harvestdor 0.0.13 → 0.0.14

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 3036c2f23661946012d909179b67effa433c2dd9
-  data.tar.gz: 149e27d97f4c9d48ca46772d550537aca8a55362
+  metadata.gz: dedc072f7e987057ecc62cd6fb174c9733ce414b
+  data.tar.gz: 452697c49d6328ffbb334666e3d9f8892ddf5e3e
 SHA512:
-  metadata.gz: 4406cda66fcb2a726564ba912c60364433b9209ac03218fc2fbf70200fe679f8c964fa416abc0f163eecdabc4db51ad7b57de148f38cf4773b84a4de3fde854e
-  data.tar.gz: b18c647163a3970ae02cb3ee79ba0df04f4d5920ccbc5740a83b19b2958297ce2b1e0c761128293ae7bbb51633e419a6689b839f33f34f602a5292b2a078a4db
+  metadata.gz: b544187441fff729f33c33bcda20f4a9d9f3c827fad46dab343d8fc4a9ac71a60a8f9a838a4c767f6e29ce23ef5e54cec843c239f1c416690fa1a12c0eeab44a
+  data.tar.gz: adf1836583a2b617321ff6b75649dd391161e31b004b2bc93bf513785ec11f6cc84570fd6d1e9bb661af379b86a291d2053c326cdac612fa28f57052ed11ffcf

data/.coveralls.yml

@@ -0,0 +1 @@
+service_name: travis-ci

data/.travis.yml

@@ -1,12 +1,10 @@
 language: ruby
 script: rake rspec
 rvm:
-  - ruby-head
   - 2.1.1
   - 2.0.0
   - 1.9.3
   - jruby-19mode # JRuby in 1.9 mode
-#  - jruby-head
 notifications:
   email:
     - ndushay@stanford.edu

data/Gemfile

@@ -2,3 +2,7 @@ source 'https://rubygems.org'
 
 # See harvestdor.gemspec for this gem's dependencies
 gemspec
+
+group :test do
+  gem 'coveralls', require: false
+end

data/README.rdoc

@@ -1,5 +1,5 @@
 = Harvestdor
-{<img src="https://travis-ci.org/sul-dlss/harvestdor.svg?branch=master" alt="Build Status" />}[https://travis-ci.org/sul-dlss/harvestdor] {<img src="https://gemnasium.com/sul-dlss/harvestdor.svg" alt="Dependency Status" />}[https://gemnasium.com/sul-dlss/harvestdor]
+{<img src="https://travis-ci.org/sul-dlss/harvestdor.svg?branch=master" alt="Build Status" />}[https://travis-ci.org/sul-dlss/harvestdor] {<img src="https://coveralls.io/repos/sul-dlss/harvestdor/badge.png" alt="Coverage Status" />}[https://coveralls.io/r/sul-dlss/harvestdor] {<img src="https://gemnasium.com/sul-dlss/harvestdor.svg" alt="Dependency Status" />}[https://gemnasium.com/sul-dlss/harvestdor] {<img src="https://badge.fury.io/rb/harvestdor.svg" alt="Gem Version" />}[http://badge.fury.io/rb/harvestdor]
 
 A Gem to harvest metadata from DOR.
 
@@ -148,6 +148,7 @@ persist this information with your app for incremental harvests
 
 == Releases
 
+* <b>0.0.14</b> Bug fix for compatibility with jruby
 * <b>0.0.13</b> Updated to work with Faraday 0.9, releases via rubygems instead of sul-gems
 * <b>0.0.11</b> better error handling, and better testing for errors
 * <b>0.0.10</b> tweak specs to test that unnec fetching isn't done.

