Issue #28379: Added sanity checks and tests for PyUnicode_CopyCharacters().

Patch by Xiang Zhang.

5 files changed, 94 insertions, 7 deletions

diff --git a/Doc/c-api/unicode.rst b/Doc/c-api/unicode.rst
index 261a43c..3f6a604 100644
--- a/Doc/c-api/unicode.rst
+++ b/Doc/c-api/unicode.rst
@@ -578,13 +578,16 @@ APIs:
    .. versionadded:: 3.3
 
 
-.. c:function:: int PyUnicode_CopyCharacters(PyObject *to, Py_ssize_t to_start, \
-                    PyObject *from, Py_ssize_t from_start, Py_ssize_t how_many)
+.. c:function:: Py_ssize_t PyUnicode_CopyCharacters(PyObject *to, \
+                                                    Py_ssize_t to_start, \
+                                                    PyObject *from, \
+                                                    Py_ssize_t from_start, \
+                                                    Py_ssize_t how_many)
 
    Copy characters from one Unicode object into another.  This function performs
    character conversion when necessary and falls back to :c:func:`memcpy` if
    possible.  Returns ``-1`` and sets an exception on error, otherwise returns
-   ``0``.
+   the number of copied characters.
 
    .. versionadded:: 3.3
 
diff --git a/Lib/test/test_unicode.py b/Lib/test/test_unicode.py
index 6086bb8..81e49d6 100644
--- a/Lib/test/test_unicode.py
+++ b/Lib/test/test_unicode.py
@@ -4,7 +4,7 @@ Written by Marc-Andre Lemburg (mal@lemburg.com).
 
 (c) Copyright CNRI, All Rights Reserved. NO WARRANTY.
 
-"""#"
+"""
 import _string
 import codecs
 import itertools
@@ -2704,6 +2704,49 @@ class CAPITest(unittest.TestCase):
             self.assertEqual(unicode_asucs4(s, len(s), 1), s+'\0')
             self.assertEqual(unicode_asucs4(s, len(s), 0), s+'\uffff')
 
+    # Test PyUnicode_CopyCharacters()
+    @support.cpython_only
+    def test_copycharacters(self):
+        from _testcapi import unicode_copycharacters
+
+        strings = [
+            'abcde', '\xa1\xa2\xa3\xa4\xa5',
+            '\u4f60\u597d\u4e16\u754c\uff01',
+            '\U0001f600\U0001f601\U0001f602\U0001f603\U0001f604'
+        ]
+
+        for idx, from_ in enumerate(strings):
+            # wide -> narrow: exceed maxchar limitation
+            for to in strings[:idx]:
+                self.assertRaises(
+                    SystemError,
+                    unicode_copycharacters, to, 0, from_, 0, 5
+                )
+            # same kind
+            for from_start in range(5):
+                self.assertEqual(
+                    unicode_copycharacters(from_, 0, from_, from_start, 5),
+                    (from_[from_start:from_start+5].ljust(5, '\0'),
+                     5-from_start)
+                )
+            for to_start in range(5):
+                self.assertEqual(
+                    unicode_copycharacters(from_, to_start, from_, to_start, 5),
+                    (from_[to_start:to_start+5].rjust(5, '\0'),
+                     5-to_start)
+                )
+            # narrow -> wide
+            # Tests omitted since this creates invalid strings.
+
+        s = strings[0]
+        self.assertRaises(IndexError, unicode_copycharacters, s, 6, s, 0, 5)
+        self.assertRaises(IndexError, unicode_copycharacters, s, -1, s, 0, 5)
+        self.assertRaises(IndexError, unicode_copycharacters, s, 0, s, 6, 5)
+        self.assertRaises(IndexError, unicode_copycharacters, s, 0, s, -1, 5)
+        self.assertRaises(SystemError, unicode_copycharacters, s, 1, s, 0, 5)
+        self.assertRaises(SystemError, unicode_copycharacters, s, 0, s, 0, -1)
+        self.assertRaises(SystemError, unicode_copycharacters, s, 0, b'', 0, 0)
+
     @support.cpython_only
     def test_encode_decimal(self):
         from _testcapi import unicode_encodedecimal
diff --git a/Misc/NEWS b/Misc/NEWS
index a89afa9..db74991 100644
--- a/Misc/NEWS
+++ b/Misc/NEWS
@@ -10,6 +10,9 @@ Release date: TBA
 Core and Builtins
 -----------------
 
