Question about BOOST_FOREACH

[plexus]

I have declared this member in a class:

std::vector<std::pair<std::string, std::string> > m_foo;

but when I do this ...

BOOST_FOREACH(std::pair<std::string, std::string>& entry, m_foo){
	// do stuff
}

... the compiler complaints with a series of confusing messages: (I translated the first line, the rest seems to be generic syntax errors resulting from the first line)

1>.\fb_master.cpp(159) : warning C4002: too many actual parameters for macro 'BOOST_FOREACH'
1>.\fb_master.cpp(159) : error C2143: Syntaxfehler: Es fehlt ')' vor '>'
1>.\fb_master.cpp(159) : error C2059: Syntaxfehler: '>'
1>.\fb_master.cpp(159) : error C2059: Syntaxfehler: ')'
1>.\fb_master.cpp(159) : error C2143: Syntaxfehler: Es fehlt ';' vor '{'
1>.\fb_master.cpp(159) : error C2181: Ungültiges 'else' ohne zugehöriges 'if'
1>.\fb_master.cpp(159) : error C2065: 'entry': nichtdeklarierter Bezeichner
1>.\fb_master.cpp(159) : error C2275: 'std::string': Ungültige Verwendung dieses Typs als Ausdruck
1>.\fb_master.cpp(159) : error C2065: 'entry': nichtdeklarierter Bezeichner
1>.\fb_master.cpp(159) : error C2275: 'std::string': Ungültige Verwendung dieses Typs als Ausdruck
1>.\fb_master.cpp(159) : error C2065: 'entry': nichtdeklarierter Bezeichner
1>.\fb_master.cpp(159) : error C2275: 'std::string': Ungültige Verwendung dieses Typs als Ausdruck
1>.\fb_master.cpp(159) : error C2065: 'entry': nichtdeklarierter Bezeichner
1>.\fb_master.cpp(159) : error C2275: 'std::string': Ungültige Verwendung dieses Typs als Ausdruck
1>.\fb_master.cpp(159) : error C2143: Syntaxfehler: Es fehlt ')' vor '>'
1>.\fb_master.cpp(159) : error C2059: Syntaxfehler: '>'
1>.\fb_master.cpp(159) : error C2059: Syntaxfehler: ')'
1>.\fb_master.cpp(159) : error C2059: Syntaxfehler: 'else'
1>.\fb_master.cpp(159) : error C2143: Syntaxfehler: Es fehlt ')' vor '>'
1>.\fb_master.cpp(159) : error C4430: Fehlender Typspezifizierer - int wird angenommen. Hinweis: "default-int" wird von C++ nicht unterstützt.
1>.\fb_master.cpp(159) : error C2244: 'encode_type': Keine Übereinstimmung für Funktionsdefinition mit vorhandener Deklaration gefunden

But this works:

typedef std::pair<std::string, std::string> foo_t;

BOOST_FOREACH(foo_t& entry, m_foo){
	// do stuff
}

Unfortunatly I have this situation a lot with different combinations of types and I don't want to typedef all of them.
Or at least I wouldn't like it because it feels like a cheap workaround.

And no, I can't use auto. I have to use a C++98 compiler.

So, the question is: what am I doing wrong?

[LB]

Macros don't know what C or C++ is. From their perspective, that comma after the first std::string separates two macro parameters, not two template parameters. Typedef is the way to go.

[plexus]

Damn. The comma thing was my first guess, too.
All right, typedef it is. Thanks.

[Moschops]

Just for completion, for anyone who isn't stuck with an old compiler, we have ranged for loops and auto now.

#include <iostream>
#include <vector>
#include <string>

int main()
{
  std::vector<std::pair<std::string, std::string> > m_foo;

  m_foo.emplace_back(std::make_pair("Beans", "cats"));
  m_foo.emplace_back(std::make_pair("Kittens", "peanuts"));

  for (const auto& item : m_foo)
    {
      std::cout << item.first << " " << item.second << std::endl;
    }

}

[plexus]

@Moschops: As I already mentioned, I have to use a C++98 compiler.
I'm well aware of state-of-the-art C++.

[Moschops]

@plexus

As I already mentioned, "for anyone who isn't stuck with an old compiler".

[plexus]

@Moschops:

Come on, that's not nice.
You edit your post and then write a reply that makes me look like an idiot.

[Moschops]

The final edit to my post was eight minutes before you replied. If you were still working on a previous version, then I apologise; it seemed safe to assume that it hadn't taken you eight minutes to write that, but I can imagine circumstances in which you were working with an old iteration.

[plexus]

Ok, I see - my bad. I guess I got sidetracked while replying.

[MikeyBoy]

Moschops wrote:

it seemed safe to assume that it hadn't taken you eight minutes to write that, but I can imagine circumstances in which you were working with an old iteration.

I frequently find myself starting a reply, then having to leave it to do something more urgent, coming back to it later. 8 minutes is nothing; I don't think it's fair to assume that at all.