feat: support column_width in xlsx format

AUTHORS

@@ -9,6 +9,7 @@ Here is a list of passed and present much-appreciated contributors:
     Bruno Soares
     Claude Paroz
     Daniel Santos
+    Egor Osokin
     Erik Youngren
     Hugo van Kemenade
     Iuri de Silvio

docs/formats.rst

@@ -234,6 +234,19 @@ The ``import_set()`` method also supports a ``skip_lines`` parameter that you
 can set to a number of lines that should be skipped before starting to read
 data.
 
+The ``export_set()`` method supports a ``column_width`` parameter. Depending on the
+value you pass, the column width will be set accordingly. It can be either ``None``, an integer, or "adaptive".
+If "adaptive" is passed, the column width will be unique for every column and will be 
+calculated based on values' length. For example:: 
+
+    data = tablib.Dataset()
+    data.export('xlsx', column_width='adaptive')
+
+
+
+.. versionchanged:: 3.3.0
+    The ``column_width`` parameter for ``export_set()`` was added.
+
 .. versionchanged:: 3.1.0
 
     The ``skip_lines`` parameter for ``import_set()`` was added.

src/tablib/formats/_xlsx.py

@@ -1,5 +1,6 @@
 """ Tablib - XLSX Support.
 """
+from __future__ import annotations
 
 import re
 from io import BytesIO
@@ -35,12 +36,17 @@ def detect(cls, stream):
             return False
 
     @classmethod
-    def export_set(cls, dataset, freeze_panes=True, invalid_char_subst="-"):
+    def export_set(
+        cls, dataset, freeze_panes=True, invalid_char_subst="-",
+        column_width: str | int | None = "adaptive"
+    ):
         """Returns XLSX representation of Dataset.
 
         If dataset.title contains characters which are considered invalid for an XLSX file
         sheet name (http://www.excelcodex.com/2012/06/worksheets-naming-conventions/), they will
         be replaced with `invalid_char_subst`.
+
+        column_width: can be None, an integer, or "adaptive"
         """
         wb = Workbook()
         ws = wb.worksheets[0]
@@ -51,6 +57,11 @@ def export_set(cls, dataset, freeze_panes=True, invalid_char_subst="-"):
         )
 
         cls.dset_sheet(dataset, ws, freeze_panes=freeze_panes)
+        if isinstance(column_width, str) and column_width != "adaptive":
+            raise ValueError(f"Unsupported value `{column_width}` passed to `column_width` "
+                             "parameter. It supports 'adaptive' or integer values")
+
+        cls._adapt_column_width(ws, column_width)
 
         stream = BytesIO()
         wb.save(stream)
@@ -166,3 +177,25 @@ def dset_sheet(cls, dataset, ws, freeze_panes=True):
                     cell.value = col
                 except (ValueError, TypeError):
                     cell.value = str(col)
+
+    @classmethod
+    def _adapt_column_width(cls, worksheet,
+                            width: str | int | None) -> None:
+        if width is None:
+            return
+
+        column_widths = []
+        if isinstance(width, str) and width == "adaptive":
+            for row in worksheet.values:
+                for i, cell in enumerate(row):
+                    cell = str(cell)
+                    if len(column_widths) > i:
+                        if len(cell) > column_widths[i]:
+                            column_widths[i] = len(cell)
+                    else:
+                        column_widths += [len(cell)]
+        else:
+            column_widths = [width] * worksheet.max_column
+
+        for i, column_width in enumerate(column_widths, 1): # start at 1
+            worksheet.column_dimensions[get_column_letter(i)].width = column_width

tests/test_tablib.py

@@ -9,9 +9,11 @@
 from collections import OrderedDict
 from io import BytesIO, StringIO
 from pathlib import Path
+from tempfile import TemporaryFile
 from uuid import uuid4
 
 from MarkupPy import markup
+from openpyxl import load_workbook
 
 import tablib
 from tablib.core import Row, detect_format
@@ -1100,6 +1102,47 @@ def test_xlsx_bad_dimensions(self):
             data = tablib.Dataset().load(fh, read_only=False)
         self.assertEqual(data.height, 3)
 
+    def _helper_export_column_width(self, input_arg):
+        """check that column width adapts to value length"""
+        def _get_width(data, input_arg):
+            xlsx_content = data.export('xlsx', column_width=input_arg)
+            wb = load_workbook(filename=BytesIO(xlsx_content))
+            ws = wb.active
+            return ws.column_dimensions['A'].width
+
+        xls_source = Path(__file__).parent / 'files' / 'xlsx_cell_values.xlsx'
+        with xls_source.open('rb') as fh:
+            data = tablib.Dataset().load(fh)
+        width_before = _get_width(data, input_arg)
+        data.append([
+            'verylongvalue-verylongvalue-verylongvalue-verylongvalue-verylongvalue-verylongvalue-verylongvalue-verylongvalue',
+        ])
+        width_after = _get_width(data, width_before)
+        return width_before, width_after
+
+    def test_xlsx_column_width_none(self):
+        """check column width with None"""
+        width_before, width_after = self._helper_export_column_width(None)
+        self.assertEqual(width_before, 13)
+        self.assertEqual(width_after, 13)
+
+    def test_xlsx_column_width_adaptive(self):
+        """check column width with 'adaptive'"""
+        width_before, width_after = self._helper_export_column_width("adaptive")
+        self.assertEqual(width_before, 11)
+        self.assertEqual(width_after, 11)
+
+    def test_xlsx_column_width_integer(self):
+        """check column width with an integer"""
+        width_before, width_after = self._helper_export_column_width(10)
+        self.assertEqual(width_before, 10)
+        self.assertEqual(width_after, 10)
+
+    def test_xlsx_column_width_value_error(self):
+        """check column width with invalid input"""
+        with self.assertRaises(ValueError):
+            self._helper_export_column_width("invalid input")
+
 
 class JSONTests(BaseTestCase):
     def test_json_format_detect(self):