1 <section xmlns="http://docbook.org/ns/docbook" version="5.0"
2 xml:id="status.iso.1998" xreflabel="ISO C++ 1998">
3 <?dbhtml filename="status_iso_cxx1998.html"?>
5 <info><title>C++ 1998/2003</title>
17 <section xml:id="iso.1998.status" xreflabel="Implementation Status">
18 <info><title>Implementation Status</title></info>
21 This status table is based on the table of contents of ISO/IEC 14882:2003.
25 This page describes the C++ support in mainline GCC SVN, not in any
29 <!-- Status is Yes or No, Broken/Partial-->
34 <?dbhtml bgcolor="#C8B0B0" ?>
36 <?dbhtml bgcolor="#B0B0B0" ?>
39 <title>C++ 1998/2003 Implementation Status</title>
41 <tgroup cols="4" align="left" colsep="0" rowsep="1">
42 <colspec colname="c1"/>
43 <colspec colname="c2"/>
44 <colspec colname="c3"/>
45 <colspec colname="c4"/>
48 <entry>Section</entry>
49 <entry>Description</entry>
51 <entry>Comments</entry>
58 <emphasis>18</emphasis>
60 <entry namest="c2" nameend="c4" align="left">
61 <emphasis>Language support</emphasis>
73 <entry>Implementation properties</entry>
79 <entry>Numeric Limits</entry>
84 <entry>18.2.1.1</entry>
85 <entry>Class template <code>numeric_limits</code></entry>
90 <entry>18.2.1.2</entry>
91 <entry><code>numeric_limits</code> members</entry>
95 <entry>18.2.1.3</entry>
96 <entry><code>float_round_style</code></entry>
100 <entry>18.2.1.4</entry>
101 <entry><code>float_denorm_style</code></entry>
105 <entry>18.2.1.5</entry>
106 <entry><code>numeric_limits</code> specializations</entry>
111 <entry>18.2.2</entry>
112 <entry>C Library</entry>
118 <entry>Start and termination</entry>
124 <entry>Dynamic memory management</entry>
130 <entry>Type identification</entry>
135 <entry>18.5.1</entry>
136 <entry>Class type_info</entry>
141 <entry>18.5.2</entry>
142 <entry>Class bad_cast</entry>
147 <entry>18.5.3</entry>
148 <entry>Class bad_typeid</entry>
154 <entry>Exception handling</entry>
159 <entry>18.6.1</entry>
160 <entry>Class exception</entry>
165 <entry>18.6.2</entry>
166 <entry>Violation exception-specifications</entry>
171 <entry>18.6.3</entry>
172 <entry>Abnormal termination</entry>
177 <entry>18.6.4</entry>
178 <entry><code>uncaught_exception</code></entry>
184 <entry>Other runtime support</entry>
190 <emphasis>19</emphasis>
192 <entry namest="c2" nameend="c4" align="left">
193 <emphasis>Diagnostics</emphasis>
198 <entry>Exception classes</entry>
204 <entry>Assertions</entry>
210 <entry>Error numbers</entry>
216 <emphasis>20</emphasis>
218 <entry namest="c2" nameend="c4" align="left">
219 <emphasis>General utilities</emphasis>
224 <entry>Requirements</entry>
230 <entry>Utility components</entry>
235 <entry>20.2.1</entry>
236 <entry>Operators</entry>
241 <entry>20.2.2</entry>
242 <entry><code>pair</code></entry>
248 <entry>Function objects</entry>
253 <entry>20.3.1</entry>
259 <entry>20.3.2</entry>
260 <entry>Arithmetic operation</entry>
265 <entry>20.3.3</entry>
266 <entry>Comparisons</entry>
271 <entry>20.3.4</entry>
272 <entry>Logical operations</entry>
277 <entry>20.3.5</entry>
278 <entry>Negators</entry>
283 <entry>20.3.6</entry>
284 <entry>Binders</entry>
289 <entry>20.3.7</entry>
290 <entry>Adaptors for pointers to functions</entry>
295 <entry>20.3.8</entry>
296 <entry>Adaptors for pointers to members</entry>
302 <entry>Memory</entry>
307 <entry>20.4.1</entry>
308 <entry>The default allocator</entry>
313 <entry>20.4.2</entry>
314 <entry>Raw storage iterator</entry>
319 <entry>20.4.3</entry>
320 <entry>Temporary buffers</entry>
325 <entry>20.4.4</entry>
326 <entry>Specialized algorithms</entry>
331 <entry>20.4.4.1</entry>
332 <entry><code>uninitialized_copy</code></entry>
337 <entry>20.4.4.2</entry>
338 <entry><code>uninitialized_fill</code></entry>
343 <entry>20.4.4.3</entry>
344 <entry><code>uninitialized_fill_n</code></entry>
349 <entry>20.4.5</entry>
350 <entry>Class template <code>auto_ptr</code></entry>
355 <entry>20.4.6</entry>
356 <entry>C library</entry>
362 <emphasis>21</emphasis>
364 <entry namest="c2" nameend="c4" align="left">