data/harvestdor.gemspec

@@ -34,8 +34,8 @@ Gem::Specification.new do |gem|
   gem.add_development_dependency "yard"
   # tests
 	gem.add_development_dependency 'rspec'
-	gem.add_development_dependency 'simplecov'
-	gem.add_development_dependency 'simplecov-rcov'
+	gem.add_development_dependency 'vcr'
+	gem.add_development_dependency 'webmock'
 	# gem.add_development_dependency 'ruby-debug19'
   
 end

data/lib/harvestdor/purl_xml.rb

@@ -111,7 +111,7 @@ module Harvestdor
     pub_xml_ng_doc = pub_xml(object, purl_url)
     begin
       # preserve namespaces, etc for the node
-      ng_doc = Nokogiri::XML(pub_xml_ng_doc.root.xpath('/publicObject/dc:dc', {'dc' => Harvestdor::OAI_DC_NAMESPACE}).to_xml)
+      ng_doc = Nokogiri::XML(pub_xml_ng_doc.root.xpath('/publicObject/dc:dc', {'dc' => Harvestdor::OAI_DC_NAMESPACE}).to_xml(:encoding => 'utf-8'))
       raise Harvestdor::Errors::MissingDC.new(object.inspect) if !ng_doc || ng_doc.children.empty?
       ng_doc
     rescue

data/lib/harvestdor/version.rb

@@ -1,3 +1,3 @@
 module Harvestdor
-  VERSION = "0.0.13"
+  VERSION = "0.0.14"
 end

