Files
iterator/doc/iterator_facade.html
Eric Niebler 5b9f37973b Merged revisions 42451-43517 via svnmerge from
https://svn.boost.org/svn/boost/trunk

................
  r42455 | johnmaddock | 2008-01-04 08:54:35 -0800 (Fri, 04 Jan 2008) | 1 line
  
  Fix typos.
................
  r42456 | eric_niebler | 2008-01-04 09:14:53 -0800 (Fri, 04 Jan 2008) | 1 line
  
  mark up borland and sun accumulators failures
................
  r42459 | andreas_huber69 | 2008-01-04 10:23:18 -0800 (Fri, 04 Jan 2008) | 1 line
  
  This should fix template parameter shadowing errors for gcc and Intel compilers.
................
  r42462 | bgubenko | 2008-01-04 15:55:43 -0800 (Fri, 04 Jan 2008) | 1 line
  
  mark Accumulators tests for gcc 4.2.1 affected by GCC Bugzilla Bug 33580
................
  r42471 | turkanis | 2008-01-04 20:51:49 -0800 (Fri, 04 Jan 2008) | 1 line
  
  merged changes from iostreams_dev, revisions 42441-42469; added 'std::' for Intel on Linux/Darwin; added STDCXX workaround for codecvt; fixed docs for invert.hpp
................
  r42473 | turkanis | 2008-01-04 23:35:44 -0800 (Fri, 04 Jan 2008) | 1 line
  
  new iostreams expected failures: stream_offset_64bit_test.cpp on Borland, and wide stream tests on gcc-3.4.2_hpux_pa_risc
................
  r42475 | andreas_huber69 | 2008-01-05 04:42:02 -0800 (Sat, 05 Jan 2008) | 1 line
  
  Added defaults for in_state_reaction template parameters and updated tests accordingly.
................
  r42476 | bemandawes | 2008-01-05 06:41:55 -0800 (Sat, 05 Jan 2008) | 1 line
  
  Disable Microsoft "secure" overloads in Dinkumware libraries since they cause compile errors with Intel versions 9 and 10
................
  r42478 | jurko | 2008-01-05 07:19:53 -0800 (Sat, 05 Jan 2008) | 1 line
  
  Typo corrections. Minor stylistic changes.
................
  r42479 | jurko | 2008-01-05 08:53:03 -0800 (Sat, 05 Jan 2008) | 1 line
  
  Updated Boost.Jam command line option texts. Minor stylistic changes.
................
  r42480 | jurko | 2008-01-05 08:55:36 -0800 (Sat, 05 Jan 2008) | 1 line
  
  Typo corrections. Minor stylistic changes.
................
  r42481 | jurko | 2008-01-05 08:56:42 -0800 (Sat, 05 Jan 2008) | 1 line
  
  Typo correction. Corrected an invalid command-line option name.
................
  r42482 | jurko | 2008-01-05 08:57:17 -0800 (Sat, 05 Jan 2008) | 1 line
  
  Corrected an invalid command-line option name reference.
................
  r42483 | jurko | 2008-01-05 08:58:01 -0800 (Sat, 05 Jan 2008) | 1 line
  
  Corrected an out-of-date comment listing all command-line options.
................
  r42484 | jurko | 2008-01-05 09:11:50 -0800 (Sat, 05 Jan 2008) | 1 line
  
  Corrected the used file suffix for the VERBATIM file type. Now the documentation is in sync with the 'customization' example. This also closes the Trac ticket 134. Minor stylistic changes.
................
  r42485 | jurko | 2008-01-05 09:46:45 -0800 (Sat, 05 Jan 2008) | 1 line
  
  Stylistic comment changes & typo corrections.
................
  r42486 | jurko | 2008-01-05 09:52:31 -0800 (Sat, 05 Jan 2008) | 1 line
  
  Stylistic comment changes & typo corrections in several files. Corrected an incorrect error message in boost-build/build/project.jam displayed when a --build-dir command-line option and a non top-level project build-dir attribute are specified.
................
  r42487 | jurko | 2008-01-05 09:54:02 -0800 (Sat, 05 Jan 2008) | 1 line
  
  Cleaned up some import rule calls.
................
  r42488 | jurko | 2008-01-05 10:02:23 -0800 (Sat, 05 Jan 2008) | 1 line
  
  Documentation wording cleaned up a bit.
................
  r42489 | jurko | 2008-01-05 10:13:10 -0800 (Sat, 05 Jan 2008) | 1 line
  
  Removed or simplified some import calls. Comment updates. Minor stylistic changes.
................
  r42490 | jurko | 2008-01-05 10:14:20 -0800 (Sat, 05 Jan 2008) | 2 lines
  
  Simplified the used make rules. Removed some dead code. Minor stylistic changes.
................
  r42492 | jurko | 2008-01-05 10:29:36 -0800 (Sat, 05 Jan 2008) | 1 line
  
  Removed trailing spaces and some empty lines.
................
  r42494 | jurko | 2008-01-05 12:02:24 -0800 (Sat, 05 Jan 2008) | 1 line
  
  Minor stylistic changes such as: comment typo corrections, wrapping lines to 80 characters, indentations, removing trailing spaces, etc.
................
  r42495 | jurko | 2008-01-05 12:06:15 -0800 (Sat, 05 Jan 2008) | 2 lines
  
  Cleaned up where the stage module is imported in tools/builtin.jam. Added a missing import in tools/stage.jam that causes errors with the previous fix. Minor stylistic changes in tools/stage.jam.
................
  r42496 | jurko | 2008-01-05 12:14:48 -0800 (Sat, 05 Jan 2008) | 1 line
  
  Comment cleanup.
................
  r42497 | jurko | 2008-01-05 12:37:44 -0800 (Sat, 05 Jan 2008) | 1 line
  
  Minor stylistic changes - cleaned up import calls, wrapped much text at 80 characters updated comments, typo corrections, removed trailing spaces, etc.
................
  r42498 | jurko | 2008-01-05 12:48:50 -0800 (Sat, 05 Jan 2008) | 1 line
  
  Renamed some Jamfile and project-root.jam references to Jamfile.jam and Jamroot.jam respectively.
................
  r42499 | jurko | 2008-01-05 13:37:15 -0800 (Sat, 05 Jan 2008) | 1 line
  
  Test code cleaned up a bit. No functional changes.
................
  r42502 | hkaiser | 2008-01-05 14:44:28 -0800 (Sat, 05 Jan 2008) | 1 line
  
  Wave: updated copyright messages to include the year 2008 (merged from release branch).
................
  r42504 | jurko | 2008-01-05 15:18:17 -0800 (Sat, 05 Jan 2008) | 1 line
  
  Refactored the test into two separate test functions. Made the 'directory names with dots' test use the expect_output_line() tool instead of a manual find, causing a better diagnostic to be displayed in case of a failure. Made one of the test Jam scripts more compact. Made the test use the Jamroot.jam file instead of project-root.jam.
................
  r42505 | bgubenko | 2008-01-05 15:35:05 -0800 (Sat, 05 Jan 2008) | 1 line
  
  mark up accumulators library test weighted_kurtosis for acc toolset
................
  r42506 | jurko | 2008-01-05 16:25:48 -0800 (Sat, 05 Jan 2008) | 1 line
  
  No functional changes but only stylistic changes such as: comment typo corrections, wrapping lines to 80 characters, indentations, removing trailing spaces, removing empty lines, made tests use Jamfile.jam and Jamroot.jam Boost Build script names, removed unnecessary module imports, etc.
................
  r42507 | jurko | 2008-01-05 16:53:09 -0800 (Sat, 05 Jan 2008) | 1 line
  
  No functional changes but only stylistic changes such as: comment typo corrections, wrapping lines to 80 characters, indentations, removing trailing spaces, removing empty lines, removed unnecessary module imports, etc.
................
  r42508 | jurko | 2008-01-05 22:15:39 -0800 (Sat, 05 Jan 2008) | 5 lines
  
  Refactored build-system.jam. Added many detailed comments.
  
  Test configuration module now gets loaded the same as all other configuration modules.
  
  In addition to being able to specify which user configuration to load, user may now also prevent loading the user configuration by specifying an empty file name for it.
................
  r42509 | jurko | 2008-01-05 22:50:56 -0800 (Sat, 05 Jan 2008) | 1 line
  
  Added a --test-config command-line option for specifying where the test configuration file should be loaded from instead of always being looked up in Boost Build's test folder. This allows non-test builds to not use test config when it exists on the system and different tests to use different test configurations when needed.
................
  r42510 | jurko | 2008-01-05 22:58:24 -0800 (Sat, 05 Jan 2008) | 1 line
  
  Specifying that a build uses test configuration no longer prevents toolset auto-configuration and therefore no longer causes --toolset options to be ignored.
................
  r42511 | jurko | 2008-01-05 23:13:28 -0800 (Sat, 05 Jan 2008) | 1 line
  
  Minor stylistic comment changes.
................
  r42512 | jurko | 2008-01-05 23:17:08 -0800 (Sat, 05 Jan 2008) | 1 line
  
  Added support for choosing the toolset and toolset version to be used 'by default' by Boost Build. Allows testing of default toolset functionality.
................
  r42513 | jurko | 2008-01-05 23:26:46 -0800 (Sat, 05 Jan 2008) | 1 line
  
  Upgraded the tool for testing that a certain line exists in the given output support so that it now also knows how to test that a certain line does not exist in the given output.
................
  r42514 | jurko | 2008-01-05 23:28:47 -0800 (Sat, 05 Jan 2008) | 1 line
  
  Added the --ignore-site-config option telling Boost.Build not to load the site configuration file.
................
  r42516 | jurko | 2008-01-06 00:16:34 -0800 (Sun, 06 Jan 2008) | 1 line
  
  Stylistic changes: typo corrections, comment alignments, output string updates, no functional changes, etc.
................
  r42527 | andreas_huber69 | 2008-01-06 05:49:31 -0800 (Sun, 06 Jan 2008) | 1 line
  
  Various doc updates.
................
  r42528 | danieljames | 2008-01-06 08:47:16 -0800 (Sun, 06 Jan 2008) | 2 lines
  
  Add Boost.Unordered and add to the documentation. Not fully integrated yet.
................
  r42529 | danieljames | 2008-01-06 08:48:36 -0800 (Sun, 06 Jan 2008) | 1 line
  
  Add more libraries that use Boost.Hash to its intro.
................
  r42530 | danieljames | 2008-01-06 08:49:11 -0800 (Sun, 06 Jan 2008) | 1 line
  
  Add Boost.Unordered to Boost.Hash's intro.
................
  r42531 | danieljames | 2008-01-06 08:59:18 -0800 (Sun, 06 Jan 2008) | 1 line
  
  Combine the 'container' and 'unordered' tests.
................
  r42532 | danieljames | 2008-01-06 08:59:49 -0800 (Sun, 06 Jan 2008) | 1 line
  
  Give the unordered exception test suite its own name.
................
  r42533 | danieljames | 2008-01-06 09:13:15 -0800 (Sun, 06 Jan 2008) | 1 line
  
  Add the contents of compile_tests.cpp to set_compile.cpp and map_compile.cpp
................
  r42534 | danieljames | 2008-01-06 09:16:51 -0800 (Sun, 06 Jan 2008) | 1 line
  
  Rename the test-suites to match other libraries' style.
................
  r42535 | danieljames | 2008-01-06 09:23:16 -0800 (Sun, 06 Jan 2008) | 2 lines
  
  Add forwarding html files for Boost.Unordered.
................
  r42536 | danieljames | 2008-01-06 09:40:32 -0800 (Sun, 06 Jan 2008) | 1 line
  
  Add Boost.Unordered to the regression tests.
................
  r42537 | eric_niebler | 2008-01-06 09:44:06 -0800 (Sun, 06 Jan 2008) | 1 line
  
  add myself as maintainer of accumulators
................
  r42538 | danieljames | 2008-01-06 09:45:18 -0800 (Sun, 06 Jan 2008) | 2 lines
  
  Add library identification.
................
  r42539 | danieljames | 2008-01-06 09:48:11 -0800 (Sun, 06 Jan 2008) | 2 lines
  
  Add the unordered library to the maintainers list.
................
  r42540 | danieljames | 2008-01-06 09:56:06 -0800 (Sun, 06 Jan 2008) | 2 lines
  
  Add unordered to the library list.
................
  r42544 | turkanis | 2008-01-06 10:32:05 -0800 (Sun, 06 Jan 2008) | 18 lines
  
  copy.hpp: 
  
      fix for Visual Age: std::min was passed arguments of different types; replaced std:: min with conditional
  
  detail/streambuf/indirect_streambuf.hpp: 
  
      removed trailing comma in enum definition
  
  test/combine_test.cpp:
  test/symmetric_filter_test.cpp:
  test/compose_test.cpp:
  test/close_test.cpp:
  test/invert_test.cpp:
  test/tee_test.cpp:
  test/restrict_test.cpp:
  test/Jamfile.v2:
  
      moved tests for close() into the test files for various adapters
................
  r42546 | jurko | 2008-01-06 12:06:33 -0800 (Sun, 06 Jan 2008) | 1 line
  
  Corrected a bug with the customized default toolset being read from the toolset module but defined in the build-system module. Effect was that customized default toolsets were getting ignored.
................
  r42547 | jurko | 2008-01-06 12:37:55 -0800 (Sun, 06 Jan 2008) | 1 line
  
  Simple code cleanup. Typo corrections.
................
  r42548 | jurko | 2008-01-06 12:43:25 -0800 (Sun, 06 Jan 2008) | 1 line
  
  Added support for tests configuring whether they want their Boost Build to ignore toolset requirements instead of always ignoring them. Minor stylistic changes.
................
  r42549 | jurko | 2008-01-06 12:56:20 -0800 (Sun, 06 Jan 2008) | 1 line
  
  Added new tests related to Boost Build's default toolset handling. They test that the correct default toolset gets use and that when it gets used that has no different status than any other explicitly specified toolset.
................
  r42563 | turkanis | 2008-01-06 17:26:52 -0800 (Sun, 06 Jan 2008) | 1 line
  
  marked restrict_test.cpp as failing on vacpp and stream_offset_64bit_test as failing on sun (iostreams)
................
  r42564 | turkanis | 2008-01-06 17:32:29 -0800 (Sun, 06 Jan 2008) | 1 line
  
  alphabetized the iostreams failures
................
  r42565 | turkanis | 2008-01-06 21:20:32 -0800 (Sun, 06 Jan 2008) | 5 lines
  
  merged changes from branches/iostreams_dev, revisions 42544-42544
  
  - added "slice" as an alias for "restrict", for platforms on which "restrict" is a keyword
  - attempted to configure file_descriptor for __IBMCPP__
  - added better error output to stream_offset_64bit_test.cpp
................
  r42566 | jurko | 2008-01-07 09:00:59 -0800 (Mon, 07 Jan 2008) | 1 line
  
  Corrected explicitly specified user-config file handling. Was not looking for the specified file in the regular path instead of the current folder and did not work with absolute paths.
................
  r42567 | jurko | 2008-01-07 10:15:13 -0800 (Mon, 07 Jan 2008) | 1 line
  
  Code cleanup. Converted to using True/False instead of 1/0. Changed to obey documented coding conventions regarding whitespace and function parameters. Renamed private member functions to use the __ name prefix.
................
  r42568 | jurko | 2008-01-07 10:38:28 -0800 (Mon, 07 Jan 2008) | 1 line
  
  Added a new configuration test making sure that the bug with not being able to process absolute user-config configuration file references never rears its ugly head again.
................
  r42570 | eric_niebler | 2008-01-07 11:06:31 -0800 (Mon, 07 Jan 2008) | 1 line
  
  s/order/tail/
................
  r42571 | eric_niebler | 2008-01-07 11:08:16 -0800 (Mon, 07 Jan 2008) | 1 line
  
  add missing includes, fix signed/unsigned warnings, clean-up trailing whitespace, fixes #1552
................
  r42572 | danieljames | 2008-01-07 11:40:32 -0800 (Mon, 07 Jan 2008) | 1 line
  
  Rename the exception tests so that they don't clash with the normal tests.
................
  r42573 | danieljames | 2008-01-07 11:41:05 -0800 (Mon, 07 Jan 2008) | 1 line
  
  Remove some development code.
................
  r42575 | danieljames | 2008-01-07 11:44:13 -0800 (Mon, 07 Jan 2008) | 3 lines
  
  Add missing 'use namespace std'. Which I should have done when I was told about
  them before. Sorry.
................
  r42576 | danieljames | 2008-01-07 11:46:27 -0800 (Mon, 07 Jan 2008) | 1 line
  
  Add support for multiple copyrights in the library info.
................
  r42578 | danieljames | 2008-01-07 11:51:02 -0800 (Mon, 07 Jan 2008) | 1 line
  
  Reneame the set and map compile tests so they'll be adjacent in the test results.
................
  r42580 | danieljames | 2008-01-07 12:06:15 -0800 (Mon, 07 Jan 2008) | 1 line
  
  Avoid some uses of an invalid pointer.
................
  r42581 | danieljames | 2008-01-07 12:07:12 -0800 (Mon, 07 Jan 2008) | 1 line
  
  Use the default location for the reference documentation.
................
  r42587 | danieljames | 2008-01-07 13:05:42 -0800 (Mon, 07 Jan 2008) | 5 lines
  
  Fix a bug which was causing the memory area stuff to fail.
  