365 <emphasis>Strings</emphasis>
370 <entry>Character traits</entry>
375 <entry>21.1.1</entry>
376 <entry>Character traits requirements</entry>
381 <entry>21.1.2</entry>
382 <entry>traits typedef</entry>
387 <entry>21.1.3</entry>
388 <entry><code>char_traits</code> specializations</entry>
393 <entry>21.1.3.1</entry>
394 <entry>struct <code>char_traits<char></code></entry>
399 <entry>21.1.3.2</entry>
400 <entry>struct <code>char_traits<wchar_t></code></entry>
406 <entry>String classes</entry>
412 <entry>Class template <code>basic_string</code></entry>
418 <entry>Null-terminated sequence utilities</entry>
420 <entry>C library dependency</entry>
424 <emphasis>22</emphasis>
426 <entry namest="c2" nameend="c4" align="left">
427 <emphasis>Localization</emphasis>
432 <entry>Locales</entry>
437 <entry>22.1.1</entry>
438 <entry>Class <code>locale</code></entry>
443 <entry>22.1.2</entry>
444 <entry><code>locale</code> globals</entry>
449 <entry>22.1.3</entry>
450 <entry>Convenience interfaces</entry>
455 <entry>22.1.3.1</entry>
456 <entry>Character classification</entry>
461 <entry>22.1.3.2</entry>
462 <entry>Character conversions</entry>
468 <entry>Standard locale categories</entry>
473 <entry>22.2.1</entry>
474 <entry><code>ctype</code></entry>
479 <entry>22.2.2</entry>
480 <entry>Numeric</entry>
485 <entry>22.2.2.1</entry>
486 <entry><code>num_get</code></entry>
491 <entry>22.2.2.2</entry>
492 <entry><code>num_put</code></entry>
497 <entry>22.2.3</entry>
498 <entry><code>num_punct</code></entry>
503 <entry>22.2.4</entry>
504 <entry><code>collate</code></entry>
509 <entry>22.2.5</entry>
515 <entry>22.2.5.1</entry>
516 <entry><code>time_get</code></entry>
521 <entry>22.2.5.2</entry>
522 <entry><code>time_get_byname</code></entry>
527 <entry>22.2.5.3</entry>
528 <entry><code>time_put</code></entry>
533 <entry>22.2.5.3</entry>
534 <entry><code>time_put_byname</code></entry>
539 <entry>22.2.6</entry>
540 <entry>Monetary</entry>
545 <entry>22.2.6.1</entry>
546 <entry><code>money_get</code></entry>
551 <entry>22.2.6.2</entry>
552 <entry><code>money_put</code></entry>
557 <entry>22.2.6.3</entry>
558 <entry><code>money_punct</code></entry>
563 <entry>22.2.6.4</entry>
564 <entry><code>money_punct_byname</code></entry>
569 <entry>22.2.7</entry>
570 <entry><code>messages</code></entry>
575 <entry>22.2.8</entry>
576 <entry>Program-defined facets</entry>
582 <entry>C Library Locales</entry>
588 <emphasis>23</emphasis>
590 <entry namest="c2" nameend="c4" align="left">
591 <emphasis>Containers</emphasis>
596 <entry>Container requirements</entry>
602 <entry>Sequence containers</entry>
607 <entry>23.2.1</entry>
608 <entry>Class template <code>deque</code></entry>
613 <entry>23.2.2</entry>
614 <entry>Class template <code>list</code></entry>
619 <entry>23.2.3</entry>
620 <entry>Adaptors</entry>
625 <entry>23.2.3.1</entry>
626 <entry>Class template <code>queue</code></entry>
631 <entry>23.2.3.2</entry>
632 <entry>Class template <code>priority_queue</code></entry>
637 <entry>23.2.3.3</entry>
638 <entry>Class template <code>stack</code></entry>
643 <entry>23.2.4</entry>
644 <entry>Class template <code>vector</code></entry>
649 <entry>23.2.5</entry>
650 <entry>Class <code>vector<bool></code></entry>
656 <entry>Associative containers</entry>
661 <entry>23.3.1</entry>
662 <entry>Class template <code>map</code></entry>
667 <entry>23.3.2</entry>
668 <entry>Class template <code>multimap</code></entry>
673 <entry>23.3.3</entry>
674 <entry>Class template <code>set</code></entry>
679 <entry>23.3.4</entry>
680 <entry>Class template <code>multiset</code></entry>
686 <emphasis>24</emphasis>
688 <entry namest="c2" nameend="c4" align="left">
689 <emphasis>Iterators</emphasis>
694 <entry>Requirements</entry>
700 <entry>Header <code><iterator></code> synopsis</entry>
706 <entry>Iterator primitives</entry>
712 <entry>Predefined iterators and Iterator adaptors</entry>
717 <entry>24.4.1</entry>
718 <entry>Reverse iterators</entry>
723 <entry>24.4.2</entry>
