[#57] Update the tests to work with newest openpyxl

There's a couple of oddities here, but I think it's important to be
using this latest version to fix
OpenDataServices/cove#176

I've flagged the potential problems as new GitHub issues:
* Python 2 ints not roundtripping properly
  #79
* Sheets with just one cell are not handled properly
  #78

Author: Bjwebb

flattentool/tests/test_init.py

@@ -3,6 +3,7 @@
 from flattentool import decimal_default, unflatten
 from decimal import Decimal
 import json
+import sys
 
 
 def test_decimal_default():
@@ -202,4 +203,4 @@ def test_unflatten_xslx_unicode(tmpdir):
         output_name=tmpdir.join('release.json').strpath,
         main_sheet_name='main')
     reloaded_json = json.load(tmpdir.join('release.json'))
-    assert reloaded_json == {'main': [{'ocid': 1, 'id': 'éαГ😼𝒞人'}]}
+    assert reloaded_json == {'main': [{'ocid': 1 if sys.version > '3' else '1', 'id': 'éαГ😼𝒞人'}]}

flattentool/tests/test_input.py

@@ -94,8 +94,12 @@ def test_csv_no_files(self, tmpdir):
 
     def test_xlsx_no_file(self, tmpdir):
         xlsxinput = XLSXInput(input_name=tmpdir.strpath.join('test.xlsx'), main_sheet_name='main')
-        with pytest.raises(openpyxl.exceptions.InvalidFileException):
-            xlsxinput.read_sheets()
+        if sys.version > '3':
+            with pytest.raises(FileNotFoundError):
+                xlsxinput.read_sheets()
+        else:
+            with pytest.raises(IOError):
+                xlsxinput.read_sheets()
 
     def test_xlsx_no_main_sheet(self):
         xlsxinput = XLSXInput(input_name='flattentool/tests/fixtures/xlsx/basic.xlsx', main_sheet_name='notmain')

flattentool/tests/test_output.py

@@ -34,8 +34,7 @@ def test_blank_sheets(tmpdir):
     wb = openpyxl.load_workbook(tmpdir.join('release.xlsx').strpath)
     assert wb.get_sheet_names() == ['release']
     assert len(wb['release'].rows) == 1
-    assert len(wb['release'].rows[0]) == 1
-    assert wb['release'].rows[0][0].value == None
+    assert len(wb['release'].rows[0]) == 0
 
     # Check CSV is Empty
     assert tmpdir.join('release').listdir() == [ tmpdir.join('release').join('release.csv') ]
@@ -47,7 +46,7 @@ def test_populated_header(tmpdir):
         subsheet = Sheet(root_id='ocid')
         subsheet.add_field('c')
         spreadsheet_output = spreadsheet_output_class(
-            parser=MockParser(['a'], {'b': subsheet}),
+            parser=MockParser(['a', 'd'], {'b': subsheet}),
             main_sheet_name='release',
             output_name=os.path.join(tmpdir.strpath, 'release'+output.FORMATS_SUFFIX[format_name]))
         spreadsheet_output.write_sheets()
@@ -56,7 +55,7 @@ def test_populated_header(tmpdir):
     wb = openpyxl.load_workbook(tmpdir.join('release.xlsx').strpath)
     assert wb.get_sheet_names() == ['release', 'b']
     assert len(wb['release'].rows) == 1
-    assert [ x.value for x in wb['release'].rows[0] ] == [ 'a' ]
+    assert [ x.value for x in wb['release'].rows[0] ] == [ 'a', 'd' ]
     assert len(wb['b'].rows) == 1
     assert [ x.value for x in wb['b'].rows[0] ] == [ 'ocid', 'c' ]
 