  I should probably try to be less clever and use memory area's lower
  bounds as the key, and do the extra work required to get that working.
................
  r42588 | danieljames | 2008-01-07 13:07:43 -0800 (Mon, 07 Jan 2008) | 1 line
  
  Fix an off by one error.
................
  r42591 | danieljames | 2008-01-07 13:47:24 -0800 (Mon, 07 Jan 2008) | 1 line
  
  Merge in spell check.
................
  r42596 | turkanis | 2008-01-07 15:13:26 -0800 (Mon, 07 Jan 2008) | 5 lines
  
  merged changes from branches/iostreams_dev, revisions 42565-42595:
  
  - Simplified implementation with the help to the C-runtime function _get_osfhandle so that on Windows only a single HANDLE is stored and the POSIX-style implementation is never needed; added the handle_type on POSIX systems (typedef for int) and a function returning the underlying handle as an instance of handle_type
  - fixed the bug described in ticket Ticket #1551 (stream_buffer::seekpos ignores openmode parameter)
  - fixed test/operation_sequence_test.cpp file description
................
  r42598 | grafik | 2008-01-07 18:49:25 -0800 (Mon, 07 Jan 2008) | 1 line
  
  Mostly fix bad path calc for direct html from doxygen doc generation. (fixes #1562)
................
  r42600 | turkanis | 2008-01-07 19:49:23 -0800 (Mon, 07 Jan 2008) | 1 line
  
  merged changes from branches/iostreams_dev revisions 42595-42599; applied Dinkumware implementation of positioning functions to IBM Visual Age; simplified and corrected implementation
................
  r42602 | turkanis | 2008-01-07 20:20:38 -0800 (Mon, 07 Jan 2008) | 1 line
  
  removed unneeded headers, one of which (restrict.hpp) causes failures on IBM Visual Age
................
  r42603 | turkanis | 2008-01-07 20:52:54 -0800 (Mon, 07 Jan 2008) | 1 line
  
  removed spurious '.cpp' extension from test names (iostreams)
................
  r42612 | danieljames | 2008-01-08 05:59:01 -0800 (Tue, 08 Jan 2008) | 3 lines
  
  Merge in latest unordered developments (revisions 42607-42611).
................
  r42613 | jurko | 2008-01-08 06:40:24 -0800 (Tue, 08 Jan 2008) | 1 line
  
  Added a new test making sure properties conditioned on multiple different feature values are handled correctly. This especially includes testing the case when one of those features is <toolset> and the value given for it includes a toolset version.
................
  r42614 | jurko | 2008-01-08 07:21:15 -0800 (Tue, 08 Jan 2008) | 1 line
  
  Made some debugging messages more consistent.
................
  r42615 | eric_niebler | 2008-01-08 09:42:20 -0800 (Tue, 08 Jan 2008) | 1 line
  
  support for LaTeX formulas in Doxygen comments
................
  r42616 | eric_niebler | 2008-01-08 09:43:34 -0800 (Tue, 08 Jan 2008) | 1 line
  
  automatically generate png files from LaTeX formulas using doxygen
................
  r42617 | jurko | 2008-01-08 09:51:30 -0800 (Tue, 08 Jan 2008) | 1 line
  
  Comment correction.
................
  r42619 | danieljames | 2008-01-08 10:15:01 -0800 (Tue, 08 Jan 2008) | 1 line
  
  Merge: Another missing 'using namespace std'
................
  r42620 | eric_niebler | 2008-01-08 10:43:43 -0800 (Tue, 08 Jan 2008) | 1 line
  
  fix signed/unsigned warnings, clean up trailing whitespace
................
  r42624 | eric_niebler | 2008-01-08 13:40:52 -0800 (Tue, 08 Jan 2008) | 1 line
  
  disable iterator debugging for all msvc versions
................
  r42625 | turkanis | 2008-01-08 14:25:01 -0800 (Tue, 08 Jan 2008) | 1 line
  
  merged changes from branches/iostreams_dev, revisions 42602-42624: adding missing included to fix #1550; fix for positioning on IBM; fix for file_descriptor on POSIX; restructed the version range for the Borland workaround in large_file_test.cpp
................
  r42626 | jurko | 2008-01-08 17:09:26 -0800 (Tue, 08 Jan 2008) | 1 line
  
  Minor stylistic code indentation changes.
................
  r42627 | jurko | 2008-01-08 17:11:03 -0800 (Tue, 08 Jan 2008) | 1 line
  
  Tried trailing spaces. Minor comment typo corrections.
................
  r42628 | jurko | 2008-01-08 17:18:38 -0800 (Tue, 08 Jan 2008) | 1 line
  
  Remove trailing spaces. Removed an empty line.
................
  r42629 | jurko | 2008-01-08 20:05:29 -0800 (Tue, 08 Jan 2008) | 8 lines
  
  Updated the __ACTION_RULE__ to not return its action command output as a single string but instead split it into a list of output lines. This allows Jam code using this output to work correctly independently of what newline character combinations are in use. This was causing problems with Boost Build unit tests which can now be updated to pass.
  
  Consequences & checks:
    * Final __ACTION_RULE__ rule parameter has changed from output ? to output-lines *.
    * Updated corresponding Jam documentation.
    * Updated the all related Boost Build code.
    * No code on the Boost trunk uses this rule except for Boost Build itself.
................
  r42630 | jurko | 2008-01-08 20:15:35 -0800 (Tue, 08 Jan 2008) | 1 line
  
  Fixed a failing test by making it access action output using the __ACTION_RULE__ rule so it would not be affected by the level of debug output given by Boost Build/Jam.
................
  r42631 | jurko | 2008-01-08 20:30:25 -0800 (Tue, 08 Jan 2008) | 1 line
  
  Fixed a failing test by making it access its action results using the __ACTION_RULE__ rule. Now it works with the default debug level settings (i.e. no action output displayed).
................
  r42632 | grafik | 2008-01-08 20:31:58 -0800 (Tue, 08 Jan 2008) | 1 line
  
  Revert various changes that break backward compatibility, and also some minor edits.
................
  r42633 | grafik | 2008-01-08 20:34:12 -0800 (Tue, 08 Jan 2008) | 1 line
  
  Revert various changes that break backward compatibility, and also some minor edits.
................
  r42637 | jurko | 2008-01-09 06:36:37 -0800 (Wed, 09 Jan 2008) | 1 line
  
  Updated the test so it passes on Windows. Had to change it to enable action output logging and compensate for the fact that now action names are sent to the output as well. Minor stylistic changes.
................
  r42639 | jurko | 2008-01-09 09:03:45 -0800 (Wed, 09 Jan 2008) | 1 line
  
  Updated the test so it passes on Windows. Had to change it to enable action output logging and compensate for the fact that now action names are sent to the output as well. Minor stylistic changes.
................
  r42641 | eric_niebler | 2008-01-09 12:21:51 -0800 (Wed, 09 Jan 2008) | 1 line
  
  copy the png files to where fop will look for them when building pdf
................
  r42642 | eric_niebler | 2008-01-09 12:33:05 -0800 (Wed, 09 Jan 2008) | 1 line
  
  fix more signed/unsigned warnings
................
  r42644 | turkanis | 2008-01-09 15:13:57 -0800 (Wed, 09 Jan 2008) | 1 line
  
  marked stream_offset_64bit_test as an expected failure on vacpp (iostreams)
................
  r42645 | turkanis | 2008-01-09 15:16:17 -0800 (Wed, 09 Jan 2008) | 1 line
  
  replaced 2 occurrences of the identifier 'restrict' with BOOST_IOSTREAMS_RESTRICT
................
  r42646 | turkanis | 2008-01-09 19:27:51 -0800 (Wed, 09 Jan 2008) | 1 line
  
  added expected failures for Sun and IBM; removed a Borland intermittent failure; removed some obsolete toolsets (iostreams)
................
  r42647 | troyer | 2008-01-10 01:49:16 -0800 (Thu, 10 Jan 2008) | 1 line
  
  Optimizations for Boost.MPI
................
  r42648 | troyer | 2008-01-10 03:54:36 -0800 (Thu, 10 Jan 2008) | 1 line
  
  Undid backward-compatibility breaking change in pair serialization
................
  r42651 | johnmaddock | 2008-01-10 04:10:37 -0800 (Thu, 10 Jan 2008) | 1 line
  
  Apply patch from Issue #1187.
................
  r42657 | anthonyw | 2008-01-10 06:19:36 -0800 (Thu, 10 Jan 2008) | 1 line
  
  removed references to NULL
................
  r42658 | turkanis | 2008-01-10 10:50:19 -0800 (Thu, 10 Jan 2008) | 1 line
  
  marked up two moe expected failures for sun-5.7-5.8 (iostreams)
................
  r42664 | bgubenko | 2008-01-10 13:13:42 -0800 (Thu, 10 Jan 2008) | 1 line
  
  marked iostreams library compose_test for gcc-3.4.6_linux_ia64 : linking exceeds 10 min. limit
................
  r42665 | danieljames | 2008-01-10 14:25:35 -0800 (Thu, 10 Jan 2008) | 2 lines
  
  Initialise svnmerge for merging changes for unordered.
................
  r42666 | danieljames | 2008-01-10 14:30:46 -0800 (Thu, 10 Jan 2008) | 6 lines
  
  Merge latest unordered developments:
  
  Make simple_test test a little more.
  Use doubles for calculating max load factor.
  Some workarounds, mostly for Borland and running the tests.
................
  r42670 | turkanis | 2008-01-10 16:02:36 -0800 (Thu, 10 Jan 2008) | 1 line
  
  merged changes from branches/iostreams_dev, revisions 42645=42660; fix for Intel-darwin; removed dependence of file_descriptor_test and mapped_file_test on library boost_iostreams
................
  r42672 | turkanis | 2008-01-10 17:35:46 -0800 (Thu, 10 Jan 2008) | 1 line
  
  force static runtime-link for intel-darwin
................
  r42674 | johnmaddock | 2008-01-11 01:55:43 -0800 (Fri, 11 Jan 2008) | 1 line
  
  Fix bugs reported by Will Drewry: certain invalid regexes can cause the library to access invalid memory, changed to ensure that the correct exception is thrown long before this happens.
................
  r42676 | jurko | 2008-01-11 05:44:51 -0800 (Fri, 11 Jan 2008) | 1 line
  
  Removed the manual boost-build.jam file creation since that file already exists in the SVN repository and this only overwrote it with the same content minus the copyright notice. Removed a reference to the no longer existing boost_build_v2.html file.
................
  r42677 | johnmaddock | 2008-01-11 05:48:57 -0800 (Fri, 11 Jan 2008) | 1 line
  
  Update for Intel-10.1.
................
  r42684 | turkanis | 2008-01-11 11:12:22 -0800 (Fri, 11 Jan 2008) | 1 line
  
  another attempt to fix linking for compression tests on intel-darwin
................
  r42688 | turkanis | 2008-01-11 14:58:21 -0800 (Fri, 11 Jan 2008) | 1 line
  
  merged changes from iostreams_dev; fixes for IBM and more documentation
................
  r42689 | turkanis | 2008-01-11 18:45:55 -0800 (Fri, 11 Jan 2008) | 1 line
  
  more detailed note for vacpp (iostreams)
................
  r42691 | eric_niebler | 2008-01-11 22:43:28 -0800 (Fri, 11 Jan 2008) | 1 line
  
  new style transforms a-la proto v3
................
  r42693 | nesotto | 2008-01-12 04:38:57 -0800 (Sat, 12 Jan 2008) | 1 line
  
  doc fixes
................
  r42698 | danieljames | 2008-01-12 06:43:40 -0800 (Sat, 12 Jan 2008) | 9 lines
  
  Merge the latest unordered changes. These are concerned with getting the tests
  working on more compilers. The biggest change is that the exception tests have
  been changed to use a very simple exception testing mechanism on top of
  lightweight_test. This was because Boost.Test exception testing isn't working
  on several platforms. I'm trying to set this up so that I can use Boost.Test on
  compilers which it completely supports, and lightweight test on others.
  Boost.Test tests more than my simple exception testing code ever will so it's
  worth using where I can.
................
  r42700 | eric_niebler | 2008-01-12 09:09:17 -0800 (Sat, 12 Jan 2008) | 1 line
  
  missing includes
................
  r42701 | eric_niebler | 2008-01-12 09:40:40 -0800 (Sat, 12 Jan 2008) | 1 line
  
  more msvc-7.1-friendly default_context implementation, more missing headers
................
  r42706 | turkanis | 2008-01-12 12:19:42 -0800 (Sat, 12 Jan 2008) | 1 line
  
  corrected use of feature detection macros for AIX; simplified implementation slightly; updated docs
................
  r42707 | turkanis | 2008-01-12 12:20:35 -0800 (Sat, 12 Jan 2008) | 1 line
  
  removed intel-darwin sepcific code, since it didn't work
................
  r42708 | eric_niebler | 2008-01-12 13:19:45 -0800 (Sat, 12 Jan 2008) | 1 line
  
  port toy_spirit example to proto v3
................
  r42713 | turkanis | 2008-01-12 23:17:07 -0800 (Sat, 12 Jan 2008) | 1 line
  
  improved docs
................
  r42715 | nesotto | 2008-01-13 03:37:41 -0800 (Sun, 13 Jan 2008) | 1 line
  
  fixed #if to #ifdef
................
  r42722 | bgubenko | 2008-01-13 07:52:12 -0800 (Sun, 13 Jan 2008) | 1 line
  
  better note for iostreams library test compose_test on gcc-3.4.6_linux_ia64
................
  r42723 | bgubenko | 2008-01-13 08:12:37 -0800 (Sun, 13 Jan 2008) | 1 line
  
  marked up random library test random_test for gcc-3.4.6_linux_ia64
................
  r42724 | danieljames | 2008-01-13 08:19:26 -0800 (Sun, 13 Jan 2008) | 3 lines
  
  Merge in latest changes to Boost.Unordered. Some compiler workarounds and
  starting to clean up the tests a little.
................
  r42729 | eric_niebler | 2008-01-13 11:39:54 -0800 (Sun, 13 Jan 2008) | 1 line
  
  work around msvc-7.1 bug
................
  r42741 | eric_niebler | 2008-01-13 13:56:56 -0800 (Sun, 13 Jan 2008) | 1 line
  
  boost template instantiation depth on darwin
................
  r42745 | johnmaddock | 2008-01-14 01:46:12 -0800 (Mon, 14 Jan 2008) | 1 line
  
  Ooops, fix broken escape sequence.
................
  r42747 | johnmaddock | 2008-01-14 01:58:36 -0800 (Mon, 14 Jan 2008) | 1 line
  
  Fix documentation typos.
................
  r42750 | chris_kohlhoff | 2008-01-14 05:13:35 -0800 (Mon, 14 Jan 2008) | 2 lines
  
  Fix concept name in comment.
................
  r42751 | johnmaddock | 2008-01-14 05:17:09 -0800 (Mon, 14 Jan 2008) | 1 line
  
  Patch regex concept checks and TR1 library to work with VC9 + MS TR1 feature pack.
................
  r42752 | chris_kohlhoff | 2008-01-14 05:20:06 -0800 (Mon, 14 Jan 2008) | 2 lines
  
  Add missing broken pipe error.
................
  r42753 | chris_kohlhoff | 2008-01-14 05:21:37 -0800 (Mon, 14 Jan 2008) | 3 lines
  
  Don't include sys/time.h when compiling with aCC, as that header does not
  supply pselect(), which is needed for HP-UX/aCC to work correctly.
................
  r42754 | chris_kohlhoff | 2008-01-14 05:22:21 -0800 (Mon, 14 Jan 2008) | 2 lines
  
  Disable noisy and incorrect /Wp64 warnings generated by MSVC.
................
  r42755 | chris_kohlhoff | 2008-01-14 05:24:28 -0800 (Mon, 14 Jan 2008) | 3 lines
  
  Don't call epoll_wait/kevent if there are no old operations (where old means
  added prior to the last epoll_wait/kevent call) needing to be demultiplexed.
................
  r42756 | chris_kohlhoff | 2008-01-14 05:25:24 -0800 (Mon, 14 Jan 2008) | 2 lines
  
  Silence some integer truncation warnings.
................
  r42758 | chris_kohlhoff | 2008-01-14 05:27:52 -0800 (Mon, 14 Jan 2008) | 8 lines
  
  Silence some integer truncation warnings.
  
  Only perform the windows-bug workaround where we use a short timeout with
  GetQueuedCompletionStatus from one thread, i.e. the timer thread.
  
  Keep track of the number of OVERLAPPED-derived operations to ensure that
  they all get cleaned up when the io_service is destroyed.
................
  r42759 | chris_kohlhoff | 2008-01-14 05:29:08 -0800 (Mon, 14 Jan 2008) | 5 lines
  
  Check for truncation when converting buffer size from size_t to openssl's
  int argument.
  
