<?xml version="1.0" encoding="UTF-8" standalone="no"?>
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"><html xmlns="http://www.w3.org/1999/xhtml"><head><meta http-equiv="Content-Type" content="text/html; charset=UTF-8" /><title>Backwards Compatibility</title><meta name="generator" content="DocBook XSL Stylesheets Vsnapshot" /><meta name="keywords" content="ISO C++, backwards" /><meta name="keywords" content="ISO C++, library" /><meta name="keywords" content="ISO C++, runtime, library" /><link rel="home" href="../index.html" title="The GNU C++ Library" /><link rel="up" href="appendix_porting.html" title="Appendix B. Porting and Maintenance" /><link rel="prev" href="api.html" title="API Evolution and Deprecation History" /><link rel="next" href="appendix_free.html" title="Appendix C. Free Software Needs Free Documentation" /></head><body><div class="navheader"><table width="100%" summary="Navigation header"><tr><th colspan="3" align="center">Backwards Compatibility</th></tr><tr><td width="20%" align="left"><a accesskey="p" href="api.html">Prev</a> </td><th width="60%" align="center">Appendix B.
Porting and Maintenance
</th><td width="20%" align="right"> <a accesskey="n" href="appendix_free.html">Next</a></td></tr></table><hr /></div><div class="section"><div class="titlepage"><div><div><h2 class="title" style="clear: both"><a id="manual.appendix.porting.backwards"></a>Backwards Compatibility</h2></div></div></div><div class="section"><div class="titlepage"><div><div><h3 class="title"><a id="backwards.first"></a>First</h3></div></div></div><p>The first generation GNU C++ library was called libg++. It was a
separate GNU project, although reliably paired with GCC. Rumors imply
that it had a working relationship with at least two kinds of
dinosaur.
</p><p>Some background: libg++ was designed and created when there was no
ISO standard to provide guidance. Classes like linked lists are now
provided for by <code class="classname">list<T></code> and do not need to be
created by <code class="function">genclass</code>. (For that matter, templates exist
now and are well-supported, whereas genclass (mostly) predates them.)
</p><p>There are other classes in libg++ that are not specified in the
ISO Standard (e.g., statistical analysis). While there are a lot of
really useful things that are used by a lot of people, the Standards
Committee couldn't include everything, and so a lot of those
<span class="quote">“<span class="quote">obvious</span>”</span> classes didn't get included.
</p><p>Known Issues include many of the limitations of its immediate ancestor.</p><p>Portability notes and known implementation limitations are as follows.</p><div class="section"><div class="titlepage"><div><div><h4 class="title"><a id="backwards.first.ios_base"></a>No <code class="code">ios_base</code></h4></div></div></div><p> At least some older implementations don't have <code class="code">std::ios_base</code>, so you should use <code class="code">std::ios::badbit</code>, <code class="code">std::ios::failbit</code> and <code class="code">std::ios::eofbit</code> and <code class="code">std::ios::goodbit</code>.
</p></div><div class="section"><div class="titlepage"><div><div><h4 class="title"><a id="backwards.first.cout_cin"></a>No <code class="code">cout</code> in <code class="filename"><ostream.h></code>, no <code class="code">cin</code> in <code class="filename"><istream.h></code></h4></div></div></div><p>
In earlier versions of the standard,
<code class="filename"><fstream.h></code>,
<code class="filename"><ostream.h></code>
and <code class="filename"><istream.h></code>
used to define
<code class="code">cout</code>, <code class="code">cin</code> and so on. ISO C++ specifies that one needs to include
<code class="filename"><iostream></code>
explicitly to get the required definitions.
</p><p> Some include adjustment may be required.</p><p>This project is no longer maintained or supported, and the sources
archived. For the desperate,
the <a class="link" href="http://gcc.gnu.org/extensions.html" target="_top">GCC extensions
page</a> describes where to find the last libg++ source. The code is
considered replaced and rewritten.
</p></div></div><div class="section"><div class="titlepage"><div><div><h3 class="title"><a id="backwards.second"></a>Second</h3></div></div></div><p>
The second generation GNU C++ library was called libstdc++, or
libstdc++-v2. It spans the time between libg++ and pre-ISO C++
standardization and is usually associated with the following GCC
releases: egcs 1.x, gcc 2.95, and gcc 2.96.
</p><p>
The STL portions of this library are based on SGI/HP STL release 3.11.
</p><p>
This project is no longer maintained or supported, and the sources
archived. The code is considered replaced and rewritten.
</p><p>
Portability notes and known implementation limitations are as follows.
</p><div class="section"><div class="titlepage"><div><div><h4 class="title"><a id="backwards.second.std"></a>Namespace <code class="code">std::</code> not supported</h4></div></div></div><p>
Some care is required to support C++ compiler and or library
implementation that do not have the standard library in
<code class="code">namespace std</code>.
</p><p>
The following sections list some possible solutions to support compilers
that cannot ignore <code class="code">std::</code>-qualified names.
</p><p>
First, see if the compiler has a flag for this. Namespace
back-portability-issues are generally not a problem for g++
compilers that do not have libstdc++ in <code class="code">std::</code>, as the
compilers use <code class="option">-fno-honor-std</code> (ignore
<code class="code">std::</code>, <code class="code">:: = std::</code>) by default. That is,
the responsibility for enabling or disabling <code class="code">std::</code> is
on the user; the maintainer does not have to care about it. This
probably applies to some other compilers as well.
