIZE_Y
+            point = (image.pixel_size ||= ImzML::Point.new)
+            point.y = element[:value].to_i
+          end
+        end
+
+        # [
+        #   {:cvRef=>"IMS", :accession=>"IMS:1000401", :name=>"top down", :value=>""},
+        #   {:cvRef=>"IMS", :accession=>"IMS:1000413", :name=>"flyback", :value=>""},
+        #   {:cvRef=>"IMS", :accession=>"IMS:1000480", :name=>"horizontal line scan", :value=>""},
+        #   {:cvRef=>"IMS", :accession=>"IMS:1000491", :name=>"linescan left right", :value=>""},
+        #   {:cvRef=>"IMS", :accession=>"IMS:1000042", :name=>"max count of pixel x", :value=>"3"},
+        #   {:cvRef=>"IMS", :accession=>"IMS:1000043", :name=>"max count of pixel y", :value=>"3"},
+        #   {:cvRef=>"IMS", :accession=>"IMS:1000044", :name=>"max dimension x", :value=>"300", :unitCvRef=>"UO", :unitAccession=>"UO:0000017", :unitName=>"micrometer"},
+        #   {:cvRef=>"IMS", :accession=>"IMS:1000045", :name=>"max dimension y", :value=>"300", :unitCvRef=>"UO", :unitAccession=>"UO:0000017", :unitName=>"micrometer"},
+        #   {:cvRef=>"IMS", :accession=>"IMS:1000046", :name=>"pixel size x", :value=>"100", :unitCvRef=>"UO", :unitAccession=>"UO:0000017", :unitName=>"micrometer"},
+        #   {:cvRef=>"IMS", :accession=>"IMS:1000047", :name=>"pixel size y", :value=>"100", :unitCvRef=>"UO", :unitAccession=>"UO:0000017", :unitName=>"micrometer"},
+        #   {:cvRef=>"MS", :accession=>"MS:1000836", :name=>"dried dropplet", :value=>""},
+        #   {:cvRef=>"MS", :accession=>"MS:1000835", :name=>"matrix solution concentration", :value=>"10"},
+        #   {:cvRef=>"MS", :accession=>"MS:1000834", :name=>"matrix solution", :value=>"DHB"}
+        # ]
+      end
+
+      # parse processing methods
+      if name == :cvParam && @stack.last == :processingMethod
+        data_processing = (@metadata.data_processing ||= Hash.new)
+        processing = (data_processing[@elements[-2][:id].to_sym] ||= DataProcessing.new)
+        processing.processing_method = @elements.last
+        (processing.processing_method[:actions] ||= Array.new) << element
+      end
+
+      # save spectrum position info
+      if name == :cvParam && @stack.last == :scan
+        spectrums = (@metadata.spectrums ||= Hash.new)
+        spectrum = (spectrums[@elements[-3][:id].to_sym] ||= Spectrum.new)
+        point = (spectrum.position ||= ImzML::Point.new)
+        
+        point.x = element[:value].to_i if element[:accession] == Spectrum::POSITION_X
+        point.y = element[:value].to_i if element[:accession] == Spectrum::POSITION_Y
+      end
+      
+      # save spectrum binary data info
+      if name == :referenceableParamGroupRef && @stack.last == :binaryDataArray
+        group = @reference_groups[element[:ref].to_sym]
+        
+        spectrum = @metadata.spectrums[@elements[-3][:id].to_sym]
+        mz_binary = (spectrum.mz_binary ||= ImzML::Spectrum::BinaryData.new)
+        intensity_binary = (spectrum.intensity_binary ||= ImzML::Spectrum::BinaryData.new)
+        
+        # detect type of the binary data info based on referenced group content
+        group.each do |param|
+          # p param
+          @binary_type = case param[:accession]
+          when ImzML::Spectrum::BinaryData::MZ_ARRAY
+            :mz_binary
+          when ImzML::Spectrum::BinaryData::INTENSITY_ARRAY
+            :intensity_binary
+          end
+          
+          break if !@binary_type.nil?
+        end
+        
+        # detect binary data type
+        number_type = nil
+        group.each do |param|
+          number_type = case param[:accession]
+          when Metadata::BINARY_TYPE_8BIT_INTEGER
+            :int8
+          when Metadata::BINARY_TYPE_16BIT_INTEGER
+            :int16
+          when Metadata::BINARY_TYPE_32BIT_INTEGER
+            :int32
+          when Metadata::BINARY_TYPE_64BIT_INTEGER
+            :int64
+          when Metadata::BINARY_TYPE_32BIT_FLOAT
+            :float32
+          when Metadata::BINARY_TYPE_64BIT_FLOAT
+            :float64
+          end
+          
+          break if !number_type.nil?
+        end
+        @metadata.send("#{@binary_type.to_s}_data_type=", number_type) if !number_type.nil?
+      end
+      
+      # save info about binary
+      if name == :cvParam && @stack.last == :binaryDataArray 
+        spectrum = @metadata.spectrums[@elements[-3][:id].to_sym]
+        
+        # convert chosen type to mz_binary/intensity_binary property selector
+        binary_data = spectrum.send(@binary_type.to_s)
+        case element[:accession]
+        when ImzML::Spectrum::BinaryData::EXTERNAL_ARRAY_LENGTH
+          binary_data.length = element[:value].to_i
+        when ImzML::Spectrum::BinaryData::EXTERNAL_OFFSET
+          binary_data.offset = element[:value].to_i
+        when ImzML::Spectrum::BinaryData::EXTERNAL_ENCODED_LENGHT
+          binary_data.encoded_length = element[:value].to_i
+        end 
+      
+      end
+      
+      # p @metadata.spectrums if name == :binaryDataArray
+
+      # p "#{name} ended #{element}"
+
+    end
+  
+  end
+
+end