  Try to fix possible thread-safety issues in SSL wrapper.
................
  r42766 | eric_niebler | 2008-01-14 08:49:32 -0800 (Mon, 14 Jan 2008) | 1 line
  
  register mpl::bool_ with typeof
................
  r42767 | dgregor | 2008-01-14 09:01:26 -0800 (Mon, 14 Jan 2008) | 1 line
  
  Improved suggestion for dealing with Qt MOC, from Niels Dekker
................
  r42771 | niels_dekker | 2008-01-14 10:17:30 -0800 (Mon, 14 Jan 2008) | 1 line
  
  Documented value_init workaround to compiler issues, added new introduction, updated to 2003 edition of C++ Standard -- reviewed by Fernando Cacciola
................
  r42773 | guwi17 | 2008-01-14 11:04:43 -0800 (Mon, 14 Jan 2008) | 2 lines
  
  - fixed typo
................
  r42776 | eric_niebler | 2008-01-14 12:26:58 -0800 (Mon, 14 Jan 2008) | 1 line
  
  add skip(), for specifying a skip regex
................
  r42778 | turkanis | 2008-01-14 12:47:17 -0800 (Mon, 14 Jan 2008) | 1 line
  
  overhaul of dual_use filters: close() is now called just once; suppressed Borland/Dinkumware warnings in mapped_file.hpp
................
  r42779 | niels_dekker | 2008-01-14 13:46:20 -0800 (Mon, 14 Jan 2008) | 1 line
  
  Minor "beautifications" of value_init documentation, inc. placing references in order of appearance
................
  r42780 | lbourdev | 2008-01-14 14:06:07 -0800 (Mon, 14 Jan 2008) | 3 lines
  
  Changed size_t to std::size_t
................
  r42781 | lbourdev | 2008-01-14 15:25:10 -0800 (Mon, 14 Jan 2008) | 3 lines
  
  GIL: Changing size_t to std::size_t
................
  r42788 | eric_niebler | 2008-01-14 22:46:39 -0800 (Mon, 14 Jan 2008) | 1 line
  
  code clean-up, begin updating the transform section in proto's docs
................
  r42789 | eric_niebler | 2008-01-14 23:46:51 -0800 (Mon, 14 Jan 2008) | 1 line
  
  try disabling iterator debugging for intel-win toolset
................
  r42797 | t_schwinger | 2008-01-15 11:46:10 -0800 (Tue, 15 Jan 2008) | 3 lines
  
  adds missing #include
................
  r42798 | niels_dekker | 2008-01-15 11:53:28 -0800 (Tue, 15 Jan 2008) | 1 line
  
  value_init doc + test: Added revision date.
................
  r42801 | jurko | 2008-01-15 13:13:52 -0800 (Tue, 15 Jan 2008) | 1 line
  
  Minor stylistic comment changes. Removed trailing spaces.
................
  r42804 | eric_niebler | 2008-01-15 14:06:51 -0800 (Tue, 15 Jan 2008) | 1 line
  
  document call<>, make<> and bind<>
................
  r42807 | turkanis | 2008-01-15 14:54:40 -0800 (Tue, 15 Jan 2008) | 1 line
  
  rewrote treatment of result_of in terms of a new (hopefully temporary) config macro BOOST_IOSTREAMS_NO_RESULT_OF; fixed docs
................
  r42811 | turkanis | 2008-01-15 17:14:04 -0800 (Tue, 15 Jan 2008) | 1 line
  
  reverted last change except for doc fixes; regression was result of test runner's local patch
................
  r42813 | turkanis | 2008-01-15 17:16:36 -0800 (Tue, 15 Jan 2008) | 1 line
  
  botched last commit
................
  r42814 | djenkins | 2008-01-15 22:39:34 -0800 (Tue, 15 Jan 2008) | 1 line
  
  missing include
................
  r42815 | niels_dekker | 2008-01-16 01:35:12 -0800 (Wed, 16 Jan 2008) | 1 line
  
  Added convenience class initialized_value, as announced at http://article.gmane.org/gmane.comp.lib.boost.devel/169833
................
  r42816 | niels_dekker | 2008-01-16 01:37:25 -0800 (Wed, 16 Jan 2008) | 1 line
  
  Added test and documentation for convenience class initialized_value, that was added with changeset [42815]
................
  r42817 | chris_kohlhoff | 2008-01-16 05:46:01 -0800 (Wed, 16 Jan 2008) | 2 lines
  
  Set the openssl callback function for getting a thread ID.
................
  r42818 | anthonyw | 2008-01-16 07:23:36 -0800 (Wed, 16 Jan 2008) | 1 line
  
  Provide tss_cleanup_implemented as a dummy function on Windows CE to allow tests to run
................
  r42821 | t_schwinger | 2008-01-16 11:16:37 -0800 (Wed, 16 Jan 2008) | 3 lines
  
  works around MSVC7.1 problems (hopefully)
................
  r42822 | t_schwinger | 2008-01-16 11:17:09 -0800 (Wed, 16 Jan 2008) | 3 lines
  
  attempts to fix Borland regressions
................
  r42823 | eric_niebler | 2008-01-16 11:24:33 -0800 (Wed, 16 Jan 2008) | 1 line
  
  document when<> and is_callable<>
................
  r42825 | turkanis | 2008-01-16 12:46:56 -0800 (Wed, 16 Jan 2008) | 1 line
  
  replaced __IBMCPP__ with _AIX
................
  r42827 | turkanis | 2008-01-16 16:50:14 -0800 (Wed, 16 Jan 2008) | 1 line
  
  added pgi to toolsets expected to fail seekawble_file_test (iostreams)
................
  r42836 | eric_niebler | 2008-01-17 14:47:54 -0800 (Thu, 17 Jan 2008) | 1 line
  
  stl_iterator does better error handling
................
  r42837 | jurko | 2008-01-17 17:14:17 -0800 (Thu, 17 Jan 2008) | 1 line
  
  Stylistic changes. Removed trailing spaces. Removed empty lines. Corrected comment typos and wording.
................
  r42839 | eric_niebler | 2008-01-17 23:56:31 -0800 (Thu, 17 Jan 2008) | 1 line
  
  tweaks for better doxygen-ated output
................
  r42840 | eric_niebler | 2008-01-17 23:56:59 -0800 (Thu, 17 Jan 2008) | 1 line
  
  updated reference section
................
  r42841 | eric_niebler | 2008-01-17 23:58:48 -0800 (Thu, 17 Jan 2008) | 1 line
  
  add back reference section, document user-defined transforms
................
  r42843 | t_schwinger | 2008-01-18 06:37:41 -0800 (Fri, 18 Jan 2008) | 3 lines
  
  makes member object support work with BCC
................
  r42851 | johnmaddock | 2008-01-18 08:56:57 -0800 (Fri, 18 Jan 2008) | 1 line
  
  Add needed <iostream> include.
................
  r42852 | johnmaddock | 2008-01-18 09:05:35 -0800 (Fri, 18 Jan 2008) | 1 line
  
  We don't have a tr1::hash functor if the std lib is the Apache version.
................
  r42853 | hkaiser | 2008-01-18 09:56:53 -0800 (Fri, 18 Jan 2008) | 1 line
  
  Wave: removed T_DEFINED token id from the library.
................
  r42855 | johnmaddock | 2008-01-18 10:18:17 -0800 (Fri, 18 Jan 2008) | 1 line
  
  Needs to #include <cstring> in order to use std::memset.
................
  r42856 | danieljames | 2008-01-18 11:35:55 -0800 (Fri, 18 Jan 2008) | 2 lines
  
  Merge in some changes to the unordered tests.
................
  r42857 | t_schwinger | 2008-01-18 12:05:56 -0800 (Fri, 18 Jan 2008) | 3 lines
  
  simplifies function_types markup
................
  r42858 | t_schwinger | 2008-01-18 12:52:06 -0800 (Fri, 18 Jan 2008) | 3 lines
  
  attempts to fix BCB 5.9 regression in synthesis/mem_func_ptr_cv_ptr_to_this test
................
  r42859 | t_schwinger | 2008-01-18 13:06:44 -0800 (Fri, 18 Jan 2008) | 3 lines
  
  attempts to fix pathscale failure 
................
  r42868 | niels_dekker | 2008-01-19 12:21:18 -0800 (Sat, 19 Jan 2008) | 1 line
  
  value_init_test now works around Borland 5.82 bug ("Error E2015: Ambiguity..." when using initialized_value), that is fixed with a newer compiler version
................
  r42869 | niels_dekker | 2008-01-19 12:52:04 -0800 (Sat, 19 Jan 2008) | 1 line
  
  Removed local named variable from     value_initialized::operator=, as Fernando Cacciola suggested me to avoid unnecessary named variables.
................
  r42873 | bemandawes | 2008-01-19 18:01:35 -0800 (Sat, 19 Jan 2008) | 1 line
  
  Remove extraneous defines since they are inherited from library build Jamfile
................
  r42877 | johnmaddock | 2008-01-20 01:42:35 -0800 (Sun, 20 Jan 2008) | 1 line
  
  Add include of <eh.h> for msvc.
................
  r42878 | igaztanaga | 2008-01-20 03:54:47 -0800 (Sun, 20 Jan 2008) | 5 lines
  
  Updated Interprocess and Intrusive:
  
  -> Added linear slist to intrusive
  -> Updated all allocators to version 2 allocators in Interprocess
  -> Optimized rbtree_best_fit size overhead to 1 std:size_t.
................
  r42881 | danieljames | 2008-01-20 09:37:21 -0800 (Sun, 20 Jan 2008) | 1 line
  
  Include <new> to get std::bad_alloc.
................
  r42882 | danieljames | 2008-01-20 10:55:57 -0800 (Sun, 20 Jan 2008) | 22 lines
  
  Merged revisions 42856-42881 via svnmerge from 
  https://svn.boost.org/svn/boost/branches/unordered/trunk
  
  ........
    r42880 | danieljames | 2008-01-20 16:10:43 +0000 (Sun, 20 Jan 2008) | 17 lines
    
    Simplify the tests a little:
    
    Add a parameter to random_values to control what sort of values it generates.
    This means that instead of using equivalent_object to test collisions (which
    was a total hack) we now just need another parameter.
    
    This requires some meta programming to act differently for maps and sets.
    Because of this pairs no longer need to be generated so remove the code for
    doing that (which doesn't work on some compilers).
    
    Remove the generator object, just call generate directly.
    
    Remove some of the tests using int containers, they didn't really add to
    anthing other than the compile time (some tests are timing out).
  ........
................
  r42884 | jurko | 2008-01-20 12:18:50 -0800 (Sun, 20 Jan 2008) | 5 lines
  
  Minor stylistic changes:
    * Removed trailing spaces.
    * Added a comment for code discovering the user's home-directories.
    * Removed a stale regex import.
    * Removed an old corpse 'identity' rule found inside the __test__ rule.
................
  r42890 | johnmaddock | 2008-01-21 01:41:17 -0800 (Mon, 21 Jan 2008) | 1 line
  
  Only disable wide character support for HP aCC: for gcc the logic is already taken care of in libstdcpp3.hpp.
................
  r42897 | hkaiser | 2008-01-21 08:13:31 -0800 (Mon, 21 Jan 2008) | 1 line
  
  Removed a duplicate entry.
................
  r42898 | rogeeff | 2008-01-21 09:02:53 -0800 (Mon, 21 Jan 2008) | 1 line
  
  changed output of booleans
................
  r42899 | johnmaddock | 2008-01-21 10:11:09 -0800 (Mon, 21 Jan 2008) | 1 line
  
  Updated type traits library so that everything compiles with -Wall -pedantic with GCC.
................
  r42904 | davedeakins | 2008-01-21 11:38:44 -0800 (Mon, 21 Jan 2008) | 1 line
  
  Don't include <eh.h> for WinCE (since WinCE does not have this header)
................
  r42906 | eric_niebler | 2008-01-21 12:39:35 -0800 (Mon, 21 Jan 2008) | 1 line
  
  minor clean-up
................
  r42909 | rogeeff | 2008-01-21 19:41:23 -0800 (Mon, 21 Jan 2008) | 1 line
  
  missing header
................
  r42911 | igaztanaga | 2008-01-22 08:49:22 -0800 (Tue, 22 Jan 2008) | 1 line
  
  Refactor some allocation code and fix instantiation problem in 64 bit platforms
................
  r42916 | eric_niebler | 2008-01-22 12:42:18 -0800 (Tue, 22 Jan 2008) | 1 line
  
  add concepts section to proto reference
................
  r42917 | hljin | 2008-01-22 14:10:48 -0800 (Tue, 22 Jan 2008) | 1 line
  
  GIL: fixed the problem with std::hex by adding #include <ios>
................
  r42918 | eric_niebler | 2008-01-22 18:23:15 -0800 (Tue, 22 Jan 2008) | 1 line
  
  proto works with boost 1.34.1
................
  r42929 | johnmaddock | 2008-01-23 08:08:44 -0800 (Wed, 23 Jan 2008) | 1 line
  
  Applies fix for issue #1598: added missing #include.
................
  r42931 | igaztanaga | 2008-01-23 11:34:39 -0800 (Wed, 23 Jan 2008) | 1 line
  
  Ticket #1593: [interprocess] 'streamoff' : is not a member of 'std'
................
  r42934 | andreas_huber69 | 2008-01-23 13:46:58 -0800 (Wed, 23 Jan 2008) | 1 line
  
  Fixes #1594
................
  r42935 | eric_niebler | 2008-01-23 13:57:47 -0800 (Wed, 23 Jan 2008) | 1 line
  
  fix dependency issue in Jamfile
................
  r42938 | jano_gaspar | 2008-01-23 15:04:57 -0800 (Wed, 23 Jan 2008) | 1 line
  
  circular_buffer: updated documentation
................
  r42939 | eric_niebler | 2008-01-23 15:25:24 -0800 (Wed, 23 Jan 2008) | 1 line
  
  fix quickbook scanner to recognize the [import ...] block
................
  r42943 | danieljames | 2008-01-23 15:39:59 -0800 (Wed, 23 Jan 2008) | 60 lines
  
  Merged revisions 42882-42941 via svnmerge from 
  https://svn.boost.org/svn/boost/branches/unordered/trunk
  
  ................
    r42887 | danieljames | 2008-01-20 21:32:04 +0000 (Sun, 20 Jan 2008) | 10 lines
    
    Merged revisions 42590-42664,42667-42697,42699-42723,42725-42855,42857-42881 via svnmerge from 
    https://svn.boost.org/svn/boost/trunk
    
    ........
      r42881 | danieljames | 2008-01-20 17:37:21 +0000 (Sun, 20 Jan 2008) | 1 line
      
      Include <new> to get std::bad_alloc.
    ........
  ................
    r42892 | danieljames | 2008-01-21 13:03:16 +0000 (Mon, 21 Jan 2008) | 1 line
    
    On some compilers the Rogue Wave/Apache stdcxx library doesn't have the normal std::distance, but instead has a variant that takes the result as the third parameter so it doesn't have to work out the type from the iterator.
  ................
    r42893 | danieljames | 2008-01-21 13:07:58 +0000 (Mon, 21 Jan 2008) | 1 line
    
    Fix a typo in the last commit.
  ................
    r42895 | danieljames | 2008-01-21 13:33:29 +0000 (Mon, 21 Jan 2008) | 1 line
    
    Remove tabs from the last checkin.
  ................
    r42896 | danieljames | 2008-01-21 15:51:40 +0000 (Mon, 21 Jan 2008) | 1 line
    
    Use Boost config to tell when we have a std::distance function. Also, no need for a macro.
  ................
    r42908 | danieljames | 2008-01-21 21:37:04 +0000 (Mon, 21 Jan 2008) | 1 line
    
    Use boost::long_long_type and boost::ulong_long_type.
  ................
    r42921 | danieljames | 2008-01-23 11:43:35 +0000 (Wed, 23 Jan 2008) | 1 line
    
    Remove some tabs.
  ................
    r42922 | danieljames | 2008-01-23 11:46:28 +0000 (Wed, 23 Jan 2008) | 2 lines
    
    Add missing include. Refs #1596
  ................
    r42923 | danieljames | 2008-01-23 11:52:47 +0000 (Wed, 23 Jan 2008) | 2 lines
    
    Always use void const* for the second parameter of allocate. Refs #1596.
  ................
    r42936 | danieljames | 2008-01-23 22:22:16 +0000 (Wed, 23 Jan 2008) | 1 line
    
    Use Boost style library name in the documentation.
  ................
    r42937 | danieljames | 2008-01-23 22:22:32 +0000 (Wed, 23 Jan 2008) | 1 line
    
    More tabs.
  ................
    r42941 | danieljames | 2008-01-23 23:35:01 +0000 (Wed, 23 Jan 2008) | 1 line
    
    Fix all the allocators.
  ................
................
  r42948 | turkanis | 2008-01-23 22:50:32 -0800 (Wed, 23 Jan 2008) | 1 line
  
  merged changes from iostreams_dev, revisions 42825-42947
................
  r42950 | t_schwinger | 2008-01-24 10:56:27 -0800 (Thu, 24 Jan 2008) | 3 lines
  
  correctsbroken compiler support for MPL
................
  r42951 | eric_niebler | 2008-01-24 13:06:23 -0800 (Thu, 24 Jan 2008) | 1 line
  
  peeker optimization looks inside independent sub-expressions
................
  r42952 | nesotto | 2008-01-24 14:22:35 -0800 (Thu, 24 Jan 2008) | 1 line
  
  test of output iterators
................
  r42953 | nesotto | 2008-01-24 14:26:36 -0800 (Thu, 24 Jan 2008) | 1 line
  
  output iterator test
................
  r42954 | nesotto | 2008-01-24 14:27:27 -0800 (Thu, 24 Jan 2008) | 1 line
  
  output iterators for ptr_containers 
................
  r42957 | t_schwinger | 2008-01-24 16:26:16 -0800 (Thu, 24 Jan 2008) | 3 lines
  
  simplifies preprocessing code
................
  r42958 | t_schwinger | 2008-01-24 16:28:15 -0800 (Thu, 24 Jan 2008) | 3 lines
  
  touched
................
  r42960 | noel_belcourt | 2008-01-24 20:41:16 -0800 (Thu, 24 Jan 2008) | 6 lines
  
  Changed the -soname and -shared options in intel-darwin.jam
  to use -dynamiclib and -install_name, as done in darwin.jam.
  Apparently the Intel compilers on the Mac support the same 
  options as gcc for setting the internal dynamic library name.
................
  r42963 | nesotto | 2008-01-24 23:52:14 -0800 (Thu, 24 Jan 2008) | 1 line
  
  renaming ...
................
  r42964 | nesotto | 2008-01-24 23:52:56 -0800 (Thu, 24 Jan 2008) | 1 line
  
  renaming
................
  r42965 | nesotto | 2008-01-24 23:54:28 -0800 (Thu, 24 Jan 2008) | 1 line
  
  renaming
................
  r42970 | turkanis | 2008-01-25 09:56:25 -0800 (Fri, 25 Jan 2008) | 1 line
  
  merged changes from iostreams_dev, revisions 42947-42962: fixed tickets 1003, 1139, 1140, 1149
................
  r42971 | noel_belcourt | 2008-01-25 11:52:47 -0800 (Fri, 25 Jan 2008) | 2 lines
  
  Fixed a typo to yesterdays patch.
................
  r42972 | dgregor | 2008-01-25 13:07:14 -0800 (Fri, 25 Jan 2008) | 2 lines
  
  Include <ios> to get std::boolalpha. Fixes #1586
................
  r42974 | igaztanaga | 2008-01-25 15:07:51 -0800 (Fri, 25 Jan 2008) | 4 lines
  
  1)Fixed gcc release mode warnings.
  2)Replaced throw with BOOST_RETHROW when BOOST_TRY is used.
  3)Fixed issues with singly linked lists
................
  r42976 | hkaiser | 2008-01-25 17:24:21 -0800 (Fri, 25 Jan 2008) | 2 lines
  
  Wave: Fixed a problem in flex_string::compare() (#include_next was non-functional).
................
  r42977 | hkaiser | 2008-01-25 17:36:20 -0800 (Fri, 25 Jan 2008) | 1 line
  
  Wave: Added new testcase.
................
  r42980 | hkaiser | 2008-01-25 17:44:32 -0800 (Fri, 25 Jan 2008) | 1 line
  
  Wave: Tweaked new testcase.
................
  r42982 | igaztanaga | 2008-01-26 03:52:25 -0800 (Sat, 26 Jan 2008) | 1 line
  
  Refactored common slist functions in a single class
................
  r42984 | noel_belcourt | 2008-01-26 10:35:59 -0800 (Sat, 26 Jan 2008) | 7 lines
  
  Fixes #416
  
  Fixed spelling of Jack Edmonds name and renamed files
  where necessary.  Updated the documentation as well.
  Tested changes by building/running tests in libs/graph/test.
................
  r42985 | noel_belcourt | 2008-01-26 10:51:28 -0800 (Sat, 26 Jan 2008) | 5 lines
  
