Fix changed URL for OPUS API

.github/workflows/python-test.yaml

@@ -3,6 +3,10 @@ name: Upload Python Package
 on:
   push:
     branches: [develop]
+  pull_request:
+    branches:
+      - master
+      - develop
 
 permissions:
   contents: read
@@ -13,14 +17,19 @@ jobs:
       run:
         working-directory: opustools_pkg
 
-    runs-on: ubuntu-latest
+    strategy:
+      fail-fast: false
+      matrix:
+        os: [ubuntu-latest]
+        python-version: ["3.10", "3.11", "3.12", "3.13", "3.14"]
+    runs-on: ${{ matrix.os }}
 
     steps:
     - uses: actions/checkout@v3
-    - name: Set up Python
+    - name: Set up Python ${{ matrix.python-version }}
       uses: actions/setup-python@v3
       with:
-        python-version: '3.x'
+        python-version: ${{ matrix.python-version }}
     - name: Install dependencies
       run: |
         python -m pip install --upgrade pip

opustools_pkg/opustools/opus_get.py

@@ -49,21 +49,23 @@ def __init__(self, source=None, target=None, directory=None,
             self.fromto = [source, target]
             self.fromto.sort()
 
-        self.url = 'http://opus.nlpl.eu/opusapi/?'
+        self.url = 'https://opus.nlpl.eu/opusapi'
         #self.url = 'http://127.0.0.1:5000/?'
 
         urlparts = [(source, 'source'), (target, 'target'),
             (directory, 'corpus'), (release, 'version'),
             (preprocess, 'preprocessing')]
         self.parameters = {}
 
+        url_parameters = ''
         for a in urlparts:
             if a[0]:
                 self.parameters[a[1]] = a[0]
                 if a[0] == ' ':
-                    self.url += a[1] + '=&'
+                    url_parameters += a[1] + '=&'
                 else:
-                    self.url += a[1] + '=' + a[0] + '&'
+                    url_parameters += a[1] + '=' + a[0] + '&'
+        self.url += '?' + url_parameters
 
         if not os.path.exists(download_dir):
             os.makedirs(download_dir)
@@ -163,7 +165,7 @@ def download(self, corpora, total_size):
                         reporthook=self.progress_status)
                     print('')
                 except urllib.error.URLError as e:
-                    print('Unable to retrieve the data.')
+                    print(f'Unable to retrieve the data from {self.url}: {e}')
                     return
 
     def get_file_info_output(self, c):
@@ -193,24 +195,23 @@ def get_files(self):
                 if self.local_db:
                     languages = self.dbo.run_languages_query(self.parameters)
                 else:
-                    languages = self.get_response(self.url+'languages=True')['languages']
+                    languages = self.get_response(self.url + '?languages=True')['languages']
                 print(', '.join([str(l) for l in languages]))
                 return
             elif self.list_corpora:
                 if self.local_db:
                     corpus_list = self.dbo.run_corpora_query(self.parameters)
                 else:
-                    corpus_list = self.get_response(self.url+'corpora=True')['corpora']
+                    corpus_list = self.get_response(self.url + '?corpora=True')['corpora']
                 print(', '.join([str(c) for c in corpus_list]))
                 return
             else:
                 corpora, total_size = self.get_corpora_data()
         except urllib.error.URLError as e:
-            print('Unable to retrieve the data.')
+            print(f'Unable to retrieve the data from {self.url}: {e}')
             return
 
         if not self.list_resources:
             self.download(corpora, total_size)
         else:
             self.print_files(corpora, total_size)
-

opustools_pkg/tests/test_opus_get.py

@@ -40,7 +40,7 @@ def test_get_files_invalid_url(self):
         opg.get_files()
         sys.stdout = old_stdout
 
-        self.assertEqual(printout.getvalue(), 'Unable to retrieve the data.\n')
+        self.assertIn('Unable to retrieve the data', printout.getvalue())
 
     @mock.patch('opustools.opus_get.input', create=True)
     def test_download_invalid_url(self, mocked_input):
@@ -55,7 +55,7 @@ def test_download_invalid_url(self, mocked_input):
         opg.download(corpora, total_size)
         sys.stdout = old_stdout
 
-        self.assertEqual(printout.getvalue(), 'Unable to retrieve the data.\n')
+        self.assertIn('Unable to retrieve the data', printout.getvalue())
 
     @mock.patch('opustools.opus_get.input', create=True)
     def test_dont_list_files_that_are_already_in_path(self, mocked_input):
@@ -136,4 +136,3 @@ def test_indonesian(self):
 
 if __name__ == '__main__':
     unittest.main()
-