New submission from Georg Brandl:
This makes the str8 constructor accept the same kinds of types as the
bytes constructor. I had to fix instances of str8("abc") to str8(b"abc")
to make tests pass again. The only remaining failure should be test_str
-- the string test suite must be thoroughly redesigned to fit all three
string-like types.
----------
assignee: gvanrossum
components: Interpreter Core
files: str_constructor.diff
keywords: py3k
messages: 56572
nosy: georg.brandl, gvanrossum
severity: normal
status: open
title: adapt str8 constructor to bytes constructor
versions: Python 3.0
Added file: http://bugs.python.org/file8570/str_constructor.diff
__________________________________
Tracker <[EMAIL PROTECTED]>
<http://bugs.python.org/issue1303>
__________________________________
Index: Objects/stringobject.c
===================================================================
--- Objects/stringobject.c (Revision 58552)
+++ Objects/stringobject.c (Arbeitskopie)
@@ -3020,16 +3020,149 @@
static PyObject *
string_new(PyTypeObject *type, PyObject *args, PyObject *kwds)
{
- PyObject *x = NULL;
- static char *kwlist[] = {"object", 0};
+ PyObject *x = NULL, *it;
+ PyObject *(*iternext)(PyObject *);
+ const char *encoding = NULL;
+ const char *errors = NULL;
+ PyObject *new = NULL;
+ Py_ssize_t i, size;
+ static char *kwlist[] = {"object", "encoding", "errors", 0};
if (type != &PyString_Type)
return str_subtype_new(type, args, kwds);
- if (!PyArg_ParseTupleAndKeywords(args, kwds, "|O:str8", kwlist, &x))
+ if (!PyArg_ParseTupleAndKeywords(args, kwds, "|Oss:str8", kwlist, &x,
+ &encoding, &errors))
return NULL;
- if (x == NULL)
+ if (x == NULL) {
+ if (encoding != NULL || errors != NULL) {
+ PyErr_SetString(PyExc_TypeError,
+ "encoding or errors without sequence "
+ "argument");
+ return NULL;
+ }
return PyString_FromString("");
- return PyObject_Str(x);
+ }
+
+ if (PyUnicode_Check(x)) {
+ /* Encode via the codec registry */
+ if (encoding == NULL) {
+ PyErr_SetString(PyExc_TypeError,
+ "string argument without an encoding");
+ return NULL;
+ }
+ new = PyCodec_Encode(x, encoding, errors);
+ if (new == NULL)
+ return NULL;
+ /* XXX(gb): must accept bytes here since codecs output bytes
+ at the moment */
+ if (PyBytes_Check(new)) {
+ PyObject *str;
+ str = PyString_FromString(PyBytes_AsString(new));
+ Py_DECREF(new);
+ if (!str)
+ return NULL;
+ return str;
+ }
+ if (!PyString_Check(new)) {
+ PyErr_Format(PyExc_TypeError,
+ "encoder did not return a str8 "
+ "object (type=%.400s)",
+ Py_Type(new)->tp_name);
+ Py_DECREF(new);
+ return NULL;
+ }
+ return new;
+ }
+
+ /* If it's not unicode, there can't be encoding or errors */
+ if (encoding != NULL || errors != NULL) {
+ PyErr_SetString(PyExc_TypeError,
+ "encoding or errors without a string argument");
+ return NULL;
+ }
+
+ /* Use the modern buffer interface */
+ if (PyObject_CheckBuffer(x)) {
+ Py_buffer view;
+ if (PyObject_GetBuffer(x, &view, PyBUF_FULL_RO) < 0)
+ return NULL;
+ new = PyString_FromStringAndSize(NULL, view.len);
+ if (!new)
+ goto fail;
