re-order the units again (back to longest first)

README.md

@@ -1452,8 +1452,6 @@ Catena Dev branch changes
 
 * Exported XML apparatus uses the n attribute for the identifier of ```<ab>``` elements rather than xml:id. The value of the attribute remains unchanged.
 
-* In exprter.py the order of overlapped units in relation to top line units has changed. The shortest unit starting at each index point now comes first (previously the the longest unit came first).
-
 * In exporter.py ```get_lemma_text()``` now takes start and end arguments as strings. This is important for dealing with joined units in inheriting exporters.
 
 * In exporter.py ```get_text()``` when om and lac are returned their string value is always returned with the full stop eg. ```om.``` 

exporter.py

@@ -357,8 +357,11 @@ def get_app_units(self, apparatus, overtext, context, missing):
                 app_list.append(app)
         return app_list
 
-    def get_overtext_data(self, structure):
-        return {'current': structure['overtext'][0]}
+    def get_overtext_data(self, entry):
+        return {'current': entry['structure']['overtext'][0]}
+
+    def sort_units(self, unit):
+        return (unit['start'], -unit['end'])
 
     def get_unit_xml(self, entry):
         """Function to turn the JSON apparatus of the collation unit into TEI XML.
@@ -428,10 +431,8 @@ def get_unit_xml(self, entry):
         # if we are ignoring the basetext add it to our missing list so it isn't listed (except in lemma)
         if self.ignore_basetext:
             missing.append(basetext_siglum)
-        # this sort will change the order of the overlap units so shortest starting at each index point comes first
-        apparatus = sorted(apparatus, key=lambda d: (d['start'], d['end']))
-
-        overtext = self.get_overtext_data(entry['structure'])
+        apparatus = sorted(apparatus, key=self.sort_units)
+        overtext = self.get_overtext_data(entry)
 
         app_units = self.get_app_units(apparatus, overtext, context, missing)
         for app in app_units:

tests_exporter.py

@@ -729,40 +729,40 @@ def test_get_lemma_text_defaults(self):
 
         # test a few ranges where there is data
         expected_text = ['ἀποκαλύψαι τὸν υἱὸν']
-        generated_text = exp.get_lemma_text(overtext, 2, 6)
+        generated_text = exp.get_lemma_text(overtext, '2', '6')
         self.assertEqual(expected_text, generated_text)
 
         expected_text = ['ἐν']
-        generated_text = exp.get_lemma_text(overtext, 10, 10)
+        generated_text = exp.get_lemma_text(overtext, '10', '10')
         self.assertEqual(expected_text, generated_text)
 
         expected_text = ['ἀποκαλύψαι']
-        generated_text = exp.get_lemma_text(overtext, 1, 3)
+        generated_text = exp.get_lemma_text(overtext, '1', '3')
         self.assertEqual(expected_text, generated_text)
 
         expected_text = ['ἔθνεσιν']
-        generated_text = exp.get_lemma_text(overtext, 24, 24)
+        generated_text = exp.get_lemma_text(overtext, '24', '24')
         self.assertEqual(expected_text, generated_text)
 
         # test where there is no data
         expected_text = ['', 'om']
-        generated_text = exp.get_lemma_text(overtext, 27, 27)
+        generated_text = exp.get_lemma_text(overtext, '27', '27')
         self.assertEqual(expected_text, generated_text)
 
         expected_text = ['', 'om']
-        generated_text = exp.get_lemma_text(overtext, 1, 1)
+        generated_text = exp.get_lemma_text(overtext, '1', '1')
         self.assertEqual(expected_text, generated_text)
 
     def test_get_lemma_text_with_punctuation(self):
         exp = Exporter(include_punctuation=True)
         overtext = {'current': self.OVERTEXT[0]}
 
         expected_text = ['ἔθνεσιν,']
-        generated_text = exp.get_lemma_text(overtext, 24, 24)
+        generated_text = exp.get_lemma_text(overtext, '24', '24')
         self.assertEqual(expected_text, generated_text)
 
         expected_text = ['(ἐν)']
-        generated_text = exp.get_lemma_text(overtext, 10, 10)
+        generated_text = exp.get_lemma_text(overtext, '10', '10')
         self.assertEqual(expected_text, generated_text)
 
     def test_get_witnesses(self):
@@ -1374,8 +1374,8 @@ def test_get_unit_xml_unit_sorting(self, mocked_get_app_units):
                                   {'start': 2, 'end': 4},
                                   {'start': 9, 'end': 9},
                                   {'start': 2, 'end': 8}]
-        expected_ordered_units = [{'start': 2, 'end': 4},
-                                  {'start': 2, 'end': 8},
+        expected_ordered_units = [{'start': 2, 'end': 8},
+                                  {'start': 2, 'end': 4},       
                                   {'start': 6, 'end': 8},
                                   {'start': 9, 'end': 9}]
         original_app = {'context': 'Gal.1.1',
@@ -1403,7 +1403,7 @@ def test_get_unit_apparatus_line_sorting(self, mocked_get_app_units):
                              'om_readings': []}
                }
         exp = Exporter()
-        expected_app_order = [{'start': 2, 'end': 4}, {'start': 2, 'end': 8}, {'start': 9, 'end': 9}]
+        expected_app_order = [{'start': 2, 'end': 8}, {'start': 2, 'end': 4}, {'start': 9, 'end': 9}]
         exp.get_unit_xml(app)
         mocked_get_app_units.assert_called_with(expected_app_order, {'current': {'id': 'basetext'}}, 'Gal.1.1', [])