data/lib/imzml/version.rb

@@ -0,0 +1,5 @@
+module ImzML
+  
+  VERSION = "0.0.2"
+  
+end

data/lib/imzml.rb

@@ -0,0 +1,23 @@
+require "obo"
+
+require "obo_ext/parser"
+require "obo_ext/stanza"
+
+require "core_ext/string"
+
+require "imzml/obo"
+require "imzml/metadata"
+require "imzml/metadata/point"
+# require "imzml/metadata/spectrum"
+require "imzml/metadata/file_description"
+require "imzml/metadata/file_description/file_content"
+require "imzml/metadata/scan_settings"
+require "imzml/metadata/scan_settings/image"
+require "imzml/metadata/data_processing"
+require "imzml/metadata/run/spectrum"
+
+require "imzml/parser"
+
+module ImzML
+  
+end

data/lib/obo_ext/parser.rb

@@ -0,0 +1,20 @@
+module Obo
+  
+  class Parser
+    
+    def stanzas
+      self.elements.to_a.keep_if { |x| x.is_a?(Obo::Stanza)}
+    end
+    
+    def stanza(id)
+      elements = self.elements.to_a.keep_if { |x| x.is_a?(Obo::Stanza) && (x.id == id)}
+      elements.first if elements.is_a?(Array)
+    end
+    
+    def children_of(id)
+      self.elements.to_a.keep_if { |x| x.is_a?(Obo::Stanza) && (x.parent?(id))}
+    end
+    
+  end
+  
+end

data/lib/obo_ext/stanza.rb

@@ -0,0 +1,19 @@
+module Obo
+  
+  class Stanza
+    
+    def id
+      self.tagvalues["id"].first
+    end
+    
+    def parent_id
+      self.tagvalues["is_a"].first
+    end
+  
+    def parent?(id)
+      self.parent_id == id
+    end
+    
+  end
+  
+end

metadata

@@ -0,0 +1,115 @@
+--- !ruby/object:Gem::Specification
+name: imzml
+version: !ruby/object:Gem::Version
+  version: 0.0.2
+platform: ruby
+authors:
+- Ondra Beneš
+autorequire: 
+bindir: bin
+cert_chain: []
+date: 2014-04-27 00:00:00.000000000 Z
+dependencies:
+- !ruby/object:Gem::Dependency
+  name: minitest
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '5.3'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '5.3'
+- !ruby/object:Gem::Dependency
+  name: ox
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '2.0'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '2.0'
+- !ruby/object:Gem::Dependency
+  name: obo
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '0.1'
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 0.1.4
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '0.1'
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 0.1.4
+description: Parser for mass spectrometry imaging standard file format. Gem does not
+  check the validity of the input file.
+email:
+- ondra.benes@gmail.com
+executables: []
+extensions: []
+extra_rdoc_files: []
+files:
+- data/Example_Continuous.ibd
+- data/Example_Continuous.imzML
+- data/Example_Processed.ibd
+- data/Example_Processed.imzML
+- data/imagingMS.obo
+- data/psi-ms.obo
+- data/unit.obo
+- lib/core_ext/string.rb
+- lib/imzml.rb
+- lib/imzml/metadata.rb
+- lib/imzml/metadata/data_processing.rb
+- lib/imzml/metadata/file_description.rb
+- lib/imzml/metadata/file_description/file_content.rb
+- lib/imzml/metadata/point.rb
+- lib/imzml/metadata/run/spectrum.rb
+- lib/imzml/metadata/scan_settings.rb
+- lib/imzml/metadata/scan_settings/image.rb
+- lib/imzml/metadata/spectrum.rb
+- lib/imzml/obo.rb
+- lib/imzml/parser.rb
+- lib/imzml/version.rb
+- lib/obo_ext/parser.rb
+- lib/obo_ext/stanza.rb
+homepage: https://github.com/beny/imzml
+licenses: []
+metadata: {}
+post_install_message: 
+rdoc_options: []
+require_paths:
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - ">="
+    - !ruby/object:Gem::Version
+      version: '0'
+required_rubygems_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - ">="
+    - !ruby/object:Gem::Version
+      version: '0'
+requirements: []
+rubyforge_project: 
+rubygems_version: 2.2.0
+signing_key: 
+specification_version: 4
+summary: Simple parser for imzML files.
+test_files: []