Parse xls booleans as booleans, not integeres (#232)

* [parsers][m]: parse xls booleans as booleans, not integeres

* pin openpyxl to versions < 2.5 because of a bug on 2.5

setup.py

@@ -38,7 +38,7 @@ def read(*paths):
     # Format: xls
     'xlrd>=1.0,<2.0',
     # Format: xlsx
-    'openpyxl>=2.4,<3.0',
+    'openpyxl>=2.4,<2.5',
 ]
 INSTALL_FORMAT_DATAPACKAGE_REQUIRES = [
     'datapackage>=1.1.3,<2.0',

tabulator/parsers/xls.py

@@ -87,9 +87,14 @@ def __iter_extended_rows(self):
             row_number = x + 1
             row = []
             for y, value in enumerate(self.__sheet.row_values(x)):
+                # hacky way to parse booleans (cell values are integers (0/1) other way)
+                if self.__sheet.cell(x, y).ctype == xlrd.XL_CELL_BOOLEAN:
+                    value = bool(value)
                 if self.__fill_merged_cells:
                     for xlo, xhi, ylo, yhi in self.__sheet.merged_cells:
                         if x in range(xlo, xhi) and y in range(ylo, yhi):
                             value = self.__sheet.cell_value(xlo, ylo)
+                            if self.__sheet.cell(xlo, ylo).ctype == xlrd.XL_CELL_BOOLEAN:
+                                value = bool(value)
                 row.append(value)
             yield (row_number, None, row)

tests/formats/test_xls.py

@@ -65,6 +65,23 @@ def test_stream_xlsx_merged_cells_fill():
         assert stream.read() == [['data', 'data'], ['data', 'data'], ['data', 'data']]
 
 
+def test_stream_xls_with_boolean():
+    with Stream('data/special/table-with-booleans.xls') as stream:
+        assert stream.headers is None
+        assert stream.read() == [['id', 'boolean'], [1.0, True], [2.0, False]]
+
+
+def test_stream_xlsx_merged_cells_boolean():
+    source = 'data/special/merged-cells-boolean.xls'
+    with Stream(source) as stream:
+        assert stream.read() == [[True, ''], ['', ''], ['', '']]
+
+
+def test_stream_xlsx_merged_cells_fill_boolean():
+    source = 'data/special/merged-cells-boolean.xls'
+    with Stream(source, fill_merged_cells=True) as stream:
+        assert stream.read() == [[True, True], [True, True], [True, True]]
+
 # Parser
 
 def test_parser_xls():