data/spec/fixtures/vcr_cassettes/content_metadata.yml

@@ -0,0 +1,98 @@
+---
+http_interactions:
+- request:
+    method: get
+    uri: http://purl-test.stanford.edu/bb375wb8869.xml
+    body:
+      encoding: US-ASCII
+      string: ''
+    headers:
+      Accept-Encoding:
+      - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+      Accept:
+      - "*/*"
+      User-Agent:
+      - Ruby
+  response:
+    status:
+      code: 200
+      message: ''
+    headers:
+      Date:
+      - Mon, 28 Jul 2014 22:23:37 GMT
+      Server:
+      - Apache/2.2.15 (Red Hat)
+      X-Powered-By:
+      - Phusion Passenger (mod_rails/mod_rack) 3.0.19
+      X-Ua-Compatible:
+      - IE=Edge
+      Etag:
+      - '"c7e2680ca3a1d497d6e46962eac8bb2d"'
+      Cache-Control:
+      - max-age=0, private, must-revalidate
+      X-Request-Id:
+      - 510a314d2ea688c5bfaff1d536204974
+      X-Runtime:
+      - '0.012861'
+      Status:
+      - '200'
+      Content-Length:
+      - '4349'
+      Content-Type:
+      - application/xml; charset=utf-8
+    body:
+      encoding: UTF-8
+      string: "<publicObject id=\"druid:bb375wb8869\" published=\"2012-01-26T16:39:44-08:00\">\n
+        \ <identityMetadata>\n    <adminPolicy>druid:xs835jp8197</adminPolicy>\n    <objectCreator>DOR</objectCreator>\n
+        \   <objectId>druid:bb375wb8869</objectId>\n    <objectLabel>Kitai_bb375wb8869</objectLabel>\n
+        \   <objectType>item</objectType>\n    <otherId name=\"uuid\">15ffc446-25f5-11e1-b6f0-0016034322e7</otherId>\n
+        \   <tag>Project : Kitai</tag>\n  </identityMetadata>\n  <contentMetadata
+        type=\"image\" objectId=\"bb375wb8869\">\n    <resource type=\"image\" sequence=\"1\"
+        id=\"bb375wb8869_1\">\n      <label>kitai_4</label>\n      <file mimetype=\"image/jp2\"
+        format=\"JPEG2000\" size=\"8116154\" id=\"36105005868828_05_0001.jp2\">\n
+        \       <imageData height=\"6007\" width=\"7175\"/>\n        <attr name=\"representation\">uncropped</attr>\n
+        \     </file>\n    </resource>\n    <resource type=\"image\" sequence=\"2\"
+        id=\"bb375wb8869_2\">\n      <label>kitai_4</label>\n      <file mimetype=\"image/jp2\"
+        format=\"JPEG2000\" size=\"8141860\" id=\"36105005868828_05_0002.jp2\">\n
+        \       <imageData height=\"6017\" width=\"7174\"/>\n        <attr name=\"representation\">uncropped</attr>\n
+        \     </file>\n    </resource>\n  </contentMetadata>\n  <rightsMetadata>\n
+        \   <access type=\"discover\">\n      <machine>\n        <world/>\n      </machine>\n
+        \   </access>\n    <access type=\"read\">\n      <machine>\n        <group>stanford</group>\n
+        \     </machine>\n    </access>\n  </rightsMetadata>\n  <rdf:RDF xmlns:fedora-model=\"info:fedora/fedora-system:def/model#\"
+        xmlns:rdf=\"http://www.w3.org/1999/02/22-rdf-syntax-ns#\" xmlns:fedora=\"info:fedora/fedora-system:def/relations-external#\"
+        xmlns:hydra=\"http://projecthydra.org/ns/relations#\">\n    <rdf:Description
+        rdf:about=\"info:fedora/druid:bb375wb8869\">\n    </rdf:Description>\n  </rdf:RDF>\n