+ if (PyBuffer_ToContiguous(((PyStringObject *)new)->ob_sval,
+ &view, view.len, 'C') < 0)
+ goto fail;
+ PyObject_ReleaseBuffer(x, &view);
+ return new;
+ fail:
+ Py_XDECREF(new);
+ PyObject_ReleaseBuffer(x, &view);
+ return NULL;
+ }
+
+ /* For the iterator version, create a string object and resize as
needed. */
+ /* XXX(gb): is 64 a good value? also, optimize this if length is known
*/
+ size = 64;
+ new = PyString_FromStringAndSize(NULL, size);
+ if (new == NULL)
+ return NULL;
+
+ /* XXX Optimize this if the arguments is a list, tuple */
+
+ /* Get the iterator */
+ it = PyObject_GetIter(x);
+ if (it == NULL)
+ goto error;
+ iternext = *Py_Type(it)->tp_iternext;
+
+ /* Run the iterator to exhaustion */
+ for (i = 0; ; i++) {
+ PyObject *item;
+ Py_ssize_t value;
+
+ /* Get the next item */
+ item = iternext(it);
+ if (item == NULL) {
+ if (PyErr_Occurred()) {
+ if
(!PyErr_ExceptionMatches(PyExc_StopIteration))
+ goto error;
+ PyErr_Clear();
+ }
+ break;
+ }
+
+ /* Interpret it as an int (__index__) */
+ value = PyNumber_AsSsize_t(item, PyExc_ValueError);
+ Py_DECREF(item);
+ if (value == -1 && PyErr_Occurred())
+ goto error;
+
+ /* Range check */
+ if (value < 0 || value >= 256) {
+ PyErr_SetString(PyExc_ValueError,
+ "bytes must be in range(0, 256)");
+ goto error;
+ }
+
+ /* Append the byte */
+ if (i >= size) {
+ size *= 2;
+ if (_PyString_Resize(&new, size) < 0)
+ goto error;
+ }
+ ((PyStringObject *)new)->ob_sval[i] = value;
+ }
+ _PyString_Resize(&new, i);
+
+ /* Clean up and return success */
+ Py_DECREF(it);
+ return new;
+
+ error:
+ /* Error handling when it != NULL */
+ Py_XDECREF(it);
+ Py_DECREF(new);
+ return NULL;
}
static PyObject *
Index: Lib/struct.py
===================================================================
--- Lib/struct.py (Revision 58552)
+++ Lib/struct.py (Arbeitskopie)
@@ -36,7 +36,7 @@
class Struct(_Struct):
def __init__(self, fmt):
if isinstance(fmt, str):
- fmt = str8(fmt)
+ fmt = str8(fmt, 'latin1')
_Struct.__init__(self, fmt)
_MAXCACHE = 100
Index: Lib/pickletools.py
===================================================================
--- Lib/pickletools.py (Revision 58552)
+++ Lib/pickletools.py (Arbeitskopie)
@@ -1978,7 +1978,7 @@
_dis_test = r"""
>>> import pickle
->>> x = [1, 2, (3, 4), {str8('abc'): "def"}]
+>>> x = [1, 2, (3, 4), {str8(b'abc'): "def"}]
>>> pkl = pickle.dumps(x, 0)
>>> dis(pkl)
0: ( MARK
Index: Lib/test/test_io.py
===================================================================
--- Lib/test/test_io.py (Revision 58552)
+++ Lib/test/test_io.py (Arbeitskopie)
@@ -88,7 +88,7 @@
self.assertEqual(f.tell(), 6)
self.assertEqual(f.seek(-1, 1), 5)
self.assertEqual(f.tell(), 5)
- self.assertEqual(f.write(str8(" world\n\n\n")), 9)
+ self.assertEqual(f.write(str8(b" world\n\n\n")), 9)
self.assertEqual(f.seek(0), 0)
self.assertEqual(f.write(b"h"), 1)
self.assertEqual(f.seek(-1, 2), 13)
Index: Lib/test/test_codeccallbacks.py
===================================================================
--- Lib/test/test_codeccallbacks.py (Revision 58552)
+++ Lib/test/test_codeccallbacks.py (Arbeitskopie)
