MAINT: Clarify supported Stata dta versions

Test against old Stata versions and remove text indicating support
for versions which do not work reliably

closes pandas-dev#26667

Author: bashtage

pandas/io/stata.py

@@ -51,9 +51,9 @@
 
 _version_error = (
     "Version of given Stata file is {version}. pandas supports importing "
-    "versions 104, 105, 108, 111 (Stata 7SE), 113 (Stata 8/9), "
-    "114 (Stata 10/11), 115 (Stata 12), 117 (Stata 13), 118 (Stata 14/15/16),"
-    "and 119 (Stata 15/16, over 32,767 variables)."
+    "versions 111 (Stata 7SE), 113 (Stata 8/9), 114 (Stata 10/11), "
+    "115 (Stata 12), 117 (Stata 13), 118 (Stata 14/15/16), and 119 "
+    "(Stata 15/16, over 32,767 variables)."
 )
 
 _statafile_processing_params1 = """\
@@ -1245,7 +1245,7 @@ def _get_time_stamp(self) -> str:
         elif self.format_version == 117:
             strlen = struct.unpack("b", self.path_or_buf.read(1))[0]
             return self._decode(self.path_or_buf.read(strlen))
-        elif self.format_version > 104:
+        elif self.format_version >= 111:
             return self._decode(self.path_or_buf.read(18))
         else:
             raise ValueError()
@@ -1264,7 +1264,7 @@ def _get_seek_variable_labels(self) -> int:
 
     def _read_old_header(self, first_char: bytes) -> None:
         self.format_version = struct.unpack("b", first_char)[0]
-        if self.format_version not in [104, 105, 108, 111, 113, 114, 115]:
+        if self.format_version not in [111, 113, 114, 115]:
             raise ValueError(_version_error.format(version=self.format_version))
         self._set_encoding()
         self.byteorder = (
@@ -1328,7 +1328,7 @@ def _read_old_header(self, first_char: bytes) -> None:
         # the size of the next read, which you discard.  You then continue
         # like this until you read 5 bytes of zeros.
 
-        if self.format_version > 104:
+        if self.format_version >= 111:
             while True:
                 data_type = struct.unpack(
                     self.byteorder + "b", self.path_or_buf.read(1)

pandas/tests/io/data/stata/stata-compat-111.dta

@@ -1853,3 +1853,13 @@ def test_writer_118_exceptions(self):
         with tm.ensure_clean() as path:
             with pytest.raises(ValueError, match="You must use version 119"):
                 StataWriterUTF8(path, df, version=118)
+
+
+@pytest.mark.parametrize("version", [111, 113, 114])
+def test_backward_compat(version, datapath):
+    data_base = datapath("io", "data", "stata")
+    ref = os.path.join(data_base, f"stata-compat-118.dta")
+    old = os.path.join(data_base, f"stata-compat-{version}.dta")
+    expected = pd.read_stata(ref)
+    old_dta = pd.read_stata(old)
+    tm.assert_frame_equal(old_dta, expected, check_dtype=False)