fwprintf does not output wide characters

Question

I want to output wide characters to a file, and fwprintf doesn't do it, even though it's described as doing just that. Sample code:

const char *testFileName = "/Users/jdmuys/wideTestFile.txt";
FILE *wideTestFile;
wideTestFile = fopen(testFileName, "w");
fwide(wideTestFile, 1);
fwprintf(wideTestFile, L"12345");
fclose(wideTestFile);

After which my file "wideTestFile.txt" contains precisely 5 bytes: 31 32 33 34 35 according to my hex dump utility.

I suspect some issue with the current locale, as perhaps fwprintf calls upon fwprintf_l, which takes a locale as an additional argument.

I have been reading on how use that last call, but I can't manage to understand what I need to pass as a locale. The documentation is rather obscure on that (or perhaps I fail to understand it).

Any explanation why fwprintf doesn't behave as documented? and any example of use for fwprintf_l?

Many thanks,

JD

This is with Xcode 4.5.1 under Mac OS X 10.8.2 targetting iOS 6.0 from Objective-C code. But none of that should really matter.

I solved similar problem on Windows, maybe you could find useful some pieces from [here](http://stackoverflow.com/q/35928843/343721). — Jan Turoň, Mar 11 '16 at 00:31

slugonamission · Answer 1 · 2012-10-15T16:58:01.137

0

I'm going to take a guess here.

It's saving in UTF-8. Now, for most ASCII characters, the representation in ASCII and the representation in UTF-8 is exactly the same. Now, in UTF-8, the upper bits of the first byte encode the length of the "character". As an example, everything up to 0x7F fits in one byte (i.e. standard ASCII), 0x7FF in two bytes, and on and on. See http://en.wikipedia.org/wiki/UTF-8#Description for more details.

To "fix" your problem, simply use a character from higher up the UTF-8 table.

To rip some examples from the aforementioned Wikipedia page:

$ should fit in a single byte
¢ in two bytes
€ in three
in four

edited Oct 15 '12 at 16:58

answered Oct 15 '12 at 16:52

slugonamission

9,562
1
34
41

There's no such thing as a UTF-16 locale. – R.. GitHub STOP HELPING ICE Oct 15 '12 at 16:56
@R.. Dammit, memory is hazy from when I did this last (although I thought there was a way to make a locale use UTF-16). Updated, thanks. – slugonamission Oct 15 '12 at 16:57

score 0 · Answer 2 · edited May 23 '17 at 12:02

0

Your locale is probably some variation of UTF-8, which means that the output will only be wide when printing wide characters (i.e. characters outside of the ascii code 0 - 127).

If you want to force wide printing (e.g. UTF-16LE) then you need to use libiconv. This answer kind of illustrates why it doesn't tend to do what you think it does.

edited May 23 '17 at 12:02

Community

1
1

answered Oct 15 '12 at 17:00

Anya Shenanigans

91,618
3
107
122

If such is the case, then `fwprintf` does the same thing as `fprintf` and there is little point in providing it as its only difference is that the format string passed as an argument is wide for `fwprintf` and narrow for `fprintf`. Your link doesn't say *why* it doesn't do what I expected. It basically says "forget the standard library and use iconv" – Jean-Denis Muys Oct 16 '12 at 08:34
No, fprintf will not deal with wide characters properly while the fwprintf will. The problem is that the end product in your example code is not going to result in non ascii characters, so the output will only use ascii characters. The example is how to use libiconv to convert from UTF-16 -> UTF-8 (i.e. a windows file-format conversion). You want to accomplish the reverse of this - you swprintf into a buffer, perform the conversion using iconv, then output the iconv output to the file. – Anya Shenanigans Oct 16 '12 at 09:42

fwprintf does not output wide characters

2 Answers2