+        \ <oai_dc:dc xmlns:dc=\"http://purl.org/dc/elements/1.1/\" xmlns:srw_dc=\"info:srw/schema/1/dc-schema\"
+        xmlns:oai_dc=\"http://www.openarchives.org/OAI/2.0/oai_dc/\" xmlns:xsi=\"http://www.w3.org/2001/XMLSchema-instance\"
+        xsi:schemaLocation=\"http://www.openarchives.org/OAI/2.0/oai_dc/ http://www.openarchives.org/OAI/2.0/oai_dc.xsd\">\n
+        \   <dc:title>Kita&#x12D;</dc:title>\n    <dc:title>China</dc:title>\n    <dc:creator>Soviet
+        Union Sovetskai&#x361;a Armii&#x361;a. General&#x2B9;ny&#x12D; shtab.</dc:creator>\n
+        \   <dc:type>map</dc:type>\n    <dc:date>1968-9999</dc:date>\n    <dc:publisher>General&#x2B9;ny&#x12D;
+        shtab</dc:publisher>\n    <dc:language>rus</dc:language>\n    <dc:format>maps
+        : col. ; 37 x 53 cm. or smaller.</dc:format>\n    <dc:format>map</dc:format>\n
+        \   <dc:description>Military topographic maps.</dc:description>\n    <dc:description>Relief
+        shown by contours and spot heights. Depths shown by contours and soundings
+        on some sheets.</dc:description>\n    <dc:description>Name of jurisdiction[s]
+        covered by sheet quadrangle appears in upper margin at left, e.g.: Provint&#x361;sii&#x361;a
+        Guandun -- Avtonomny&#x12D; ra&#x12D;on Vnutrenni&#x361;ai&#x361;a Mongolii&#x361;a,
+        Provint&#x361;sii Kh&#x117;b&#x117;&#x12D;, Shan&#x2B9;si. May include name
+        and province of neighboring country.</dc:description>\n    <dc:description>Common
+        title precedes jurisdiction name at upper left on each sheet.</dc:description>\n
+        \   <dc:description>Sheets individually subtitled and numbered (with roman-alphabet
+        lettering) in upper margin, e.g.: Ch&#x117;nma&#x12D;, E-49-III -- Sich&#x117;n,
+        K-50-XXXI.</dc:description>\n    <dc:description>Ed./publication date in upper
+        border at right, e.g.: Izdanie 1970 g. -- Izdanie 1986 g. -- Izdanie 1991
+        g.</dc:description>\n    <dc:description>Former security classification in
+        upper margin at right: Sekretno.</dc:description>\n    <dc:description>Includes
+        notes, diagrams, compilation date[s], and sheet compiler's/editor's names
+        in lower margin.</dc:description>\n    <dc:description>Blue sheet no. in upper
+        margin precedes black sheet no., e.g.: 05-49-01 -- 11-50-31.</dc:description>\n
+        \   <dc:description>Text and ancillary map, \"Skhema gruntov,\" on verso.</dc:description>\n
+        \   <dc:description>In Russian.</dc:description>\n    <dc:subject/>\n    <dc:coverage>Scale
+        1:200,000. 1 cm. to 2 km.</dc:coverage>\n    <dc:coverage>116.0,27.333333
+        \n117.0,27.333333 \n117.0,26.666667 \n116.0,26.666667 \n116.0,27.333333\n</dc:coverage>\n
+        \   <dc:coverage>Pulkovo 1942</dc:coverage>\n    <dc:subject>Military maps</dc:subject>\n
+        \   <dc:subject/>\n    <dc:coverage>China</dc:coverage>\n  </oai_dc:dc>\n</publicObject>\n"
+    http_version: 
+  recorded_at: Mon, 28 Jul 2014 22:23:37 GMT
+recorded_with: VCR 2.9.2