+- Issue #28379: Added sanity checks and tests for PyUnicode_CopyCharacters().
+  Patch by Xiang Zhang.
+
 - Issue #28376: The type of long range iterator is now registered as Iterator.
   Patch by Oren Milman.
 
diff --git a/Modules/_testcapimodule.c b/Modules/_testcapimodule.c
index f4a1e97..4ecd48a 100644
--- a/Modules/_testcapimodule.c
+++ b/Modules/_testcapimodule.c
@@ -1860,6 +1860,39 @@ unicode_asucs4(PyObject *self, PyObject *args)
 }
 
 static PyObject *
+unicode_copycharacters(PyObject *self, PyObject *args)
+{
+    PyObject *from, *to, *to_copy;
+    Py_ssize_t from_start, to_start, how_many, copied;
+
+    if (!PyArg_ParseTuple(args, "UnOnn:unicode_copycharacters", &to, &to_start,
+                          &from, &from_start, &how_many)) {
+        return NULL;
+    }
+
+    if (PyUnicode_READY(to) < 0) {
+        return NULL;
+    }
+
+    if (!(to_copy = PyUnicode_New(PyUnicode_GET_LENGTH(to),
+                                  PyUnicode_MAX_CHAR_VALUE(to)))) {
+        return NULL;
+    }
+    if (PyUnicode_Fill(to_copy, 0, PyUnicode_GET_LENGTH(to_copy), 0U) < 0) {
+        Py_DECREF(to_copy);
+        return NULL;
+    }
+
+    if ((copied = PyUnicode_CopyCharacters(to_copy, to_start, from,
+                                           from_start, how_many)) < 0) {
+        Py_DECREF(to_copy);
+        return NULL;
+    }
+
+    return Py_BuildValue("(Nn)", to_copy, copied);
+}
+
+static PyObject *
 unicode_encodedecimal(PyObject *self, PyObject *args)
 {
     Py_UNICODE *unicode;
@@ -3915,6 +3948,7 @@ static PyMethodDef TestMethods[] = {
     {"unicode_aswidechar",      unicode_aswidechar,              METH_VARARGS},
     {"unicode_aswidecharstring",unicode_aswidecharstring,        METH_VARARGS},
     {"unicode_asucs4",          unicode_asucs4,                  METH_VARARGS},
+    {"unicode_copycharacters",  unicode_copycharacters,          METH_VARARGS},
     {"unicode_encodedecimal",   unicode_encodedecimal,           METH_VARARGS},
     {"unicode_transformdecimaltoascii", unicode_transformdecimaltoascii, METH_VARARGS},
     {"unicode_legacy_string",   unicode_legacy_string,           METH_VARARGS},
diff --git a/Objects/unicodeobject.c b/Objects/unicodeobject.c
index 176ec13..b734eec 100644
--- a/Objects/unicodeobject.c
+++ b/Objects/unicodeobject.c
@@ -1366,15 +1366,19 @@ PyUnicode_CopyCharacters(PyObject *to, Py_ssize_t to_start,
     if (PyUnicode_READY(to) == -1)
         return -1;
 
-    if (from_start < 0) {
+    if ((size_t)from_start > (size_t)PyUnicode_GET_LENGTH(from)) {
         PyErr_SetString(PyExc_IndexError, "string index out of range");
         return -1;
     }
-    if (to_start < 0) {
+    if ((size_t)to_start > (size_t)PyUnicode_GET_LENGTH(to)) {
         PyErr_SetString(PyExc_IndexError, "string index out of range");
         return -1;
     }
-    how_many = Py_MIN(PyUnicode_GET_LENGTH(from), how_many);
+    if (how_many < 0) {
+        PyErr_SetString(PyExc_SystemError, "how_many cannot be negative");
+        return -1;
+    }
+    how_many = Py_MIN(PyUnicode_GET_LENGTH(from)-from_start, how_many);
     if (to_start + how_many > PyUnicode_GET_LENGTH(to)) {
         PyErr_Format(PyExc_SystemError,
                      "Cannot write %zi characters at %zi "