Skip to content

Commit 8d3d465

Browse files
fangererfangerer
committed
Add test for native string subclass.
1 parent 3ef2ed9 commit 8d3d465

File tree

1 file changed

+178
-0
lines changed
  • graalpython/com.oracle.graal.python.test/src/tests/cpyext

1 file changed

+178
-0
lines changed

‎graalpython/com.oracle.graal.python.test/src/tests/cpyext/test_object.py

Lines changed: 178 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -384,6 +384,184 @@ def test_reverse_ops(self):
384384
assert 1 @ tester == 123, "__rmatmul__ failed"
385385

386386

387+
def test_str_subclass(self):
388+
TestStrSubclass = CPyExtType("TestStrSubclass",
389+
r"""
390+
static PyTypeObject* testStrSubclassPtr = NULL;
391+
392+
#define MAX_UNICODE 0x10ffff
393+
394+
#define _PyUnicode_UTF8(op) \
395+
(((PyCompactUnicodeObject*)(op))->utf8)
396+
#define PyUnicode_UTF8(op) \
397+
(assert(_PyUnicode_CHECK(op)), \
398+
assert(PyUnicode_IS_READY(op)), \
399+
PyUnicode_IS_COMPACT_ASCII(op) ? \
400+
((char*)((PyASCIIObject*)(op) + 1)) : \
401+
_PyUnicode_UTF8(op))
402+
#define _PyUnicode_UTF8_LENGTH(op) \
403+
(((PyCompactUnicodeObject*)(op))->utf8_length)
404+
#define PyUnicode_UTF8_LENGTH(op) \
405+
(assert(_PyUnicode_CHECK(op)), \
406+
assert(PyUnicode_IS_READY(op)), \
407+
PyUnicode_IS_COMPACT_ASCII(op) ? \
408+
((PyASCIIObject*)(op))->length : \
409+
_PyUnicode_UTF8_LENGTH(op))
410+
#define _PyUnicode_WSTR(op) \
411+
(((PyASCIIObject*)(op))->wstr)
412+
#define _PyUnicode_WSTR_LENGTH(op) \
413+
(((PyCompactUnicodeObject*)(op))->wstr_length)
414+
#define _PyUnicode_LENGTH(op) \
415+
(((PyASCIIObject *)(op))->length)
416+
#define _PyUnicode_STATE(op) \
417+
(((PyASCIIObject *)(op))->state)
418+
#define _PyUnicode_HASH(op) \
419+
(((PyASCIIObject *)(op))->hash)
420+
#define _PyUnicode_KIND(op) \
421+
(assert(_PyUnicode_CHECK(op)), \
422+
((PyASCIIObject *)(op))->state.kind)
423+
#define _PyUnicode_GET_LENGTH(op) \
424+
(assert(_PyUnicode_CHECK(op)), \
425+
((PyASCIIObject *)(op))->length)
426+
#define _PyUnicode_DATA_ANY(op) \
427+
(((PyUnicodeObject*)(op))->data.any)
428+
429+
// that's taken from CPython's 'PyUnicode_New'
430+
static PyUnicodeObject * new_empty_unicode(Py_ssize_t size, Py_UCS4 maxchar) {
431+
PyUnicodeObject *obj;
432+
PyCompactUnicodeObject *unicode;
433+
void *data;
434+
enum PyUnicode_Kind kind;
435+
int is_sharing, is_ascii;
436+
Py_ssize_t char_size;
437+
Py_ssize_t struct_size;
438+
439+
is_ascii = 0;
440+
is_sharing = 0;
441+
struct_size = sizeof(PyCompactUnicodeObject);
442+
if (maxchar < 128) {
443+
kind = PyUnicode_1BYTE_KIND;
444+
char_size = 1;
445+
is_ascii = 1;
446+
struct_size = sizeof(PyASCIIObject);
447+
}
448+
else if (maxchar < 256) {
449+
kind = PyUnicode_1BYTE_KIND;
450+
char_size = 1;
451+
}
452+
else if (maxchar < 65536) {
453+
kind = PyUnicode_2BYTE_KIND;
454+
char_size = 2;
455+
if (sizeof(wchar_t) == 2)
456+
is_sharing = 1;
457+
}
458+
else {
459+
if (maxchar > MAX_UNICODE) {
460+
PyErr_SetString(PyExc_SystemError,
461+
"invalid maximum character passed to PyUnicode_New");
462+
return NULL;
463+
}
464+
kind = PyUnicode_4BYTE_KIND;
465+
char_size = 4;
466+
if (sizeof(wchar_t) == 4)
467+
is_sharing = 1;
468+
}
469+
470+
/* Ensure we won't overflow the size. */
471+
if (size < 0) {
472+
PyErr_SetString(PyExc_SystemError,