data/spec/fixtures/vcr_cassettes/dc.yml

@@ -0,0 +1,98 @@
+---
+http_interactions:
+- request:
+    method: get
+    uri: http://purl-test.stanford.edu/bb375wb8869.xml
+    body:
+      encoding: US-ASCII
+      string: ''
+    headers:
+      Accept-Encoding:
+      - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+      Accept:
+      - "*/*"
+      User-Agent:
+      - Ruby
+  response:
+    status:
+      code: 200
+      message: ''
+    headers:
+      Date:
+      - Mon, 28 Jul 2014 22:26:22 GMT
+      Server:
+      - Apache/2.2.15 (Red Hat)
+      X-Powered-By:
+      - Phusion Passenger (mod_rails/mod_rack) 3.0.19
+      X-Ua-Compatible:
+      - IE=Edge
+      Etag:
+      - '"c7e2680ca3a1d497d6e46962eac8bb2d"'
+      Cache-Control:
+      - max-age=0, private, must-revalidate
+      X-Request-Id:
+      - 494d94bd96cf627c1eff6e27957cddd6
+      X-Runtime:
+      - '0.011981'
+      Status:
+      - '200'
+      Content-Length:
+      - '4349'
+      Content-Type:
+      - application/xml; charset=utf-8
+    body:
+      encoding: UTF-8
+      string: "<publicObject id=\"druid:bb375wb8869\" published=\"2012-01-26T16:39:44-08:00\">\n
+        \ <identityMetadata>\n    <adminPolicy>druid:xs835jp8197</adminPolicy>\n    <objectCreator>DOR</objectCreator>\n
+        \   <objectId>druid:bb375wb8869</objectId>\n    <objectLabel>Kitai_bb375wb8869</objectLabel>\n
+        \   <objectType>item</objectType>\n    <otherId name=\"uuid\">15ffc446-25f5-11e1-b6f0-0016034322e7</otherId>\n
+        \   <tag>Project : Kitai</tag>\n  </identityMetadata>\n  <contentMetadata
+        type=\"image\" objectId=\"bb375wb8869\">\n    <resource type=\"image\" sequence=\"1\"
+        id=\"bb375wb8869_1\">\n      <label>kitai_4</label>\n      <file mimetype=\"image/jp2\"
+        format=\"JPEG2000\" size=\"8116154\" id=\"36105005868828_05_0001.jp2\">\n
+        \       <imageData height=\"6007\" width=\"7175\"/>\n        <attr name=\"representation\">uncropped</attr>\n
+        \     </file>\n    </resource>\n    <resource type=\"image\" sequence=\"2\"
+        id=\"bb375wb8869_2\">\n      <label>kitai_4</label>\n      <file mimetype=\"image/jp2\"
+        format=\"JPEG2000\" size=\"8141860\" id=\"36105005868828_05_0002.jp2\">\n
+        \       <imageData height=\"6017\" width=\"7174\"/>\n        <attr name=\"representation\">uncropped</attr>\n
+        \     </file>\n    </resource>\n  </contentMetadata>\n  <rightsMetadata>\n
+        \   <access type=\"discover\">\n      <machine>\n        <world/>\n      </machine>\n
+        \   </access>\n    <access type=\"read\">\n      <machine>\n        <group>stanford</group>\n
+        \     </machine>\n    </access>\n  </rightsMetadata>\n  <rdf:RDF xmlns:fedora-model=\"info:fedora/fedora-system:def/model#\"
+        xmlns:rdf=\"http://www.w3.org/1999/02/22-rdf-syntax-ns#\" xmlns:fedora=\"info:fedora/fedora-system:def/relations-external#\"
+        xmlns:hydra=\"http://projecthydra.org/ns/relations#\">\n    <rdf:Description
+        rdf:about=\"info:fedora/druid:bb375wb8869\">\n    </rdf:Description>\n  </rdf:RDF>\n
+        \ <oai_dc:dc xmlns:dc=\"http://purl.org/dc/elements/1.1/\" xmlns:srw_dc=\"info:srw/schema/1/dc-schema\"
+        xmlns:oai_dc=\"http://www.openarchives.org/OAI/2.0/oai_dc/\" xmlns:xsi=\"http://www.w3.org/2001/XMLSchema-instance\"
+        xsi:schemaLocation=\"http://www.openarchives.org/OAI/2.0/oai_dc/ http://www.openarchives.org/OAI/2.0/oai_dc.xsd\">\n
+        \   <dc:title>Kita&#x12D;</dc:title>\n    <dc:title>China</dc:title>\n    <dc:creator>Soviet
+        Union Sovetskai&#x361;a Armii&#x361;a. General&#x2B9;ny&#x12D; shtab.</dc:creator>\n
+        \   <dc:type>map</dc:type>\n    <dc:date>1968-9999</dc:date>\n    <dc:publisher>General&#x2B9;ny&#x12D;
+        shtab</dc:publisher>\n    <dc:language>rus</dc:language>\n    <dc:format>maps
+        : col. ; 37 x 53 cm. or smaller.</dc:format>\n    <dc:format>map</dc:format>\n
+        \   <dc:description>Military topographic maps.</dc:description>\n    <dc:description>Relief
+        shown by contours and spot heights. Depths shown by contours and soundings
+        on some sheets.</dc:description>\n    <dc:description>Name of jurisdiction[s]
+        covered by sheet quadrangle appears in upper margin at left, e.g.: Provint&#x361;sii&#x361;a
+        Guandun -- Avtonomny&#x12D; ra&#x12D;on Vnutrenni&#x361;ai&#x361;a Mongolii&#x361;a,
+        Provint&#x361;sii Kh&#x117;b&#x117;&#x12D;, Shan&#x2B9;si. May include name
+        and province of neighboring country.</dc:description>\n    <dc:description>Common
+        title precedes jurisdiction name at upper left on each sheet.</dc:description>\n
+        \   <dc:description>Sheets individually subtitled and numbered (with roman-alphabet
+        lettering) in upper margin, e.g.: Ch&#x117;nma&#x12D;, E-49-III -- Sich&#x117;n,
+        K-50-XXXI.</dc:description>\n    <dc:description>Ed./publication date in upper
+        border at right, e.g.: Izdanie 1970 g. -- Izdanie 1986 g. -- Izdanie 1991
+        g.</dc:description>\n    <dc:description>Former security classification in
+        upper margin at right: Sekretno.</dc:description>\n    <dc:description>Includes
+        notes, diagrams, compilation date[s], and sheet compiler's/editor's names
+        in lower margin.</dc:description>\n    <dc:description>Blue sheet no. in upper
+        margin precedes black sheet no., e.g.: 05-49-01 -- 11-50-31.</dc:description>\n
+        \   <dc:description>Text and ancillary map, \"Skhema gruntov,\" on verso.</dc:description>\n
+        \   <dc:description>In Russian.</dc:description>\n    <dc:subject/>\n    <dc:coverage>Scale
+        1:200,000. 1 cm. to 2 km.</dc:coverage>\n    <dc:coverage>116.0,27.333333
+        \n117.0,27.333333 \n117.0,26.666667 \n116.0,26.666667 \n116.0,27.333333\n</dc:coverage>\n
+        \   <dc:coverage>Pulkovo 1942</dc:coverage>\n    <dc:subject>Military maps</dc:subject>\n
+        \   <dc:subject/>\n    <dc:coverage>China</dc:coverage>\n  </oai_dc:dc>\n</publicObject>\n"
+    http_version: 
+  recorded_at: Mon, 28 Jul 2014 22:26:22 GMT
+recorded_with: VCR 2.9.2