</p><p>
Second, experiment with a variety of pre-processor tricks.
</p><p>
By defining <code class="code">std</code> as a macro, fully-qualified namespace
calls become global. Volia.
</p><pre class="programlisting">
#ifdef WICKEDLY_OLD_COMPILER
# define std
#endif
</pre><p>
Thanks to Juergen Heinzl who posted this solution on gnu.gcc.help.
</p><p>
Another pre-processor based approach is to define a macro
<code class="code">NAMESPACE_STD</code>, which is defined to either
<span class="quote">“<span class="quote"> </span>”</span> or <span class="quote">“<span class="quote">std</span>”</span> based on a compile-type
test. On GNU systems, this can be done with autotools by means of
an autoconf test (see below) for <code class="code">HAVE_NAMESPACE_STD</code>,
then using that to set a value for the <code class="code">NAMESPACE_STD</code>
macro. At that point, one is able to use
<code class="code">NAMESPACE_STD::string</code>, which will evaluate to
<code class="code">std::string</code> or <code class="code">::string</code> (i.e., in the
global namespace on systems that do not put <code class="code">string</code> in
<code class="code">std::</code>).
</p><pre class="programlisting">
dnl @synopsis AC_CXX_NAMESPACE_STD
dnl
dnl If the compiler supports namespace std, define
dnl HAVE_NAMESPACE_STD.
dnl
dnl @category Cxx
dnl @author Todd Veldhuizen
dnl @author Luc Maisonobe <luc@spaceroots.org>
dnl @version 2004-02-04
dnl @license AllPermissive
AC_DEFUN([AC_CXX_NAMESPACE_STD], [
AC_CACHE_CHECK(if g++ supports namespace std,
ac_cv_cxx_have_std_namespace,
[AC_LANG_SAVE
AC_LANG_CPLUSPLUS
AC_TRY_COMPILE([#include <iostream>
std::istream& is = std::cin;],,
ac_cv_cxx_have_std_namespace=yes, ac_cv_cxx_have_std_namespace=no)
AC_LANG_RESTORE
])
if test "$ac_cv_cxx_have_std_namespace" = yes; then
AC_DEFINE(HAVE_NAMESPACE_STD,,[Define if g++ supports namespace std. ])
fi
])
</pre></div><div class="section"><div class="titlepage"><div><div><h4 class="title"><a id="backwards.second.iterators"></a>Illegal iterator usage</h4></div></div></div><p>
The following illustrate implementation-allowed illegal iterator
use, and then correct use.
</p><div class="itemizedlist"><ul class="itemizedlist" style="list-style-type: disc; "><li class="listitem"><p>
you cannot do <code class="code">ostream::operator<<(iterator)</code>
to print the address of the iterator => use
<code class="code">operator<< &*iterator</code> instead
</p></li><li class="listitem"><p>
you cannot clear an iterator's reference (<code class="code">iterator =
0</code>) => use <code class="code">iterator = iterator_type();</code>
</p></li><li class="listitem"><p>
<code class="code">if (iterator)</code> won't work any more => use
<code class="code">if (iterator != iterator_type())</code>
</p></li></ul></div></div><div class="section"><div class="titlepage"><div><div><h4 class="title"><a id="backwards.second.isspace"></a><code class="code">isspace</code> from <code class="filename"><cctype></code> is a macro
</h4></div></div></div><p>
Glibc 2.0.x and 2.1.x define <code class="filename"><ctype.h></code> functionality as macros
(isspace, isalpha etc.).
</p><p>
This implementations of libstdc++, however, keep these functions
as macros, and so it is not back-portable to use fully qualified
names. For example:
</p><pre class="programlisting">
#include <cctype>
int main() { std::isspace('X'); }
</pre><p>
Results in something like this:
</p><pre class="programlisting">
std:: (__ctype_b[(int) ( ( 'X' ) )] & (unsigned short int) _ISspace ) ;
</pre><p>
A solution is to modify a header-file so that the compiler tells
<code class="filename"><ctype.h></code> to define functions
instead of macros:
</p><pre class="programlisting">
// This keeps isalnum, et al from being propagated as macros.
#if __linux__
# define __NO_CTYPE 1
#endif
</pre><p>
Then, include <code class="filename"><ctype.h></code>
</p><p>
Another problem arises if you put a <code class="code">using namespace
std;</code> declaration at the top, and include
<code class="filename"><ctype.h></code>. This will
result in ambiguities between the definitions in the global namespace
(<code class="filename"><ctype.h></code>) and the
definitions in namespace <code class="code">std::</code>
(<code class="code"><cctype></code>).
</p></div><div class="section"><div class="titlepage"><div><div><h4 class="title"><a id="backwards.second.at"></a>No <code class="code">vector::at</code>, <code class="code">deque::at</code>, <code class="code">string::at</code></h4></div></div></div><p>
One solution is to add an autoconf-test for this:
</p><pre class="programlisting">
AC_MSG_CHECKING(for container::at)
AC_TRY_COMPILE(
[
#include <vector>
#include <deque>
#include <string>
using namespace std;
],
[
deque<int> test_deque(3);
test_deque.at(2);
vector<int> test_vector(2);
test_vector.at(1);
string test_string(<span class="quote">“<span class="quote">test_string</span>”</span>);
test_string.at(3);
],
[AC_MSG_RESULT(yes)
AC_DEFINE(HAVE_CONTAINER_AT)],
[AC_MSG_RESULT(no)])
</pre><p>
If you are using other (non-GNU) compilers it might be a good idea
to check for <code class="code">string::at</code> separately.
