summaryrefslogtreecommitdiff
path: root/dpm/strdup16to8.cpp
diff options
context:
space:
mode:
authorArian <arian.kulmer@web.de>2020-09-22 17:32:07 +0200
committerMichael Bestas <mkbestas@lineageos.org>2020-12-20 18:55:01 +0200
commitd7853458d0f848282480b01b8c106c2a7979c218 (patch)
treee84b937e99a0acb885f712a477e11b551f1b6e0c /dpm/strdup16to8.cpp
parent088a145a3a900a2d46481073c928445fec9db81a (diff)
sdm660-common: Add a shim for dpmframework
* libdpmframework depends on strndup16to8 and strdup8to16 methods. Change-Id: I39d252b5830b752728a0c8607d32d8a279524ce1 Co-authored-by: Demon000 <demonsingur@gmail.com>
Diffstat (limited to 'dpm/strdup16to8.cpp')
-rw-r--r--dpm/strdup16to8.cpp167
1 files changed, 167 insertions, 0 deletions
diff --git a/dpm/strdup16to8.cpp b/dpm/strdup16to8.cpp
new file mode 100644
index 0000000..eb9e012
--- /dev/null
+++ b/dpm/strdup16to8.cpp
@@ -0,0 +1,167 @@
+/* libs/cutils/strdup16to8.c
+**
+** Copyright 2006, The Android Open Source Project
+**
+** Licensed under the Apache License, Version 2.0 (the "License");
+** you may not use this file except in compliance with the License.
+** You may obtain a copy of the License at
+**
+** http://www.apache.org/licenses/LICENSE-2.0
+**
+** Unless required by applicable law or agreed to in writing, software
+** distributed under the License is distributed on an "AS IS" BASIS,
+** WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+** See the License for the specific language governing permissions and
+** limitations under the License.
+*/
+
+#include "jstring.h"
+
+#include <assert.h>
+#include <limits.h>
+#include <stdlib.h>
+
+/**
+ * Convert a UTF-16 string to UTF-8.
+ *
+ */
+char * strndup16to8 (const char16_t* s, size_t n)
+{
+ if (s == NULL) {
+ return NULL;
+ }
+
+ size_t len = strnlen16to8(s, n);
+
+ /* We are paranoid, and we check for SIZE_MAX-1
+ * too since it is an overflow value for our
+ * strnlen16to8 implementation.
+ */
+ if (len >= SIZE_MAX-1)
+ return NULL;
+
+ char* ret = static_cast<char*>(malloc(len + 1));
+ if (ret == NULL)
+ return NULL;
+
+ strncpy16to8 (ret, s, n);
+
+ return ret;
+}
+
+/**
+ * Given a UTF-16 string, compute the length of the corresponding UTF-8
+ * string in bytes.
+ */
+extern size_t strnlen16to8(const char16_t* utf16Str, size_t len)
+{
+ size_t utf8Len = 0;
+
+ /* A small note on integer overflow. The result can
+ * potentially be as big as 3*len, which will overflow
+ * for len > SIZE_MAX/3.
+ *
+ * Moreover, the result of a strnlen16to8 is typically used
+ * to allocate a destination buffer to strncpy16to8 which
+ * requires one more byte to terminate the UTF-8 copy, and
+ * this is generally done by careless users by incrementing
+ * the result without checking for integer overflows, e.g.:
+ *
+ * dst = malloc(strnlen16to8(utf16,len)+1)
+ *
+ * Due to this, the following code will try to detect
+ * overflows, and never return more than (SIZE_MAX-1)
+ * when it detects one. A careless user will try to malloc
+ * SIZE_MAX bytes, which will return NULL which can at least
+ * be detected appropriately.
+ *
+ * As far as I know, this function is only used by strndup16(),
+ * but better be safe than sorry.
+ */
+
+ /* Fast path for the usual case where 3*len is < SIZE_MAX-1.
+ */
+ if (len < (SIZE_MAX-1)/3) {
+ while (len != 0) {
+ len--;
+ unsigned int uic = *utf16Str++;
+
+ if (uic > 0x07ff)
+ utf8Len += 3;
+ else if (uic > 0x7f || uic == 0)
+ utf8Len += 2;
+ else
+ utf8Len++;
+ }
+ return utf8Len;
+ }
+
+ /* The slower but paranoid version */
+ while (len != 0) {
+ len--;
+ unsigned int uic = *utf16Str++;
+ size_t utf8Cur = utf8Len;
+
+ if (uic > 0x07ff)
+ utf8Len += 3;
+ else if (uic > 0x7f || uic == 0)
+ utf8Len += 2;
+ else
+ utf8Len++;
+
+ if (utf8Len < utf8Cur) /* overflow detected */
+ return SIZE_MAX-1;
+ }
+
+ /* don't return SIZE_MAX to avoid common user bug */
+ if (utf8Len == SIZE_MAX)
+ utf8Len = SIZE_MAX-1;
+
+ return utf8Len;
+}
+
+
+/**
+ * Convert a Java-Style UTF-16 string + length to a JNI-Style UTF-8 string.
+ *
+ * This basically means: embedded \0's in the UTF-16 string are encoded
+ * as "0xc0 0x80"
+ *
+ * Make sure you allocate "utf8Str" with the result of strlen16to8() + 1,
+ * not just "len".
+ *
+ * Please note, a terminated \0 is always added, so your result will always
+ * be "strlen16to8() + 1" bytes long.
+ */
+extern char* strncpy16to8(char* utf8Str, const char16_t* utf16Str, size_t len)
+{
+ char* utf8cur = utf8Str;
+
+ /* Note on overflows: We assume the user did check the result of
+ * strnlen16to8() properly or at a minimum checked the result of
+ * its malloc(SIZE_MAX) in case of overflow.
+ */
+ while (len != 0) {
+ len--;
+ unsigned int uic = *utf16Str++;
+
+ if (uic > 0x07ff) {
+ *utf8cur++ = (uic >> 12) | 0xe0;
+ *utf8cur++ = ((uic >> 6) & 0x3f) | 0x80;
+ *utf8cur++ = (uic & 0x3f) | 0x80;
+ } else if (uic > 0x7f || uic == 0) {
+ *utf8cur++ = (uic >> 6) | 0xc0;
+ *utf8cur++ = (uic & 0x3f) | 0x80;
+ } else {
+ *utf8cur++ = uic;
+
+ if (uic == 0) {
+ break;
+ }
+ }
+ }
+
+ *utf8cur = '\0';
+
+ return utf8Str;
+}