summaryrefslogtreecommitdiffstats
path: root/src/StringUtils.cpp
diff options
context:
space:
mode:
Diffstat (limited to 'src/StringUtils.cpp')
-rw-r--r--src/StringUtils.cpp41
1 files changed, 25 insertions, 16 deletions
diff --git a/src/StringUtils.cpp b/src/StringUtils.cpp
index 34f2da682..5febf5d6c 100644
--- a/src/StringUtils.cpp
+++ b/src/StringUtils.cpp
@@ -416,24 +416,30 @@ static bool isLegalUTF8(const unsigned char * source, int length)
{
default: return false;
// Everything else falls through when "true"...
- case 4: if ((a = (*--srcptr)) < 0x80 || a > 0xBF) return false;
- case 3: if ((a = (*--srcptr)) < 0x80 || a > 0xBF) return false;
+ case 4: if (((a = (*--srcptr)) < 0x80) || (a > 0xbf)) return false;
+ case 3: if (((a = (*--srcptr)) < 0x80) || (a > 0xbf)) return false;
case 2:
{
- if ((a = (*--srcptr)) > 0xBF) return false;
+ if ((a = (*--srcptr)) > 0xbf)
+ {
+ return false;
+ }
switch (*source)
{
// no fall-through in this inner switch
- case 0xE0: if (a < 0xA0) return false; break;
- case 0xED: if (a > 0x9F) return false; break;
- case 0xF0: if (a < 0x90) return false; break;
- case 0xF4: if (a > 0x8F) return false; break;
+ case 0xe0: if (a < 0xa0) return false; break;
+ case 0xed: if (a > 0x9f) return false; break;
+ case 0xf0: if (a < 0x90) return false; break;
+ case 0xf4: if (a > 0x8f) return false; break;
default: if (a < 0x80) return false;
}
}
- case 1: if (*source >= 0x80 && *source < 0xC2) return false;
+ case 1: if ((*source >= 0x80) && (*source < 0xc2)) return false;
+ }
+ if (*source > 0xf4)
+ {
+ return false;
}
- if (*source > 0xF4) return false;
return true;
}
@@ -446,11 +452,11 @@ AString UTF8ToRawBEUTF16(const char * a_UTF8, size_t a_UTF8Length)
AString UTF16;
UTF16.reserve(a_UTF8Length * 3);
- const unsigned char * source = (const unsigned char*)a_UTF8;
+ const unsigned char * source = (const unsigned char *)a_UTF8;
const unsigned char * sourceEnd = source + a_UTF8Length;
const int halfShift = 10; // used for shifting by 10 bits
const unsigned int halfBase = 0x0010000UL;
- const unsigned int halfMask = 0x3FFUL;
+ const unsigned int halfMask = 0x3ffUL;
while (source < sourceEnd)
{
@@ -481,7 +487,7 @@ AString UTF8ToRawBEUTF16(const char * a_UTF8, size_t a_UTF8Length)
if (ch <= UNI_MAX_BMP)
{
// Target is a character <= 0xFFFF
- if (ch >= UNI_SUR_HIGH_START && ch <= UNI_SUR_LOW_END)
+ if ((ch >= UNI_SUR_HIGH_START) && (ch <= UNI_SUR_LOW_END))
{
// UTF-16 surrogate values are illegal in UTF-32
ch = ' ';
@@ -520,7 +526,10 @@ are equivalent to the following loop:
{
ch += *source++;
--tmpBytesToRead;
- if (tmpBytesToRead) ch <<= 6;
+ if (tmpBytesToRead)
+ {
+ ch <<= 6;
+ }
} while (tmpBytesToRead > 0);
}
---------------------------------------------------------------------
@@ -723,15 +732,15 @@ AString ReplaceAllCharOccurrences(const AString & a_String, char a_From, char a_
/// Converts one Hex character in a Base64 encoding into the data value
static inline int UnBase64(char c)
{
- if (c >='A' && c <= 'Z')
+ if ((c >='A') && (c <= 'Z'))
{
return c - 'A';
}
- if (c >='a' && c <= 'z')
+ if ((c >='a') && (c <= 'z'))
{
return c - 'a' + 26;
}
- if (c >= '0' && c <= '9')
+ if ((c >= '0') && (c <= '9'))
{
return c - '0' + 52;
}