GH-98363: Add itertools.batched() (GH-98364)

Doc/library/itertools.rst

@@ -48,6 +48,7 @@ Iterator            Arguments               Results
 Iterator                        Arguments                       Results                                             Example
 ============================    ============================    =================================================   =============================================================
 :func:`accumulate`              p [,func]                       p0, p0+p1, p0+p1+p2, ...                            ``accumulate([1,2,3,4,5]) --> 1 3 6 10 15``
+:func:`batched`                 p, n                            [p0, p1, ..., p_n-1], ...                           ``batched('ABCDEFG', n=3) --> ABC DEF G``
 :func:`chain`                   p, q, ...                       p0, p1, ... plast, q0, q1, ...                      ``chain('ABC', 'DEF') --> A B C D E F``
 :func:`chain.from_iterable`     iterable                        p0, p1, ... plast, q0, q1, ...                      ``chain.from_iterable(['ABC', 'DEF']) --> A B C D E F``
 :func:`compress`                data, selectors                 (d[0] if s[0]), (d[1] if s[1]), ...                 ``compress('ABCDEF', [1,0,1,0,1,1]) --> A C E F``
@@ -170,6 +171,44 @@ loops that truncate the stream.
     .. versionchanged:: 3.8
        Added the optional *initial* parameter.
 
+
+.. function:: batched(iterable, n)
+
+   Batch data from the *iterable* into lists of length *n*. The last
+   batch may be shorter than *n*.
+
+   Loops over the input iterable and accumulates data into lists up to
+   size *n*.  The input is consumed lazily, just enough to fill a list.
+   The result is yielded as soon as the batch is full or when the input
+   iterable is exhausted:
+
+   .. doctest::
+
+      >>> flattened_data = ['roses', 'red', 'violets', 'blue', 'sugar', 'sweet']
+      >>> unflattened = list(batched(flattened_data, 2))
+      >>> unflattened
+      [['roses', 'red'], ['violets', 'blue'], ['sugar', 'sweet']]
+
+      >>> for batch in batched('ABCDEFG', 3):
+      ...     print(batch)
+      ...
+      ['A', 'B', 'C']
+      ['D', 'E', 'F']
+      ['G']
+
+   Roughly equivalent to::
+
+      def batched(iterable, n):
+          # batched('ABCDEFG', 3) --> ABC DEF G
+          if n < 1:
+              raise ValueError('n must be at least one')
+          it = iter(iterable)
+          while (batch := list(islice(it, n))):
+              yield batch
+
+    .. versionadded:: 3.12
+
+
 .. function:: chain(*iterables)
 
    Make an iterator that returns elements from the first iterable until it is
@@ -858,13 +897,6 @@ which incur interpreter overhead.
        else:
            raise ValueError('Expected fill, strict, or ignore')
 
-   def batched(iterable, n):
-       "Batch data into lists of length n. The last batch may be shorter."
-       # batched('ABCDEFG', 3) --> ABC DEF G
-       it = iter(iterable)
-       while (batch := list(islice(it, n))):
-           yield batch
-
    def triplewise(iterable):
        "Return overlapping triplets from an iterable"
        # triplewise('ABCDEFG') --> ABC BCD CDE DEF EFG
@@ -1211,36 +1243,6 @@ which incur interpreter overhead.
     >>> list(grouper('abcdefg', n=3, incomplete='ignore'))
     [('a', 'b', 'c'), ('d', 'e', 'f')]
 
-    >>> list(batched('ABCDEFG', 3))
-    [['A', 'B', 'C'], ['D', 'E', 'F'], ['G']]
-    >>> list(batched('ABCDEF', 3))
-    [['A', 'B', 'C'], ['D', 'E', 'F']]
-    >>> list(batched('ABCDE', 3))
-    [['A', 'B', 'C'], ['D', 'E']]
-    >>> list(batched('ABCD', 3))
-    [['A', 'B', 'C'], ['D']]
-    >>> list(batched('ABC', 3))
-    [['A', 'B', 'C']]
-    >>> list(batched('AB', 3))
-    [['A', 'B']]
-    >>> list(batched('A', 3))
-    [['A']]
-    >>> list(batched('', 3))
-    []
-    >>> list(batched('ABCDEFG', 2))
-    [['A', 'B'], ['C', 'D'], ['E', 'F'], ['G']]
-    >>> list(batched('ABCDEFG', 1))
-    [['A'], ['B'], ['C'], ['D'], ['E'], ['F'], ['G']]
-    >>> list(batched('ABCDEFG', 0))
-    []
-    >>> list(batched('ABCDEFG', -1))
-    Traceback (most recent call last):
-      ...
-    ValueError: Stop argument for islice() must be None or an integer: 0 <= x <= sys.maxsize.
-    >>> s = 'ABCDEFGHIJKLMNOPQRSTUVWXYZ'
-    >>> all(list(flatten(batched(s[:n], 5))) == list(s[:n]) for n in range(len(s)))
-    True
-
     >>> list(triplewise('ABCDEFG'))
     [('A', 'B', 'C'), ('B', 'C', 'D'), ('C', 'D', 'E'), ('D', 'E', 'F'), ('E', 'F', 'G')]
 