@@ -65,14 +64,14 @@ def test_populated_header(tmpdir):
         tmpdir.join('release').join('release.csv'),
         tmpdir.join('release').join('b.csv')
     ])
-    assert tmpdir.join('release', 'release.csv').read().strip('\r\n') == 'a'
+    assert tmpdir.join('release', 'release.csv').read().strip('\r\n') == 'a,d'
     assert tmpdir.join('release', 'b.csv').read().strip('\r\n') == 'ocid,c'
 
 
 def test_empty_lines(tmpdir):
     subsheet = Sheet(root_id='ocid')
     subsheet.add_field('c')
-    parser = MockParser(['a'], {'b': subsheet})
+    parser = MockParser(['a', 'd'], {'b': subsheet})
     parser.main_sheet.lines = []
     for format_name, spreadsheet_output_class in output.FORMATS.items():
         spreadsheet_output = spreadsheet_output_class(
@@ -85,7 +84,7 @@ def test_empty_lines(tmpdir):
     wb = openpyxl.load_workbook(tmpdir.join('release.xlsx').strpath)
     assert wb.get_sheet_names() == ['release', 'b']
     assert len(wb['release'].rows) == 1
-    assert [ x.value for x in wb['release'].rows[0] ] == [ 'a' ]
+    assert [ x.value for x in wb['release'].rows[0] ] == [ 'a', 'd' ]
     assert len(wb['b'].rows) == 1
     assert [ x.value for x in wb['b'].rows[0] ] == [ 'ocid', 'c' ]
 
@@ -94,7 +93,7 @@ def test_empty_lines(tmpdir):
         tmpdir.join('release').join('release.csv'),
         tmpdir.join('release').join('b.csv')
     ])
-    assert tmpdir.join('release', 'release.csv').read().strip('\r\n') == 'a'
+    assert tmpdir.join('release', 'release.csv').read().strip('\r\n') == 'a,d'
     assert tmpdir.join('release', 'b.csv').read().strip('\r\n') == 'ocid,c'
 
 

flattentool/tests/test_roundtrip.py

@@ -1,6 +1,7 @@
 from flattentool import unflatten, flatten
 import json
 import pytest
+import sys
 
 
 @pytest.mark.parametrize('output_format', ['xlsx', 'csv'])
@@ -9,12 +10,12 @@ def test_roundtrip(tmpdir, output_format):
     base_name = 'flattentool/tests/fixtures/tenders_releases_base.json'
     flatten(
         input_name=input_name,
-        output_name=tmpdir.join('flattened').strpath,
+        output_name=tmpdir.join('flattened').strpath+'.'+output_format,
         output_format=output_format,
         schema='flattentool/tests/fixtures/release-schema.json',
         main_sheet_name='releases')
     unflatten(
-        input_name=tmpdir.join('flattened').strpath,
+        input_name=tmpdir.join('flattened').strpath+'.'+output_format,
         output_name=tmpdir.join('roundtrip.json').strpath,
         input_format=output_format,
         base_json=base_name,
@@ -37,15 +38,15 @@ def test_roundtrip_360(tmpdir, output_format, use_titles):
     input_name = 'flattentool/tests/fixtures/WellcomeTrust-grants_fixed_2_grants.json'
     flatten(
         input_name=input_name,
-        output_name=tmpdir.join('flattened').strpath,
+        output_name=tmpdir.join('flattened').strpath+'.'+output_format,
         output_format=output_format,
         schema='flattentool/tests/fixtures/360-giving-schema.json',
         main_sheet_name='grants',
         root_list_path='grants',
         root_id='',
         use_titles=use_titles)
     unflatten(
-        input_name=tmpdir.join('flattened').strpath,
+        input_name=tmpdir.join('flattened').strpath+'.'+output_format,
         output_name=tmpdir.join('roundtrip.json').strpath,
         input_format=output_format,
         schema='flattentool/tests/fixtures/360-giving-schema.json',
@@ -58,7 +59,8 @@ def test_roundtrip_360(tmpdir, output_format, use_titles):
 
     # Currently not enough information to successfully roundtrip that values
     # are numbers, when this is not required by the schema
-    if output_format == 'csv':
+    # for CSV, and for openpyxl under Python 2
+    if output_format == 'csv' or sys.version_info < (3, 0):
         for grant in original_json['grants']:
             grant['plannedDates'][0]['duration'] = str(grant['plannedDates'][0]['duration'])