  Fixes #640
  
  Corrected the mpl push_front html documentation.
................
  r42986 | eric_niebler | 2008-01-26 11:38:44 -0800 (Sat, 26 Jan 2008) | 1 line
  
  optimize repeated searches with patterns that have leading repeats
................
  r42987 | t_schwinger | 2008-01-26 13:50:14 -0800 (Sat, 26 Jan 2008) | 3 lines
  
  attempts to allow some preprocessing with VACPP (IBM)
................
  r42988 | noel_belcourt | 2008-01-26 14:21:57 -0800 (Sat, 26 Jan 2008) | 5 lines
  
  Fixes #1539
  
  Fixed typo in the random documentation.
................
  r42989 | noel_belcourt | 2008-01-26 15:06:24 -0800 (Sat, 26 Jan 2008) | 6 lines
  
  Fixes #965
  
  Patched the XML and will check to ensure the html
  page reflects this change.
................
  r42990 | eric_niebler | 2008-01-26 21:56:46 -0800 (Sat, 26 Jan 2008) | 1 line
  
  updated vcproj
................
  r42991 | eric_niebler | 2008-01-26 21:57:08 -0800 (Sat, 26 Jan 2008) | 1 line
  
  fix typo
................
  r42992 | johnmaddock | 2008-01-27 10:43:35 -0800 (Sun, 27 Jan 2008) | 1 line
  
  Extended leading repeat optimization to more cases.
................
  r42997 | vladimir_prus | 2008-01-28 09:59:27 -0800 (Mon, 28 Jan 2008) | 1 line
  
  Correct speliing of --build-dir in --help output
................
  r43000 | eric_niebler | 2008-01-28 12:03:41 -0800 (Mon, 28 Jan 2008) | 1 line
  
  update acknowledgement of john maddock
................
  r43001 | bgubenko | 2008-01-28 13:27:13 -0800 (Mon, 28 Jan 2008) | 1 line
  
  marked 2 asio library tests for gcc-4.2.1_hpux_ia64 (HP-UX 11.23 with gcc)
................
  r43002 | eric_niebler | 2008-01-28 14:55:30 -0800 (Mon, 28 Jan 2008) | 1 line
  
  doc more concepts, misc clean-up
................
  r43003 | eric_niebler | 2008-01-28 14:56:46 -0800 (Mon, 28 Jan 2008) | 1 line
  
  proto doxygen comments, misc clean-up
................
  r43006 | eric_niebler | 2008-01-28 18:20:45 -0800 (Mon, 28 Jan 2008) | 1 line
  
  more proto doxygen comments, update copyright
................
  r43007 | vladimir_prus | 2008-01-28 22:28:09 -0800 (Mon, 28 Jan 2008) | 1 line
  
  Retain top-level boost-build.jam
................
  r43008 | vladimir_prus | 2008-01-28 22:40:06 -0800 (Mon, 28 Jan 2008) | 4 lines
  
  Disable relinking when <target-os> is either windows or cygwin.
  
  Fixes #1062.
................
  r43009 | eric_niebler | 2008-01-28 23:03:03 -0800 (Mon, 28 Jan 2008) | 1 line
  
  add tests for deep_copy, make_expr, unpack_expr; fix bugs; update more copyrights
................
  r43012 | djenkins | 2008-01-29 08:41:12 -0800 (Tue, 29 Jan 2008) | 1 line
  
  fix typo
................
  r43013 | djenkins | 2008-01-29 08:43:51 -0800 (Tue, 29 Jan 2008) | 1 line
  
  update copyright and misc cleanup
................
  r43014 | bgubenko | 2008-01-29 09:47:01 -0800 (Tue, 29 Jan 2008) | 1 line
  
  marked interprocess library unusable on gcc-4.2.1_hpux_ia64 (until it is ported to HP-UX platform)
................
  r43016 | eric_niebler | 2008-01-29 13:02:52 -0800 (Tue, 29 Jan 2008) | 1 line
  
  make_expr and unpack_expr improvements, fix scary transform::arg_c bug
................
  r43018 | djenkins | 2008-01-29 19:39:02 -0800 (Tue, 29 Jan 2008) | 1 line
  
  use skip directive to simplify example
................
  r43023 | eric_niebler | 2008-01-30 14:10:13 -0800 (Wed, 30 Jan 2008) | 1 line
  
  finally, a make_expr() I can live with
................
  r43024 | eric_niebler | 2008-01-30 14:26:34 -0800 (Wed, 30 Jan 2008) | 1 line
  
  regenerated boostbook reference
................
  r43025 | niels_dekker | 2008-01-30 14:42:23 -0800 (Wed, 30 Jan 2008) | 1 line
  
  value_init: Removed aligned_storage::address() calls, to improve TR1 compatibility, as confirmed by John Maddock. Added internal helper function, wrapper_address(), as discussed with Fernando.
................
  r43026 | eric_niebler | 2008-01-30 15:03:36 -0800 (Wed, 30 Jan 2008) | 1 line
  
  minor tweak to make_expr result_of return type calculation
................
  r43031 | eric_niebler | 2008-01-30 23:36:28 -0800 (Wed, 30 Jan 2008) | 1 line
  
  simplify make_expr.hpp, user docs for make_expr()
................
  r43035 | eric_niebler | 2008-01-31 10:44:17 -0800 (Thu, 31 Jan 2008) | 1 line
  
  minor tweak to fusion value_of and value_at for expressions, for better interop with proto::unpack_expr
................
  r43037 | vladimir_prus | 2008-01-31 11:47:12 -0800 (Thu, 31 Jan 2008) | 1 line
  
  Build in MT mode (as long as wave links to boost.thread)
................
  r43038 | hkaiser | 2008-01-31 12:57:47 -0800 (Thu, 31 Jan 2008) | 1 line
  
  Wave: fixed expanding_function_like_macro()
................
  r43040 | eric_niebler | 2008-01-31 13:12:44 -0800 (Thu, 31 Jan 2008) | 1 line
  
  finish documentation for expression construction utilities
................
  r43041 | hkaiser | 2008-01-31 14:33:43 -0800 (Thu, 31 Jan 2008) | 1 line
  
  Wave: Added additional configuration possibility to allow control threading support.
................
  r43042 | hkaiser | 2008-01-31 14:48:56 -0800 (Thu, 31 Jan 2008) | 1 line
  
  Wave: Added additional configuration possibility to allow control threading support. Updated the documentation.
................
  r43043 | jurko | 2008-01-31 16:27:31 -0800 (Thu, 31 Jan 2008) | 1 line
  
  Reverted changes made in rev 43038 which seem to have been committed by mistake and include some user specific settings in it local to the comitter's environment while this file is intended to be used as generic template for actual user-config.jam files and do nothing in case user does not specify his own settings there.
................
  r43044 | jurko | 2008-01-31 16:44:23 -0800 (Thu, 31 Jan 2008) | 1 line
  
  Minor stylistic spacing changes. Remove trailing spaces.
................
  r43045 | jurko | 2008-01-31 16:46:50 -0800 (Thu, 31 Jan 2008) | 1 line
  
  Corrected outputting native Windows paths so that it works correctly for absolute paths without the drive letter being explicitly specified, e.g. \aaa\bbb or /aaa/bbb.
................
  r43046 | jurko | 2008-01-31 17:49:16 -0800 (Thu, 31 Jan 2008) | 1 line
  
  Added the missing end-of-line character when outputting DEBUG_SEARCH debug messages from file_build1(). This cleans up the -d+6 bjam output a lot.
................
  r43050 | eric_niebler | 2008-02-01 12:30:29 -0800 (Fri, 01 Feb 2008) | 1 line
  
  add future group example
................
  r43052 | noel_belcourt | 2008-02-01 18:41:23 -0800 (Fri, 01 Feb 2008) | 4 lines
  
  Fix a typo in pgi.jam that prevented shared libraries
  from being built correctly.
................
  r43054 | chris_kohlhoff | 2008-02-02 03:37:45 -0800 (Sat, 02 Feb 2008) | 4 lines
  
  Ensure that the workaround for the MSVC secure iterator problem is only
  used when compiling with MSVC. The workaround causes g++'s library debug
  mode to report errors due to the assignment from a singular iterator.
................
  r43055 | chris_kohlhoff | 2008-02-02 03:39:17 -0800 (Sat, 02 Feb 2008) | 2 lines
  
  Fix "possible loss of data" warning when building for Windows 2000 targets.
................
  r43056 | chris_kohlhoff | 2008-02-02 04:02:23 -0800 (Sat, 02 Feb 2008) | 3 lines
  
  The latest Windows SDKs don't support IPv6 when building for Windows 2000,
  so we need to use the SDK emulation in that case.
................
  r43057 | eric_niebler | 2008-02-02 04:27:16 -0800 (Sat, 02 Feb 2008) | 1 line
  
  port test to boost version 1.34.1
................
  r43061 | turkanis | 2008-02-02 14:10:46 -0800 (Sat, 02 Feb 2008) | 1 line
  
  merged changes from iostreams_dev, revisions 42962-43059: updated copyright notices
................
  r43080 | eric_niebler | 2008-02-03 10:40:03 -0800 (Sun, 03 Feb 2008) | 1 line
  
  fix bug found by L. Evans re: fusion and stateful function objects
................
  r43083 | johnmaddock | 2008-02-04 01:13:36 -0800 (Mon, 04 Feb 2008) | 1 line
  
  Added missing file.
................
  r43085 | johnmaddock | 2008-02-04 01:17:35 -0800 (Mon, 04 Feb 2008) | 1 line
  
  Removed dead file.
................
  r43087 | johnmaddock | 2008-02-04 01:20:46 -0800 (Mon, 04 Feb 2008) | 1 line
  
  Removed dead files.
................
  r43089 | johnmaddock | 2008-02-04 01:23:28 -0800 (Mon, 04 Feb 2008) | 1 line
  
  Removed dead files.
................
  r43094 | anthonyw | 2008-02-04 05:16:32 -0800 (Mon, 04 Feb 2008) | 1 line
  
  added test for duration overloads of timed_lock, and added missing implementation to win32 version
................
  r43101 | hkaiser | 2008-02-04 11:21:46 -0800 (Mon, 04 Feb 2008) | 1 line
  
  Wave: trying to fix stdcxx_gcc regression.
................
  r43103 | matias | 2008-02-04 13:01:06 -0800 (Mon, 04 Feb 2008) | 1 line
  
  hooking --> additional information in html docs
................
  r43106 | eric_niebler | 2008-02-04 18:09:51 -0800 (Mon, 04 Feb 2008) | 1 line
  
  fleshing out evaluation.qbk, document transforms of if_, not_, and_ and or_
................
  r43107 | eric_niebler | 2008-02-04 21:33:12 -0800 (Mon, 04 Feb 2008) | 1 line
  
  eliminate warnings under msvc's -W4
................
  r43111 | eric_niebler | 2008-02-04 22:03:01 -0800 (Mon, 04 Feb 2008) | 1 line
  
  fix typo
................
  r43112 | marshall | 2008-02-05 08:07:19 -0800 (Tue, 05 Feb 2008) | 1 line
  
  Fix typo (bug #1434)
................
  r43113 | marshall | 2008-02-05 08:15:35 -0800 (Tue, 05 Feb 2008) | 1 line
  
  Applied patch (fixes bug #1307)
................
  r43117 | dgregor | 2008-02-05 12:51:23 -0800 (Tue, 05 Feb 2008) | 1 line
  
  Fix add_vertex and add_vertices when the CSR graph has vertex properties
................
  r43118 | danieljames | 2008-02-05 12:57:02 -0800 (Tue, 05 Feb 2008) | 13 lines
  
  Merged revisions 42942-43116 via svnmerge from 
  https://svn.boost.org/svn/boost/branches/unordered/trunk
  
  ........
    r42975 | danieljames | 2008-01-26 00:29:32 +0000 (Sat, 26 Jan 2008) | 1 line
    
    Typedef some types before using them, to make life easier for Borland.
  ........
    r43116 | danieljames | 2008-02-05 20:47:44 +0000 (Tue, 05 Feb 2008) | 1 line
    
    Some compilers and libraries combinations have problems with deques of non-assingable types. Using a list instead.
  ........
................
  r43120 | eric_niebler | 2008-02-05 13:07:31 -0800 (Tue, 05 Feb 2008) | 1 line
  
  add missing #include
................
  r43121 | bemandawes | 2008-02-05 18:01:46 -0800 (Tue, 05 Feb 2008) | 1 line
  
  Add circular_buffer to the alphabetic list
................
  r43125 | t_schwinger | 2008-02-06 05:00:08 -0800 (Wed, 06 Feb 2008) | 3 lines
  
  attempts to make synthesis metafunctions work with sun compiler
................
  r43129 | danieljames | 2008-02-06 11:02:38 -0800 (Wed, 06 Feb 2008) | 2 lines
  
  In the boostbook navbar, link FAQ and people to the website.
................
  r43130 | eric_niebler | 2008-02-06 11:57:51 -0800 (Wed, 06 Feb 2008) | 1 line
  
  untabify
................
  r43132 | nesotto | 2008-02-06 14:46:19 -0800 (Wed, 06 Feb 2008) | 1 line
  
  cleanup to pass inspection report
................
  r43133 | nesotto | 2008-02-06 14:46:31 -0800 (Wed, 06 Feb 2008) | 1 line
  
  cleanup to pass inspection report
................
  r43134 | eric_niebler | 2008-02-06 14:57:57 -0800 (Wed, 06 Feb 2008) | 1 line
  
  add handy get() accessors on literal<> wrapper
................
  r43135 | nesotto | 2008-02-06 15:12:21 -0800 (Wed, 06 Feb 2008) | 1 line
  
  cleanup to pass inspection tool
................
  r43136 | eric_niebler | 2008-02-06 16:05:01 -0800 (Wed, 06 Feb 2008) | 1 line
  
  reasonably complete user docs for expression evaluation
................
  r43138 | eric_niebler | 2008-02-07 00:06:29 -0800 (Thu, 07 Feb 2008) | 1 line
  
  tweaks for doxygen 1.5.4, document matches<>
................
  r43141 | johnmaddock | 2008-02-07 01:55:41 -0800 (Thu, 07 Feb 2008) | 1 line
  
  Fix last checked version.
................
  r43143 | johnmaddock | 2008-02-07 02:03:16 -0800 (Thu, 07 Feb 2008) | 1 line
  
  Remove tabs.
................
  r43145 | johnmaddock | 2008-02-07 02:13:31 -0800 (Thu, 07 Feb 2008) | 1 line
  
  Fix min/max usage violation.
................
  r43147 | vladimir_prus | 2008-02-07 02:17:03 -0800 (Thu, 07 Feb 2008) | 2 lines
  
  Attempt to unbreak <library-file>
................
  r43148 | johnmaddock | 2008-02-07 02:24:29 -0800 (Thu, 07 Feb 2008) | 1 line
  
  Added comment to suppress inspect warning.
................
  r43150 | johnmaddock | 2008-02-07 02:29:59 -0800 (Thu, 07 Feb 2008) | 1 line
  
  Added fix for inspection report.
................
  r43152 | vladimir_prus | 2008-02-07 03:04:30 -0800 (Thu, 07 Feb 2008) | 4 lines
  
  Fix <framework> with no path.
  
  Patch from Jon Olsson.
................
  r43154 | bemandawes | 2008-02-07 05:22:34 -0800 (Thu, 07 Feb 2008) | 1 line
  
  Remove obsolete CVS scripts, add 1.35.0 SVN scripts, beginning of docs page
................
  r43155 | nesotto | 2008-02-07 06:41:04 -0800 (Thu, 07 Feb 2008) | 6 lines
  
  
  iterator_range disables msvc warning 4996
      <http://svn.boost.org/trac/boost/ticket/1565>
  
  [range] sub_range assignment issue
      <http://svn.boost.org/trac/boost/ticket/1284> 
................
  r43156 | nesotto | 2008-02-07 06:46:19 -0800 (Thu, 07 Feb 2008) | 1 line
  
  test
................
  r43157 | joaquin | 2008-02-07 08:29:27 -0800 (Thu, 07 Feb 2008) | 1 line
  
  updated according to latest regression tests results, fixed a broken link, typo
................
  r43159 | turkanis | 2008-02-07 09:07:28 -0800 (Thu, 07 Feb 2008) | 1 line
  
  added missing 'self.' qualification
................
  r43165 | dgregor | 2008-02-07 13:08:09 -0800 (Thu, 07 Feb 2008) | 1 line
  
  Support for non-blocking MPI operations in Python, from Andreas Kloeckner
................
  r43166 | dgregor | 2008-02-07 13:09:38 -0800 (Thu, 07 Feb 2008) | 1 line
  
  Note addition of nonblocking operations to the Python interface
................
  r43171 | nesotto | 2008-02-08 01:58:35 -0800 (Fri, 08 Feb 2008) | 1 line
  
  silence of warnings for unused arguments
................
  r43175 | nesotto | 2008-02-08 07:25:01 -0800 (Fri, 08 Feb 2008) | 1 line
  
  missing ) fixed
................
  r43176 | noel_belcourt | 2008-02-08 08:32:35 -0800 (Fri, 08 Feb 2008) | 14 lines
  
  Force PPC Darwin to use fork instead of vfork.  This change
  requires both the parent and child process to explicitly set
  the process group id. Vfork guarantees the child process 
  runs to the exec before it releases the parent process.  
  Now that we use fork instead of vfork, it's possible for the 
  parent to wait on the child process without having the child 
  setpgid on itself.  This eliminates spurious hangs on ppc
  darwin caused by either a race condition between vfork and
  execvp, or a bug in the vfork implementation.
  