</p></div><div class="section"><div class="titlepage"><div><div><h4 class="title"><a id="backwards.second.eof"></a>No <code class="code">std::char_traits<char>::eof</code></h4></div></div></div><p>
Use some kind of autoconf test, plus this:
</p><pre class="programlisting">
#ifdef HAVE_CHAR_TRAITS
#define CPP_EOF std::char_traits<char>::eof()
#else
#define CPP_EOF EOF
#endif
</pre></div><div class="section"><div class="titlepage"><div><div><h4 class="title"><a id="backwards.second.stringclear"></a>No <code class="code">string::clear</code></h4></div></div></div><p>
There are two functions for deleting the contents of a string:
<code class="code">clear</code> and <code class="code">erase</code> (the latter returns the
string).
</p><pre class="programlisting">
void
clear() { _M_mutate(0, this->size(), 0); }
</pre><pre class="programlisting">
basic_string&
erase(size_type __pos = 0, size_type __n = npos)
{
return this->replace(_M_check(__pos), _M_fold(__pos, __n),
_M_data(), _M_data());
}
</pre><p>
Unfortunately, <code class="code">clear</code> is not implemented in this
version, so you should use <code class="code">erase</code> (which is probably
faster than <code class="code">operator=(charT*)</code>).
</p></div><div class="section"><div class="titlepage"><div><div><h4 class="title"><a id="backwards.second.ostreamform_istreamscan"></a>
Removal of <code class="code">ostream::form</code> and <code class="code">istream::scan</code>
extensions
</h4></div></div></div><p>
These are no longer supported. Please use stringstreams instead.
</p></div><div class="section"><div class="titlepage"><div><div><h4 class="title"><a id="backwards.second.stringstreams"></a>No <code class="code">basic_stringbuf</code>, <code class="code">basic_stringstream</code></h4></div></div></div><p>
Although the ISO standard <code class="code">i/ostringstream</code>-classes are
provided, (<code class="filename"><sstream></code>), for
compatibility with older implementations the pre-ISO
<code class="code">i/ostrstream</code> (<code class="filename"><strstream></code>) interface is also provided,
with these caveats:
</p><div class="itemizedlist"><ul class="itemizedlist" style="list-style-type: disc; "><li class="listitem"><p>
<code class="code">strstream</code> is considered to be deprecated
</p></li><li class="listitem"><p>
<code class="code">strstream</code> is limited to <code class="code">char</code>
</p></li><li class="listitem"><p>
with <code class="code">ostringstream</code> you don't have to take care of
terminating the string or freeing its memory
</p></li><li class="listitem"><p>
<code class="code">istringstream</code> can be re-filled (clear();
str(input);)
</p></li></ul></div><p>
You can then use output-stringstreams like this:
</p><pre class="programlisting">
#ifdef HAVE_SSTREAM
# include <sstream>
#else
# include <strstream>
#endif
#ifdef HAVE_SSTREAM
std::ostringstream oss;
#else
std::ostrstream oss;
#endif
oss << "Name=" << m_name << ", number=" << m_number << std::endl;
...
#ifndef HAVE_SSTREAM
oss << std::ends; // terminate the char*-string
#endif
// str() returns char* for ostrstream and a string for ostringstream
// this also causes ostrstream to think that the buffer's memory
// is yours
m_label.set_text(oss.str());
#ifndef HAVE_SSTREAM
// let the ostrstream take care of freeing the memory
oss.freeze(false);
#endif
</pre><p>
Input-stringstreams can be used similarly:
</p><pre class="programlisting">
std::string input;
...
#ifdef HAVE_SSTREAM
std::istringstream iss(input);
#else
std::istrstream iss(input.c_str());
#endif
int i;
iss >> i;
</pre><p> One (the only?) restriction is that an istrstream cannot be re-filled:
</p><pre class="programlisting">
std::istringstream iss(numerator);
iss >> m_num;
// this is not possible with istrstream
iss.clear();
iss.str(denominator);
iss >> m_den;
</pre><p>
If you don't care about speed, you can put these conversions in
a template-function:
</p><pre class="programlisting">
template <class X>
void fromString(const string& input, X& any)
{
#ifdef HAVE_SSTREAM
std::istringstream iss(input);
#else
std::istrstream iss(input.c_str());
#endif
X temp;
iss >> temp;
if (iss.fail())
throw runtime_error(..)
any = temp;
}
</pre><p>
Another example of using stringstreams is in <a class="link" href="strings.html#strings.string.shrink" title="Shrink to Fit">this howto</a>.
</p><p> There is additional information in the libstdc++-v2 info files, in
particular <span class="quote">“<span class="quote">info iostream</span>”</span>.
</p></div><div class="section"><div class="titlepage"><div><div><h4 class="title"><a id="backwards.second.wchar"></a>Little or no wide character support</h4></div></div></div><p>
Classes <code class="classname">wstring</code> and
<code class="classname">char_traits<wchar_t></code> are
not supported.
</p></div><div class="section"><div class="titlepage"><div><div><h4 class="title"><a id="backwards.second.iostream_templates"></a>No templatized iostreams</h4></div></div></div><p>
Classes <code class="classname">wfilebuf</code> and
<code class="classname">wstringstream</code> are not supported.
