dali/bacnet_stack
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Exposed utf8_valid as a public function to reduce duplicity. Allow utf-8 string length of 0 to be valid.

Browse Source
This commit is contained in:
skarg
2012-05-11 15:03:35 +00:00
parent 35e953ba8b
commit 54c4ce342c
3 changed files with 21 additions and 117 deletions
Download Patch File Download Diff File Expand all files Collapse all files
bacnet-stack/include/bacstr.h
+3
View File
@@ -145,6 +145,9 @@ extern "C" {
BACNET_CHARACTER_STRING * char_string);
bool characterstring_valid(
BACNET_CHARACTER_STRING * char_string);
bool utf8_isvalid(
const char *str,
size_t length);
/* returns false if the string exceeds capacity
initialize by using length=0 */
bacnet-stack/ports/bdk-atxx4-mstp/bname.c
+2 -105
View File
@@ -35,107 +35,6 @@
#include "device.h"
#include "bname.h"
/* Basic UTF-8 manipulation routines
by Jeff Bezanson
placed in the public domain Fall 2005 */
static const char trailingBytesForUTF8[256] = {
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0,
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
1, 1, 1, 1, 1, 1, 1,
2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 3, 3, 3, 3, 3, 3, 3, 3, 4,
4, 4, 4, 5, 5, 5, 5
};
/* based on the valid_utf8 routine from the PCRE library by Philip Hazel
length is in bytes, since without knowing whether the string is valid
it's hard to know how many characters there are! */
static int utf8_isvalid(
const char *str,
int length)
{
const unsigned char *p, *pend = (unsigned char *) str + length;
unsigned char c;
int ab;
for (p = (unsigned char *) str; p < pend; p++) {
c = *p;
/* null in middle of string */
if (c == 0) {
return 0;
}
/* ASCII character */
if (c < 128) {
continue;
}
if ((c & 0xc0) != 0xc0) {
return 0;
}
ab = trailingBytesForUTF8[c];
if (length < ab) {
return 0;
}
length -= ab;
p++;
/* Check top bits in the second byte */
if ((*p & 0xc0) != 0x80) {
return 0;
}
/* Check for overlong sequences for each different length */
switch (ab) {
/* Check for xx00 000x */
case 1:
if ((c & 0x3e) == 0)
return 0;
continue; /* We know there aren't any more bytes to check */
/* Check for 1110 0000, xx0x xxxx */
case 2:
if (c == 0xe0 && (*p & 0x20) == 0)
return 0;
break;
/* Check for 1111 0000, xx00 xxxx */
case 3:
if (c == 0xf0 && (*p & 0x30) == 0)
return 0;
break;
/* Check for 1111 1000, xx00 0xxx */
case 4:
if (c == 0xf8 && (*p & 0x38) == 0)
return 0;
break;
/* Check for leading 0xfe or 0xff,
and then for 1111 1100, xx00 00xx */
case 5:
if (c == 0xfe || c == 0xff || (c == 0xfc && (*p & 0x3c) == 0))
return 0;
break;
}
/* Check for valid bytes after the 2nd, if any; all must start 10 */
while (--ab > 0) {
if ((*(++p) & 0xc0) != 0x80)
return 0;
}
}
return 1;
}
static bool bacnet_name_isvalid(
uint8_t encoding,
uint8_t length,
@@ -145,10 +44,8 @@ static bool bacnet_name_isvalid(
if ((encoding < MAX_CHARACTER_STRING_ENCODING) &&
(length <= NV_EEPROM_NAME_SIZE)) {
if (encoding == CHARACTER_ANSI_X34) {
if (utf8_isvalid(str, length)) {
valid = true;
}
if (encoding == CHARACTER_UTF8) {
valid = utf8_isvalid(str, length);
} else {
valid = true;
}
bacnet-stack/src/bacstr.c
+16 -12
View File
@@ -521,7 +521,7 @@ static const char trailingBytesForUTF8[256] = {
/* based on the valid_utf8 routine from the PCRE library by Philip Hazel
length is in bytes, since without knowing whether the string is valid
it's hard to know how many characters there are! */
static int utf8_isvalid(
bool utf8_isvalid(
const char *str,
size_t length)
{
@@ -529,72 +529,76 @@ static int utf8_isvalid(
unsigned char c;
size_t ab;
/* empty string is valid */
if (length == 0) {
return true;
}
for (p = (unsigned char *) str; p < pend; p++) {
c = *p;
/* null in middle of string */
if (c == 0) {
return 0;
return false;
}
/* ASCII character */
if (c < 128) {
continue;
}
if ((c & 0xc0) != 0xc0) {
return 0;
return false;
}
ab = (size_t)trailingBytesForUTF8[c];
if (length < ab) {
return 0;
return false;
}
length -= ab;
p++;
/* Check top bits in the second byte */
if ((*p & 0xc0) != 0x80) {
return 0;
return false;
}
/* Check for overlong sequences for each different length */
switch (ab) {
/* Check for xx00 000x */
case 1:
if ((c & 0x3e) == 0)
return 0;
return false;
continue; /* We know there aren't any more bytes to check */
/* Check for 1110 0000, xx0x xxxx */
case 2:
if (c == 0xe0 && (*p & 0x20) == 0)
return 0;
return false;
break;
/* Check for 1111 0000, xx00 xxxx */
case 3:
if (c == 0xf0 && (*p & 0x30) == 0)
return 0;
return false;
break;
/* Check for 1111 1000, xx00 0xxx */
case 4:
if (c == 0xf8 && (*p & 0x38) == 0)
return 0;
return false;
break;
/* Check for leading 0xfe or 0xff,
and then for 1111 1100, xx00 00xx */
case 5:
if (c == 0xfe || c == 0xff || (c == 0xfc && (*p & 0x3c) == 0))
return 0;
return false;
break;
}
/* Check for valid bytes after the 2nd, if any; all must start 10 */
while (--ab > 0) {
if ((*(++p) & 0xc0) != 0x80)
return 0;
return false;
}
}
return 1;
return true;
}
bool characterstring_valid(