  Added a test to ensure we don't try to read from the
  stderr pipe descriptor if the descriptor's not valid.
................
  r43177 | eric_niebler | 2008-02-08 09:11:57 -0800 (Fri, 08 Feb 2008) | 1 line
  
  reserve some c_type bits for dinkumware on windows, fixes #1625
................
  r43179 | noel_belcourt | 2008-02-08 09:53:50 -0800 (Fri, 08 Feb 2008) | 13 lines
  
  I've added the -single_module option to the intel-darwin.link.dll
  action to fix this linker error when linking dylibs:
  
  ld: common symbols not allowed with MH_DYLIB output format with the -multi_module option
  boost/bin.v2/libs/system/build/intel-darwin-9.1/debug/macosx-version-10.4/error_code.o 
    definition of common __ZGVZNK5boost6system14error_category7messageEiE1s (size 16)
  boost/bin.v2/libs/system/build/intel-darwin-9.1/debug/macosx-version-10.4/error_code.o 
    definition of common __ZZNK5boost6system14error_category7messageEiE1s (size 16)
  
  though I would note that the common symbols problem occurs in a number of
  other libraries (test, graph, spirit, ...) as well.
................
  r43188 | danieljames | 2008-02-09 04:29:02 -0800 (Sat, 09 Feb 2008) | 2 lines
  
  Fix a link in the intrusive redirect.
................
  r43189 | danieljames | 2008-02-09 04:37:00 -0800 (Sat, 09 Feb 2008) | 1 line
  
  Fix another redirect link.
................
  r43190 | danieljames | 2008-02-09 04:38:19 -0800 (Sat, 09 Feb 2008) | 1 line
  
  Update link to Jamfile, to link to the version 2 jamfile.
................
  r43191 | danieljames | 2008-02-09 04:39:06 -0800 (Sat, 09 Feb 2008) | 1 line
  
  Fix a link.
................
  r43192 | danieljames | 2008-02-09 04:45:32 -0800 (Sat, 09 Feb 2008) | 2 lines
  
  Add a forwarding header for hash/custom.html as Boost.Bimap links to it.
................
  r43193 | danieljames | 2008-02-09 05:02:45 -0800 (Sat, 09 Feb 2008) | 1 line
  
  Fix the link to the license.
................
  r43199 | eric_niebler | 2008-02-09 12:32:27 -0800 (Sat, 09 Feb 2008) | 1 line
  
  more doxygen comments, const-correctness tweak for fusion::at() on proto expression
................
  r43200 | eric_niebler | 2008-02-09 12:34:33 -0800 (Sat, 09 Feb 2008) | 1 line
  
  document how to access children of proto expressions
................
  r43204 | eric_niebler | 2008-02-09 22:57:24 -0800 (Sat, 09 Feb 2008) | 1 line
  
  fix oops in proto fusion interface
................
  r43205 | eric_niebler | 2008-02-09 23:02:54 -0800 (Sat, 09 Feb 2008) | 1 line
  
  suppress msvc warning
................
  r43206 | danieljames | 2008-02-10 01:55:03 -0800 (Sun, 10 Feb 2008) | 1 line
  
  Fix some broken links.
................
  r43207 | vladimir_prus | 2008-02-10 05:13:41 -0800 (Sun, 10 Feb 2008) | 4 lines
  
  Tolerate argc being zero.
  
  Patch from C. K. Jester-Young.
................
  r43209 | danieljames | 2008-02-10 06:56:22 -0800 (Sun, 10 Feb 2008) | 1 line
  
  Link to people pages on the website, as they've been removed from the download.
................
  r43210 | danieljames | 2008-02-10 07:02:17 -0800 (Sun, 10 Feb 2008) | 1 line
  
  Point links to the pages that used to be in 'more' to the site.
................
  r43212 | danieljames | 2008-02-10 08:10:16 -0800 (Sun, 10 Feb 2008) | 1 line
  
  Fix links on the home page as well.
................
  r43213 | danieljames | 2008-02-10 08:21:22 -0800 (Sun, 10 Feb 2008) | 1 line
  
  Generated documentation which is no longer generated.
................
  r43220 | eric_niebler | 2008-02-10 19:48:41 -0800 (Sun, 10 Feb 2008) | 1 line
  
  include config.hpp and workaround.hpp before uses of BOOST_WORKAROUND and BOOST_MSVC
................
  r43221 | chris_kohlhoff | 2008-02-11 05:59:44 -0800 (Mon, 11 Feb 2008) | 2 lines
  
  Need to define _XOPEN_SOURCE_EXTENDED when compiling for HP-UX.
................
  r43226 | djenkins | 2008-02-11 12:49:19 -0800 (Mon, 11 Feb 2008) | 1 line
  
  cleanup using local<> and skip()
................
  r43239 | turkanis | 2008-02-12 21:43:39 -0800 (Tue, 12 Feb 2008) | 1 line
  
  fixed return value of read(), to correctly handle eof
................
  r43240 | turkanis | 2008-02-12 21:47:44 -0800 (Tue, 12 Feb 2008) | 1 line
  
  merged changes from iostreams_dev, revisions 43059-43238: better debug output for mapped file; fixed large_file_test.cpp under UNICODE on Windows
................
  r43241 | turkanis | 2008-02-13 11:38:52 -0800 (Wed, 13 Feb 2008) | 1 line
  
  added markup for stdcxx failures (iostreams)
................
  r43243 | turkanis | 2008-02-13 11:42:10 -0800 (Wed, 13 Feb 2008) | 1 line
  
  switched from <wchar.h> to <cwchar>, for stdcxx (which is conforming in this case)
................
  r43246 | matias | 2008-02-14 09:33:12 -0800 (Thu, 14 Feb 2008) | 1 line
  
  remove local admonitions
................
  r43247 | matias | 2008-02-14 09:43:52 -0800 (Thu, 14 Feb 2008) | 1 line
  
  optional docs fixes
................
  r43248 | matias | 2008-02-14 09:44:21 -0800 (Thu, 14 Feb 2008) | 1 line
  
  redirect optional docs to new version
................
  r43251 | matias | 2008-02-14 10:08:16 -0800 (Thu, 14 Feb 2008) | 1 line
  
  conversion docs fixes
................
  r43252 | matias | 2008-02-14 10:09:34 -0800 (Thu, 14 Feb 2008) | 1 line
  
  redirect to new conversion docs
................
  r43253 | matias | 2008-02-14 10:19:34 -0800 (Thu, 14 Feb 2008) | 1 line
  
  redirect optional and numeric/conversion docs to new version
................
  r43254 | matias | 2008-02-14 11:03:55 -0800 (Thu, 14 Feb 2008) | 1 line
  
  bimap doc fixes
................
  r43255 | matias | 2008-02-14 11:05:04 -0800 (Thu, 14 Feb 2008) | 1 line
  
  fix tabs in files
................
  r43256 | matias | 2008-02-14 11:22:15 -0800 (Thu, 14 Feb 2008) | 1 line
  
  fix tabs in files
................
  r43260 | matias | 2008-02-14 13:24:11 -0800 (Thu, 14 Feb 2008) | 1 line
  
  add missing images
................
  r43262 | hkaiser | 2008-02-14 14:01:54 -0800 (Thu, 14 Feb 2008) | 1 line
  
  Fixed a whitespace insertion glitch, where whitespace got inserted unconditionally between two operators even if one of these was a comma.
................
  r43264 | hkaiser | 2008-02-14 15:52:33 -0800 (Thu, 14 Feb 2008) | 1 line
  
  Wave: More fixes to whitespace insertion engine.
................
  r43266 | hkaiser | 2008-02-15 06:35:36 -0800 (Fri, 15 Feb 2008) | 1 line
  
  Wave: More fixes to whitespace insertion engine.
................
  r43269 | pdimov | 2008-02-15 10:40:36 -0800 (Fri, 15 Feb 2008) | 1 line
  
  Added support for &&, ||
................
  r43272 | andreas_huber69 | 2008-02-16 02:13:08 -0800 (Sat, 16 Feb 2008) | 1 line
  
  Updated statechart markup
................
  r43274 | andreas_huber69 | 2008-02-16 02:19:49 -0800 (Sat, 16 Feb 2008) | 2 lines
  
  Silenced GCC 4.0.1 warning (patch supplied by Euan)
  <http://thread.gmane.org/gmane.comp.lib.boost.devel/171071>
................
  r43280 | jurko | 2008-02-16 08:50:42 -0800 (Sat, 16 Feb 2008) | 1 line
  
  Corrected comments related to the allowed linker & linker-type values. Minor stylistic changes.
................
  r43281 | jurko | 2008-02-16 08:53:33 -0800 (Sat, 16 Feb 2008) | 1 line
  
  Added support for compiling C++ programs without RTTI support using the gcc toolset.
................
  r43282 | jurko | 2008-02-16 09:03:54 -0800 (Sat, 16 Feb 2008) | 1 line
  
  Made the msvc toolset always explicitly enable or disable rtti support based on the <rtti> feature value instead of only setting it if <rtti>on and depending on it being disabled by default. The original behaviour did not work well with msvc 8.0 for which there was not way to disable rtti support as that compiler enables rtti support by default.
................
  r43283 | bemandawes | 2008-02-16 18:01:32 -0800 (Sat, 16 Feb 2008) | 1 line
  
  Show output of example program
................
  r43290 | hkaiser | 2008-02-17 08:45:08 -0800 (Sun, 17 Feb 2008) | 1 line
  
  Wave: Extended a workaround to newest Intel compiler version (Linux V10.1)
................
  r43292 | nesotto | 2008-02-17 08:49:38 -0800 (Sun, 17 Feb 2008) | 1 line
  
  support for comparinson operators
................
  r43293 | nesotto | 2008-02-17 08:50:02 -0800 (Sun, 17 Feb 2008) | 1 line
  
  support for comparison operators
................
  r43294 | hkaiser | 2008-02-17 09:00:20 -0800 (Sun, 17 Feb 2008) | 1 line
  
  Wave: Fixed test cases to reflect recent changes to whitespace insertion.
................
  r43296 | eric_niebler | 2008-02-17 12:53:18 -0800 (Sun, 17 Feb 2008) | 1 line
  
  proto documentation improvements
................
  r43299 | turkanis | 2008-02-17 21:48:13 -0800 (Sun, 17 Feb 2008) | 1 line
  
  merged changes from iostreams_dev, revisions 43243-43298: overhaul of category_of and close(): stringstream is now dual_seekable; standard file streams and string streams are closable; public Boost.Iostreams streams and streambufs are closable; close() pops filtering streams and streambufs
................
  r43300 | eric_niebler | 2008-02-17 22:16:27 -0800 (Sun, 17 Feb 2008) | 1 line
  
  remove dependence on boost.lambda, make numeric function objects work with std binders
................
  r43301 | chris_kohlhoff | 2008-02-18 05:31:26 -0800 (Mon, 18 Feb 2008) | 2 lines
  
  Fix printing of error messages.
................
  r43302 | chris_kohlhoff | 2008-02-18 05:33:23 -0800 (Mon, 18 Feb 2008) | 2 lines
  
  Only define _XOPEN_SOURCE_EXTENDED when building with gcc on HP-UX.
................
  r43303 | chris_kohlhoff | 2008-02-18 05:35:15 -0800 (Mon, 18 Feb 2008) | 3 lines
  
  Add missing #include of socket_types.hpp needed for the SSL unit tests
  to compile successfully on Windows.
................
  r43306 | eric_niebler | 2008-02-18 10:29:29 -0800 (Mon, 18 Feb 2008) | 1 line
  
  remove post_construct docs, fix link to boost.parameter library
................
  r43308 | niels_dekker | 2008-02-18 14:11:19 -0800 (Mon, 18 Feb 2008) | 1 line
  
  Fixed the assignment of value_initialized<T> for T being a C-style array. (The previous version would trigger a compile error in this case.)
................
  r43309 | niels_dekker | 2008-02-18 14:13:21 -0800 (Mon, 18 Feb 2008) | 1 line
  
  Tested the assignment of value_initialized<T>, for T being a C-style array. Related to the fix of changeset [43308]
................
  r43310 | eric_niebler | 2008-02-18 15:03:23 -0800 (Mon, 18 Feb 2008) | 1 line
  
  some doxygen comments for proto/traits.hpp
................
  r43311 | eric_niebler | 2008-02-18 21:56:52 -0800 (Mon, 18 Feb 2008) | 1 line
  
  more proto documentation tweaks, remove unnecessary result_of::arg_c instantiation
................
  r43312 | eric_niebler | 2008-02-18 23:14:37 -0800 (Mon, 18 Feb 2008) | 1 line
  
  fix droppable accumulators
................
  r43314 | eric_niebler | 2008-02-18 23:33:30 -0800 (Mon, 18 Feb 2008) | 1 line
  
  darn, back out bad droppable changes
................
  r43316 | pdimov | 2008-02-19 05:18:58 -0800 (Tue, 19 Feb 2008) | 1 line
  
  Fixes #1590.
................
  r43317 | pdimov | 2008-02-19 06:01:13 -0800 (Tue, 19 Feb 2008) | 1 line
  
  Fixes #1444.
................
  r43318 | pdimov | 2008-02-19 06:26:36 -0800 (Tue, 19 Feb 2008) | 1 line
  
  Fix #398, as long as the macros BOOST_NO_STD_TYPEINFO and BOOST_NO_IOSTREAM are defined. I don't know how Boost.Config needs to be changed to autodetect eVC4 and set these on its own.
................
  r43319 | pdimov | 2008-02-19 06:51:10 -0800 (Tue, 19 Feb 2008) | 1 line
  
  Fix #1641.
................
  r43320 | pdimov | 2008-02-19 06:59:28 -0800 (Tue, 19 Feb 2008) | 1 line
  
  Fix #1646.
................
  r43321 | pdimov | 2008-02-19 07:09:10 -0800 (Tue, 19 Feb 2008) | 1 line
  
  Fix #1642.
................
  r43322 | nesotto | 2008-02-19 07:10:05 -0800 (Tue, 19 Feb 2008) | 1 line
  
  fixed problem with operator()() when the value_type was abstract.
................
  r43323 | pdimov | 2008-02-19 07:40:58 -0800 (Tue, 19 Feb 2008) | 1 line
  
  Fix #1643.
................
  r43325 | turkanis | 2008-02-19 11:34:07 -0800 (Tue, 19 Feb 2008) | 1 line
  
  stringstreams are no longer closable; the semantics of close() for these devices was illconsidered
................
  r43328 | turkanis | 2008-02-19 16:09:06 -0800 (Tue, 19 Feb 2008) | 1 line
  
  menu fix from iostreams_dev
................
  r43329 | turkanis | 2008-02-19 19:20:17 -0800 (Tue, 19 Feb 2008) | 1 line
  
  markup for pgi-7.0 (iostreams)
................
  r43330 | bemandawes | 2008-02-20 05:46:49 -0800 (Wed, 20 Feb 2008) | 1 line
  
  Add .z7 archive generation
................
  r43332 | grafik | 2008-02-20 09:32:09 -0800 (Wed, 20 Feb 2008) | 1 line
  
  New readme page for now standalone release.
................
  r43334 | grafik | 2008-02-20 11:15:16 -0800 (Wed, 20 Feb 2008) | 1 line
  
  Fix link to getting started docs.
................
  r43335 | grafik | 2008-02-20 14:50:03 -0800 (Wed, 20 Feb 2008) | 1 line
  
  Add the <python.interpreter> to all requirements to allow other toolsets to use the currently configured python instead of relying on python being in the path.
................
  r43336 | grafik | 2008-02-20 15:01:43 -0800 (Wed, 20 Feb 2008) | 1 line
  
  Use the configured python interpreter instead of assuming it's in the path.
................
  r43337 | grafik | 2008-02-20 15:03:28 -0800 (Wed, 20 Feb 2008) | 1 line
  
  Add missing, and assumed, white background for screen rendering.
................
  r43338 | grafik | 2008-02-20 15:26:58 -0800 (Wed, 20 Feb 2008) | 1 line
  
  Doc cleanups.
................
  r43344 | eric_niebler | 2008-02-20 23:18:24 -0800 (Wed, 20 Feb 2008) | 1 line
  
  More Proto documentation
................
  r43346 | johnmaddock | 2008-02-21 02:37:59 -0800 (Thu, 21 Feb 2008) | 3 lines
  
  Fix typo in example.
  Added links to PDF versions of the docs.
  Regenerated all the docs to fix people links.
................
  r43347 | johnmaddock | 2008-02-21 03:53:59 -0800 (Thu, 21 Feb 2008) | 1 line
  
  Update main overview page.
................
  r43349 | bemandawes | 2008-02-21 04:46:11 -0800 (Thu, 21 Feb 2008) | 1 line
  
  Fix typo; .z7 should be .7z
................
  r43351 | johnmaddock | 2008-02-21 04:58:15 -0800 (Thu, 21 Feb 2008) | 1 line
  
  Added link to PDF docs, and regenerated.
................
  r43354 | johnmaddock | 2008-02-21 05:51:18 -0800 (Thu, 21 Feb 2008) | 1 line
  
  Added link to PDF docs, and regenerated.
................
  r43357 | johnmaddock | 2008-02-21 08:49:59 -0800 (Thu, 21 Feb 2008) | 1 line
  
  Added link to PDF docs.
................
  r43359 | johnmaddock | 2008-02-21 09:01:26 -0800 (Thu, 21 Feb 2008) | 1 line
  
  Regenerated docs to fix links.
................
  r43361 | bemandawes | 2008-02-21 12:11:32 -0800 (Thu, 21 Feb 2008) | 1 line
  
  Fix still another typo
................
  r43362 | eric_niebler | 2008-02-21 12:12:02 -0800 (Thu, 21 Feb 2008) | 1 line
  
  doxygen comments for proto::when<>
................
  r43363 | eric_niebler | 2008-02-21 16:42:12 -0800 (Thu, 21 Feb 2008) | 1 line
  
  fix crash when actions are in keep() expressions
................
  r43364 | eric_niebler | 2008-02-21 18:01:46 -0800 (Thu, 21 Feb 2008) | 1 line
  
  doxygen comments
................
  r43365 | grafik | 2008-02-21 21:26:39 -0800 (Thu, 21 Feb 2008) | 1 line
  
  Rename readme.html to index.html, and add forwarding index.htm for backward compatibility.
................
  r43368 | danieljames | 2008-02-22 01:21:22 -0800 (Fri, 22 Feb 2008) | 2 lines
  
  Update the index.html link in the navbar.
................
  r43371 | grafik | 2008-02-22 08:25:21 -0800 (Fri, 22 Feb 2008) | 1 line
  
  Remove obsolete getting started files. They where replaced by more/getting_started/*.
................
  r43377 | chris_kohlhoff | 2008-02-22 14:43:54 -0800 (Fri, 22 Feb 2008) | 2 lines
  
  Use the correct vector of timer queues when dispatching timers.
................
  r43390 | turkanis | 2008-02-22 16:05:49 -0800 (Fri, 22 Feb 2008) | 2 lines
  