</p></div><div class="section"><div class="titlepage"><div><div><h4 class="title"><a id="backwards.second.thread_safety"></a>Thread safety issues</h4></div></div></div><p>
Earlier GCC releases had a somewhat different approach to
threading configuration and proper compilation. Before GCC 3.0,
configuration of the threading model was dictated by compiler
command-line options and macros (both of which were somewhat
thread-implementation and port-specific). There were no
guarantees related to being able to link code compiled with one
set of options and macro setting with another set.
</p><p>
For GCC 3.0, configuration of the threading model used with
libraries and user-code is performed when GCC is configured and
built using the --enable-threads and --disable-threads options.
The ABI is stable for symbol name-mangling and limited functional
compatibility exists between code compiled under different
threading models.
</p><p>
The libstdc++ library has been designed so that it can be used in
multithreaded applications (with libstdc++-v2 this was only true
of the STL parts.) The first problem is finding a
<span class="emphasis"><em>fast</em></span> method of implementation portable to
all platforms. Due to historical reasons, some of the library is
written against per-CPU-architecture spinlocks and other parts
against the gthr.h abstraction layer which is provided by gcc. A
minor problem that pops up every so often is different
interpretations of what "thread-safe" means for a
library (not a general program). We currently use the <a class="link" href="https://web.archive.org/web/20171225062613/http://www.sgi.com/tech/stl/thread_safety.html" target="_top">same
definition that SGI</a> uses for their STL subset. However,
the exception for read-only containers only applies to the STL
components. This definition is widely-used and something similar
will be used in the next version of the C++ standard library.
</p><p>
Here is a small link farm to threads (no pun) in the mail
archives that discuss the threading problem. Each link is to the
first relevant message in the thread; from there you can use
"Thread Next" to move down the thread. This farm is in
latest-to-oldest order.
</p><div class="itemizedlist"><ul class="itemizedlist" style="list-style-type: disc; "><li class="listitem"><p>
Our threading expert Loren gives a breakdown of <a class="link" href="http://gcc.gnu.org/ml/libstdc++/2001-10/msg00024.html" target="_top">the
six situations involving threads</a> for the 3.0
release series.
</p></li><li class="listitem"><p>
<a class="link" href="http://gcc.gnu.org/ml/libstdc++/2001-05/msg00384.html" target="_top">
This message</a> inspired a recent updating of issues with
threading and the SGI STL library. It also contains some
example POSIX-multithreaded STL code.
</p></li></ul></div><p>
(A large selection of links to older messages has been removed;
many of the messages from 1999 were lost in a disk crash, and the
few people with access to the backup tapes have been too swamped
with work to restore them. Many of the points have been
superseded anyhow.)
</p></div></div><div class="section"><div class="titlepage"><div><div><h3 class="title"><a id="backwards.third"></a>Third</h3></div></div></div><p> The third generation GNU C++ library is called libstdc++, or
libstdc++-v3.
</p><p>The subset commonly known as the Standard Template Library
(clauses 23 through 25, mostly) is adapted from the final release
of the SGI STL (version 3.3), with extensive changes.
</p><p>A more formal description of the V3 goals can be found in the
official <a class="link" href="source_design_notes.html" title="Design Notes">design document</a>.
</p><p>Portability notes and known implementation limitations are as follows.</p><div class="section"><div class="titlepage"><div><div><h4 class="title"><a id="backwards.third.headers"></a>Pre-ISO headers removed</h4></div></div></div><p> The pre-ISO C++ headers
(<code class="filename"><iostream.h></code>,
<code class="filename"><defalloc.h></code> etc.) are
not supported.
</p><p>For those of you new to ISO C++ (welcome, time travelers!), the
ancient pre-ISO headers have new names.
The C++ FAQ has a good explanation in <a class="link" href="https://isocpp.org/wiki/faq/coding-standards#std-headers" target="_top">What's
the difference between <xxx> and <xxx.h> headers?</a>.
</p><p>Porting between pre-ISO headers and ISO headers is simple: headers
like <code class="filename"><vector.h></code> can be replaced with <code class="filename"><vector></code> and a using
directive <code class="code">using namespace std;</code> can be put at the global
scope. This should be enough to get this code compiling, assuming the
other usage is correct.
</p></div><div class="section"><div class="titlepage"><div><div><h4 class="title"><a id="backwards.third.hash"></a>Extension headers hash_map, hash_set moved to ext or backwards</h4></div></div></div><p>At this time most of the features of the SGI STL extension have been
replaced by standardized libraries.
In particular, the <code class="classname">unordered_map</code> and
<code class="classname">unordered_set</code> containers of TR1 and C++ 2011
are suitable replacements for the non-standard
<code class="classname">hash_map</code> and <code class="classname">hash_set</code>
containers in the SGI STL.
</p><p> Header files <code class="filename"><hash_map></code> and <code class="filename"><hash_set></code> moved
to <code class="filename"><ext/hash_map></code> and <code class="filename"><ext/hash_set></code>,
respectively. At the same time, all types in these files are enclosed
in <code class="code">namespace __gnu_cxx</code>. Later versions deprecate
these files, and suggest using TR1's <code class="filename"><unordered_map></code>
and <code class="filename"><unordered_set></code> instead.