724 <entry>Insert iterators</entry>
730 <entry>Stream iterators</entry>
735 <entry>24.5.1</entry>
736 <entry>Class template <code>istream_iterator</code></entry>
741 <entry>24.5.2</entry>
742 <entry>Class template <code>ostream_iterator</code></entry>
747 <entry>24.5.3</entry>
748 <entry>Class template <code>istreambuf_iterator</code></entry>
753 <entry>24.5.4</entry>
754 <entry>Class template <code>ostreambuf_iterator</code></entry>
760 <emphasis>25</emphasis>
762 <entry namest="c2" nameend="c4" align="left">
763 <emphasis>Algorithms</emphasis>
768 <entry>Non-modifying sequence operations</entry>
774 <entry>Mutating sequence operations</entry>
780 <entry>Sorting and related operations</entry>
786 <entry>C library algorithms</entry>
792 <emphasis>26</emphasis>
794 <entry namest="c2" nameend="c4" align="left">
795 <emphasis>Numerics</emphasis>
800 <entry>Numeric type requirements</entry>
806 <entry>Complex numbers</entry>
812 <entry>Numeric arrays</entry>
817 <entry>26.3.1</entry>
818 <entry>Header <code><valarray></code> synopsis</entry>
823 <entry>26.3.2</entry>
824 <entry>Class template <code>valarray</code></entry>
829 <entry>26.3.3</entry>
830 <entry><code>valarray</code> non-member operations</entry>
835 <entry>26.3.4</entry>
836 <entry>Class <code>slice</code></entry>
841 <entry>26.3.5</entry>
842 <entry>Class template <code>slice_array</code></entry>
847 <entry>26.3.6</entry>
848 <entry>Class <code>gslice</code></entry>
853 <entry>26.3.7</entry>
854 <entry>Class template <code>gslice_array</code></entry>
859 <entry>26.3.8</entry>
860 <entry>Class template <code>mask_array</code></entry>
865 <entry>26.3.9</entry>
866 <entry>Class template <code>indirect_array</code></entry>
872 <entry>Generalized numeric operations</entry>
877 <entry>26.4.1</entry>
878 <entry><code>accumulate</code></entry>
883 <entry>26.4.2</entry>
884 <entry><code>inner_product</code></entry>
889 <entry>26.4.3</entry>
890 <entry><code>partial_sum</code></entry>
895 <entry>26.4.4</entry>
896 <entry><code>adjacent_difference</code></entry>
901 <entry>26.4.5</entry>
908 <entry>C Library</entry>
914 <emphasis>27</emphasis>
916 <entry namest="c2" nameend="c4" align="left">
917 <emphasis>Input/output</emphasis>
922 <entry>Requirements</entry>
928 <entry>Forward declarations</entry>
934 <entry>Standard iostream objects</entry>
939 <entry>27.3.1</entry>
940 <entry>Narrow stream objects</entry>
945 <entry>27.3.2</entry>
946 <entry>Wide stream objects</entry>
952 <entry>Iostreams base classes</entry>
958 <entry>Stream buffers</entry>
964 <entry>Formatting and manipulators</entry>
970 <entry>String-based streams</entry>
976 <entry>File-based streams</entry>
982 <emphasis>Appendix D</emphasis>
984 <entry namest="c2" nameend="c4" align="left">
985 <emphasis>Compatibility features</emphasis>
990 <entry>Increment operator with bool operand</entry>
996 <entry><code>static</code> keyword</entry>
1002 <entry>Access declarations</entry>
1008 <entry>Implicit conversion from const strings</entry>
1014 <entry>C standard library headers</entry>
1020 <entry>Old iostreams members</entry>
1026 <entry>char* streams</entry>
1036 <section xml:id="iso.1998.specific" xreflabel="Implementation Specific"><info><title>Implementation Specific Behavior</title></info>
1040 The ISO standard defines the following phrase:
1046 <code>[1.3.5] implementation-defined behavior</code>
1050 Behavior, for a well-formed program construct and correct data, that
1051 depends on the implementation <emphasis>and that each implementation
1052 shall document</emphasis>.
1059 We do so here, for the C++ library only. Behavior of the
1060 compiler, linker, runtime loader, and other elements of "the
1061 implementation" are documented elsewhere. Everything listed
1062 in Annex B, Implementation Qualities, are also part of the
1063 compiler, not the library.
1066 For each entry, we give the section number of the standard, when
1067 applicable. This list is probably incomplet and inkorrekt.
