Temporarily fix openpyxl read_worksheets issue

Orange/__init__.py

@@ -1,5 +1,23 @@
 # This module is a mixture of imports and code, so we allow import anywhere
 # pylint: disable=wrong-import-position,wrong-import-order
+import warnings
+
+from openpyxl.cell import MergedCell
+from openpyxl.comments.comment_sheet import CommentSheet
+from openpyxl.drawing.spreadsheet_drawing import SpreadsheetDrawing
+from openpyxl.packaging.relationship import (
+    get_rels_path,
+    RelationshipList,
+    get_dependents,
+)
+from openpyxl.pivot.table import TableDefinition
+from openpyxl.reader.drawings import find_images
+from openpyxl.reader.excel import ExcelReader
+from openpyxl.worksheet._read_only import ReadOnlyWorksheet
+from openpyxl.worksheet._reader import WorksheetReader
+from openpyxl.xml.constants import COMMENTS_NS
+from openpyxl.xml.functions import fromstring
+from openpyxl.worksheet.table import Table
 
 from Orange import data
 
@@ -46,3 +64,90 @@
     pass
 finally:
     del ctypes
+
+
+# temporary fix for file not closed issue until openpyxl prepare release
+# https://foss.heptapod.net/openpyxl/openpyxl/-/merge_requests/
+# 436#7922bd5f66e11e4ca4539f093b2680a25c1f80db
+def read_worksheets(self):
+    # pylint: too-many-branches
+    comment_warning = (
+        "Cell '{0}':{1} is part of a merged range but has a comment which will "
+        "be removed because merged cells cannot contain any data."
+    )
+    for sheet, rel in self.parser.find_sheets():
+        if rel.target not in self.valid_files:
+            continue
+
+        if "chartsheet" in rel.Type:
+            self.read_chartsheet(sheet, rel)
+            continue
+
+        rels_path = get_rels_path(rel.target)
+        rels = RelationshipList()
+        if rels_path in self.valid_files:
+            rels = get_dependents(self.archive, rels_path)
+
+        if self.read_only:
+            ws = ReadOnlyWorksheet(self.wb, sheet.name, rel.target, self.shared_strings)
+            ws.sheet_state = sheet.state
+            self.wb._sheets.append(ws)  # pylint: disable=protected-access
+            continue  # pylint: disable=no-else-continue
+        else:
+            fh = self.archive.open(rel.target)
+            ws = self.wb.create_sheet(sheet.name)
+            ws._rels = rels  # pylint: disable=protected-access
+            ws_parser = WorksheetReader(
+                ws, fh, self.shared_strings, self.data_only, self.rich_text
+            )
+            ws_parser.bind_all()
+            fh.close()
+
+        # assign any comments to cells
+        for r in rels.find(COMMENTS_NS):
+            src = self.archive.read(r.target)
+            comment_sheet = CommentSheet.from_tree(fromstring(src))
+            for ref, comment in comment_sheet.comments:
+                try:
+                    ws[ref].comment = comment
+                except AttributeError:
+                    c = ws[ref]
+                    if isinstance(c, MergedCell):
+                        warnings.warn(comment_warning.format(ws.title, c.coordinate))
+                        continue
+
+        # preserve link to VML file if VBA
+        if self.wb.vba_archive and ws.legacy_drawing:
+            ws.legacy_drawing = rels.get(ws.legacy_drawing).target
+        else:
+            ws.legacy_drawing = None
+
+        for t in ws_parser.tables:
+            src = self.archive.read(t)
+            xml = fromstring(src)
+            table = Table.from_tree(xml)
+            ws.add_table(table)
+
+        #pylint: disable=protected-access
+        drawings = rels.find(SpreadsheetDrawing._rel_type)
+        for rel in drawings:
+            charts, images = find_images(self.archive, rel.target)
+            for c in charts:
+                ws.add_chart(c, c.anchor)
+            for im in images:
+                ws.add_image(im, im.anchor)
+
+        pivot_rel = rels.find(TableDefinition.rel_type)
+        pivot_caches = self.parser.pivot_caches
+        for r in pivot_rel:
+            pivot_path = r.Target
+            src = self.archive.read(pivot_path)
+            tree = fromstring(src)
+            pivot = TableDefinition.from_tree(tree)
+            pivot.cache = pivot_caches[pivot.cacheId]
+            ws.add_pivot(pivot)
+
+        ws.sheet_state = sheet.state
+
+
+ExcelReader.read_worksheets = read_worksheets

Orange/tests/test_base.py

@@ -4,6 +4,8 @@
 import unittest
 from distutils.version import LooseVersion
 
+import openpyxl
+
 import Orange
 
 from Orange.base import SklLearner, Learner, Model
@@ -155,5 +157,15 @@ def test_pickle(self):
         self.assertEqual(model2.original_data, None)
 
 
+class TestRemoveTemporarySolution(unittest.TestCase):
+    def test_remove_openpyxl_temp_solution(self):
+        """
+        When this test starts to fail revert https://github.com/biolab/orange3/pull/6737
+        """
+        self.assertLessEqual(
+            [int(x) for x in openpyxl.__version__.split(".")], [3, 1, 2]
+        )
+
+
 if __name__ == "__main__":
     unittest.main()