</p><p>The extensions are no longer in the global or <code class="code">std</code>
namespaces, instead they are declared in the <code class="code">__gnu_cxx</code>
namespace. For maximum portability, consider defining a namespace
alias to use to talk about extensions, e.g.:
</p><pre class="programlisting">
#ifdef __GNUC__
#if __GNUC__ < 3
#include <hash_map.h>
namespace extension { using ::hash_map; }; // inherit globals
#else
#include <backward/hash_map>
#if __GNUC__ == 3 && __GNUC_MINOR__ == 0
namespace extension = std; // GCC 3.0
#else
namespace extension = ::__gnu_cxx; // GCC 3.1 and later
#endif
#endif
#else // ... there are other compilers, right?
namespace extension = std;
#endif
extension::hash_map<int,int> my_map;
</pre><p>This is a bit cleaner than defining typedefs for all the
instantiations you might need.
</p><p>The following autoconf tests check for working HP/SGI hash containers.
</p><pre class="programlisting">
# AC_HEADER_EXT_HASH_MAP
AC_DEFUN([AC_HEADER_EXT_HASH_MAP], [
AC_CACHE_CHECK(for ext/hash_map,
ac_cv_cxx_ext_hash_map,
[AC_LANG_SAVE
AC_LANG_CPLUSPLUS
ac_save_CXXFLAGS="$CXXFLAGS"
CXXFLAGS="$CXXFLAGS -Werror"
AC_TRY_COMPILE([#include <ext/hash_map>], [using __gnu_cxx::hash_map;],
ac_cv_cxx_ext_hash_map=yes, ac_cv_cxx_ext_hash_map=no)
CXXFLAGS="$ac_save_CXXFLAGS"
AC_LANG_RESTORE
])
if test "$ac_cv_cxx_ext_hash_map" = yes; then
AC_DEFINE(HAVE_EXT_HASH_MAP,,[Define if ext/hash_map is present. ])
fi
])
</pre><pre class="programlisting">
# AC_HEADER_EXT_HASH_SET
AC_DEFUN([AC_HEADER_EXT_HASH_SET], [
AC_CACHE_CHECK(for ext/hash_set,
ac_cv_cxx_ext_hash_set,
[AC_LANG_SAVE
AC_LANG_CPLUSPLUS
ac_save_CXXFLAGS="$CXXFLAGS"
CXXFLAGS="$CXXFLAGS -Werror"
AC_TRY_COMPILE([#include <ext/hash_set>], [using __gnu_cxx::hash_set;],
ac_cv_cxx_ext_hash_set=yes, ac_cv_cxx_ext_hash_set=no)
CXXFLAGS="$ac_save_CXXFLAGS"
AC_LANG_RESTORE
])
if test "$ac_cv_cxx_ext_hash_set" = yes; then
AC_DEFINE(HAVE_EXT_HASH_SET,,[Define if ext/hash_set is present. ])
fi
])
</pre></div><div class="section"><div class="titlepage"><div><div><h4 class="title"><a id="backwards.third.nocreate_noreplace"></a>No <code class="code">ios::nocreate/ios::noreplace</code>.
</h4></div></div></div><p>Historically these flags were used with iostreams to control whether
new files are created or not when opening a file stream, similar to the
<code class="code">O_CREAT</code> and <code class="code">O_EXCL</code> flags for the
<code class="function">open(2)</code> system call. Because iostream modes correspond
to <code class="function">fopen(3)</code> modes these flags are not supported.
For input streams a new file will not be created anyway, so
<code class="code">ios::nocreate</code> is not needed.
For output streams, a new file will be created if it does not exist, which is
consistent with the behaviour of <code class="function">fopen</code>.
</p><p>When one of these flags is needed a possible alternative is to attempt
to open the file using <span class="type">std::ifstream</span> first to determine whether
the file already exists or not. This may not be reliable however, because
whether the file exists or not could change between opening the
<span class="type">std::istream</span> and re-opening with an output stream. If you need
to check for existence and open a file as a single operation then you will
need to use OS-specific facilities outside the C++ standard library, such
as <code class="function">open(2)</code>.
</p></div><div class="section"><div class="titlepage"><div><div><h4 class="title"><a id="backwards.third.streamattach"></a>
No <code class="code">stream::attach(int fd)</code>
</h4></div></div></div><p>
Phil Edwards writes: It was considered and rejected for the ISO
standard. Not all environments use file descriptors. Of those
that do, not all of them use integers to represent them.
</p><p>
For a portable solution (among systems which use
file descriptors), you need to implement a subclass of
<code class="code">std::streambuf</code> (or
<code class="code">std::basic_streambuf<..></code>) which opens a file
given a descriptor, and then pass an instance of this to the
stream-constructor.
</p><p>
An extension is available that implements this.
<code class="filename"><ext/stdio_filebuf.h></code>
contains a derived class called
<code class="classname">__gnu_cxx::stdio_filebuf</code>.
This class can be constructed from a C <code class="code">FILE*</code> or a file
descriptor, and provides the <code class="code">fd()</code> function.
</p><p>
For another example of this, refer to
<a class="link" href="http://www.josuttis.com/cppcode/fdstream.html" target="_top">fdstream example</a>
by Nicolai Josuttis.
</p></div><div class="section"><div class="titlepage"><div><div><h4 class="title"><a id="backwards.third.support_cxx98"></a>
Support for C++98 dialect.
</h4></div></div></div><p>Check for complete library coverage of the C++1998/2003 standard.
