Add support for dates

Author: roberthdevries

pandas/io/excel/_odfreader.py

@@ -118,6 +118,7 @@ def get_sheet_data(self, sheet, convert_float: bool) -> List[List[Scalar]]:
             if len(row) < max_row_len:
                 row.extend([self.empty_value] * (max_row_len - len(row)))
 
+        # print(table)
         return table
 
     def _get_row_repeat(self, row) -> int:
@@ -148,7 +149,11 @@ def _is_empty_row(self, row) -> bool:
     def _get_cell_value(self, cell, convert_float: bool) -> Scalar:
         from odf.namespaces import OFFICENS
 
+        # print("cell: ", cell, convert_float)
+
         cell_type = cell.attributes.get((OFFICENS, "value-type"))
+        cell_value = cell.attributes.get((OFFICENS, "value"))
+        # print("type=", cell_type, "value=", repr(cell_value))
         if cell_type == "boolean":
             if str(cell) == "TRUE":
                 return True
@@ -157,14 +162,16 @@ def _get_cell_value(self, cell, convert_float: bool) -> Scalar:
             return self.empty_value
         elif cell_type == "float":
             # GH5394
-            cell_value = float(cell.attributes.get((OFFICENS, "value")))
-
-            if cell_value == 0.0:  # NA handling
-                return str(cell)
 
+            value = cell.attributes.get((OFFICENS, "value"))
+            if value == "":  # NA handling
+                return ""
+            cell_value = float(cell_value)
             if convert_float:
+                # print("convert", cell_value, int(cell_value))
                 val = int(cell_value)
                 if val == cell_value:
+                    # print("return the int")
                     return val
             return cell_value
         elif cell_type == "percentage":

pandas/io/excel/_odswriter.py

@@ -1,14 +1,16 @@
 from collections import defaultdict
+import datetime
 
-import pandas._libs.json as json
+# import pandas._libs.json as json
 
 from pandas.io.excel._base import ExcelWriter
-from pandas.io.excel._util import _validate_freeze_panes
+# from pandas.io.excel._util import _validate_freeze_panes
 
 from odf.opendocument import OpenDocumentSpreadsheet
 from odf.table import Table, TableRow, TableCell
 from odf.text import P
 
+
 class _ODSWriter(ExcelWriter):
     engine = "odf"
     supported_extensions = (".ods",)
@@ -24,8 +26,9 @@ def __init__(self, path, engine=None, encoding=None, mode="w", **engine_kwargs):
         if encoding is None:
             encoding = "ascii"
         self.book = OpenDocumentSpreadsheet()
-#        self.fm_datetime = xlwt.easyxf(num_format_str=self.datetime_format)
-#        self.fm_date = xlwt.easyxf(num_format_str=self.date_format)
+
+    #        self.fm_datetime = xlwt.easyxf(num_format_str=self.datetime_format)
+    #        self.fm_date = xlwt.easyxf(num_format_str=self.date_format)
 
     def save(self):
         """
@@ -46,31 +49,49 @@ def write_cells(
             wks = self.sheets[sheet_name]
         else:
             wks = Table(name=sheet_name)
-#            wks = self.book.add_sheet(sheet_name) (do at the end or immediately? FIXME)
+            #            wks = self.book.add_sheet(sheet_name) (do at the end or immediately? FIXME)
             self.sheets[sheet_name] = wks
 
-#        if _validate_freeze_panes(freeze_panes):
-#            wks.set_panes_frozen(True)
-#            wks.set_horz_split_pos(freeze_panes[0])
-#            wks.set_vert_split_pos(freeze_panes[1])
+        #        if _validate_freeze_panes(freeze_panes):
+        #            wks.set_panes_frozen(True)
+        #            wks.set_horz_split_pos(freeze_panes[0])
+        #            wks.set_vert_split_pos(freeze_panes[1])
 
         style_dict = {}
 
         rows = defaultdict(TableRow)
         col_count = defaultdict(int)
 
         for cell in cells:
-            print(cell.row, cell.col, cell.val)
+            # print(cell.row, cell.col, cell.val)
             # fill with empty cells if needed
             for _ in range(cell.col - col_count[cell.row]):
                 rows[cell.row].addElement(TableCell())
                 col_count[cell.row] += 1
-            class_to_cell_type = { str: "string", int: "float", float: "float", bool: "boolean" }
+            class_to_cell_type = {
+                str: "string",
+                int: "float",
+                float: "float",
+                bool: "boolean",
+            }
             val, fmt = self._value_with_fmt(cell.val)
-            tc = TableCell(valuetype=class_to_cell_type[type(val)], value=val)
+            # print("type", type(val), "value", val)
+            value = val
+            if isinstance(val, bool):
+                value = str(val).lower()
+                #            if isinstance(val, datetime.date):
+                #                tc = TableCell(valuetype="date",
+            if isinstance(val, datetime.date):
+                print('date', val.strftime("%Y-%m-%d"), val.strftime("%x"))
+                value = val.strftime("%Y-%m-%d")
+                tc = TableCell(valuetype="date", datevalue=value)
+            else:
+                tc = TableCell(valuetype=class_to_cell_type[type(val)], value=value)
             rows[cell.row].addElement(tc)
             col_count[cell.row] += 1
-            p = P(text=val)
+            if isinstance(val, bool):
+                value = str(val).upper()
+            p = P(text=value)
             tc.addElement(p)
             """
             stylekey = json.dumps(cell.style)

pandas/io/excel/_util.py

@@ -35,7 +35,12 @@ def _get_default_writer(ext):
     str
         The default engine for the extension.
     """
-    _default_writers = {"xlsx": "openpyxl", "xlsm": "openpyxl", "xls": "xlwt", "ods": "odf"}
+    _default_writers = {
+        "xlsx": "openpyxl",
+        "xlsm": "openpyxl",
+        "xls": "xlwt",
+        "ods": "odf",
+    }
     xlsxwriter = import_optional_dependency(
         "xlsxwriter", raise_on_missing=False, on_version="warn"
     )

pandas/tests/io/excel/test_writers.py

@@ -323,7 +323,7 @@ def test_excel_sheet_size(self, path):
         with pytest.raises(ValueError, match=msg):
             col_df.to_excel(path)
 
-    def test_excel_sheet_by_name_raise(self, path):
+    def test_excel_sheet_by_name_raise(self, path, engine):
         import xlrd
 
         gt = DataFrame(np.random.randn(10, 2))
@@ -334,9 +334,14 @@ def test_excel_sheet_by_name_raise(self, path):
 
         tm.assert_frame_equal(gt, df)
 
-        msg = "No sheet named <'0'>"
-        with pytest.raises(xlrd.XLRDError, match=msg):
-            pd.read_excel(xl, "0")
+        if engine == "odf":
+            msg = "sheet 0 not found"
+            with pytest.raises(ValueError, match=msg):
+                pd.read_excel(xl, "0")
+        else:
+            msg = "No sheet named <'0'>"
+            with pytest.raises(xlrd.XLRDError, match=msg):
+                pd.read_excel(xl, "0")
 
     def test_excel_writer_context_manager(self, frame, path):
         with ExcelWriter(path) as writer: