octoleo/qpdf
2
1
Fork 0
mirror of https://github.com/qpdf/qpdf.git synced 2024-12-22 19:08:59 +00:00
Code Issues Releases Wiki Activity

Refactor QUtil::utf8_to_ascii

Browse Source
This commit is contained in:
Jay Berkenbilt 2019-01-05 13:04:05 -05:00
parent 089ce5902e
commit 3ef1b77304
3 changed files with 38 additions and 29 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

53
libqpdf/QUtil.cc
View File

@ -893,22 +893,34 @@ QUtil::parse_numrange(char const* range, int max)
return result;
}
enum encoding_e { e_utf16 };
enum encoding_e { e_utf16, e_ascii };
static
std::string
transcode_utf8(std::string const& utf8_val, encoding_e encoding)
transcode_utf8(std::string const& utf8_val, encoding_e encoding,
char unknown)
{
std::string result = "\xfe\xff";
std::string result;
if (encoding == e_utf16)
{
result += "\xfe\xff";
}
size_t len = utf8_val.length();
for (size_t i = 0; i < len; ++i)
{
unsigned char ch = static_cast<unsigned char>(utf8_val.at(i));
if (ch < 128)
{
if (encoding == e_utf16)
{
result += QUtil::toUTF16(ch);
}
else
{
result.append(1, ch);
}
}
else
{
size_t bytes_needed = 0;
unsigned bit_check = 0x40;
@ -922,10 +934,17 @@ transcode_utf8(std::string const& utf8_val, encoding_e encoding)
if (((bytes_needed > 5) || (bytes_needed < 1)) ||
((i + bytes_needed) >= len))
{
if (encoding == e_utf16)
{
result += "\xff\xfd";
}
else
{
result.append(1, unknown);
}
}
else
{
unsigned long codepoint = (ch & ~to_clear);
while (bytes_needed > 0)
@ -941,8 +960,15 @@ transcode_utf8(std::string const& utf8_val, encoding_e encoding)
codepoint <<= 6;
codepoint += (ch & 0x3f);
}
if (encoding == e_utf16)
{
result += QUtil::toUTF16(codepoint);
}
else
{
result.append(1, unknown);
}
}
}
}
return result;
@ -951,28 +977,11 @@ transcode_utf8(std::string const& utf8_val, encoding_e encoding)
std::string
QUtil::utf8_to_utf16(std::string const& utf8)
{
return transcode_utf8(utf8, e_utf16);
return transcode_utf8(utf8, e_utf16, 0);
}
std::string
QUtil::utf8_to_ascii(std::string const& utf8, char unknown_char)
{
std::string ascii_value;
for (size_t i = 0; i < utf8.length(); ++i)
{
unsigned char ch = static_cast<unsigned char>(utf8.at(i));
if (ch < 128)
{
ascii_value.append(1, ch);
}
else if ((ch & 0xc0) == 0x80)
{
// Ignore subsequent byte of UTF-8 encoded character
}
else
{
ascii_value.append(1, unknown_char);
}
}
return ascii_value;
return transcode_utf8(utf8, e_ascii, unknown_char);
}

6
libtests/qtest/qutil/qutil.out
View File

@ -48,9 +48,9 @@ HAGOOGAMAGOOGLE: 0
0x7fffffff -> ff fd
0x80000000 -> ff fd
---- utf8_to_ascii
Does π have fingers?
Does ? have fingers?
Does * have fingers?
¿Does π have fingers?
?Does ? have fingers?
*Does * have fingers?
---- whoami
quack1
quack2

2
libtests/qutil.cc
View File

@ -222,7 +222,7 @@ void to_utf16_test()
void utf8_to_ascii_test()
{
char const* input = "Does \317\200 have fingers?";
char const* input = "\302\277Does \317\200 have fingers?";
std::cout << input
<< std::endl
<< QUtil::utf8_to_ascii(input)