</p><pre class="programlisting">
# AC_HEADER_STDCXX_98
AC_DEFUN([AC_HEADER_STDCXX_98], [
AC_CACHE_CHECK(for ISO C++ 98 include files,
ac_cv_cxx_stdcxx_98,
[AC_LANG_SAVE
AC_LANG_CPLUSPLUS
AC_TRY_COMPILE([
#include <cassert>
#include <cctype>
#include <cerrno>
#include <cfloat>
#include <ciso646>
#include <climits>
#include <clocale>
#include <cmath>
#include <csetjmp>
#include <csignal>
#include <cstdarg>
#include <cstddef>
#include <cstdio>
#include <cstdlib>
#include <cstring>
#include <ctime>
#include <algorithm>
#include <bitset>
#include <complex>
#include <deque>
#include <exception>
#include <fstream>
#include <functional>
#include <iomanip>
#include <ios>
#include <iosfwd>
#include <iostream>
#include <istream>
#include <iterator>
#include <limits>
#include <list>
#include <locale>
#include <map>
#include <memory>
#include <new>
#include <numeric>
#include <ostream>
#include <queue>
#include <set>
#include <sstream>
#include <stack>
#include <stdexcept>
#include <streambuf>
#include <string>
#include <typeinfo>
#include <utility>
#include <valarray>
#include <vector>
],,
ac_cv_cxx_stdcxx_98=yes, ac_cv_cxx_stdcxx_98=no)
AC_LANG_RESTORE
])
if test "$ac_cv_cxx_stdcxx_98" = yes; then
AC_DEFINE(STDCXX_98_HEADERS,,[Define if ISO C++ 1998 header files are present. ])
fi
])
</pre></div><div class="section"><div class="titlepage"><div><div><h4 class="title"><a id="backwards.third.support_tr1"></a>
Support for C++TR1 dialect.
</h4></div></div></div><p>Check for library coverage of the TR1 standard.
</p><pre class="programlisting">
# AC_HEADER_STDCXX_TR1
AC_DEFUN([AC_HEADER_STDCXX_TR1], [
AC_CACHE_CHECK(for ISO C++ TR1 include files,
ac_cv_cxx_stdcxx_tr1,
[AC_LANG_SAVE
AC_LANG_CPLUSPLUS
AC_TRY_COMPILE([
#include <tr1/array>
#include <tr1/ccomplex>
#include <tr1/cctype>
#include <tr1/cfenv>
#include <tr1/cfloat>
#include <tr1/cinttypes>
#include <tr1/climits>
#include <tr1/cmath>
#include <tr1/complex>
#include <tr1/cstdarg>
#include <tr1/cstdbool>
#include <tr1/cstdint>
#include <tr1/cstdio>
#include <tr1/cstdlib>
#include <tr1/ctgmath>
#include <tr1/ctime>
#include <tr1/cwchar>
#include <tr1/cwctype>
#include <tr1/functional>
#include <tr1/memory>
#include <tr1/random>
#include <tr1/regex>
#include <tr1/tuple>
#include <tr1/type_traits>
#include <tr1/unordered_set>
#include <tr1/unordered_map>
#include <tr1/utility>
],,
ac_cv_cxx_stdcxx_tr1=yes, ac_cv_cxx_stdcxx_tr1=no)
AC_LANG_RESTORE
])
if test "$ac_cv_cxx_stdcxx_tr1" = yes; then
AC_DEFINE(STDCXX_TR1_HEADERS,,[Define if ISO C++ TR1 header files are present. ])
fi
])
</pre><p>An alternative is to check just for specific TR1 includes, such as <unordered_map> and <unordered_set>.
</p><pre class="programlisting">
# AC_HEADER_TR1_UNORDERED_MAP
AC_DEFUN([AC_HEADER_TR1_UNORDERED_MAP], [
AC_CACHE_CHECK(for tr1/unordered_map,
ac_cv_cxx_tr1_unordered_map,
[AC_LANG_SAVE
AC_LANG_CPLUSPLUS
AC_TRY_COMPILE([#include <tr1/unordered_map>], [using std::tr1::unordered_map;],
ac_cv_cxx_tr1_unordered_map=yes, ac_cv_cxx_tr1_unordered_map=no)
AC_LANG_RESTORE
])
if test "$ac_cv_cxx_tr1_unordered_map" = yes; then
AC_DEFINE(HAVE_TR1_UNORDERED_MAP,,[Define if tr1/unordered_map is present. ])
fi
])
</pre><pre class="programlisting">
# AC_HEADER_TR1_UNORDERED_SET
AC_DEFUN([AC_HEADER_TR1_UNORDERED_SET], [
AC_CACHE_CHECK(for tr1/unordered_set,
ac_cv_cxx_tr1_unordered_set,
[AC_LANG_SAVE
AC_LANG_CPLUSPLUS
AC_TRY_COMPILE([#include <tr1/unordered_set>], [using std::tr1::unordered_set;],
ac_cv_cxx_tr1_unordered_set=yes, ac_cv_cxx_tr1_unordered_set=no)
AC_LANG_RESTORE
])
if test "$ac_cv_cxx_tr1_unordered_set" = yes; then
AC_DEFINE(HAVE_TR1_UNORDERED_SET,,[Define if tr1/unordered_set is present. ])
fi
])
</pre></div><div class="section"><div class="titlepage"><div><div><h4 class="title"><a id="backwards.third.support_cxx11"></a>
Support for C++11 dialect.
</h4></div></div></div><p>Check for baseline language coverage in the compiler for the C++11 standard.