  Ported change from iostreams_dev
................
  r43391 | turkanis | 2008-02-22 16:06:24 -0800 (Fri, 22 Feb 2008) | 2 lines
  
  Ported changes from iostreams_dev
................
  r43392 | turkanis | 2008-02-22 16:07:13 -0800 (Fri, 22 Feb 2008) | 1 line
  
  merged changes from iostreams_dev, revisions 43327-43389
................
  r43393 | turkanis | 2008-02-22 16:11:07 -0800 (Fri, 22 Feb 2008) | 1 line
  
  merged changes from iostreams_dev
................
  r43395 | turkanis | 2008-02-22 22:07:59 -0800 (Fri, 22 Feb 2008) | 1 line
  
  updated to test close() on filtering streambufs
................
  r43399 | turkanis | 2008-02-22 23:44:58 -0800 (Fri, 22 Feb 2008) | 1 line
  
  merged changes from iostreams_dev
................
  r43402 | bemandawes | 2008-02-23 06:04:02 -0800 (Sat, 23 Feb 2008) | 1 line
  
  Give the 1st and 2nd level index.html files a common look-and-feel.
................
  r43405 | vladimir_prus | 2008-02-24 04:59:04 -0800 (Sun, 24 Feb 2008) | 3 lines
  
  Recognize that fact, for that for intel-win, <runtime-debuggin>
  matters and should be added to the library name.
................
  r43409 | bemandawes | 2008-02-24 16:53:26 -0800 (Sun, 24 Feb 2008) | 1 line
  
  Fix html boo boo
................
  r43410 | schoepflin | 2008-02-25 00:37:10 -0800 (Mon, 25 Feb 2008) | 1 line
  
  Added missing template keyword.
................
  r43411 | t_schwinger | 2008-02-25 03:45:51 -0800 (Mon, 25 Feb 2008) | 3 lines
  
  removes unnecessary escaping
................
  r43412 | t_schwinger | 2008-02-25 03:47:59 -0800 (Mon, 25 Feb 2008) | 3 lines
  
  removes unnecessary comment
................
  r43416 | hkaiser | 2008-02-26 11:25:05 -0800 (Tue, 26 Feb 2008) | 1 line
  
  Wave: Fixed expanding_function_like_macro preprocessing hook.
................
  r43417 | danieljames | 2008-02-26 14:04:55 -0800 (Tue, 26 Feb 2008) | 2 lines
  
  Fix a link to Boost.Bimap.
................
  r43418 | danieljames | 2008-02-26 14:07:25 -0800 (Tue, 26 Feb 2008) | 2 lines
  
  Change another link that's no longer in the repository to link to the website.
................
  r43419 | hkaiser | 2008-02-26 14:36:36 -0800 (Tue, 26 Feb 2008) | 1 line
  
  Fixed a compilation problem on pathscale
................
  r43421 | eric_niebler | 2008-02-27 10:48:22 -0800 (Wed, 27 Feb 2008) | 1 line
  
  partially revert breaking change to independent sub-expressions until I can make a proper fix
................
  r43422 | danieljames | 2008-02-27 10:51:14 -0800 (Wed, 27 Feb 2008) | 1 line
  
  Fix broken copyright urls. Fixes #1573.
................
  r43423 | danieljames | 2008-02-27 11:22:01 -0800 (Wed, 27 Feb 2008) | 1 line
  
  Fix incorrect links to copyright of the form 'http:#www.boost.org
................
  r43424 | eric_niebler | 2008-02-27 11:39:43 -0800 (Wed, 27 Feb 2008) | 1 line
  
  fix bug in use_simple_repeat calculation
................
  r43428 | eric_niebler | 2008-02-27 16:03:15 -0800 (Wed, 27 Feb 2008) | 1 line
  
  add test case for use_simple_repeat fix
................
  r43433 | eric_niebler | 2008-02-28 14:47:12 -0800 (Thu, 28 Feb 2008) | 1 line
  
  fix oops
................
  r43434 | johnmaddock | 2008-02-29 01:49:42 -0800 (Fri, 29 Feb 2008) | 1 line
  
  Apply fixes to issue #1658 which fixes some broken URL's.
................
  r43435 | johnmaddock | 2008-02-29 01:58:30 -0800 (Fri, 29 Feb 2008) | 1 line
  
  Fix broken link as per report #1658.
................
  r43437 | chris_kohlhoff | 2008-02-29 04:57:57 -0800 (Fri, 29 Feb 2008) | 2 lines
  
  Add missing tie().
................
  r43438 | schoepflin | 2008-02-29 07:13:41 -0800 (Fri, 29 Feb 2008) | 2 lines
  
  Added expected failure markup for the test weighted_tail_variate_means on Tru64/CXX.
................
  r43441 | eric_niebler | 2008-03-01 11:32:56 -0800 (Sat, 01 Mar 2008) | 1 line
  
  add map_assign example
................
  r43458 | turkanis | 2008-03-02 22:20:14 -0800 (Sun, 02 Mar 2008) | 1 line
  
  merged changes from iostreams_dev, revisions 43399-43457
................
  r43461 | anthonyw | 2008-03-03 00:44:42 -0800 (Mon, 03 Mar 2008) | 1 line
  
  Test and fix for issue #1665
................
  r43464 | anthonyw | 2008-03-03 02:52:44 -0800 (Mon, 03 Mar 2008) | 1 line
  
  thread constructor now accepts up to three additional arguments to pass to thread function
................
  r43467 | danieljames | 2008-03-03 04:10:35 -0800 (Mon, 03 Mar 2008) | 2 lines
  
  Tell subversion that date_time.doc is a text file, not a word document.
................
  r43468 | danieljames | 2008-03-03 04:11:25 -0800 (Mon, 03 Mar 2008) | 1 line
  
  Fix license link in date_time.doc
................
  r43469 | chris_kohlhoff | 2008-03-03 05:21:05 -0800 (Mon, 03 Mar 2008) | 4 lines
  
  Disable use of CancelIo by default, due to the possibility of silent
  failure on some system configurations. Swallow error returned by CancelIoEx
  if there are no operations to be cancelled.
................
  r43470 | chris_kohlhoff | 2008-03-03 05:27:06 -0800 (Mon, 03 Mar 2008) | 2 lines
  
  Add missing 'boost_' prefix to helper namespace.
................
  r43471 | chris_kohlhoff | 2008-03-03 05:36:35 -0800 (Mon, 03 Mar 2008) | 2 lines
  
  Regenerate documentation.
................
  r43472 | chris_kohlhoff | 2008-03-03 06:05:35 -0800 (Mon, 03 Mar 2008) | 1 line
  
  Update copyright notices.
................
  r43473 | chris_kohlhoff | 2008-03-03 06:13:01 -0800 (Mon, 03 Mar 2008) | 2 lines
  
  Update copyright notices.
................
  r43476 | eric_niebler | 2008-03-03 11:44:54 -0800 (Mon, 03 Mar 2008) | 1 line
  
  add Map Assign example to documentation
................
  r43478 | eric_niebler | 2008-03-03 11:47:47 -0800 (Mon, 03 Mar 2008) | 1 line
  
  second attempt at fixing actions in independent expressions
................
  r43484 | eric_niebler | 2008-03-03 15:48:17 -0800 (Mon, 03 Mar 2008) | 1 line
  
  handle static regexes with actions nested in dynamic independent subexpressions
................
  r43485 | emildotchevski | 2008-03-03 17:41:17 -0800 (Mon, 03 Mar 2008) | 1 line
  
  boost exception
................
  r43496 | eric_niebler | 2008-03-04 10:51:07 -0800 (Tue, 04 Mar 2008) | 1 line
  
  rename numeric::empty to numeric::default_, fixes #1650
................
  r43501 | eric_niebler | 2008-03-04 11:31:57 -0800 (Tue, 04 Mar 2008) | 1 line
  
  eliminate msvc level 4 warnings, fixes #1631
................
  r43502 | eric_niebler | 2008-03-04 11:42:36 -0800 (Tue, 04 Mar 2008) | 1 line
  
  add Dave Jenkin's evil static/dynamic actions in keep test case
................
  r43503 | eric_niebler | 2008-03-04 13:09:47 -0800 (Tue, 04 Mar 2008) | 1 line
  
  work around msvc bug 331418, fixes #1652
................
  r43506 | eric_niebler | 2008-03-04 15:01:17 -0800 (Tue, 04 Mar 2008) | 1 line
  
  fix bad interaction between boyer-moore optimization and partial match feature, fixes #1564
................
  r43508 | eric_niebler | 2008-03-04 22:32:39 -0800 (Tue, 04 Mar 2008) | 1 line
  
  add BOOST_REVERSE_FOREACH, fixes #1071
................
  r43509 | eric_niebler | 2008-03-04 23:12:03 -0800 (Tue, 04 Mar 2008) | 1 line
  
  fix bug iterating over abstract base
................


