ERR: Clarify location of EOF on unbalanced quotes (pandas-dev#22814)

Closes pandas-devgh-22789.

Author: gfyoung

pandas/_libs/src/parser/tokenizer.c

@@ -1150,7 +1150,7 @@ static int parser_handle_eof(parser_t *self) {
         case IN_QUOTED_FIELD:
             self->error_msg = (char *)malloc(bufsize);
             snprintf(self->error_msg, bufsize,
-                    "EOF inside string starting at line %lld",
+                    "EOF inside string starting at row %lld",
                     (long long)self->file_lines);
             return -1;
 

pandas/io/parsers.py

@@ -2727,9 +2727,6 @@ def _next_iter_line(self, row_num):
                            'cannot be processed in Python\'s '
                            'native csv library at the moment, '
                            'so please pass in engine=\'c\' instead')
-                elif 'newline inside string' in msg:
-                    msg = ('EOF inside string starting with '
-                           'line ' + str(row_num))
 
                 if self.skipfooter > 0:
                     reason = ('Error could possibly be due to '

pandas/tests/io/parser/common.py

@@ -198,20 +198,6 @@ def test_malformed(self):
                                 header=1, comment='#',
                                 skipfooter=1)
 
-    def test_quoting(self):
-        bad_line_small = """printer\tresult\tvariant_name
-Klosterdruckerei\tKlosterdruckerei <Salem> (1611-1804)\tMuller, Jacob
-Klosterdruckerei\tKlosterdruckerei <Salem> (1611-1804)\tMuller, Jakob
-Klosterdruckerei\tKlosterdruckerei <Kempten> (1609-1805)\t"Furststiftische Hofdruckerei,  <Kempten""
-Klosterdruckerei\tKlosterdruckerei <Kempten> (1609-1805)\tGaller, Alois
-Klosterdruckerei\tKlosterdruckerei <Kempten> (1609-1805)\tHochfurstliche Buchhandlung <Kempten>"""  # noqa
-        pytest.raises(Exception, self.read_table, StringIO(bad_line_small),
-                      sep='\t')
-
-        good_line_small = bad_line_small + '"'
-        df = self.read_table(StringIO(good_line_small), sep='\t')
-        assert len(df) == 3
-
     def test_unnamed_columns(self):
         data = """A,B,C,,
 1,2,3,4,5

pandas/tests/io/parser/quoting.py

@@ -9,6 +9,7 @@
 import pandas.util.testing as tm
 
 from pandas import DataFrame
+from pandas.errors import ParserError
 from pandas.compat import PY3, StringIO, u
 
 
@@ -151,3 +152,19 @@ def test_quotechar_unicode(self):
         if PY3:
             result = self.read_csv(StringIO(data), quotechar=u('\u0001'))
             tm.assert_frame_equal(result, expected)
+
+    def test_unbalanced_quoting(self):
+        # see gh-22789.
+        data = "a,b,c\n1,2,\"3"
+
+        if self.engine == "c":
+            regex = "EOF inside string starting at row 1"
+        else:
+            regex = "unexpected end of data"
+
+        with tm.assert_raises_regex(ParserError, regex):
+            self.read_csv(StringIO(data))
+
+        expected = DataFrame([[1, 2, 3]], columns=["a", "b", "c"])
+        data = self.read_csv(StringIO(data + '"'))
+        tm.assert_frame_equal(data, expected)