473+
"Negative size passed to PyUnicode_New");
474+
return NULL;
475+
}
476+
if (size > ((PY_SSIZE_T_MAX - struct_size) / char_size - 1))
477+
return NULL;
478+
479+
/* Duplicated allocation code from _PyObject_New() instead of a call to
480+
* PyObject_New() so we are able to allocate space for the object and
481+
* it's data buffer.
482+
*/
483+
obj = (PyUnicodeObject *) malloc(struct_size + (size + 1) * char_size);
484+
if (obj == NULL)
485+
return NULL;
486+
obj = PyObject_INIT(obj, &PyUnicode_Type);
487+
if (obj == NULL)
488+
return NULL;
489+
490+
unicode = (PyCompactUnicodeObject *)obj;
491+
if (is_ascii)
492+
data = ((PyASCIIObject*)obj) + 1;
493+
else
494+
data = unicode + 1;
495+
_PyUnicode_LENGTH(unicode) = size;
496+
_PyUnicode_HASH(unicode) = -1;
497+
_PyUnicode_STATE(unicode).interned = 0;
498+
_PyUnicode_STATE(unicode).kind = kind;
499+
_PyUnicode_STATE(unicode).compact = 1;
500+
_PyUnicode_STATE(unicode).ready = 1;
501+
_PyUnicode_STATE(unicode).ascii = is_ascii;
502+
if (is_ascii) {
503+
((char*)data)[size] = 0;
504+
_PyUnicode_WSTR(unicode) = NULL;
505+
}
506+
else if (kind == PyUnicode_1BYTE_KIND) {
507+
((char*)data)[size] = 0;
508+
_PyUnicode_WSTR(unicode) = NULL;
509+
_PyUnicode_WSTR_LENGTH(unicode) = 0;
510+
unicode->utf8 = NULL;
511+
unicode->utf8_length = 0;
512+
}
513+
else {
514+
unicode->utf8 = NULL;
515+
unicode->utf8_length = 0;
516+
if (kind == PyUnicode_2BYTE_KIND)
517+
((Py_UCS2*)data)[size] = 0;
518+
else /* kind == PyUnicode_4BYTE_KIND */
519+
((Py_UCS4*)data)[size] = 0;
520+
if (is_sharing) {
521+
_PyUnicode_WSTR_LENGTH(unicode) = size;
522+
_PyUnicode_WSTR(unicode) = (wchar_t *)data;
523+
}
524+
else {
525+
_PyUnicode_WSTR_LENGTH(unicode) = 0;
526+
_PyUnicode_WSTR(unicode) = NULL;
527+
}
528+
}
529+
return obj;
530+
}
531+
532+
static PyObject* nstr_tpnew(PyTypeObject* type, PyObject* args, PyObject* kwargs) {
533+
char *ascii_data = NULL;
534+
Py_XINCREF(args);
535+
if (!PyArg_ParseTuple(args, "s", &ascii_data)) {{
536+
return NULL;
537+
}}
538+
Py_ssize_t len = strlen(ascii_data);
539+
PyUnicodeObject* strObj = new_empty_unicode(len, (Py_UCS4) 127);
540+
memcpy(PyUnicode_1BYTE_DATA(strObj), (Py_UCS1*)ascii_data, len);
541+
return (PyObject*) strObj;
542+
}
543+
""",
544+
cmembers="""PyUnicodeObject base;
545+
int marker;""",
546+
tp_base="&PyUnicode_Type",
547+
tp_new="nstr_tpnew",
548+
post_ready_code="testStrSubclassPtr = &TestStrSubclassType; Py_INCREF(testStrSubclassPtr);"
549+
)
550+
tester = TestStrSubclass("hello\nworld")
551+
assert tester == "hello\nworld"
552+
assert str(tester) == "hello\nworld"
553+
assert tester.splitlines() == ['hello', 'world']
554+
assert tester >= "hello"
555+
assert not (tester >= "helloasdfasdfasdf")
556+
assert tester <= "helloasdfasdfasdf"
557+
assert not (tester <= "hello")
558+
assert tester.startswith("hello")
559+
assert tester.endswith("rld")
560+
assert tester.join(["a", "b"]) == "ahello\nworldb"
561+
assert tester.upper() == "HELLO\nWORLD"
562+
assert tester.replace("o", "uff") == "helluff\nwuffrld"
563+
assert tester.replace("o", "uff", 1) == "helluff\nworld"
564+
387565
class TestObjectFunctions(CPyExtTestCase):
388566
def compile_module(self, name):
389567
type(self).mro()[1].__dict__["test_%s" % name].create_module(name)

0 commit comments

Comments
 (0)