1070 <emphasis>[1.9]/11 #3</emphasis> If <code>isatty(3)</code> is true, then
1071 interactive stream support is implied.
1074 <emphasis>[17.4.4.5]</emphasis> Non-reentrant functions are probably best
1075 discussed in the various sections on multithreading (see above).
1077 <!-- [17.4.4.8]/3 says any function that doesn't have an exception-spec
1078 can throw whatever we want; see also its footnote. Let's list those
1079 in the sections where the function itself occurs.
1081 <para><emphasis>[18.1]/4</emphasis> The type of <code>NULL</code> is described
1082 <link linkend="std.support.types.null">here</link>.
1084 <para><emphasis>[18.3]/8</emphasis> Even though it's listed in the library
1085 sections, libstdc++ has zero control over what the cleanup code hands
1086 back to the runtime loader. Talk to the compiler people. :-)
1088 <para><emphasis>[18.4.2.1]/5</emphasis> (bad_alloc),
1089 <emphasis>[18.5.2]/5</emphasis> (bad_cast),
1090 <emphasis>[18.5.3]/5</emphasis> (bad_typeid),
1091 <emphasis>[18.6.1]/8</emphasis> (exception),
1092 <emphasis>[18.6.2.1]/5</emphasis> (bad_exception): The <code>what()</code>
1093 member function of class <code>std::exception</code>, and these other
1094 classes publicly derived from it, simply returns the name of the
1095 class. But they are the <emphasis>mangled</emphasis> names; you will need to call
1096 <code>c++filt</code> and pass the names as command-line parameters to
1097 demangle them, or call a
1098 <link linkend="manual.ext.demangle">runtime demangler function</link>.
1099 (The classes in <code><stdexcept></code> have constructors which
1100 require an argument to use later for <code>what()</code> calls, so the
1101 problem of <code>what()</code>'s value does not arise in most
1102 user-defined exceptions.)
1104 <para><emphasis>[18.5.1]/7</emphasis> The return value of
1105 <code>std::type_info::name()</code> is the mangled type name (see the
1106 previous entry for more).
1108 <para><emphasis>[20.1.5]/5</emphasis> <emphasis>"Implementors are encouraged to
1109 supply libraries that can accept allocators that encapsulate more
1110 general memory models and that support non-equal instances. In such
1111 implementations, any requirements imposed on allocators by containers
1112 beyond those requirements that appear in Table 32, and the semantics
1113 of containers and algorithms when allocator instances compare
1114 non-equal, are implementation-defined."</emphasis> As yet we don't
1115 have any allocators which compare non-equal, so we can't describe how
1118 <para><emphasis>[21.1.3.1]/3,4</emphasis>,
1119 <emphasis>[21.1.3.2]/2</emphasis>,
1120 <emphasis>[23.*]'s foo::iterator</emphasis>,
1121 <emphasis>[27.*]'s foo::*_type</emphasis>,
1122 <emphasis>others...</emphasis>
1123 Nope, these types are called implementation-defined because you
1124 shouldn't be taking advantage of their underlying types. Listing them
1125 here would defeat the purpose. :-)
1127 <para><emphasis>[21.1.3.1]/5</emphasis> I don't really know about
1128 the mbstate_t stuff... see
1129 the <link linkend="std.localization.facet.codecvt">chapter 22
1130 notes</link> for what does exist.
1132 <para><emphasis>[22.*]</emphasis> Anything and everything we have on locale
1133 implementation will be described
1134 <link linkend="std.localization.locales.locale">over here</link>.
1136 <para><emphasis>[26.2.8]/9</emphasis> I have no idea what
1137 <code>complex<T></code>'s pow(0,0) returns.
1139 <para><emphasis>[27.4.2.4]/2</emphasis> Calling
1140 <code>std::ios_base::sync_with_stdio</code> after I/O has already been
1141 performed on the standard stream objects will
1142 flush the buffers, and <!-- this line might go away -->
1143 destroy and recreate the underlying buffer instances. Whether or not
1144 the previously-written I/O is destroyed in this process depends mostly
1145 on the --enable-libio choice: for stdio, if the written data is
1146 already in the stdio buffer, the data may be completely safe!
1148 <para><emphasis>[27.6.1.1.2]</emphasis>,
1149 <emphasis>[27.6.2.3]</emphasis> The I/O sentry ctor and dtor can perform
1150 additional work than the minimum required. We are not currently taking
1151 advantage of this yet.
1153 <para><emphasis>[27.7.1.3]/16</emphasis>,
1154 <emphasis>[27.8.1.4]/10</emphasis>
1155 The effects of <code>pubsetbuf/setbuf</code> are described
1156 <link linkend="std.io">in this chapter</link>.
1158 <para><emphasis>[27.8.1.4]/16</emphasis> Calling <code>fstream::sync</code> when
1159 a get area exists will... whatever <code>fflush()</code> does, I think.