Patch 520694: arraymodule.c improvements:

- make array.array a type
- add Py_UNICODE arrays
- support +=, *=

Doc/lib/libarray.tex

@@ -5,7 +5,7 @@
 \modulesynopsis{Efficient arrays of uniformly typed numeric values.}
 
 
-This module defines a new object type which can efficiently represent
+This module defines an object type which can efficiently represent
 an array of basic values: characters, integers, floating point
 numbers.  Arrays\index{arrays} are sequence types and behave very much
 like lists, except that the type of objects stored in them is
@@ -17,6 +17,7 @@ codes are defined:
 \lineiii{'c'}{character}{1}
 \lineiii{'b'}{signed int}{1}
 \lineiii{'B'}{unsigned int}{1}
+\lineiii{'u'}{Unicode character}{2}
 \lineiii{'h'}{signed int}{2}
 \lineiii{'H'}{unsigned int}{2}
 \lineiii{'i'}{signed int}{2}
@@ -35,23 +36,25 @@ Python long integers when retrieved, because Python's plain integer
 type cannot represent the full range of C's unsigned (long) integers.
 
 
-The module defines the following function and type object:
+The module defines the following type:
 
 \begin{funcdesc}{array}{typecode\optional{, initializer}}
-Return a new array whose items are restricted by \var{typecode}, and
-initialized from the optional \var{initializer} value, which must be a
-list or a string.  The list or string is passed to the new array's
-\method{fromlist()} or \method{fromstring()} method (see below) to add
-initial items to the array.
+Return a new array whose items are restricted by \var{typecode},
+and initialized from the optional \var{initializer} value, which
+must be a list or a string.  The list or string is passed to the
+new array's \method{fromlist()}, \method{fromstring()}, or
+\method{fromunicode()} method (see below) to add initial items to
+the array.
 \end{funcdesc}
 
 \begin{datadesc}{ArrayType}
-Type object corresponding to the objects returned by
-\function{array()}.
+Obsolete alias for \function{array}.
 \end{datadesc}
 
 
-Array objects support the following data items and methods:
+Array objects support the ordinary sequence operations of
+indexing, slicing, concatenation, and multiplication.  The
+following data items and methods are also supported:
 
 \begin{memberdesc}[array]{typecode}
 The typecode character used to create the array.
@@ -121,6 +124,13 @@ array of machine values (as if it had been read from a
 file using the \method{fromfile()} method).
 \end{methoddesc}
 
+\begin{methoddesc}[array]{fromunicode}{s}
+Extends this array with data from the given unicode string.
+The array must be a type 'u' array; otherwise a ValueError
+is raised.  Use \samp{array.fromstring(ustr.decode(enc))} to
+append Unicode data to an array of some other type.
+\end{methoddesc}
+
 \begin{methoddesc}[array]{index}{x}
 Return the smallest \var{i} such that \var{i} is the index of
 the first occurence of \var{x} in the array.
@@ -134,7 +144,7 @@ Insert a new item with value \var{x} in the array before position
 \begin{methoddesc}[array]{pop}{\optional{i}}
 Removes the item with the index \var{i} from the array and returns
 it. The optional argument defaults to \code{-1}, so that by default
-the last item is removed and returned. 
+the last item is removed and returned.
 \end{methoddesc}
 
 \begin{methoddesc}[array]{read}{f, n}
@@ -170,6 +180,13 @@ string representation (the same sequence of bytes that would
 be written to a file by the \method{tofile()} method.)
 \end{methoddesc}
 
+\begin{methoddesc}[array]{tounicode}{}
+Convert the array to a unicode string.  The array must be
+a type 'u' array; otherwise a ValueError is raised.  Use
+array.tostring().decode(enc) to obtain a unicode string
+from an array of some other type.
+\end{methoddesc}
+
 \begin{methoddesc}[array]{write}{f}
 \deprecated {1.5.1}
   {Use the \method{tofile()} method.}
@@ -188,6 +205,7 @@ imported using \code{from array import array}.  Examples:
 \begin{verbatim}
 array('l')
 array('c', 'hello world')
+array('u', u'hello \textbackslash u2641')
 array('l', [1, 2, 3, 4, 5])
 array('d', [1.0, 2.0, 3.14])
 \end{verbatim}

Lib/test/test_array.py

@@ -6,14 +6,87 @@ import array
 from test_support import verbose, TESTFN, unlink, TestFailed
 
 def main():
-
     testtype('c', 'c')
-
+    testtype('u', u'\u263a')
     for type in (['b', 'h', 'i', 'l', 'f', 'd']):
         testtype(type, 1)
-
+    testunicode()
+    testsubclassing()
     unlink(TESTFN)
 
