Revert 2819d02 completely & a954641's test changes

inspirehep · Mar 5, 2018 · 9062db9 · 9062db9
1 parent 2819d02
commit 9062db9
Show file tree

Hide file tree

Showing 9 changed files with 14 additions and 15 deletions.
diff --git a/hepcrawl/spiders/edp_spider.py b/hepcrawl/spiders/edp_spider.py
@@ -16,7 +16,6 @@
 import tarfile
 from tempfile import mkdtemp
 
-from re import sub
 from scrapy import Request
 from scrapy.spiders import XMLFeedSpider
 
@@ -508,7 +507,7 @@ def _get_external_links(self, ref):
         urls = []
         for ext_link in ext_links:
             if "doi" in ext_link:
-                doi = "doi:" + ext_link.sub("https?://(dx\.)?doi.org/", "")
+                doi = "doi:" + ext_link.replace("http://dx.doi.org/", "")
             else:
                 urls.append(ext_link)
 

diff --git a/tests/unit/responses/aps/PhysRevD.96.095036.xml b/tests/unit/responses/aps/PhysRevD.96.095036.xml
diff --git a/tests/unit/responses/aps/PhysRevX.7.021022.xml b/tests/unit/responses/aps/PhysRevX.7.021022.xml
@@ -96,7 +96,7 @@
         <award-group award-type="unspecified">
           <funding-source>
             <named-content content-type="name">National Science Foundation</named-content>
-            <named-content content-type="fundref:id">https://doi.org/10.13039/100000001</named-content>
+            <named-content content-type="fundref:id">http://dx.doi.org/10.13039/100000001</named-content>
             <named-content content-type="fundref:alternate-name">NSF</named-content>
             <named-content content-type="country">http://sws.geonames.org/6252001/</named-content>
             <named-content content-type="aps-code">NSF</named-content>
@@ -106,22 +106,22 @@
         <award-group award-type="unspecified">
           <funding-source>
             <named-content content-type="name">Gordon and Betty Moore Foundation</named-content>
-            <named-content content-type="fundref:id">https://doi.org/10.13039/100000936</named-content>
+            <named-content content-type="fundref:id">http://dx.doi.org/10.13039/100000936</named-content>
             <named-content content-type="fundref:alternate-name">Gordon E. and Betty I. Moore Foundation</named-content>
             <named-content content-type="country">http://sws.geonames.org/6252001/</named-content>
           </funding-source>
         </award-group>
         <award-group award-type="unspecified">
           <funding-source>
             <named-content content-type="name">Simons Foundation</named-content>
-            <named-content content-type="fundref:id">https://doi.org/10.13039/100000893</named-content>
+            <named-content content-type="fundref:id">http://dx.doi.org/10.13039/100000893</named-content>
             <named-content content-type="country">http://sws.geonames.org/6252001/</named-content>
           </funding-source>
         </award-group>
         <award-group award-type="unspecified">
           <funding-source>
             <named-content content-type="name">H2020 European Research Council</named-content>
-            <named-content content-type="fundref:id">https://doi.org/10.13039/100010663</named-content>
+            <named-content content-type="fundref:id">http://dx.doi.org/10.13039/100010663</named-content>
             <named-content content-type="fundref:alternate-name">H2020 Excellent Science - European Research Council</named-content>
             <named-content content-type="fundref:alternate-name">European Research Council</named-content>
             <named-content content-type="fundref:alternate-name">ERC</named-content>

diff --git a/tests/unit/responses/aps/aps_single_response.json b/tests/unit/responses/aps/aps_single_response.json
@@ -42,14 +42,14 @@
          },
          "fundings":[
             {
-               "funderId":"https://doi.org/10.13039/100000001",
+               "funderId":"http://dx.doi.org/10.13039/100000001",
                "awards":[
                   "IIS\u20131247696"
                ],
                "funderName":"National Science Foundation"
             },
             {
-               "funderId":"https://doi.org/10.13039/100007231",
+               "funderId":"http://dx.doi.org/10.13039/100007231",
                "awards":[
 
                ],

diff --git a/tests/unit/responses/elsevier/sample_consyn_record.xml b/tests/unit/responses/elsevier/sample_consyn_record.xml
@@ -16,7 +16,7 @@
 	 xmlns:sb="http://www.elsevier.com/xml/common/struct-bib/schema"
 	 xmlns:xlink="http://www.w3.org/1999/xlink">
 	 <rdf:RDF>
-	 	<rdf:Description rdf:about="https://doi.org/10.1016/0370-2693(88)91603-6">
+	 	<rdf:Description rdf:about="http://dx.doi.org/10.1016/0370-2693(88)91603-6">
 	 		<dct:format>application/xml</dct:format>
 	 		<dct:title>Toward classification of conformal theories</dct:title>
 	 		<dct:creator>Cumrun Vafa</dct:creator>
@@ -34,7 +34,7 @@
 	 		<prism:startingPage>421</prism:startingPage>
 	 		<prism:endingPage>426</prism:endingPage>
 	 		<prism:doi>10.1016/0370-2693(88)91603-6</prism:doi>
-	 		<prism:url>https://doi.org/10.1016/0370-2693(88)91603-6</prism:url>
+	 		<prism:url>http://dx.doi.org/10.1016/0370-2693(88)91603-6</prism:url>
 	 		<dct:identifier>doi:10.1016/0370-2693(88)91603-6</dct:identifier>
 	 	</rdf:Description>
 	 </rdf:RDF>

diff --git a/tests/unit/responses/hindawi/test_1.xml b/tests/unit/responses/hindawi/test_1.xml
@@ -74,7 +74,7 @@ detection and observation of the prompt optical emission of the &#x201C;naked-ey
             <subfield code="n">fulltext</subfield>
           </datafield>
           <datafield tag="856" ind1="4" ind2="0">
-            <subfield code="u">https://doi.org/10.1155/2010/194946</subfield>
+            <subfield code="u">http://dx.doi.org/10.1155/2010/194946</subfield>
             <subfield code="z">Full text URL</subfield>
           </datafield>
         </marc:record>

diff --git a/tests/unit/test_edp.py b/tests/unit/test_edp.py
@@ -555,7 +555,7 @@ def record_references_only():
         <title>References</title>
             <ref id="R5"><label>5.</label><mixed-citation publication-type="journal" id="a"><string-name><given-names>R.V.</given-names> <surname>Krishnan</surname></string-name>, <string-name><given-names>G.</given-names> <surname>Panneerselvam</surname></string-name>, <string-name><given-names>P.</given-names> <surname>Manikandan</surname></string-name> <string-name><given-names>M.P.</given-names> <surname>Antony</surname></string-name>, <string-name><given-names>K.</given-names> <surname>Nagarajan</surname></string-name>, <source>J. Nucl. Radiochem. Sci.</source>, <volume>10</volume>.<issue>1</issue>, <fpage>19</fpage>–<lpage>26</lpage> (<year>2009</year>).</mixed-citation></ref>
 
-            <ref id="R44"><label>44.</label><mixed-citation publication-type="journal"><string-name><given-names>L.</given-names> <surname>Cronin</surname></string-name>, <string-name><given-names>P.</given-names> <surname>Sojka</surname></string-name>, <string-name><given-names>A.</given-names> <surname>Lefebvre</surname></string-name>, <source>SAE Technical Paper</source>, DOI: <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.4271/852086">10.4271/852086</ext-link>, (<year>1985</year>)</mixed-citation></ref>
+            <ref id="R44"><label>44.</label><mixed-citation publication-type="journal"><string-name><given-names>L.</given-names> <surname>Cronin</surname></string-name>, <string-name><given-names>P.</given-names> <surname>Sojka</surname></string-name>, <string-name><given-names>A.</given-names> <surname>Lefebvre</surname></string-name>, <source>SAE Technical Paper</source>, DOI: <ext-link ext-link-type="uri" xlink:href="http://dx.doi.org/10.4271/852086">10.4271/852086</ext-link>, (<year>1985</year>)</mixed-citation></ref>
 
             <ref id="R3"><label>3.</label><mixed-citation publication-type="book"><string-name><given-names>T.</given-names> <surname>Aliyev</surname></string-name>, <string-name><given-names>Т.</given-names> <surname>Belyaev</surname></string-name>, <string-name><given-names>S.</given-names> <surname>Gallagher</surname></string-name> <article-title>Simulation in ANSYS flow to the gas purification section of the multicomponent gas mixture through the dust cyclone CKBN GP-628</article-title>. <source>Mechanical engineering</source>, <publisher-loc>Moscow</publisher-loc>, №<issue>10</issue>, (<year>2014</year>).</mixed-citation></ref>
 

diff --git a/tests/unit/test_elsevier.py b/tests/unit/test_elsevier.py
@@ -77,7 +77,7 @@ def parsed_node():
                 </oa:sponsor>
                 <oa:userLicense xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance">http://creativecommons.org/licenses/by/3.0/</oa:userLicense>
             </oa:openAccessInformation>
-            <rdf:Description rdf:about="https://doi.org/10.1016/0370-2693(88)91603-6">
+            <rdf:Description rdf:about="http://dx.doi.org/10.1016/0370-2693(88)91603-6">
                 <dct:title>Toward classification of conformal theories</dct:title>
                 <prism:doi>10.1016/0370-2693(88)91603-6</prism:doi>
                 <prism:startingPage>421</prism:startingPage>

diff --git a/tests/unit/test_hindawi.py b/tests/unit/test_hindawi.py
@@ -82,7 +82,7 @@ def test_files(record):
 
 def test_urls(record):
     """Test url in record."""
-    urls = ["https://doi.org/10.1155/2010/194946"]
+    urls = ["http://dx.doi.org/10.1155/2010/194946"]
     assert "urls" in record
     assert len(record["urls"]) == 1