Buffer input files. wireservice/csvkit#707.

CHANGELOG.rst

@@ -1,7 +1,7 @@
 1.5.3
 -----
 
-
+* :meth:`.Table.from_csv` now buffers input files to prevent issues with using STDIN as an input.
 
 1.5.2 - December 24, 2016
 -------------------------

agate/table/from_csv.py

@@ -62,19 +62,17 @@ def from_csv(cls, path, column_names=None, column_types=None, row_names=None, sk
     else:
         raise ValueError('skip_lines argument must be an int')
 
-    start = f.tell()
+    contents = six.StringIO(f.read())
 
     if sniff_limit is None:
-        kwargs['dialect'] = csv.Sniffer().sniff(f.read())
+        kwargs['dialect'] = csv.Sniffer().sniff(contents.getvalue())
     elif sniff_limit > 0:
-        kwargs['dialect'] = csv.Sniffer().sniff(f.read(sniff_limit))
+        kwargs['dialect'] = csv.Sniffer().sniff(contents.getvalue()[:sniff_limit])
 
     if six.PY2:
         kwargs['encoding'] = encoding
 
-    f.seek(start)
-
-    reader = csv.reader(f, header=header, **kwargs)
+    reader = csv.reader(contents, header=header, **kwargs)
 
     if header:
         if column_names is None:

tests/test_table/test_from_csv.py

@@ -98,7 +98,23 @@ def test_from_csv_no_header_columns(self):
         self.assertColumnNames(table, self.column_names)
         self.assertColumnTypes(table, [Number, Text, Boolean, Date, DateTime, TimeDelta])
 
-    def test_from_csv_sniff_limit(self):
+    def test_from_csv_sniff_limit_0(self):
+        table1 = Table(self.rows, self.column_names, self.column_types)
+        table2 = Table.from_csv('examples/test_csv_sniff.csv', sniff_limit=0)
+
+        self.assertColumnNames(table2, ['number|text|boolean|date|datetime|timedelta'])
+        self.assertColumnTypes(table2, [Text])
+
+    def test_from_csv_sniff_limit_200(self):
+        table1 = Table(self.rows, self.column_names, self.column_types)
+        table2 = Table.from_csv('examples/test_csv_sniff.csv', sniff_limit=200)
+
+        self.assertColumnNames(table2, table1.column_names)
+        self.assertColumnTypes(table2, [Number, Text, Boolean, Date, DateTime, TimeDelta])
+
+        self.assertRows(table2, table1.rows)
+
+    def test_from_csv_sniff_limit_none(self):
         table1 = Table(self.rows, self.column_names, self.column_types)
         table2 = Table.from_csv('examples/test_csv_sniff.csv', sniff_limit=None)