+def testunicode():
+    try:
+        array.array('b', u'foo')
+    except TypeError:
+        pass
+    else:
+        raise TestFailed("creating a non-unicode array from "
+                         "a Unicode string should fail")
+
+    x = array.array('u', u'\xa0\xc2\u1234')
+    x.fromunicode(u' ')
+    x.fromunicode(u'')
+    x.fromunicode(u'')
+    x.fromunicode(u'\x11abc\xff\u1234')
+    s = x.tounicode()
+    if s != u'\xa0\xc2\u1234 \x11abc\xff\u1234':
+        raise TestFailed("fromunicode()/tounicode()")
+
+    s = u'\x00="\'a\\b\x80\xff\u0000\u0001\u1234'
+    a = array.array('u', s)
+    if verbose:
+        print "repr of type 'u' array:", repr(a)
+        print "              expected: array('u', %r)" % s
+
+def testsubclassing():
+    class EditableString(array.array):
+        def __new__(cls, s, *args, **kwargs):
+            return array.array.__new__(cls, 'c', s)
+
+        def __init__(self, s, color='blue'):
+            array.array.__init__(self, 'c', s)
+            self.color = color
+
+        def strip(self):
+            self[:] = array.array('c', self.tostring().strip())
+
+        def __repr__(self):
+            return 'EditableString(%r)' % self.tostring()
+
+    s = EditableString("\ttest\r\n")
+    s.strip()
+    if s.tostring() != 'test':
+        raise TestFailed, "subclassing array.array failed somewhere"
+    if s.color != 'blue':
+        raise TestFailed, "assigning attributes to instance of array subclass"
+    s.color = 'red'
+    if s.color != 'red':
+        raise TestFailed, "assigning attributes to instance of array subclass"
+    if s.__dict__.keys() != ['color']:
+        raise TestFailed, "array subclass __dict__"
+
+    class ExaggeratingArray(array.array):
+        __slots__ = ['offset']
+
+        def __new__(cls, typecode, data, offset):
+            return array.array.__new__(cls, typecode, data)
+
+        def __init__(self, typecode, data, offset):
+            self.offset = offset
+
+        def __getitem__(self, i):
+            return array.array.__getitem__(self, i) + self.offset
+
+    a = ExaggeratingArray('i', [3, 6, 7, 11], 4)
+    if a[0] != 7:
+        raise TestFailed, "array subclass overriding __getitem__"
+    try:
+        a.color = 'blue'
+    except AttributeError:
+        pass
+    else:
+        raise TestFailed, "array subclass __slots__ was ignored"
+
 
 def testoverflow(type, lowerLimit, upperLimit):
         # should not overflow assigning lower limit
@@ -54,7 +127,6 @@ def testoverflow(type, lowerLimit, upperLimit):
 
 
 def testtype(type, example):
-
     a = array.array(type)
     a.append(example)
     if verbose:
@@ -97,6 +169,33 @@ def testtype(type, example):
         print 'array of %s converted to a string: ' \
                % a.typecode, `a.tostring()`
 
+    # Try out inplace addition and multiplication
+    a = array.array(type, [example])
+    b = a
+    a += array.array(type, [example]*2)
+    if a is not b:
+        raise TestFailed, "array(%s) inplace addition" % `type`
+    if a != array.array(type, [example] * 3):
+        raise TestFailed, "array(%s) inplace addition" % `type`
+
+    a *= 5
+    if a is not b:
+        raise TestFailed, "array(%s) inplace multiplication" % `type`
+    if a != array.array(type, [example] * 15):
+        raise TestFailed, "array(%s) inplace multiplication" % `type`
+
+    a *= 0
+    if a is not b:
+        raise TestFailed, "array(%s) inplace multiplication by 0" % `type`
+    if a != array.array(type, []):
+        raise TestFailed, "array(%s) inplace multiplication by 0" % `type`
+
+    a *= 1000
+    if a is not b:
+        raise TestFailed, "empty array(%s) inplace multiplication" % `type`
+    if a != array.array(type, []):
+        raise TestFailed, "empty array(%s) inplace multiplication" % `type`
+
     if type == 'c':
         a = array.array(type, "abcde")
         a[:-1] = a
@@ -135,6 +234,44 @@ def testtype(type, example):
         a.reverse()
         if a != array.array(type, "dca"):
             raise TestFailed, "array(%s) reverse-test" % `type`
+    elif type == 'u':
+        a = array.array(type, u"abcde")
+        a[:-1] = a
+        if a != array.array(type, u"abcdee"):
+            raise TestFailed, "array(%s) self-slice-assign (head)" % `type`
+        a = array.array(type, u"abcde")
+        a[1:] = a
+        if a != array.array(type, u"aabcde"):
+            raise TestFailed, "array(%s) self-slice-assign (tail)" % `type`
+        a = array.array(type, u"abcde")
+        a[1:-1] = a
+        if a != array.array(type, u"aabcdee"):
+            raise TestFailed, "array(%s) self-slice-assign (cntr)" % `type`
+        if a.index(u"e") != 5:
+            raise TestFailed, "array(%s) index-test" % `type`
+        if a.count(u"a") != 2:
+            raise TestFailed, "array(%s) count-test" % `type`
+        a.remove(u"e")
+        if a != array.array(type, u"aabcde"):
+            raise TestFailed, "array(%s) remove-test" % `type`
+        if a.pop(0) != u"a":
+            raise TestFailed, "array(%s) pop-test" % `type`
+        if a.pop(1) != u"b":
+            raise TestFailed, "array(%s) pop-test" % `type`
+        a.extend(array.array(type, u"xyz"))
+        if a != array.array(type, u"acdexyz"):
+            raise TestFailed, "array(%s) extend-test" % `type`
+        a.pop()
+        a.pop()
+        a.pop()
+        x = a.pop()
+        if x != u'e':
+            raise TestFailed, "array(%s) pop-test" % `type`
+        if a != array.array(type, u"acd"):
+            raise TestFailed, "array(%s) pop-test" % `type`
+        a.reverse()
+        if a != array.array(type, u"dca"):
+            raise TestFailed, "array(%s) reverse-test" % `type`
     else:
         a = array.array(type, [1, 2, 3, 4, 5])
         a[:-1] = a

Misc/NEWS

@@ -17,6 +17,11 @@ Core and builtins
 
 Extension modules
 
+- array.array is now a type object. A new format character
+  'u' indicates Py_UNICODE arrays. For those, .tounicode and
+  .fromunicode methods are available. Arrays now support __iadd__
+  and __imul__.
+
 - dl now builds on every system that has dlfcn.h.  Failure in case
   of sizeof(int)!=sizeof(long)!=sizeof(void*) is delayed until dl.open
   is called.