</p><pre class="programlisting">
# AC_COMPILE_STDCXX_11
AC_DEFUN([AC_COMPILE_STDCXX_11], [
AC_CACHE_CHECK(if g++ supports C++11 features without additional flags,
ac_cv_cxx_compile_cxx11_native,
[AC_LANG_SAVE
AC_LANG_CPLUSPLUS
AC_TRY_COMPILE([
template <typename T>
struct check final
{
static constexpr T value{ __cplusplus };
};
typedef check<check<bool>> right_angle_brackets;
int a;
decltype(a) b;
typedef check<int> check_type;
check_type c{};
check_type&& cr = static_cast<check_type&&>(c);
static_assert(check_type::value == 201103L, "C++11 compiler");],,
ac_cv_cxx_compile_cxx11_native=yes, ac_cv_cxx_compile_cxx11_native=no)
AC_LANG_RESTORE
])
AC_CACHE_CHECK(if g++ supports C++11 features with -std=c++11,
ac_cv_cxx_compile_cxx11_cxx,
[AC_LANG_SAVE
AC_LANG_CPLUSPLUS
ac_save_CXXFLAGS="$CXXFLAGS"
CXXFLAGS="$CXXFLAGS -std=c++11"
AC_TRY_COMPILE([
template <typename T>
struct check final
{
static constexpr T value{ __cplusplus };
};
typedef check<check<bool>> right_angle_brackets;
int a;
decltype(a) b;
typedef check<int> check_type;
check_type c{};
check_type&& cr = static_cast<check_type&&>(c);
static_assert(check_type::value == 201103L, "C++11 compiler");],,
ac_cv_cxx_compile_cxx11_cxx=yes, ac_cv_cxx_compile_cxx11_cxx=no)
CXXFLAGS="$ac_save_CXXFLAGS"
AC_LANG_RESTORE
])
AC_CACHE_CHECK(if g++ supports C++11 features with -std=gnu++11,
ac_cv_cxx_compile_cxx11_gxx,
[AC_LANG_SAVE
AC_LANG_CPLUSPLUS
ac_save_CXXFLAGS="$CXXFLAGS"
CXXFLAGS="$CXXFLAGS -std=gnu++11"
AC_TRY_COMPILE([
template <typename T>
struct check final
{
static constexpr T value{ __cplusplus };
};
typedef check<check<bool>> right_angle_brackets;
int a;
decltype(a) b;
typedef check<int> check_type;
check_type c{};
check_type&& cr = static_cast<check_type&&>(c);
static_assert(check_type::value == 201103L, "C++11 compiler");],,
ac_cv_cxx_compile_cxx11_gxx=yes, ac_cv_cxx_compile_cxx11_gxx=no)
CXXFLAGS="$ac_save_CXXFLAGS"
AC_LANG_RESTORE
])
if test "$ac_cv_cxx_compile_cxx11_native" = yes ||
test "$ac_cv_cxx_compile_cxx11_cxx" = yes ||
test "$ac_cv_cxx_compile_cxx11_gxx" = yes; then
AC_DEFINE(HAVE_STDCXX_11,,[Define if g++ supports C++11 features. ])
fi
])
</pre><p>Check for library coverage of the C++2011 standard.
(Some library headers are commented out in this check, they are
not currently provided by libstdc++).
</p><pre class="programlisting">
# AC_HEADER_STDCXX_11
AC_DEFUN([AC_HEADER_STDCXX_11], [
AC_CACHE_CHECK(for ISO C++11 include files,
ac_cv_cxx_stdcxx_11,
[AC_REQUIRE([AC_COMPILE_STDCXX_11])
AC_LANG_SAVE
AC_LANG_CPLUSPLUS
ac_save_CXXFLAGS="$CXXFLAGS"
CXXFLAGS="$CXXFLAGS -std=gnu++11"
AC_TRY_COMPILE([
#include <cassert>
#include <ccomplex>
#include <cctype>
#include <cerrno>
#include <cfenv>
#include <cfloat>
#include <cinttypes>
#include <ciso646>
#include <climits>
#include <clocale>
#include <cmath>
#include <csetjmp>
#include <csignal>
#include <cstdalign>
#include <cstdarg>
#include <cstdbool>
#include <cstddef>
#include <cstdint>
#include <cstdio>
#include <cstdlib>
#include <cstring>
#include <ctgmath>
#include <ctime>
// #include <cuchar>
#include <cwchar>
#include <cwctype>
#include <algorithm>
#include <array>
#include <atomic>
#include <bitset>
#include <chrono>
// #include <codecvt>
#include <complex>
#include <condition_variable>
#include <deque>
#include <exception>
#include <forward_list>
#include <fstream>
#include <functional>
#include <future>
#include <initializer_list>
#include <iomanip>
#include <ios>
#include <iosfwd>
#include <iostream>
#include <istream>
#include <iterator>
#include <limits>
#include <list>
#include <locale>
#include <map>
#include <memory>
#include <mutex>
#include <new>
#include <numeric>
#include <ostream>
#include <queue>
#include <random>
#include <ratio>
#include <regex>
#include <scoped_allocator>
#include <set>
#include <sstream>
#include <stack>
#include <stdexcept>
#include <streambuf>
#include <string>
#include <system_error>
#include <thread>
#include <tuple>
#include <typeindex>
#include <typeinfo>
#include <type_traits>
#include <unordered_map>
#include <unordered_set>
#include <utility>
#include <valarray>
#include <vector>
],,
ac_cv_cxx_stdcxx_11=yes, ac_cv_cxx_stdcxx_11=no)
AC_LANG_RESTORE