[SVN r43519]
2008-03-05 20:37:04 +00:00

1349 lines
74 KiB
HTML

<?xml version="1.0" encoding="utf-8" ?>
<!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" xml:lang="en" lang="en">
<head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8" />
<meta name="generator" content="Docutils 0.5: http://docutils.sourceforge.net/" />
<title>Iterator Facade</title>
<meta name="author" content="David Abrahams, Jeremy Siek, Thomas Witt" />
<meta name="organization" content="Boost Consulting, Indiana University Open Systems Lab, University of Hanover Institute for Transport Railway Operation and Construction" />
<meta name="date" content="2006-09-11" />
<meta name="copyright" content="Copyright David Abrahams, Jeremy Siek, and Thomas Witt 2003." />
<link rel="stylesheet" href="../../../rst.css" type="text/css" />
</head>
<body>
<div class="document" id="iterator-facade">
<h1 class="title">Iterator Facade</h1>
<table class="docinfo" frame="void" rules="none">
<col class="docinfo-name" />
<col class="docinfo-content" />
<tbody valign="top">
<tr><th class="docinfo-name">Author:</th>
<td>David Abrahams, Jeremy Siek, Thomas Witt</td></tr>
<tr><th class="docinfo-name">Contact:</th>
<td><a class="first reference external" href="mailto:dave&#64;boost-consulting.com">dave&#64;boost-consulting.com</a>, <a class="reference external" href="mailto:jsiek&#64;osl.iu.edu">jsiek&#64;osl.iu.edu</a>, <a class="last reference external" href="mailto:witt&#64;ive.uni-hannover.de">witt&#64;ive.uni-hannover.de</a></td></tr>
<tr><th class="docinfo-name">Organization:</th>
<td><a class="first reference external" href="http://www.boost-consulting.com">Boost Consulting</a>, Indiana University <a class="reference external" href="http://www.osl.iu.edu">Open Systems
Lab</a>, University of Hanover <a class="last reference external" href="http://www.ive.uni-hannover.de">Institute for Transport
Railway Operation and Construction</a></td></tr>
<tr><th class="docinfo-name">Date:</th>
<td>2006-09-11</td></tr>
<tr><th class="docinfo-name">Copyright:</th>
<td>Copyright David Abrahams, Jeremy Siek, and Thomas Witt 2003.</td></tr>
</tbody>
</table>
<!-- Distributed under the Boost -->
<!-- Software License, Version 1.0. (See accompanying -->
<!-- file LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) -->
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field"><th class="field-name">abstract:</th><td class="field-body"><!-- Copyright David Abrahams 2006. Distributed under the Boost -->
<!-- Software License, Version 1.0. (See accompanying -->
<!-- file LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) -->
<tt class="docutils literal"><span class="pre">iterator_facade</span></tt> is a base class template that implements the
interface of standard iterators in terms of a few core functions
and associated types, to be supplied by a derived iterator class.</td>
</tr>
</tbody>
</table>
<div class="contents topic" id="table-of-contents">
<p class="topic-title first">Table of Contents</p>
<ul class="simple">
<li><a class="reference internal" href="#overview" id="id23">Overview</a><ul>
<li><a class="reference internal" href="#usage" id="id24">Usage</a></li>
<li><a class="reference internal" href="#iterator-core-access" id="id25">Iterator Core Access</a></li>
<li><a class="reference internal" href="#operator" id="id26"><tt class="docutils literal"><span class="pre">operator[]</span></tt></a></li>
<li><a class="reference internal" href="#id2" id="id27"><tt class="docutils literal"><span class="pre">operator-&gt;</span></tt></a></li>
</ul>
</li>
<li><a class="reference internal" href="#reference" id="id28">Reference</a><ul>
<li><a class="reference internal" href="#iterator-facade-requirements" id="id29"><tt class="docutils literal"><span class="pre">iterator_facade</span></tt> Requirements</a></li>
<li><a class="reference internal" href="#iterator-facade-operations" id="id30"><tt class="docutils literal"><span class="pre">iterator_facade</span></tt> operations</a></li>
</ul>
</li>
<li><a class="reference internal" href="#tutorial-example" id="id31">Tutorial Example</a><ul>
<li><a class="reference internal" href="#the-problem" id="id32">The Problem</a></li>
<li><a class="reference internal" href="#a-basic-iterator-using-iterator-facade" id="id33">A Basic Iterator Using <tt class="docutils literal"><span class="pre">iterator_facade</span></tt></a><ul>
<li><a class="reference internal" href="#template-arguments-for-iterator-facade" id="id34">Template Arguments for <tt class="docutils literal"><span class="pre">iterator_facade</span></tt></a><ul>
<li><a class="reference internal" href="#derived" id="id35"><tt class="docutils literal"><span class="pre">Derived</span></tt></a></li>
<li><a class="reference internal" href="#value" id="id36"><tt class="docutils literal"><span class="pre">Value</span></tt></a></li>
<li><a class="reference internal" href="#categoryortraversal" id="id37"><tt class="docutils literal"><span class="pre">CategoryOrTraversal</span></tt></a></li>
<li><a class="reference internal" href="#id12" id="id38"><tt class="docutils literal"><span class="pre">Reference</span></tt></a></li>
<li><a class="reference internal" href="#difference" id="id39"><tt class="docutils literal"><span class="pre">Difference</span></tt></a></li>
</ul>
</li>
<li><a class="reference internal" href="#constructors-and-data-members" id="id40">Constructors and Data Members</a></li>
<li><a class="reference internal" href="#implementing-the-core-operations" id="id41">Implementing the Core Operations</a></li>
</ul>
</li>
<li><a class="reference internal" href="#a-constant-node-iterator" id="id42">A constant <tt class="docutils literal"><span class="pre">node_iterator</span></tt></a></li>
<li><a class="reference internal" href="#interoperability" id="id43">Interoperability</a></li>
<li><a class="reference internal" href="#telling-the-truth" id="id44">Telling the Truth</a></li>
<li><a class="reference internal" href="#wrap-up" id="id45">Wrap Up</a></li>
</ul>
</li>
</ul>
</div>
<div class="section" id="overview">
<h1><a class="toc-backref" href="#id23">Overview</a></h1>
<!-- Distributed under the Boost -->
<!-- Software License, Version 1.0. (See accompanying -->
<!-- file LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) -->
<!-- Version 1.1 of this ReStructuredText document corresponds to
n1530_, the paper accepted by the LWG for TR1. -->
<!-- Copyright David Abrahams, Jeremy Siek, and Thomas Witt 2003. -->
<p>While the iterator interface is rich, there is a core subset of the
interface that is necessary for all the functionality. We have
identified the following core behaviors for iterators:</p>
<ul class="simple">
<li>dereferencing</li>
<li>incrementing</li>
<li>decrementing</li>
<li>equality comparison</li>
<li>random-access motion</li>
<li>distance measurement</li>
</ul>
<p>In addition to the behaviors listed above, the core interface elements
include the associated types exposed through iterator traits:
<tt class="docutils literal"><span class="pre">value_type</span></tt>, <tt class="docutils literal"><span class="pre">reference</span></tt>, <tt class="docutils literal"><span class="pre">difference_type</span></tt>, and
<tt class="docutils literal"><span class="pre">iterator_category</span></tt>.</p>
<p>Iterator facade uses the Curiously Recurring Template
Pattern (CRTP) <a class="citation-reference" href="#cop95" id="id1">[Cop95]</a> so that the user can specify the behavior
of <tt class="docutils literal"><span class="pre">iterator_facade</span></tt> in a derived class. Former designs used
policy objects to specify the behavior, but that approach was
discarded for several reasons:</p>
<blockquote>
<ol class="arabic simple">
<li>the creation and eventual copying of the policy object may create
overhead that can be avoided with the current approach.</li>
<li>The policy object approach does not allow for custom constructors
on the created iterator types, an essential feature if
<tt class="docutils literal"><span class="pre">iterator_facade</span></tt> should be used in other library
implementations.</li>
<li>Without the use of CRTP, the standard requirement that an
iterator's <tt class="docutils literal"><span class="pre">operator++</span></tt> returns the iterator type itself
would mean that all iterators built with the library would
have to be specializations of <tt class="docutils literal"><span class="pre">iterator_facade&lt;...&gt;</span></tt>, rather
than something more descriptive like
<tt class="docutils literal"><span class="pre">indirect_iterator&lt;T*&gt;</span></tt>. Cumbersome type generator
metafunctions would be needed to build new parameterized
iterators, and a separate <tt class="docutils literal"><span class="pre">iterator_adaptor</span></tt> layer would be
impossible.</li>
</ol>
</blockquote>
<div class="section" id="usage">
<h2><a class="toc-backref" href="#id24">Usage</a></h2>
<p>The user of <tt class="docutils literal"><span class="pre">iterator_facade</span></tt> derives his iterator class from a
specialization of <tt class="docutils literal"><span class="pre">iterator_facade</span></tt> and passes the derived
iterator class as <tt class="docutils literal"><span class="pre">iterator_facade</span></tt>'s first template parameter.
The order of the other template parameters have been carefully
chosen to take advantage of useful defaults. For example, when
defining a constant lvalue iterator, the user can pass a
const-qualified version of the iterator's <tt class="docutils literal"><span class="pre">value_type</span></tt> as
<tt class="docutils literal"><span class="pre">iterator_facade</span></tt>'s <tt class="docutils literal"><span class="pre">Value</span></tt> parameter and omit the
<tt class="docutils literal"><span class="pre">Reference</span></tt> parameter which follows.</p>
<p>The derived iterator class must define member functions implementing
the iterator's core behaviors. The following table describes
expressions which are required to be valid depending on the category
of the derived iterator type. These member functions are described
briefly below and in more detail in the iterator facade
requirements.</p>
<blockquote>
<table border="1" class="docutils">
<colgroup>
<col width="44%" />
<col width="56%" />
</colgroup>
<thead valign="bottom">
<tr><th class="head">Expression</th>
<th class="head">Effects</th>
</tr>
</thead>
<tbody valign="top">
<tr><td><tt class="docutils literal"><span class="pre">i.dereference()</span></tt></td>
<td>Access the value referred to</td>
</tr>
<tr><td><tt class="docutils literal"><span class="pre">i.equal(j)</span></tt></td>
<td>Compare for equality with <tt class="docutils literal"><span class="pre">j</span></tt></td>
</tr>
<tr><td><tt class="docutils literal"><span class="pre">i.increment()</span></tt></td>
<td>Advance by one position</td>
</tr>
<tr><td><tt class="docutils literal"><span class="pre">i.decrement()</span></tt></td>
<td>Retreat by one position</td>
</tr>
<tr><td><tt class="docutils literal"><span class="pre">i.advance(n)</span></tt></td>
<td>Advance by <tt class="docutils literal"><span class="pre">n</span></tt> positions</td>
</tr>
<tr><td><tt class="docutils literal"><span class="pre">i.distance_to(j)</span></tt></td>
<td>Measure the distance to <tt class="docutils literal"><span class="pre">j</span></tt></td>
</tr>
</tbody>
</table>
</blockquote>
<!-- Should we add a comment that a zero overhead implementation of iterator_facade
is possible with proper inlining? -->
<p>In addition to implementing the core interface functions, an iterator
derived from <tt class="docutils literal"><span class="pre">iterator_facade</span></tt> typically defines several
constructors. To model any of the standard iterator concepts, the
iterator must at least have a copy constructor. Also, if the iterator
type <tt class="docutils literal"><span class="pre">X</span></tt> is meant to be automatically interoperate with another
iterator type <tt class="docutils literal"><span class="pre">Y</span></tt> (as with constant and mutable iterators) then
there must be an implicit conversion from <tt class="docutils literal"><span class="pre">X</span></tt> to <tt class="docutils literal"><span class="pre">Y</span></tt> or from <tt class="docutils literal"><span class="pre">Y</span></tt>
to <tt class="docutils literal"><span class="pre">X</span></tt> (but not both), typically implemented as a conversion
constructor. Finally, if the iterator is to model Forward Traversal
Iterator or a more-refined iterator concept, a default constructor is
required.</p>
</div>
<div class="section" id="iterator-core-access">
<h2><a class="toc-backref" href="#id25">Iterator Core Access</a></h2>
<p><tt class="docutils literal"><span class="pre">iterator_facade</span></tt> and the operator implementations need to be able
to access the core member functions in the derived class. Making the
core member functions public would expose an implementation detail to
the user. The design used here ensures that implementation details do
not appear in the public interface of the derived iterator type.</p>
<p>Preventing direct access to the core member functions has two
advantages. First, there is no possibility for the user to accidently
use a member function of the iterator when a member of the value_type
was intended. This has been an issue with smart pointer
implementations in the past. The second and main advantage is that
library implementers can freely exchange a hand-rolled iterator
implementation for one based on <tt class="docutils literal"><span class="pre">iterator_facade</span></tt> without fear of
breaking code that was accessing the public core member functions
directly.</p>
<p>In a naive implementation, keeping the derived class' core member
functions private would require it to grant friendship to
<tt class="docutils literal"><span class="pre">iterator_facade</span></tt> and each of the seven operators. In order to
reduce the burden of limiting access, <tt class="docutils literal"><span class="pre">iterator_core_access</span></tt> is
provided, a class that acts as a gateway to the core member functions
in the derived iterator class. The author of the derived class only
needs to grant friendship to <tt class="docutils literal"><span class="pre">iterator_core_access</span></tt> to make his core
member functions available to the library.</p>
<!-- This is no long uptodate -thw -->
<!-- Yes it is; I made sure of it! -DWA -->
<p><tt class="docutils literal"><span class="pre">iterator_core_access</span></tt> will be typically implemented as an empty
class containing only private static member functions which invoke the
iterator core member functions. There is, however, no need to
standardize the gateway protocol. Note that even if
<tt class="docutils literal"><span class="pre">iterator_core_access</span></tt> used public member functions it would not
open a safety loophole, as every core member function preserves the
invariants of the iterator.</p>
</div>
<div class="section" id="operator">
<h2><a class="toc-backref" href="#id26"><tt class="docutils literal"><span class="pre">operator[]</span></tt></a></h2>
<p>The indexing operator for a generalized iterator presents special
challenges. A random access iterator's <tt class="docutils literal"><span class="pre">operator[]</span></tt> is only
required to return something convertible to its <tt class="docutils literal"><span class="pre">value_type</span></tt>.
Requiring that it return an lvalue would rule out currently-legal
random-access iterators which hold the referenced value in a data
member (e.g. <a class="reference external" href="counting_iterator.html"><tt class="docutils literal"><span class="pre">counting_iterator</span></tt></a>), because <tt class="docutils literal"><span class="pre">*(p+n)</span></tt> is a reference
into the temporary iterator <tt class="docutils literal"><span class="pre">p+n</span></tt>, which is destroyed when
<tt class="docutils literal"><span class="pre">operator[]</span></tt> returns.</p>
<p>Writable iterators built with <tt class="docutils literal"><span class="pre">iterator_facade</span></tt> implement the
semantics required by the preferred resolution to <a class="reference external" href="http://anubis.dkuug.dk/jtc1/sc22/wg21/docs/lwg-active.html#299">issue 299</a> and
adopted by proposal <a class="reference external" href="http://anubis.dkuug.dk/JTC1/SC22/WG21/docs/papers/2003/n1550.html">n1550</a>: the result of <tt class="docutils literal"><span class="pre">p[n]</span></tt> is an object
convertible to the iterator's <tt class="docutils literal"><span class="pre">value_type</span></tt>, and <tt class="docutils literal"><span class="pre">p[n]</span> <span class="pre">=</span> <span class="pre">x</span></tt> is
equivalent to <tt class="docutils literal"><span class="pre">*(p</span> <span class="pre">+</span> <span class="pre">n)</span> <span class="pre">=</span> <span class="pre">x</span></tt> (Note: This result object may be
implemented as a proxy containing a copy of <tt class="docutils literal"><span class="pre">p+n</span></tt>). This approach
will work properly for any random-access iterator regardless of the
other details of its implementation. A user who knows more about
the implementation of her iterator is free to implement an
<tt class="docutils literal"><span class="pre">operator[]</span></tt> that returns an lvalue in the derived iterator
class; it will hide the one supplied by <tt class="docutils literal"><span class="pre">iterator_facade</span></tt> from
clients of her iterator.</p>
</div>
<div class="section" id="id2">
<span id="operator-arrow"></span><h2><a class="toc-backref" href="#id27"><tt class="docutils literal"><span class="pre">operator-&gt;</span></tt></a></h2>
<p>The <tt class="docutils literal"><span class="pre">reference</span></tt> type of a readable iterator (and today's input
iterator) need not in fact be a reference, so long as it is
convertible to the iterator's <tt class="docutils literal"><span class="pre">value_type</span></tt>. When the <tt class="docutils literal"><span class="pre">value_type</span></tt>
is a class, however, it must still be possible to access members
through <tt class="docutils literal"><span class="pre">operator-&gt;</span></tt>. Therefore, an iterator whose <tt class="docutils literal"><span class="pre">reference</span></tt>
type is not in fact a reference must return a proxy containing a copy
of the referenced value from its <tt class="docutils literal"><span class="pre">operator-&gt;</span></tt>.</p>
<p>The return types for <tt class="docutils literal"><span class="pre">iterator_facade</span></tt>'s <tt class="docutils literal"><span class="pre">operator-&gt;</span></tt> and
<tt class="docutils literal"><span class="pre">operator[]</span></tt> are not explicitly specified. Instead, those types
are described in terms of a set of requirements, which must be
satisfied by the <tt class="docutils literal"><span class="pre">iterator_facade</span></tt> implementation.</p>
<table class="docutils citation" frame="void" id="cop95" rules="none">
<colgroup><col class="label" /><col /></colgroup>
<tbody valign="top">
<tr><td class="label">[Cop95]</td><td><em>(<a class="fn-backref" href="#id1">1</a>, <a class="fn-backref" href="#id10">2</a>)</em> [Coplien, 1995] Coplien, J., Curiously Recurring Template
Patterns, C++ Report, February 1995, pp. 24-27.</td></tr>
</tbody>
</table>
</div>
</div>
<div class="section" id="reference">
<h1><a class="toc-backref" href="#id28">Reference</a></h1>
<!-- Distributed under the Boost -->
<!-- Software License, Version 1.0. (See accompanying -->
<!-- file LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) -->
<!-- Version 1.3 of this ReStructuredText document corresponds to
n1530_, the paper accepted by the LWG for TR1. -->
<!-- Copyright David Abrahams, Jeremy Siek, and Thomas Witt 2003. -->
<pre class="literal-block">
template &lt;
class Derived
, class Value
, class CategoryOrTraversal
, class Reference = Value&amp;
, class Difference = ptrdiff_t
&gt;
class iterator_facade {
public:
typedef remove_const&lt;Value&gt;::type value_type;
typedef Reference reference;
typedef Value* pointer;
typedef Difference difference_type;
typedef /* see <a class="reference internal" href="#iterator-category">below</a> */ iterator_category;
reference operator*() const;
/* see <a class="reference internal" href="#operator-arrow">below</a> */ operator-&gt;() const;
/* see <a class="reference internal" href="#brackets">below</a> */ operator[](difference_type n) const;
Derived&amp; operator++();
Derived operator++(int);
Derived&amp; operator--();
Derived operator--(int);
Derived&amp; operator+=(difference_type n);
Derived&amp; operator-=(difference_type n);
Derived operator-(difference_type n) const;
protected:
typedef iterator_facade iterator_facade_;
};
// Comparison operators
template &lt;class Dr1, class V1, class TC1, class R1, class D1,
class Dr2, class V2, class TC2, class R2, class D2&gt;
typename enable_if_interoperable&lt;Dr1,Dr2,bool&gt;::type // exposition
operator ==(iterator_facade&lt;Dr1,V1,TC1,R1,D1&gt; const&amp; lhs,
iterator_facade&lt;Dr2,V2,TC2,R2,D2&gt; const&amp; rhs);
template &lt;class Dr1, class V1, class TC1, class R1, class D1,
class Dr2, class V2, class TC2, class R2, class D2&gt;
typename enable_if_interoperable&lt;Dr1,Dr2,bool&gt;::type
operator !=(iterator_facade&lt;Dr1,V1,TC1,R1,D1&gt; const&amp; lhs,
iterator_facade&lt;Dr2,V2,TC2,R2,D2&gt; const&amp; rhs);
template &lt;class Dr1, class V1, class TC1, class R1, class D1,
class Dr2, class V2, class TC2, class R2, class D2&gt;
typename enable_if_interoperable&lt;Dr1,Dr2,bool&gt;::type
operator &lt;(iterator_facade&lt;Dr1,V1,TC1,R1,D1&gt; const&amp; lhs,
iterator_facade&lt;Dr2,V2,TC2,R2,D2&gt; const&amp; rhs);
template &lt;class Dr1, class V1, class TC1, class R1, class D1,
class Dr2, class V2, class TC2, class R2, class D2&gt;
typename enable_if_interoperable&lt;Dr1,Dr2,bool&gt;::type
operator &lt;=(iterator_facade&lt;Dr1,V1,TC1,R1,D1&gt; const&amp; lhs,
iterator_facade&lt;Dr2,V2,TC2,R2,D2&gt; const&amp; rhs);
template &lt;class Dr1, class V1, class TC1, class R1, class D1,
class Dr2, class V2, class TC2, class R2, class D2&gt;
typename enable_if_interoperable&lt;Dr1,Dr2,bool&gt;::type
operator &gt;(iterator_facade&lt;Dr1,V1,TC1,R1,D1&gt; const&amp; lhs,
iterator_facade&lt;Dr2,V2,TC2,R2,D2&gt; const&amp; rhs);
template &lt;class Dr1, class V1, class TC1, class R1, class D1,
class Dr2, class V2, class TC2, class R2, class D2&gt;
typename enable_if_interoperable&lt;Dr1,Dr2,bool&gt;::type