@@ -181,7 +181,7 @@
# mapped through the encoding again. This means, that
# to be able to use e.g. the "replace" handler, the
# charmap has to have a mapping for "?".
- charmap = dict((ord(c), str8(2*c.upper())) for c in "abcdefgh")
+ charmap = dict((ord(c), str8(2*c.upper(), 'ascii')) for c in
"abcdefgh")
sin = "abc"
sout = b"AABBCC"
self.assertEquals(codecs.charmap_encode(sin, "strict", charmap)[0],
sout)
@@ -189,7 +189,7 @@
sin = "abcA"
self.assertRaises(UnicodeError, codecs.charmap_encode, sin, "strict",
charmap)
- charmap[ord("?")] = str8("XYZ")
+ charmap[ord("?")] = str8(b"XYZ")
sin = "abcDEF"
sout = b"AABBCCXYZXYZXYZ"
self.assertEquals(codecs.charmap_encode(sin, "replace", charmap)[0],
sout)
@@ -309,7 +309,7 @@
# check with one argument too much
self.assertRaises(TypeError, exctype, *(args + ["too much"]))
# check with one argument of the wrong type
- wrongargs = [ "spam", str8("eggs"), b"spam", 42, 1.0, None ]
+ wrongargs = [ "spam", str8(b"eggs"), b"spam", 42, 1.0, None ]
for i in range(len(args)):
for wrongarg in wrongargs:
if type(wrongarg) is type(args[i]):
Index: Lib/test/test_locale.py
===================================================================
--- Lib/test/test_locale.py (Revision 58552)
+++ Lib/test/test_locale.py (Arbeitskopie)
@@ -82,7 +82,7 @@
# Test BSD Rune locale's bug for isctype functions.
def teststrop(s, method, output):
- s = str8(s)
+ s = str8(s, 'latin1') # XXX
if verbose:
print("%s.%s() =? %s ..." % (repr(s), method, repr(output)), end=' ')
result = getattr(s, method)()
Index: Lib/test/testcodec.py
===================================================================
--- Lib/test/testcodec.py (Revision 58552)
+++ Lib/test/testcodec.py (Arbeitskopie)
@@ -36,7 +36,7 @@
decoding_map = codecs.make_identity_dict(range(256))
decoding_map.update({
0x78: "abc", # 1-n decoding mapping
- str8("abc"): 0x0078,# 1-n encoding mapping
+ str8(b"abc"): 0x0078,# 1-n encoding mapping
0x01: None, # decoding mapping to <undefined>
0x79: "", # decoding mapping to <remove character>
})
Index: Lib/test/test_builtin.py
===================================================================
--- Lib/test/test_builtin.py (Revision 58552)
+++ Lib/test/test_builtin.py (Arbeitskopie)
@@ -580,7 +580,8 @@
self.assertEqual(hash(1), hash(1))
self.assertEqual(hash(1), hash(1.0))
hash('spam')
- self.assertEqual(hash('spam'), hash(str8('spam')))
+ self.assertEqual(hash('spam'), hash(str8(b'spam'))) # remove str8()
+ # when b"" is immutable
hash((0,1,2,3))
def f(): pass
self.assertRaises(TypeError, hash, [])
Index: Lib/test/test_sys.py
===================================================================
--- Lib/test/test_sys.py (Revision 58552)
+++ Lib/test/test_sys.py (Arbeitskopie)
@@ -300,7 +300,7 @@
def test_intern(self):
self.assertRaises(TypeError, sys.intern)
- s = str8("never interned before")
+ s = str8(b"never interned before")
self.assert_(sys.intern(s) is s)
s2 = s.swapcase().swapcase()
self.assert_(sys.intern(s2) is s)
@@ -314,7 +314,7 @@
def __hash__(self):
return 123
- self.assertRaises(TypeError, sys.intern, S("abc"))
+ self.assertRaises(TypeError, sys.intern, S(b"abc"))
s = "never interned as unicode before"
self.assert_(sys.intern(s) is s)
Index: Lib/test/test_format.py
===================================================================
--- Lib/test/test_format.py (Revision 58552)
+++ Lib/test/test_format.py (Arbeitskopie)
@@ -40,7 +40,7 @@
print('yes')
def testboth(formatstr, *args):
- testformat(str8(formatstr), *args)
+ testformat(str8(formatstr, 'ascii'), *args)
testformat(formatstr, *args)
_______________________________________________
Python-bugs-list mailing list
Unsubscribe:
http://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