Lib/test/test_itertools.py

@@ -159,6 +159,44 @@ def test_accumulate(self):
         with self.assertRaises(TypeError):
             list(accumulate([10, 20], 100))
 
+    def test_batched(self):
+        self.assertEqual(list(batched('ABCDEFG', 3)),
+                             [['A', 'B', 'C'], ['D', 'E', 'F'], ['G']])
+        self.assertEqual(list(batched('ABCDEFG', 2)),
+                             [['A', 'B'], ['C', 'D'], ['E', 'F'], ['G']])
+        self.assertEqual(list(batched('ABCDEFG', 1)),
+                             [['A'], ['B'], ['C'], ['D'], ['E'], ['F'], ['G']])
+
+        with self.assertRaises(TypeError):          # Too few arguments
+            list(batched('ABCDEFG'))
+        with self.assertRaises(TypeError):
+            list(batched('ABCDEFG', 3, None))       # Too many arguments
+        with self.assertRaises(TypeError):
+            list(batched(None, 3))                  # Non-iterable input
+        with self.assertRaises(TypeError):
+            list(batched('ABCDEFG', 'hello'))       # n is a string
+        with self.assertRaises(ValueError):
+            list(batched('ABCDEFG', 0))             # n is zero
+        with self.assertRaises(ValueError):
+            list(batched('ABCDEFG', -1))            # n is negative
+
+        data = 'ABCDEFGHIJKLMNOPQRSTUVWXYZ'
+        for n in range(1, 6):
+            for i in range(len(data)):
+                s = data[:i]
+                batches = list(batched(s, n))
+                with self.subTest(s=s, n=n, batches=batches):
+                    # Order is preserved and no data is lost
+                    self.assertEqual(''.join(chain(*batches)), s)
+                    # Each batch is an exact list
+                    self.assertTrue(all(type(batch) is list for batch in batches))
+                    # All but the last batch is of size n
+                    if batches:
+                        last_batch = batches.pop()
+                        self.assertTrue(all(len(batch) == n for batch in batches))
+                        self.assertTrue(len(last_batch) <= n)
+                        batches.append(last_batch)
+
     def test_chain(self):
 
         def chain2(*iterables):
@@ -1737,6 +1775,31 @@ def test_takewhile(self):
 
 class TestPurePythonRoughEquivalents(unittest.TestCase):
 
+    def test_batched_recipe(self):
+        def batched_recipe(iterable, n):
+            "Batch data into lists of length n. The last batch may be shorter."
+            # batched('ABCDEFG', 3) --> ABC DEF G
+            if n < 1:
+                raise ValueError('n must be at least one')
+            it = iter(iterable)
+            while (batch := list(islice(it, n))):
+                yield batch
+
+        for iterable, n in product(
+                ['', 'a', 'ab', 'abc', 'abcd', 'abcde', 'abcdef', 'abcdefg', None],
+                [-1, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, None]):
+            with self.subTest(iterable=iterable, n=n):
+                try:
+                    e1, r1 = None, list(batched(iterable, n))
+                except Exception as e:
+                    e1, r1 = type(e), None
+                try:
+                    e2, r2 = None, list(batched_recipe(iterable, n))
+                except Exception as e:
+                    e2, r2 = type(e), None
+                self.assertEqual(r1, r2)
+                self.assertEqual(e1, e2)
+
     @staticmethod
     def islice(iterable, *args):
         s = slice(*args)
@@ -1788,6 +1851,10 @@ def test_accumulate(self):
         a = []
         self.makecycle(accumulate([1,2,a,3]), a)
 
+    def test_batched(self):
+        a = []
+        self.makecycle(batched([1,2,a,3], 2), a)
+
     def test_chain(self):
         a = []
         self.makecycle(chain(a), a)
@@ -1972,6 +2039,18 @@ def test_accumulate(self):
         self.assertRaises(TypeError, accumulate, N(s))
         self.assertRaises(ZeroDivisionError, list, accumulate(E(s)))
 
+    def test_batched(self):
+        s = 'abcde'
+        r = [['a', 'b'], ['c', 'd'], ['e']]
+        n = 2
+        for g in (G, I, Ig, L, R):
+            with self.subTest(g=g):
+                self.assertEqual(list(batched(g(s), n)), r)
+        self.assertEqual(list(batched(S(s), 2)), [])
+        self.assertRaises(TypeError, batched, X(s), 2)
+        self.assertRaises(TypeError, batched, N(s), 2)
+        self.assertRaises(ZeroDivisionError, list, batched(E(s), 2))
+
     def test_chain(self):
         for s in ("123", "", range(1000), ('do', 1.2), range(2000,2200,5)):
             for g in (G, I, Ig, S, L, R):

Misc/NEWS.d/next/Library/2022-10-17-12-49-02.gh-issue-98363.aFmSP-.rst

@@ -0,0 +1,2 @@
+Added itertools.batched() to batch data into lists of a given length with
+the last list possibly being shorter than the others.