CXXFLAGS="$ac_save_CXXFLAGS"
])
if test "$ac_cv_cxx_stdcxx_11" = yes; then
AC_DEFINE(STDCXX_11_HEADERS,,[Define if ISO C++11 header files are present. ])
fi
])
</pre><p>As is the case for TR1 support, these autoconf macros can be made for a finer-grained, per-header-file check. For
<code class="filename"><unordered_map></code>
</p><pre class="programlisting">
# AC_HEADER_UNORDERED_MAP
AC_DEFUN([AC_HEADER_UNORDERED_MAP], [
AC_CACHE_CHECK(for unordered_map,
ac_cv_cxx_unordered_map,
[AC_REQUIRE([AC_COMPILE_STDCXX_11])
AC_LANG_SAVE
AC_LANG_CPLUSPLUS
ac_save_CXXFLAGS="$CXXFLAGS"
CXXFLAGS="$CXXFLAGS -std=gnu++11"
AC_TRY_COMPILE([#include <unordered_map>], [using std::unordered_map;],
ac_cv_cxx_unordered_map=yes, ac_cv_cxx_unordered_map=no)
CXXFLAGS="$ac_save_CXXFLAGS"
AC_LANG_RESTORE
])
if test "$ac_cv_cxx_unordered_map" = yes; then
AC_DEFINE(HAVE_UNORDERED_MAP,,[Define if unordered_map is present. ])
fi
])
</pre><pre class="programlisting">
# AC_HEADER_UNORDERED_SET
AC_DEFUN([AC_HEADER_UNORDERED_SET], [
AC_CACHE_CHECK(for unordered_set,
ac_cv_cxx_unordered_set,
[AC_REQUIRE([AC_COMPILE_STDCXX_11])
AC_LANG_SAVE
AC_LANG_CPLUSPLUS
ac_save_CXXFLAGS="$CXXFLAGS"
CXXFLAGS="$CXXFLAGS -std=gnu++11"
AC_TRY_COMPILE([#include <unordered_set>], [using std::unordered_set;],
ac_cv_cxx_unordered_set=yes, ac_cv_cxx_unordered_set=no)
CXXFLAGS="$ac_save_CXXFLAGS"
AC_LANG_RESTORE
])
if test "$ac_cv_cxx_unordered_set" = yes; then
AC_DEFINE(HAVE_UNORDERED_SET,,[Define if unordered_set is present. ])
fi
])
</pre><p>
Some C++11 features first appeared in GCC 4.3 and could be enabled by
<code class="option">-std=c++0x</code> and <code class="option">-std=gnu++0x</code> for GCC
releases which pre-date the 2011 standard. Those C++11 features and GCC's
support for them were still changing until the 2011 standard was finished,
but the autoconf checks above could be extended to test for incomplete
C++11 support with <code class="option">-std=c++0x</code> and
<code class="option">-std=gnu++0x</code>.
</p></div><div class="section"><div class="titlepage"><div><div><h4 class="title"><a id="backwards.third.iterator_type"></a>
<code class="code">Container::iterator_type</code> is not necessarily <code class="code">Container::value_type*</code>
</h4></div></div></div><p>
This is a change in behavior from older versions. Now, most
<span class="type">iterator_type</span> typedefs in container classes are POD
objects, not <span class="type">value_type</span> pointers.
</p></div></div><div class="bibliography"><div class="titlepage"><div><div><h3 class="title"><a id="backwards.biblio"></a>Bibliography</h3></div></div></div><div class="biblioentry"><a id="id-1.3.6.3.8.5.2"></a><p><span class="title"><em>
<a class="link" href="http://www.kegel.com/gcc/gcc4.html" target="_top">
Migrating to GCC 4.1
</a>
</em>. </span><span class="author"><span class="firstname">Dan</span> <span class="surname">Kegel</span>. </span></p></div><div class="biblioentry"><a id="id-1.3.6.3.8.5.3"></a><p><span class="title"><em>
<a class="link" href="https://lists.debian.org/debian-gcc/2006/03/msg00405.html" target="_top">
Building the Whole Debian Archive with GCC 4.1: A Summary
</a>
</em>. </span><span class="author"><span class="firstname">Martin</span> <span class="surname">Michlmayr</span>. </span></p></div><div class="biblioentry"><a id="id-1.3.6.3.8.5.4"></a><p><span class="title"><em>
<a class="link" href="http://annwm.lbl.gov/~leggett/Atlas/gcc-3.2.html" target="_top">
Migration guide for GCC-3.2
</a>
</em>. </span></p></div></div></div><div class="navfooter"><hr /><table width="100%" summary="Navigation footer"><tr><td width="40%" align="left"><a accesskey="p" href="api.html">Prev</a> </td><td width="20%" align="center"><a accesskey="u" href="appendix_porting.html">Up</a></td><td width="40%" align="right"> <a accesskey="n" href="appendix_free.html">Next</a></td></tr><tr><td width="40%" align="left" valign="top">API Evolution and Deprecation History </td><td width="20%" align="center"><a accesskey="h" href="../index.html">Home</a></td><td width="40%" align="right" valign="top"> Appendix C.
Free Software Needs Free Documentation
</td></tr></table></div></body></html>