[3.10] gh-98820: Fix quadratic time in csv.Sniffer for quoted fields (GH-154867) (#155539)

Author: serhiy-storchaka

Lib/csv.py

@@ -215,12 +215,18 @@ def _guess_quote_and_delimiter(self, data, delimiters):
         this way.
         """
 
+        # The body of a quoted field ends at the first quote which is
+        # not doubled, as it does for a reader.  A lazy ".*?" scans to
+        # the end of the sample instead, from every start: quadratically.
+        # As an unrolled loop it is unambiguous, so it does not backtrack.
+        other = r'(?:(?!(?P=quote)).)*'
+        body = r'%s(?:(?P=quote){2}%s)*' % (other, other)
         matches = []
-        for restr in (r'(?P<delim>[^\w\n"\'])(?P<space> ?)(?P<quote>["\']).*?(?P=quote)(?P=delim)', # ,".*?",
-                      r'(?:^|\n)(?P<quote>["\']).*?(?P=quote)(?P<delim>[^\w\n"\'])(?P<space> ?)',   #  ".*?",
-                      r'(?P<delim>[^\w\n"\'])(?P<space> ?)(?P<quote>["\']).*?(?P=quote)(?:$|\n)',   # ,".*?"
-                      r'(?:^|\n)(?P<quote>["\']).*?(?P=quote)(?:$|\n)'):                            #  ".*?" (no delim, no space)
-            regexp = re.compile(restr, re.DOTALL | re.MULTILINE)
+        for restr in (r'(?P<delim>[^\w\n"\'])(?P<space> ?)(?P<quote>["\'])%s(?P=quote)(?P=delim)',   # ,"...",
+                      r'(?:^|\n)(?P<quote>["\'])%s(?P=quote)(?P<delim>[^\w\n"\'])(?P<space> ?)',     #  "...",
+                      r'(?P<delim>[^\w\n"\'])(?P<space> ?)(?P<quote>["\'])%s(?P=quote)(?:$|\n)',  # ,"..."
+                      r'(?:^|\n)(?P<quote>["\'])%s(?P=quote)(?:$|\n)'):                           #  "..." (no delim, no space)
+            regexp = re.compile(restr % body, re.DOTALL | re.MULTILINE)
             matches = regexp.findall(data)
             if matches:
                 break

Lib/test/test_csv.py

@@ -1232,6 +1232,13 @@ def test_sniff_space_delimiter(self):
         self.assertEqual(dialect.delimiter, ' ')
         self.assertIs(dialect.doublequote, False)
 
+    def test_sniff_quoted_single_column(self):
+        # gh-98820: this sample used to take minutes.
+        sniffer = csv.Sniffer()
+        sample = '"abcdefghijklmnopqrstuvwxyz"\n' * 10000
+        with self.assertRaisesRegex(csv.Error, "Could not determine delimiter"):
+            sniffer.sniff(sample, delimiters=',:|\t')
+
 
 class NUL:
     def write(s, *args):

Misc/NEWS.d/next/Library/2026-07-29-11-20-00.gh-issue-98820.Qm7Hs4.rst

@@ -0,0 +1,2 @@
+Fix quadratic time in :meth:`csv.Sniffer.sniff` for a sample which contains
+quoted fields, in particular for a single column of quoted fields.