llt-tei_handler 0.0.1 → 0.0.2

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: a4c68ed6f6c5df8c0e7d02e44261c41ba25599e1
-  data.tar.gz: 7528aead71bc6217a8ba3aa4e2d7e9c2f817c6ab
+  metadata.gz: e59ea02658a4ad35e4ad4621532d8af69be04fda
+  data.tar.gz: 531f99e7c1b0e1c91ba7eb85e289f16392f8e22b
 SHA512:
-  metadata.gz: bf561a4a09e51e5de98a8d00c28df1abf64a6cc7d4838291971bc884b769258f79e939d6045ca25bd7415718ee43ecb4e5b069df6ea9945ee22254c12b9687ce
-  data.tar.gz: a600bcaef8a7a51525f75a1d7c4b20746bcfcc111eb1e84fa03a613663c080a9d836661f8dbcded6919eda6a62c8a85ccf0a039b00430977c147cbca5c01b58a
+  metadata.gz: de353304255b5b1a3da61cc27538716b01929688ac90eb98f5eb70267c0a17d2c1d222d6047ce82107534fdc1abb16292e310f3b6b7d85a3305eed9812a1d65d
+  data.tar.gz: c3c184ce65d6b735b98d18c3886ca9c3e17d0919756dd268ddc75c698488aa2bb93cbdd9a737110d76e01817be84d8f0bd5649ae3e0078f4e2bd9d2d9c2bd20e

data/Gemfile

@@ -2,4 +2,4 @@ source 'https://rubygems.org'
 
 # Specify your gem's dependencies in llt-tei_handler.gemspec
 gemspec
-gem 'coveralls', require: false
+gem 'coveralls', require: false

data/lib/llt/tei_handler/pre_processor.rb

@@ -5,7 +5,7 @@ module LLT
 
       def initialize(document)
         @document = Nokogiri::XML(document)
-        raise ArgumentError.new('Document is no TEI XML') unless is_tei?
+        try_to_find_tei_root unless is_tei?
       end
 
       def to_xml
@@ -13,7 +13,7 @@ module LLT
       end
 
       def is_tei?
-        @document.root.name == 'TEI'
+        @document.root.name =~ /^TEI/
       end
 
       def ignore_nodes(*nodes)
@@ -33,6 +33,27 @@ module LLT
           doc
         end
       end
+
+      def try_to_find_tei_root
+        tei = @document.xpath('//*[name() = "TEI" or name() = "TEI.2"]').first
+        if tei
+          if RUBY_ENGINE == "jruby"
+            # Pretty unnecessarily reparses the document fragment, but
+            # nokogiri-java seems to have problems with handling namespaces
+            # when assigning the root node of a document by hand.
+            #
+            # The issue has been reported on the nokogiri-talk Google group
+            # and is described in detail there (currently awaiting approval)
+            @document = Nokogiri::XML(tei.to_s)
+          else
+            @document = Nokogiri::XML::Document.new
+            @document.root = tei
+          end
+          @document.encoding = "UTF-8"
+        else
+          raise ArgumentError.new('Document is no TEI XML')
+        end
+      end
     end
   end
 end

data/lib/llt/tei_handler/version.rb

@@ -1,5 +1,5 @@
 module LLT
   module TeiHandler
-    VERSION = "0.0.1"
+    VERSION = "0.0.2"
   end
 end

data/spec/lib/llt/tei_handler/pre_processor_spec.rb

@@ -17,15 +17,36 @@ describe LLT::TeiHandler::PreProcessor do
       expect { pre_processor.new(doc) }.not_to raise_error
     end
 
+    it "tries to find the TEI root element" do
+      doc = <<-EOF
+        <?xml version="1.0" encoding="utf-8"?>
+        <reply>
+          <TEI xmlns="http://www.tei-c.org/ns/1.0">
+            <teiHeader></teiHeader>
+          </TEI>
+        </reply>
+      EOF
+      expect { pre_processor.new(doc) }.not_to raise_error
+    end
+
     it "throws an error when the document is NOT TEI" do
       doc = <<-EOF
         <?xml version="1.0" encoding="utf-8"?>
         <doc>
         </doc>
       EOF
-
       expect { pre_processor.new(doc) }.to raise_error ArgumentError
     end
+
+    it "allows different TEI versions" do
+      doc = <<-EOF
+        <?xml version="1.0" encoding="utf-8"?>
+        <TEI.2 xmlns="http://www.tei-c.org/ns/1.0">
+        </TEI.2>
+      EOF
+      expect { pre_processor.new(doc) }.not_to raise_error
+
+    end
   end
 
   let(:tei_doc) do

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: llt-tei_handler
 version: !ruby/object:Gem::Version
-  version: 0.0.1
+  version: 0.0.2
 platform: ruby
 authors:
 - LFDM
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2014-05-01 00:00:00.000000000 Z
+date: 2014-08-12 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler