fixed eurovoc tag loader for arbitrary URLs

ckanext/dgu/bin/tag_loader.py

@@ -4,6 +4,8 @@ import zipfile
 import urllib2
 import xml.etree.ElementTree as eTree
 
+EUROVOC_FILENAME = "eurovoc_xml.zip"
+GEMET_FILENAME = "gemet-definitions.rdf"
 
 class TagLoader:
     def __init__(self, name, url):
@@ -14,13 +16,15 @@ class TagLoader:
     def loader(self):
         response = urllib2.urlopen(self._url)
         data = response.read()
-        filename = self._url.split('/')[-1]
-        file_ = open(filename, 'w')
+        if self._name == 'eurovoc':
+            file_ = open(EUROVOC_FILENAME, 'w')
+        else:
+            file_ = open(GEMET_FILENAME, 'w')
         file_.write(data)
         file_.close()
 
         if self._name == 'eurovoc':
-            with zipfile.ZipFile("eurovoc_xml.zip", "r") as zip_ref:
+            with zipfile.ZipFile(EUROVOC_FILENAME, "r") as zip_ref:
                 zip_ref.extractall("eurovoc_xml")
 
     def parse_xml(self, file_name, to_find):
@@ -50,7 +54,7 @@ class TagLoader:
 
     def clean_up(self):
         if self._name == 'eurovoc':
-            os.remove('eurovoc_xml.zip')
+            os.remove(EUROVOC_FILENAME)
             shutil.rmtree('eurovoc_xml')
         elif self._name == 'gemet':
             os.remove('gemet-definitions.rdf')