gnustep/libs-base
0
0
Fork 0
mirror of https://github.com/gnustep/libs-base.git synced 2025-04-23 00:41:02 +00:00
Code Issues Projects Releases Packages Wiki Activity

tentative/partial improvement of 'precision' of c-strings (%s format)

Browse source 
git-svn-id: svn+ssh://svn.gna.org/svn/gnustep/libs/base/trunk@35625 72102866-910b-0410-8b05-ffd578937521
This commit is contained in:
Richard Frith-MacDonald 2012-10-03 08:40:48 +00:00
parent f1fb0822c9
commit 1d7830df31
2 changed files with 33 additions and 21 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

9
ChangeLog
View file

@ -1,3 +1,12 @@
2012-10-03 Richard Frith-Macdonald <rfm@gnu.org>
* Source/GSFormat.m: for C-string formatting, avoid reading to the
nul terminator ... limit to a number of bytes equal to the precision
(if any) given in the format. This may be the wrong thing to do when
handling multibyte encodings (where a natural expectation is to have
precision mean number of characters), but seems legitimate going by
current xopen printf documentation.
2012-10-03 Wolfgang Lux <wolfgang.lux@gmail.com>
* SSL/GSSSLHandle.m (-sslSetOptions): Apply string name change

45
Source/GSFormat.m
View file

@ -1758,37 +1758,40 @@ NSDictionary *locale)
byteEncoding = GSPrivateIsByteEncoding(enc);
}
len = strlen(str); // Number of bytes to convert.
blen = len; // Size of unichar output buffer.
if (prec != -1)
if (-1 == prec)
{
len = strlen(str); // Number of bytes to convert.
blen = len; // Size of unichar output buffer.
}
else
{
if (prec < len)
{
/* We don't neeed an output buffer bigger than the
* precision specifies.
*/
blen = prec;
}
if (byteEncoding == YES)
{
/* Where the external encoding is one byte per character,
* we know we don't need to convert more bytes than the
* precision required for output.
*/
if (prec < len)
{
len = prec;
len = 0;
while (len < prec && str[0] != 0)
{
len++;
}
blen = len;
}
else if (prec * 4 < len)
else
{
/* We assume no multibyte encoding is going to use more
* than the maximum four bytes used by utf-8 for any
* unicode code point. So we do not need to convert
* more than four times the precision.
*/
len = prec * 4;
/* FIXME ... it looks like modern standards mean that
* the number of *bytes* in an input string may not
* exceed the precision ... but that's unintuitive for
* input strings with multibyte characters, so we need
* to check and emulate OSX behavior.
*/
len = 0;
while (len < prec && str[0] != 0)
{
len++;
}
blen = len;
}
}