simple_bioc 0.0.4 → 0.0.5

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: b3fa4051460fa744f5d6c1f48fc3d10350662750
-  data.tar.gz: 280387a719ad30f1b9eb4477a53a20fbab59cc01
+  metadata.gz: 7f41d51c5fdefb00b99d4fc0dfcd79e3f2f7738b
+  data.tar.gz: 6ca95848028aa0b04b2b44fb3563111218cd5c26
 SHA512:
-  metadata.gz: b3c76d3fc8ef4d863fdd589bd3bf4b6d81da1dade9d18a35b4e7cebd7bb59a77c2e4c6eb7dcc4897829898e5993e71605b3b6f06fbbd60a49750e6eab822e787
-  data.tar.gz: a1e6c5242c0d097efe64256276c27a293977906bb397be240852fd6a9552ee7c36dee56e1989790477326ce3201b2929cb31868bfa8f19f48f6964eed5be59e6
+  metadata.gz: fb18148c22f361d66ca12413e1f0c34cbfcc1d03cc65ede1e6f08a44978e0406455ddbadf727b8802b4bde9d6360f1dce2a42d45e7cedae44db5818ea847b939
+  data.tar.gz: cb892612b41da3f73e9edc58836f290ff04a348e86869a487e92af577274e058f51b6b1eedcddab194926e7fcf3bc3f10634a01510baa33a7d3da62b83e926af

data/lib/simple_bioc/bioc_reader.rb

@@ -7,20 +7,28 @@ module BioCReader
   def read(path, options) 
     collection = nil
     File.open(path) do |file|
-      xml_doc  = Nokogiri::XML(file) do |config|
-        config.noent.strict.noblanks
-      end
-      xml = xml_doc.at_xpath("//collection")
-      if xml.nil?
-        fail 'Wrong format'
-      end
-      collection = SimpleBioC::Collection.new
-      read_collection(xml, collection, options)
+      collection = read_from_file_or_string(file, options)
     end
 
     collection
   end
 
+  def read_from_file_or_string(file_or_string, options)
+    collection = nil
+
+    xml_doc  = Nokogiri::XML(file_or_string) do |config|
+      config.noent.strict.noblanks
+    end
+    xml = xml_doc.at_xpath("//collection")
+    if xml.nil?
+      fail 'Wrong format'
+    end
+    collection = SimpleBioC::Collection.new
+    read_collection(xml, collection, options)
+
+    collection
+  end
+
   def read_text(xml, name)
     node = xml.at_xpath(name)
     node && node.content

data/lib/simple_bioc/bioc_writer.rb

@@ -1,4 +1,6 @@
 require 'nokogiri'
+require 'json_builder'
+
 Dir[File.dirname(__FILE__) + '/*.rb'].each {|file| require file }
 
 module BioCWriter

data/lib/simple_bioc/version.rb

@@ -1,3 +1,3 @@
 module SimpleBioC
-  VERSION = "0.0.4"
+  VERSION = "0.0.5"
 end

data/lib/simple_bioc.rb

@@ -24,6 +24,47 @@ module SimpleBioC
     BioCReader.read(file_path, options)
   end
 
+  # parse a BioC XML file and convert it into a collection instance
+  # 
+  # ==== Arguments
+  # * +file_path+ - file object for parse
+  # * +options+ - (optional) additional options
+  #
+  # ==== Options
+  # * +documents+ - specify IDs of documents to parse. The result will include only the specified documents
+  #
+  # ==== Examples
+  #    file = File.open(path)
+  #    collection = SimpleBioC.from_xml(file)
+  #    collection = SimpleBioC.from_xml(file, {documents:[21785578, 21488974]})
+  def from_xml_file(file, options = {})
+    options[:documents] = options[:documents].map{|e| e.to_s} if options[:documents].kind_of?(Array)
+    BioCReader.read_from_file_or_string(file, options)
+  end
+
+  # parse a BioC XML string and convert it into a collection instance
+  # 
+  # ==== Arguments
+  # * +string+ - xml string (text) for parse
+  # * +options+ - (optional) additional options
+  #
+  # ==== Options
+  # * +documents+ - specify IDs of documents to parse. The result will include only the specified documents
+  #
+  # ==== Examples
+  #    content = "<?xml version=\"1.0\" encoding=\"UTF-8\"?> ..."
+  #    collection = SimpleBioC.from_xml(content)
+  #    collection = SimpleBioC.from_xml(content, {documents:[21785578, 21488974]})
+  def from_xml_string(string, options = {})
+    options[:documents] = options[:documents].map{|e| e.to_s} if options[:documents].kind_of?(Array)
+    BioCReader.read_from_file_or_string(string, options)
+  end
+
+  def from_json(json, options = {})
+    options[:documents] = options[:documents].map{|e| e.to_s} if options[:documents].kind_of?(Array)
+    BioCReader.read_from_json(json, options)
+  end
+
   # convert a collection instance to a BioC XML text. Output will return as string
   # 
   # ==== Arguments

data/simple_bioc.gemspec

@@ -19,6 +19,7 @@ Gem::Specification.new do |spec|
   spec.require_paths = ["lib"]
 
   spec.add_dependency "nokogiri", "~> 1.6"
+  spec.add_dependency "json_builder", "~> 3.1"
 
   spec.add_development_dependency "bundler", "~> 1.3"
   spec.add_development_dependency "rake", "~> 10.3"

data/spec/simple_bioc_spec.rb

@@ -28,4 +28,21 @@ describe "Simple function" do
     output = SimpleBioC.to_xml(collection, {save_with: 0})
     expect(output).to include('<location offset="1" length="2"/><text>annotation text</text>')
   end
+
+  it "should read bioc from file" do    
+    f = File.open("./xml/everything.xml")
+    collection = SimpleBioC.from_xml_file(f)
+    f.close
+    output = SimpleBioC.to_xml(collection)
+    expect(output).to include('<!DOCTYPE collection SYSTEM "BioC.dtd">')
+  end
+
+  it "should read bioc from string" do    
+    f = File.open("./xml/everything.xml")
+    contents = f.read
+    collection = SimpleBioC.from_xml_file(contents)
+    f.close
+    output = SimpleBioC.to_xml(collection)
+    expect(output).to include('<!DOCTYPE collection SYSTEM "BioC.dtd">')
+  end
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: simple_bioc
 version: !ruby/object:Gem::Version
-  version: 0.0.4
+  version: 0.0.5
 platform: ruby
 authors:
 - Dongseop Kwon
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2015-02-17 00:00:00.000000000 Z
+date: 2015-06-22 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: nokogiri
@@ -24,6 +24,20 @@ dependencies:
     - - "~>"
       - !ruby/object:Gem::Version
         version: '1.6'
+- !ruby/object:Gem::Dependency
+  name: json_builder
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '3.1'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '3.1'
 - !ruby/object:Gem::Dependency
   name: bundler
   requirement: !ruby/object:Gem::Requirement