Thanks to visit codestin.com
Credit goes to github.com

Skip to content

Commit 09095f3

Browse files
gvanrossumgvanrossum
committed
Marc-Andre Lemburg: added new builtin functions unicode() and
unichr(); changed ord() to support Unicode strings; added new exception UnicodeError; fixed a typo in doc string for buffer().
1 parent feee4b9 commit 09095f3

1 file changed

Lines changed: 78 additions & 5 deletions

File tree

‎Python/bltinmodule.c‎

Lines changed: 78 additions & 5 deletions
Original file line numberDiff line numberDiff line change
@@ -152,14 +152,38 @@ builtin_buffer(self, args)
152152
}
153153

154154
static char buffer_doc[] =
155-
"buffer(object [, offset[, size]) -> object\n\
155+
"buffer(object [, offset[, size]]) -> object\n\
156156
\n\
157157
Creates a new buffer object which references the given object.\n\
158158
The buffer will reference a slice of the target object from the\n\
159159
start of the object (or at the specified offset). The slice will\n\
160160
extend to the end of the target object (or with the specified size).";
161161

162162

163+
static PyObject *
164+
builtin_unicode(self, args)
165+
PyObject *self;
166+
PyObject *args;
167+
{
168+
char *s;
169+
int len;
170+
char *encoding = NULL;
171+
char *errors = NULL;
172+
173+
if ( !PyArg_ParseTuple(args, "s#|ss:unicode", &s, &len,
174+
&encoding, &errors) )
175+
return NULL;
176+
return PyUnicode_Decode(s, len, encoding, errors);
177+
}
178+
179+
static char unicode_doc[] =
180+
"unicode(string [, encoding[, errors]]) -> object\n\
181+
\n\
182+
Creates a new unicode object from the given encoded string.\n\
183+
encoding defaults to 'utf-8' and errors, defining the error handling,\n\
184+
to 'strict'.";
185+
186+
163187
static PyObject *
164188
builtin_callable(self, args)
165189
PyObject *self;
@@ -311,6 +335,31 @@ static char chr_doc[] =
311335
Return a string of one character with ordinal i; 0 <= i < 256.";
312336

313337

338+
static PyObject *
339+
builtin_unichr(self, args)
340+
PyObject *self;
341+
PyObject *args;
342+
{
343+
long x;
344+
Py_UNICODE s[1];
345+
346+
if (!PyArg_ParseTuple(args, "l:unichr", &x))
347+
return NULL;
348+
if (x < 0 || x >= 65536) {
349+
PyErr_SetString(PyExc_ValueError,
350+
"unichr() arg not in range(65536)");
351+
return NULL;
352+
}
353+
s[0] = (Py_UNICODE)x;
354+
return PyUnicode_FromUnicode(s, 1);
355+
}
356+
357+
static char unichr_doc[] =
358+
"unichr(i) -> unicode character\n\
359+
\n\
360+
Return a unicode string of one character with ordinal i; 0 <= i < 65536.";
361+
362+
314363
static PyObject *
315364
builtin_cmp(self, args)
316365
PyObject *self;
@@ -1541,17 +1590,29 @@ builtin_ord(self, args)
15411590
PyObject *self;
15421591
PyObject *args;
15431592
{
1544-
char c;
1593+
PyObject *obj;
1594+
long ord;
15451595

1546-
if (!PyArg_ParseTuple(args, "c:ord", &c))
1596+
if (!PyArg_ParseTuple(args, "O:ord", &obj))
15471597
return NULL;
1548-
return PyInt_FromLong((long)(c & 0xff));
1598+
1599+
if (PyString_Check(obj) && PyString_GET_SIZE(obj) == 1)
1600+
ord = (long)((unsigned char)*PyString_AS_STRING(obj));
1601+
else if (PyUnicode_Check(obj) && PyUnicode_GET_SIZE(obj) == 1)
1602+
ord = (long)*PyUnicode_AS_UNICODE(obj);
1603+
else {
1604+
PyErr_SetString(PyExc_TypeError,
1605+
"expected a string or unicode character");
1606+
return NULL;
1607+
}
1608+
1609+
return PyInt_FromLong(ord);
15491610
}
15501611

15511612
static char ord_doc[] =
15521613
"ord(c) -> integer\n\
15531614
\n\
1554-
Return the integer ordinal of a one character string.";
1615+
Return the integer ordinal of a one character [unicode] string.";
15551616

15561617

15571618
static PyObject *
@@ -2227,6 +2288,8 @@ static PyMethodDef builtin_methods[] = {
22272288
{"str", builtin_str, 1, str_doc},
22282289
{"tuple", builtin_tuple, 1, tuple_doc},
22292290
{"type", builtin_type, 1, type_doc},
2291+
{"unicode", builtin_unicode, 1, unicode_doc},
2292+
{"unichr", builtin_unichr, 1, unichr_doc},
22302293
{"vars", builtin_vars, 1, vars_doc},
22312294
{"xrange", builtin_xrange, 1, xrange_doc},
22322295
{NULL, NULL},
@@ -2259,6 +2322,7 @@ PyObject *PyExc_SyntaxError;
22592322
PyObject *PyExc_SystemError;
22602323
PyObject *PyExc_SystemExit;
22612324
PyObject *PyExc_UnboundLocalError;
2325+
PyObject *PyExc_UnicodeError;
22622326
PyObject *PyExc_TypeError;
22632327
PyObject *PyExc_ValueError;
22642328
PyObject *PyExc_ZeroDivisionError;
@@ -2304,6 +2368,7 @@ bltin_exc[] = {
23042368
{"SystemError", &PyExc_SystemError, 1},
23052369
{"SystemExit", &PyExc_SystemExit, 1},
23062370
{"UnboundLocalError", &PyExc_UnboundLocalError, 1},
2371+
{"UnicodeError", &PyExc_UnicodeError, 1},
23072372
{"TypeError", &PyExc_TypeError, 1},
23082373
{"ValueError", &PyExc_ValueError, 1},
23092374
#ifdef MS_WINDOWS
@@ -2467,6 +2532,14 @@ initerrors(dict)
24672532
PyExc_NameError) != 0)
24682533
Py_FatalError("Cannot create string-based exceptions");
24692534

2535+
/* Make UnicodeError an alias for ValueError */
2536+
Py_INCREF(PyExc_ValueError);
2537+
Py_DECREF(PyExc_UnicodeError);
2538+
PyExc_UnicodeError = PyExc_ValueError;
2539+
if (PyDict_SetItemString(dict, "UnicodeError",
2540+
PyExc_ValueError) != 0)
2541+
Py_FatalError("Cannot create string-based exceptions");
2542+
24702543
/* missing from the StandardError tuple: Exception, StandardError,
24712544
* and SystemExit
24722545
*/

0 commit comments

Comments
 (0)