tags:

views:

1374

answers:

10

I read somewhere that snprintf is faster than ostringstream. Has anyone has any experiences with it? If yes why is it faster.

A: 

It's quite possible that because sprintf is part of the CRT that is written in assembly. The ostringstream is part of the STL, and probably a little more generically written, and has OOP code/overhead to deal with.

LarryF
+12  A: 

std::ostringstream is not designed to be slower, but it generally is slower when implemented. FastFormat's website has some benchmarks.

I've never pinned down why it's slower; my guess is that stream operations basically need to use growable buffers internally and that implies memory allocation which can be slow.

Max Lybbert
Wow - I haven't been to Matthew Wilson's website in a little while - looks like he's had a busy last few months...
Michael Burr
The FastFormat tests are biased. Downloading the code, you'll see the snprintf() benchmarks call a function called fastformat_util_snprintf(), which includes some wrapper code that probably throws the tests off. But you're right, the memory management probably is what makes ostringstream slower.
Max
Thanks for looking at the code. I had only looked at the pretty graphs.
Max Lybbert
Max, you're not correct about the snprintf() benchmarks. Check the code again
dcw
I didn't do the benchmarks, and I'm not able to update Matt Wilson's website or his code. I do stand by the original statement (that they aren't designed to be slower, but generally are implemented slower). If you don't like Matt Wilson's benchmarks, I'm sure a little Googling will find others.
Max Lybbert
Matt Wilson, by the way, wrote the STLSoft libraries, so I'm pretty willing to accept his benchmarks on name alone.
Max Lybbert
I think dcw meant the other Max. :)Anyway, FastFormat looks nice, and the std IOStreams library is hopelessly flawed. But the convoluted installation of the library makes it a bit of a nonstarter for me. If I have to mess with env variables, I'll find another library.
jalf
LOL. Thanks jalf. I had forgotten about this question; and re-reading I think you're right.
Max Lybbert
One reason is that string streams usually are thread safe, and snprintf is not, so there is some overhead there.
kotlinski
+1  A: 

Absolutely this is implementation-specific.

But if you really want to know, write two small programs, and compare them. You would need to include typical usage for what you have in mind, the two programs would need to generate the same string, and you would use a profiler to look at the timing information.

Then you would know.

Paul Beckingham
A: 

One reason I know that the printf family of functions are faster than the corresponding C++ functions (cout, cin, and other streams) is that the latter do typechecking. As this usually involves some requests to overloaded operators, it can take some time.

In fact, in programming competitions it is often recommended that you use printf et al rather than cout/cin for precisely this reason.

sykora
This kind of typechecking as well as operator overloads are resolved in compile time. The reduced runtime performance is not explained by them.
pyrtsa
In fact, the printf() family has to parse the format at runtime, which does not match the observation
MSalters
C++ typechecking isn't a runtime cost. Unless you're talking about *compilation* speed, in which case `<iostream>` is awful.
Tom
+3  A: 

Some guys would possibly tell you about that the functions can't be faster than each other, but their implementation can. That's right i think i would agree.

You are unlikely to ever notice a difference in other than benchmarks. The reason that c++ streams generally tend to be slower is that they are much more flexible. Flexibility most often comes at the cost of either time or code growth.

In this case, C++ streams are based on stream-buffers. In itself, streams are just the hull that keep formatting and error flags in place, and call the right i/o facets of the c++ standard library (for example, num_put to print numbers), that print the values, well formatted, into the underlying stream-buffer connected to the c++ stream.

All this mechanisms - the facets, and the buffers, are implemented by virtual functions. While there is indeed no mark note, those functions must be implemented to be slower than c stdio pendants that fact will make them somewhat slower than using c stdio functions normally (i benchmark'ed that some time ago with gcc/libstdc++ and in fact noticed a slowdown - but which you hardly notice in day-by-day usage).

Johannes Schaub - litb
+1  A: 

One issue would probably be that the type safety added by ostringstream carries extra overhead. I've not done any measurements, though.

Johann Gerell
+4  A: 

We replaced some stringstreams in inner loops with sprintf (using statically allocated buffers), and this made a big difference, both in msvc and gcc. I imagine that the dynamic memory management of this code:

{
  char buf[100];
  int i = 100;
  sprintf(buf, "%d", i);
  // do something with buf
}

is much simpler than

{
  std::stringstream ss;
  int i = 100;
  ss << i;
  std::string s = ss.str();
  // do something with s
}

but i am very happy with the overall performance of stringstreams.

+2  A: 

Using the printf family of functions explicitly forces you to manage the type coercion instead of relying on the type system to do that for you.

Martin Wickman
A: 

As litb said, standard streams support many things we don't always need. Some streams implementation get rid of this never used flexibility, see FAStream for instance.

Luc Hermitte
A: 

Yes, if you run the function below on a few million numbers with Visual C++ 5.0, the first version takes about twice as long as the second and produces the same output.

Compiling tight loops into a .exe and running the Windows timethis something.exe' or the Linux time something' is how I investigate most of my performance curiosities. (`timethis' is available on the web somewhere)

void Hex32Bit(unsigned int n, string &result)
{
#if 0
    stringstream ss;
    ss
        << hex
        << setfill('0')
        << "0x" << setw(8) << n
    ;
    result = ss.str();
#else
    const size_t len = 11;
    char temp[len];
    _snprintf(temp, len, "0x%08x", n);
    temp[len - 1] = '\0';
    result = temp;
#endif
}
Matthew