BUG: pandas-dev#7757 Fix CSV parsing of singleton list header

Write test for header with len 1 that fails Fix header list manipulation causing problems with singleton lists
threecgreen · Jul 27, 2017 · 4648a7c · 4648a7c
1 parent f9a552d
commit 4648a7c
Show file tree

Hide file tree

Showing 3 changed files with 32 additions and 14 deletions.
diff --git a/pandas/_libs/parsers.pyx b/pandas/_libs/parsers.pyx
@@ -535,18 +535,25 @@ cdef class TextReader:
             self.parser_start = 0
             self.header = []
         else:
-            if isinstance(header, list) and len(header):
-                # need to artifically skip the final line
-                # which is still a header line
-                header = list(header)
-                header.append(header[-1] + 1)
-
-                self.parser.header_start = header[0]
-                self.parser.header_end = header[-1]
-                self.parser.header = header[0]
-                self.parser_start = header[-1] + 1
-                self.has_mi_columns = 1
-                self.header = header
+            if isinstance(header, list):
+                if len(header) > 1:
+                    # need to artifically skip the final line
+                    # which is still a header line
+                    header = list(header)
+                    header.append(header[-1] + 1)
+
+                    self.parser.header_start = header[0]
+                    self.parser.header_end = header[-1]
+                    self.parser.header = header[0]
+                    self.parser_start = header[-1] + 1
+                    self.has_mi_columns = 1
+                    self.header = header
+                else:
+                    self.parser.header_start = header[0]
+                    self.parser.header_end = header[0]
+                    self.parser.header = header[0]
+                    self.parser_start = header[0] + 1
+                    self.header = header
             else:
                 self.parser.header_start = header
                 self.parser.header_end = header

diff --git a/pandas/io/parsers.py b/pandas/io/parsers.py
@@ -2281,8 +2281,12 @@ def _infer_columns(self):
 
             # we have a mi columns, so read an extra line
             if isinstance(header, (list, tuple, np.ndarray)):
-                have_mi_columns = True
-                header = list(header) + [header[-1] + 1]
+                if len(header) > 1:
+                    have_mi_columns = True
+                    header = list(header) + [header[-1] + 1]
+                # Don't add second element to header if it's of length 1
+                else:
+                    have_mi_columns = False
             else:
                 have_mi_columns = False
                 header = [header]

diff --git a/pandas/tests/io/parser/header.py b/pandas/tests/io/parser/header.py
@@ -286,3 +286,10 @@ def test_non_int_header(self):
             self.read_csv(StringIO(data), sep=',', header=['a', 'b'])
         with tm.assert_raises_regex(ValueError, msg):
             self.read_csv(StringIO(data), sep=',', header='string_header')
+
+    def test_singleton_header(self):
+        # See GH #7757
+        data = """a,b,c\n0,1,2\n1,2,3"""
+        df = self.read_csv(StringIO(data), header=[0])
+        expected = pd.DataFrame({'a': [0, 1], 'b': [1, 2], 'c': [2, 3]})
+        tm.assert_frame_equal(df, expected)