Backport PR pandas-dev#39482: ERR: Unify error message for bad excel sheetnames (pandas-dev#39536)

Co-authored-by: Richard Shadrach <[email protected]>

Author: simonjayhawkins

doc/source/whatsnew/v1.2.2.rst

@@ -29,7 +29,7 @@ Fixed regressions
 Bug fixes
 ~~~~~~~~~
 
--
+- :func:`pandas.read_excel` error message when a specified ``sheetname`` does not exist is now uniform across engines (:issue:`39250`)
 -
 
 .. ---------------------------------------------------------------------------

pandas/io/excel/_base.py

@@ -425,6 +425,17 @@ def get_sheet_by_index(self, index):
     def get_sheet_data(self, sheet, convert_float):
         pass
 
+    def raise_if_bad_sheet_by_index(self, index: int) -> None:
+        n_sheets = len(self.sheet_names)
+        if index >= n_sheets:
+            raise ValueError(
+                f"Worksheet index {index} is invalid, {n_sheets} worksheets found"
+            )
+
+    def raise_if_bad_sheet_by_name(self, name: str) -> None:
+        if name not in self.sheet_names:
+            raise ValueError(f"Worksheet named '{name}' not found")
+
     def parse(
         self,
         sheet_name=0,

pandas/io/excel/_odfreader.py

@@ -57,12 +57,14 @@ def sheet_names(self) -> List[str]:
     def get_sheet_by_index(self, index: int):
         from odf.table import Table
 
+        self.raise_if_bad_sheet_by_index(index)
         tables = self.book.getElementsByType(Table)
         return tables[index]
 
     def get_sheet_by_name(self, name: str):
         from odf.table import Table
 
+        self.raise_if_bad_sheet_by_name(name)
         tables = self.book.getElementsByType(Table)
 
         for table in tables:

pandas/io/excel/_openpyxl.py

@@ -492,9 +492,11 @@ def sheet_names(self) -> List[str]:
         return self.book.sheetnames
 
     def get_sheet_by_name(self, name: str):
+        self.raise_if_bad_sheet_by_name(name)
         return self.book[name]
 
     def get_sheet_by_index(self, index: int):
+        self.raise_if_bad_sheet_by_index(index)
         return self.book.worksheets[index]
 
     def _convert_cell(self, cell, convert_float: bool) -> Scalar:

pandas/io/excel/_pyxlsb.py

@@ -47,9 +47,11 @@ def sheet_names(self) -> List[str]:
         return self.book.sheets
 
     def get_sheet_by_name(self, name: str):
+        self.raise_if_bad_sheet_by_name(name)
         return self.book.get_sheet(name)
 
     def get_sheet_by_index(self, index: int):
+        self.raise_if_bad_sheet_by_index(index)
         # pyxlsb sheets are indexed from 1 onwards
         # There's a fix for this in the source, but the pypi package doesn't have it
         return self.book.get_sheet(index + 1)

pandas/io/excel/_xlrd.py

@@ -44,9 +44,11 @@ def sheet_names(self):
         return self.book.sheet_names()
 
     def get_sheet_by_name(self, name):
+        self.raise_if_bad_sheet_by_name(name)
         return self.book.sheet_by_name(name)
 
     def get_sheet_by_index(self, index):
+        self.raise_if_bad_sheet_by_index(index)
         return self.book.sheet_by_index(index)
 
     def get_sheet_data(self, sheet, convert_float):

pandas/tests/io/excel/test_odf.py

@@ -42,5 +42,5 @@ def test_nonexistent_sheetname_raises(read_ext):
     # GH-27676
     # Specifying a non-existent sheet_name parameter should throw an error
     # with the sheet name.
-    with pytest.raises(ValueError, match="sheet xyz not found"):
+    with pytest.raises(ValueError, match="Worksheet named 'xyz' not found"):
         pd.read_excel("blank.ods", sheet_name="xyz")

pandas/tests/io/excel/test_readers.py

@@ -622,6 +622,16 @@ def test_bad_engine_raises(self, read_ext):
         with pytest.raises(ValueError, match="Unknown engine: foo"):
             pd.read_excel("", engine=bad_engine)
 
+    @pytest.mark.parametrize(
+        "sheet_name",
+        [3, [0, 3], [3, 0], "Sheet4", ["Sheet1", "Sheet4"], ["Sheet4", "Sheet1"]],
+    )
+    def test_bad_sheetname_raises(self, read_ext, sheet_name):
+        # GH 39250
+        msg = "Worksheet index 3 is invalid|Worksheet named 'Sheet4' not found"
+        with pytest.raises(ValueError, match=msg):
+            pd.read_excel("blank" + read_ext, sheet_name=sheet_name)
+
     def test_missing_file_raises(self, read_ext):
         bad_file = f"foo{read_ext}"
         # CI tests with zh_CN.utf8, translates to "No such file or directory"
@@ -1159,6 +1169,17 @@ def test_sheet_name(self, read_ext, df_ref):
         tm.assert_frame_equal(df1_parse, df_ref, check_names=False)
         tm.assert_frame_equal(df2_parse, df_ref, check_names=False)
 
+    @pytest.mark.parametrize(
+        "sheet_name",
+        [3, [0, 3], [3, 0], "Sheet4", ["Sheet1", "Sheet4"], ["Sheet4", "Sheet1"]],
+    )
+    def test_bad_sheetname_raises(self, read_ext, sheet_name):
+        # GH 39250
+        msg = "Worksheet index 3 is invalid|Worksheet named 'Sheet4' not found"
+        with pytest.raises(ValueError, match=msg):
+            with pd.ExcelFile("blank" + read_ext) as excel:
+                excel.parse(sheet_name=sheet_name)
+
     def test_excel_read_buffer(self, engine, read_ext):
         pth = "test1" + read_ext
         expected = pd.read_excel(pth, sheet_name="Sheet1", index_col=0, engine=engine)

pandas/tests/io/excel/test_writers.py

@@ -347,19 +347,9 @@ def test_excel_sheet_by_name_raise(self, path, engine):
 
         tm.assert_frame_equal(gt, df)
 
-        if engine == "odf":
-            msg = "sheet 0 not found"
-            with pytest.raises(ValueError, match=msg):
-                pd.read_excel(xl, "0")
-        elif engine == "xlwt":
-            import xlrd
-
-            msg = "No sheet named <'0'>"
-            with pytest.raises(xlrd.XLRDError, match=msg):
-                pd.read_excel(xl, sheet_name="0")
-        else:
-            with pytest.raises(KeyError, match="Worksheet 0 does not exist."):
-                pd.read_excel(xl, sheet_name="0")
+        msg = "Worksheet named '0' not found"
+        with pytest.raises(ValueError, match=msg):
+            pd.read_excel(xl, "0")
 
     def test_excel_writer_context_manager(self, frame, path):
         with ExcelWriter(path) as writer:

pandas/tests/io/excel/test_xlrd.py

@@ -43,9 +43,10 @@ def test_read_xlrd_book(read_ext, frame):
 # TODO: test for openpyxl as well
 def test_excel_table_sheet_by_index(datapath, read_ext):
     path = datapath("io", "data", "excel", f"test1{read_ext}")
+    msg = "Worksheet named 'invalid_sheet_name' not found"
     with ExcelFile(path, engine="xlrd") as excel:
-        with pytest.raises(xlrd.XLRDError):
-            pd.read_excel(excel, sheet_name="asdf")
+        with pytest.raises(ValueError, match=msg):
+            pd.read_excel(excel, sheet_name="invalid_sheet_name")
 
 
 def test_excel_file_warning_with_xlsx_file(datapath):