OcrdMets: fix (return) types when caching is on, #957

OCR-D · Nov 29, 2022 · d0b5513 · d0b5513
1 parent 862ece1
commit d0b5513
Show file tree

Hide file tree

Showing 2 changed files with 14 additions and 13 deletions.
diff --git a/ocrd_models/ocrd_models/ocrd_mets.py b/ocrd_models/ocrd_models/ocrd_mets.py
@@ -557,7 +557,7 @@ def physical_pages(self):
         List all page IDs (the ``@ID`` of each physical ``mets:structMap`` ``mets:div``)
         """
         if self._cache_flag:
-            return self._page_cache.values()
+            return list(self._page_cache.keys())
 
         return self._tree.getroot().xpath(
             'mets:structMap[@TYPE="PHYSICAL"]/mets:div[@TYPE="physSequence"]/mets:div[@TYPE="page"]/@ID',
@@ -685,7 +685,7 @@ def remove_physical_page(self, ID):
             mets_div = self._tree.getroot().xpath(
                 'mets:structMap[@TYPE="PHYSICAL"]/mets:div[@TYPE="physSequence"]/mets:div[@TYPE="page"][@ID="%s"]' % ID,
                 namespaces=NS)
-        if mets_div is not None:
+        if mets_div:
             mets_div[0].getparent().remove(mets_div[0])
             if self._cache_flag:
                 del self._page_cache[ID]

diff --git a/tests/model/test_ocrd_mets.py b/tests/model/test_ocrd_mets.py
@@ -37,6 +37,15 @@ def _fixture(request):
     yield mets
 
 
+@pytest.fixture(name='sbb_directory_ocrd_mets', params=CACHING_ENABLED)
+def _fixture_sbb(tmp_path, request):
+    src_path = assets.path_to('SBB0000F29300010000/data')
+    dst_path = tmp_path / 'SBB_directory'
+    shutil.copytree(src_path, dst_path)
+    mets_path = str(join(dst_path, 'mets.xml'))
+    yield OcrdMets(filename=mets_path, cache_flag=request.param)
+
+
 def test_unique_identifier():
     mets = OcrdMets(filename=assets.url_of('SBB0000F29300010000/data/mets.xml'))
     assert mets.unique_identifier == 'http://resolver.staatsbibliothek-berlin.de/SBB0000F29300010000', 'Right identifier'
@@ -89,7 +98,8 @@ def test_find_all_files_local_only(sbb_sample_01):
 
 def test_physical_pages(sbb_sample_01):
     assert len(sbb_sample_01.physical_pages) == 3, '3 physical pages'
-
+    assert isinstance(sbb_sample_01.physical_pages, list)
+    assert isinstance(sbb_sample_01.physical_pages[0], str)
 
 def test_physical_pages_from_empty_mets():
     mets = OcrdMets(content="<mets></mets>")
@@ -98,15 +108,6 @@ def test_physical_pages_from_empty_mets():
     assert len(mets.physical_pages) == 1, '1 physical page'
 
 
-@pytest.fixture(name='sbb_directory_ocrd_mets')
-def _fixture_sbb(tmp_path):
-    src_path = assets.path_to('SBB0000F29300010000/data')
-    dst_path = tmp_path / 'SBB_directory'
-    shutil.copytree(src_path, dst_path)
-    mets_path = str(join(dst_path, 'mets.xml'))
-    yield OcrdMets(filename=mets_path)
-
-
 def test_physical_pages_for_fileids(sbb_directory_ocrd_mets):
     assert sbb_directory_ocrd_mets.get_physical_pages(
         for_fileIds=['FILE_0002_IMAGE']) == ['PHYS_0002']
@@ -265,6 +266,7 @@ def test_remove_page(sbb_directory_ocrd_mets):
 def test_remove_physical_page_fptr(sbb_directory_ocrd_mets):
     assert sbb_directory_ocrd_mets.get_physical_pages(for_fileIds=['FILE_0002_IMAGE']), ['PHYS_0002']
     sbb_directory_ocrd_mets.remove_physical_page_fptr('FILE_0002_IMAGE')
+    sbb_directory_ocrd_mets.remove_physical_page_fptr('FILE_0002_IMAGE')
     assert sbb_directory_ocrd_mets.get_physical_pages(for_fileIds=['FILE_0002_IMAGE']), [None]
 
 
@@ -346,7 +348,6 @@ def test_merge(sbb_sample_01):
     sbb_sample_01.merge(other_mets, fileGrp_mapping={'OCR-D-IMG': 'FOO'})
     assert len(sbb_sample_01.file_groups) == 18
 
-
 def test_invalid_filegrp():
     """addresses https://github.com/OCR-D/core/issues/746"""