data/spec/fixtures/vcr_cassettes/headers.yml

@@ -0,0 +1,42 @@
+---
+http_interactions:
+- request:
+    method: get
+    uri: https://dor-oaiprovider-test.stanford.edu/oai?metadataPrefix=mods&set=is_governed_by_hy787xj5878&verb=ListIdentifiers
+    body:
+      encoding: US-ASCII
+      string: ''
+    headers:
+      User-Agent:
+      - Faraday v0.9.0
+      Accept-Encoding:
+      - gzip;q=1.0,deflate;q=0.6,identity;q=0.3
+      Accept:
+      - "*/*"
+  response:
+    status:
+      code: 200
+      message: OK
+    headers:
+      Date:
+      - Mon, 28 Jul 2014 22:02:50 GMT
+      Server:
+      - Apache/2.2.15 (Red Hat)
+      X-Content-Type-Options:
+      - nosniff
+      X-Powered-By:
+      - Phusion Passenger 4.0.10
+      Content-Length:
+      - '1402'
+      Status:
+      - 200 OK
+      Content-Type:
+      - application/xml;charset=utf-8
+    body:
+      encoding: UTF-8
+      string: |-
+        <?xml version="1.0" encoding="UTF-8" ?>
+        <?xml-stylesheet type="text/xsl" href="https://dor-oaiprovider-test.stanford.edu/oai2.xsl" ?><OAI-PMH xmlns="http://www.openarchives.org/OAI/2.0/" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://www.openarchives.org/OAI/2.0/ http://www.openarchives.org/OAI/2.0/OAI-PMH.xsd"><responseDate>2014-07-28T22:02:50Z</responseDate><request metadataPrefix="mods" set="is_governed_by_hy787xj5878" verb="ListIdentifiers">https://dor-oaiprovider-test.stanford.edu/oai</request><ListIdentifiers><header><identifier>oai:searchworks.stanford.edu/druid:jt959wc5586</identifier><datestamp>2012-02-27T23:11:50Z</datestamp><setSpec>is_member_of_jh957jy1101</setSpec><setSpec>is_member_of_collection_jh957jy1101</setSpec><setSpec>is_governed_by_hy787xj5878</setSpec></header><header><identifier>oai:searchworks.stanford.edu/druid:pd817cv1860</identifier><datestamp>2012-03-28T21:25:56Z</datestamp><setSpec>is_member_of_jh957jy1101</setSpec><setSpec>is_member_of_collection_jh957jy1101</setSpec><setSpec>is_governed_by_hy787xj5878</setSpec></header><header><identifier>oai:searchworks.stanford.edu/druid:qk487nd2690</identifier><datestamp>2013-05-22T23:38:57Z</datestamp><setSpec>is_member_of_jh957jy1101</setSpec><setSpec>is_member_of_collection_jh957jy1101</setSpec><setSpec>is_governed_by_hy787xj5878</setSpec></header></ListIdentifiers></OAI-PMH>
+    http_version: 
+  recorded_at: Mon, 28 Jul 2014 22:02:50 GMT
+recorded_with: VCR 2.9.2