operator &gt;=(iterator_facade&lt;Dr1,V1,TC1,R1,D1&gt; const&amp; lhs,
iterator_facade&lt;Dr2,V2,TC2,R2,D2&gt; const&amp; rhs);
// Iterator difference
template &lt;class Dr1, class V1, class TC1, class R1, class D1,
class Dr2, class V2, class TC2, class R2, class D2&gt;
/* see <a class="reference internal" href="#minus">below</a> */
operator-(iterator_facade&lt;Dr1,V1,TC1,R1,D1&gt; const&amp; lhs,
iterator_facade&lt;Dr2,V2,TC2,R2,D2&gt; const&amp; rhs);
// Iterator addition
template &lt;class Dr, class V, class TC, class R, class D&gt;
Derived operator+ (iterator_facade&lt;Dr,V,TC,R,D&gt; const&amp;,
typename Derived::difference_type n);
template &lt;class Dr, class V, class TC, class R, class D&gt;
Derived operator+ (typename Derived::difference_type n,
iterator_facade&lt;Dr,V,TC,R,D&gt; const&amp;);
</pre>
<p id="iterator-category">The <tt class="docutils literal"><span class="pre">iterator_category</span></tt> member of <tt class="docutils literal"><span class="pre">iterator_facade</span></tt> is</p>
<pre class="literal-block">
<em>iterator-category</em>(CategoryOrTraversal, value_type, reference)
</pre>
<p>where <em>iterator-category</em> is defined as follows:</p>
<pre class="literal-block" id="id7">
<em>iterator-category</em>(C,R,V) :=
if (C is convertible to std::input_iterator_tag
|| C is convertible to std::output_iterator_tag
)
return C
else if (C is not convertible to incrementable_traversal_tag)
<em>the program is ill-formed</em>
else return a type X satisfying the following two constraints:
1. X is convertible to X1, and not to any more-derived
type, where X1 is defined by:
if (R is a reference type
&amp;&amp; C is convertible to forward_traversal_tag)
{
if (C is convertible to random_access_traversal_tag)
X1 = random_access_iterator_tag
else if (C is convertible to bidirectional_traversal_tag)
X1 = bidirectional_iterator_tag
else
X1 = forward_iterator_tag
}
else
{
if (C is convertible to single_pass_traversal_tag
&amp;&amp; R is convertible to V)
X1 = input_iterator_tag
else
X1 = C
}
2. <a class="reference external" href="new-iter-concepts.html#category-to-traversal"><em>category-to-traversal</em></a>(X) is convertible to the most
derived traversal tag type to which X is also
convertible, and not to any more-derived traversal tag
type.
</pre>
<p>[Note: the intention is to allow <tt class="docutils literal"><span class="pre">iterator_category</span></tt> to be one of
the five original category tags when convertibility to one of the
traversal tags would add no information]</p>
<!-- Copyright David Abrahams 2004. Use, modification and distribution is -->
<!-- subject to the Boost Software License, Version 1.0. (See accompanying -->
<!-- file LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) -->
<p>The <tt class="docutils literal"><span class="pre">enable_if_interoperable</span></tt> template used above is for exposition
purposes. The member operators should only be in an overload set
provided the derived types <tt class="docutils literal"><span class="pre">Dr1</span></tt> and <tt class="docutils literal"><span class="pre">Dr2</span></tt> are interoperable,
meaning that at least one of the types is convertible to the other. The
<tt class="docutils literal"><span class="pre">enable_if_interoperable</span></tt> approach uses SFINAE to take the operators
out of the overload set when the types are not interoperable.
The operators should behave <em>as-if</em> <tt class="docutils literal"><span class="pre">enable_if_interoperable</span></tt>
were defined to be:</p>
<pre class="literal-block">
template &lt;bool, typename&gt; enable_if_interoperable_impl
{};
template &lt;typename T&gt; enable_if_interoperable_impl&lt;true,T&gt;
{ typedef T type; };
template&lt;typename Dr1, typename Dr2, typename T&gt;
struct enable_if_interoperable
: enable_if_interoperable_impl&lt;
is_convertible&lt;Dr1,Dr2&gt;::value || is_convertible&lt;Dr2,Dr1&gt;::value
, T
&gt;
{};
</pre>
<div class="section" id="iterator-facade-requirements">
<h2><a class="toc-backref" href="#id29"><tt class="docutils literal"><span class="pre">iterator_facade</span></tt> Requirements</a></h2>
<p>The following table describes the typical valid expressions on
<tt class="docutils literal"><span class="pre">iterator_facade</span></tt>'s <tt class="docutils literal"><span class="pre">Derived</span></tt> parameter, depending on the
iterator concept(s) it will model. The operations in the first
column must be made accessible to member functions of class
<tt class="docutils literal"><span class="pre">iterator_core_access</span></tt>. In addition,
<tt class="docutils literal"><span class="pre">static_cast&lt;Derived*&gt;(iterator_facade*)</span></tt> shall be well-formed.</p>
<p>In the table below, <tt class="docutils literal"><span class="pre">F</span></tt> is <tt class="docutils literal"><span class="pre">iterator_facade&lt;X,V,C,R,D&gt;</span></tt>, <tt class="docutils literal"><span class="pre">a</span></tt> is an
object of type <tt class="docutils literal"><span class="pre">X</span></tt>, <tt class="docutils literal"><span class="pre">b</span></tt> and <tt class="docutils literal"><span class="pre">c</span></tt> are objects of type <tt class="docutils literal"><span class="pre">const</span> <span class="pre">X</span></tt>,
<tt class="docutils literal"><span class="pre">n</span></tt> is an object of <tt class="docutils literal"><span class="pre">F::difference_type</span></tt>, <tt class="docutils literal"><span class="pre">y</span></tt> is a constant
object of a single pass iterator type interoperable with <tt class="docutils literal"><span class="pre">X</span></tt>, and <tt class="docutils literal"><span class="pre">z</span></tt>
is a constant object of a random access traversal iterator type
interoperable with <tt class="docutils literal"><span class="pre">X</span></tt>.</p>
<div class="topic" id="core-operations">
<p class="topic-title first"><tt class="docutils literal"><span class="pre">iterator_facade</span></tt> Core Operations</p>
<table border="1" class="docutils">
<colgroup>
<col width="21%" />
<col width="23%" />
<col width="27%" />
<col width="29%" />
</colgroup>
<thead valign="bottom">
<tr><th class="head">Expression</th>
<th class="head">Return Type</th>
<th class="head">Assertion/Note</th>
<th class="head">Used to implement Iterator
Concept(s)</th>
</tr>
</thead>
<tbody valign="top">
<tr><td><tt class="docutils literal"><span class="pre">c.dereference()</span></tt></td>
<td><tt class="docutils literal"><span class="pre">F::reference</span></tt></td>
<td>&nbsp;</td>
<td>Readable Iterator, Writable
Iterator</td>
</tr>
<tr><td><tt class="docutils literal"><span class="pre">c.equal(y)</span></tt></td>
<td>convertible to bool</td>
<td>true iff <tt class="docutils literal"><span class="pre">c</span></tt> and <tt class="docutils literal"><span class="pre">y</span></tt>
refer to the same
position.</td>
<td>Single Pass Iterator</td>
</tr>
<tr><td><tt class="docutils literal"><span class="pre">a.increment()</span></tt></td>
<td>unused</td>
<td>&nbsp;</td>
<td>Incrementable Iterator</td>
</tr>
<tr><td><tt class="docutils literal"><span class="pre">a.decrement()</span></tt></td>
<td>unused</td>
<td>&nbsp;</td>
<td>Bidirectional Traversal
Iterator</td>
</tr>
<tr><td><tt class="docutils literal"><span class="pre">a.advance(n)</span></tt></td>
<td>unused</td>
<td>&nbsp;</td>
<td>Random Access Traversal
Iterator</td>
</tr>
<tr><td><tt class="docutils literal"><span class="pre">c.distance_to(z)</span></tt></td>
<td>convertible to
<tt class="docutils literal"><span class="pre">F::difference_type</span></tt></td>
<td>equivalent to
<tt class="docutils literal"><span class="pre">distance(c,</span> <span class="pre">X(z))</span></tt>.</td>
<td>Random Access Traversal
Iterator</td>
</tr>
</tbody>
</table>
</div>
</div>
<div class="section" id="iterator-facade-operations">
<h2><a class="toc-backref" href="#id30"><tt class="docutils literal"><span class="pre">iterator_facade</span></tt> operations</a></h2>
<p>The operations in this section are described in terms of operations on
the core interface of <tt class="docutils literal"><span class="pre">Derived</span></tt> which may be inaccessible
(i.e. private). The implementation should access these operations
through member functions of class <tt class="docutils literal"><span class="pre">iterator_core_access</span></tt>.</p>
<p><tt class="docutils literal"><span class="pre">reference</span> <span class="pre">operator*()</span> <span class="pre">const;</span></tt></p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field"><th class="field-name">Returns:</th><td class="field-body"><tt class="docutils literal"><span class="pre">static_cast&lt;Derived</span> <span class="pre">const*&gt;(this)-&gt;dereference()</span></tt></td>
</tr>
</tbody>
</table>
<p><tt class="docutils literal"><span class="pre">operator-&gt;()</span> <span class="pre">const;</span></tt> (see <a class="reference internal" href="#operator-arrow">below</a>)</p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">If <tt class="docutils literal"><span class="pre">reference</span></tt> is a reference type, an object
of type <tt class="docutils literal"><span class="pre">pointer</span></tt> equal to:</p>
<pre class="literal-block">
&amp;static_cast&lt;Derived const*&gt;(this)-&gt;dereference()
</pre>
<p class="last">Otherwise returns an object of unspecified type such that,
<tt class="docutils literal"><span class="pre">(*static_cast&lt;Derived</span> <span class="pre">const*&gt;(this))-&gt;m</span></tt> is equivalent to <tt class="docutils literal"><span class="pre">(w</span> <span class="pre">=</span> <span class="pre">**static_cast&lt;Derived</span> <span class="pre">const*&gt;(this),</span>
<span class="pre">w.m)</span></tt> for some temporary object <tt class="docutils literal"><span class="pre">w</span></tt> of type <tt class="docutils literal"><span class="pre">value_type</span></tt>.</p>
</td>
</tr>
</tbody>
</table>
<p id="brackets"><em>unspecified</em> <tt class="docutils literal"><span class="pre">operator[](difference_type</span> <span class="pre">n)</span> <span class="pre">const;</span></tt></p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field"><th class="field-name">Returns:</th><td class="field-body">an object convertible to <tt class="docutils literal"><span class="pre">value_type</span></tt>. For constant
objects <tt class="docutils literal"><span class="pre">v</span></tt> of type <tt class="docutils literal"><span class="pre">value_type</span></tt>, and <tt class="docutils literal"><span class="pre">n</span></tt> of type
<tt class="docutils literal"><span class="pre">difference_type</span></tt>, <tt class="docutils literal"><span class="pre">(*this)[n]</span> <span class="pre">=</span> <span class="pre">v</span></tt> is equivalent to
<tt class="docutils literal"><span class="pre">*(*this</span> <span class="pre">+</span> <span class="pre">n)</span> <span class="pre">=</span> <span class="pre">v</span></tt>, and <tt class="docutils literal"><span class="pre">static_cast&lt;value_type</span>
<span class="pre">const&amp;&gt;((*this)[n])</span></tt> is equivalent to
<tt class="docutils literal"><span class="pre">static_cast&lt;value_type</span> <span class="pre">const&amp;&gt;(*(*this</span> <span class="pre">+</span> <span class="pre">n))</span></tt></td>
</tr>
</tbody>
</table>
<p><tt class="docutils literal"><span class="pre">Derived&amp;</span> <span class="pre">operator++();</span></tt></p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field"><th class="field-name">Effects:</th><td class="field-body"><pre class="first last literal-block">
static_cast&lt;Derived*&gt;(this)-&gt;increment();
return *static_cast&lt;Derived*&gt;(this);
</pre>
</td>
</tr>
</tbody>
</table>
<p><tt class="docutils literal"><span class="pre">Derived</span> <span class="pre">operator++(int);</span></tt></p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field"><th class="field-name">Effects:</th><td class="field-body"><pre class="first last literal-block">
Derived tmp(static_cast&lt;Derived const*&gt;(this));
++*this;
return tmp;
</pre>
</td>
</tr>
</tbody>
</table>
<p><tt class="docutils literal"><span class="pre">Derived&amp;</span> <span class="pre">operator--();</span></tt></p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field"><th class="field-name">Effects:</th><td class="field-body"><pre class="first last literal-block">
static_cast&lt;Derived*&gt;(this)-&gt;decrement();
return *static_cast&lt;Derived*&gt;(this);
</pre>
</td>
</tr>
</tbody>
</table>
<p><tt class="docutils literal"><span class="pre">Derived</span> <span class="pre">operator--(int);</span></tt></p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field"><th class="field-name">Effects:</th><td class="field-body"><pre class="first last literal-block">
Derived tmp(static_cast&lt;Derived const*&gt;(this));
--*this;
return tmp;
</pre>
</td>
</tr>
</tbody>
</table>
<p><tt class="docutils literal"><span class="pre">Derived&amp;</span> <span class="pre">operator+=(difference_type</span> <span class="pre">n);</span></tt></p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field"><th class="field-name">Effects:</th><td class="field-body"><pre class="first last literal-block">
static_cast&lt;Derived*&gt;(this)-&gt;advance(n);
return *static_cast&lt;Derived*&gt;(this);
</pre>
</td>
</tr>
</tbody>
</table>
<p><tt class="docutils literal"><span class="pre">Derived&amp;</span> <span class="pre">operator-=(difference_type</span> <span class="pre">n);</span></tt></p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field"><th class="field-name">Effects:</th><td class="field-body"><pre class="first last literal-block">
static_cast&lt;Derived*&gt;(this)-&gt;advance(-n);
return *static_cast&lt;Derived*&gt;(this);
</pre>
</td>
</tr>
</tbody>
</table>
<p><tt class="docutils literal"><span class="pre">Derived</span> <span class="pre">operator-(difference_type</span> <span class="pre">n)</span> <span class="pre">const;</span></tt></p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field"><th class="field-name">Effects:</th><td class="field-body"><pre class="first last literal-block">
Derived tmp(static_cast&lt;Derived const*&gt;(this));
return tmp -= n;
</pre>
</td>
</tr>
</tbody>
</table>
<pre class="literal-block">
template &lt;class Dr, class V, class TC, class R, class D&gt;
Derived operator+ (iterator_facade&lt;Dr,V,TC,R,D&gt; const&amp;,
typename Derived::difference_type n);
template &lt;class Dr, class V, class TC, class R, class D&gt;
Derived operator+ (typename Derived::difference_type n,
iterator_facade&lt;Dr,V,TC,R,D&gt; const&amp;);
</pre>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field"><th class="field-name">Effects:</th><td class="field-body"><pre class="first last literal-block">
Derived tmp(static_cast&lt;Derived const*&gt;(this));
return tmp += n;
</pre>
</td>
</tr>
</tbody>
</table>
<pre class="literal-block">
template &lt;class Dr1, class V1, class TC1, class R1, class D1,
class Dr2, class V2, class TC2, class R2, class D2&gt;
typename enable_if_interoperable&lt;Dr1,Dr2,bool&gt;::type
operator ==(iterator_facade&lt;Dr1,V1,TC1,R1,D1&gt; const&amp; lhs,
iterator_facade&lt;Dr2,V2,TC2,R2,D2&gt; const&amp; rhs);
</pre>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">if <tt class="docutils literal"><span class="pre">is_convertible&lt;Dr2,Dr1&gt;::value</span></tt></p>
<dl class="last docutils">
<dt>then</dt>
<dd><p class="first last"><tt class="docutils literal"><span class="pre">((Dr1</span> <span class="pre">const&amp;)lhs).equal((Dr2</span> <span class="pre">const&amp;)rhs)</span></tt>.</p>
</dd>
<dt>Otherwise,</dt>
<dd><p class="first last"><tt class="docutils literal"><span class="pre">((Dr2</span> <span class="pre">const&amp;)rhs).equal((Dr1</span> <span class="pre">const&amp;)lhs)</span></tt>.</p>
</dd>
</dl>
</td>
</tr>
</tbody>
</table>
<pre class="literal-block">
template &lt;class Dr1, class V1, class TC1, class R1, class D1,
class Dr2, class V2, class TC2, class R2, class D2&gt;
typename enable_if_interoperable&lt;Dr1,Dr2,bool&gt;::type
operator !=(iterator_facade&lt;Dr1,V1,TC1,R1,D1&gt; const&amp; lhs,
iterator_facade&lt;Dr2,V2,TC2,R2,D2&gt; const&amp; rhs);
</pre>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">if <tt class="docutils literal"><span class="pre">is_convertible&lt;Dr2,Dr1&gt;::value</span></tt></p>
<dl class="last docutils">
<dt>then</dt>
<dd><p class="first last"><tt class="docutils literal"><span class="pre">!((Dr1</span> <span class="pre">const&amp;)lhs).equal((Dr2</span> <span class="pre">const&amp;)rhs)</span></tt>.</p>
</dd>
<dt>Otherwise,</dt>
<dd><p class="first last"><tt class="docutils literal"><span class="pre">!((Dr2</span> <span class="pre">const&amp;)rhs).equal((Dr1</span> <span class="pre">const&amp;)lhs)</span></tt>.</p>
</dd>
</dl>
</td>
</tr>
</tbody>
</table>
<pre class="literal-block">
template &lt;class Dr1, class V1, class TC1, class R1, class D1,
class Dr2, class V2, class TC2, class R2, class D2&gt;
typename enable_if_interoperable&lt;Dr1,Dr2,bool&gt;::type
operator &lt;(iterator_facade&lt;Dr1,V1,TC1,R1,D1&gt; const&amp; lhs,
iterator_facade&lt;Dr2,V2,TC2,R2,D2&gt; const&amp; rhs);
</pre>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">if <tt class="docutils literal"><span class="pre">is_convertible&lt;Dr2,Dr1&gt;::value</span></tt></p>
<dl class="last docutils">
<dt>then</dt>
<dd><p class="first last"><tt class="docutils literal"><span class="pre">((Dr1</span> <span class="pre">const&amp;)lhs).distance_to((Dr2</span> <span class="pre">const&amp;)rhs)</span> <span class="pre">&lt;</span> <span class="pre">0</span></tt>.</p>
</dd>
<dt>Otherwise,</dt>
<dd><p class="first last"><tt class="docutils literal"><span class="pre">((Dr2</span> <span class="pre">const&amp;)rhs).distance_to((Dr1</span> <span class="pre">const&amp;)lhs)</span> <span class="pre">&gt;</span> <span class="pre">0</span></tt>.</p>
</dd>
</dl>
</td>
</tr>
</tbody>
</table>
<pre class="literal-block">
template &lt;class Dr1, class V1, class TC1, class R1, class D1,
class Dr2, class V2, class TC2, class R2, class D2&gt;
typename enable_if_interoperable&lt;Dr1,Dr2,bool&gt;::type
operator &lt;=(iterator_facade&lt;Dr1,V1,TC1,R1,D1&gt; const&amp; lhs,
iterator_facade&lt;Dr2,V2,TC2,R2,D2&gt; const&amp; rhs);
</pre>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">if <tt class="docutils literal"><span class="pre">is_convertible&lt;Dr2,Dr1&gt;::value</span></tt></p>
<dl class="last docutils">
<dt>then</dt>
<dd><p class="first last"><tt class="docutils literal"><span class="pre">((Dr1</span> <span class="pre">const&amp;)lhs).distance_to((Dr2</span> <span class="pre">const&amp;)rhs)</span> <span class="pre">&lt;=</span> <span class="pre">0</span></tt>.</p>
</dd>
<dt>Otherwise,</dt>
<dd><p class="first last"><tt class="docutils literal"><span class="pre">((Dr2</span> <span class="pre">const&amp;)rhs).distance_to((Dr1</span> <span class="pre">const&amp;)lhs)</span> <span class="pre">&gt;=</span> <span class="pre">0</span></tt>.</p>
</dd>
</dl>
</td>
</tr>
</tbody>
</table>
<pre class="literal-block">
template &lt;class Dr1, class V1, class TC1, class R1, class D1,
class Dr2, class V2, class TC2, class R2, class D2&gt;
typename enable_if_interoperable&lt;Dr1,Dr2,bool&gt;::type
operator &gt;(iterator_facade&lt;Dr1,V1,TC1,R1,D1&gt; const&amp; lhs,
iterator_facade&lt;Dr2,V2,TC2,R2,D2&gt; const&amp; rhs);
</pre>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">if <tt class="docutils literal"><span class="pre">is_convertible&lt;Dr2,Dr1&gt;::value</span></tt></p>
<dl class="last docutils">
<dt>then</dt>
<dd><p class="first last"><tt class="docutils literal"><span class="pre">((Dr1</span> <span class="pre">const&amp;)lhs).distance_to((Dr2</span> <span class="pre">const&amp;)rhs)</span> <span class="pre">&gt;</span> <span class="pre">0</span></tt>.</p>
</dd>
<dt>Otherwise,</dt>
<dd><p class="first last"><tt class="docutils literal"><span class="pre">((Dr2</span> <span class="pre">const&amp;)rhs).distance_to((Dr1</span> <span class="pre">const&amp;)lhs)</span> <span class="pre">&lt;</span> <span class="pre">0</span></tt>.</p>
</dd>
</dl>
</td>
</tr>
</tbody>
</table>
<pre class="literal-block">
template &lt;class Dr1, class V1, class TC1, class R1, class D1,
class Dr2, class V2, class TC2, class R2, class D2&gt;
typename enable_if_interoperable&lt;Dr1,Dr2,bool&gt;::type
operator &gt;=(iterator_facade&lt;Dr1,V1,TC1,R1,D1&gt; const&amp; lhs,
iterator_facade&lt;Dr2,V2,TC2,R2,D2&gt; const&amp; rhs);
</pre>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">if <tt class="docutils literal"><span class="pre">is_convertible&lt;Dr2,Dr1&gt;::value</span></tt></p>
<dl class="last docutils">
<dt>then</dt>
<dd><p class="first last"><tt class="docutils literal"><span class="pre">((Dr1</span> <span class="pre">const&amp;)lhs).distance_to((Dr2</span> <span class="pre">const&amp;)rhs)</span> <span class="pre">&gt;=</span> <span class="pre">0</span></tt>.</p>
</dd>
<dt>Otherwise,</dt>
<dd><p class="first last"><tt class="docutils literal"><span class="pre">((Dr2</span> <span class="pre">const&amp;)rhs).distance_to((Dr1</span> <span class="pre">const&amp;)lhs)</span> <span class="pre">&lt;=</span> <span class="pre">0</span></tt>.</p>
</dd>
</dl>
</td>
</tr>
</tbody>
</table>
<pre class="literal-block" id="minus">
template &lt;class Dr1, class V1, class TC1, class R1, class D1,
class Dr2, class V2, class TC2, class R2, class D2&gt;
typename enable_if_interoperable&lt;Dr1,Dr2,difference&gt;::type
operator -(iterator_facade&lt;Dr1,V1,TC1,R1,D1&gt; const&amp; lhs,
iterator_facade&lt;Dr2,V2,TC2,R2,D2&gt; const&amp; rhs);
</pre>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field"><th class="field-name">Return Type:</th><td class="field-body"><p class="first">if <tt class="docutils literal"><span class="pre">is_convertible&lt;Dr2,Dr1&gt;::value</span></tt></p>
<blockquote>
<dl class="docutils">
<dt>then</dt>
<dd><p class="first last"><tt class="docutils literal"><span class="pre">difference</span></tt> shall be
<tt class="docutils literal"><span class="pre">iterator_traits&lt;Dr1&gt;::difference_type</span></tt>.</p>
</dd>
<dt>Otherwise</dt>
<dd><p class="first last"><tt class="docutils literal"><span class="pre">difference</span></tt> shall be <tt class="docutils literal"><span class="pre">iterator_traits&lt;Dr2&gt;::difference_type</span></tt></p>
</dd>
</dl>
</blockquote>
</td>
</tr>
<tr class="field"><th class="field-name">Returns:</th><td class="field-body"><p class="first">if <tt class="docutils literal"><span class="pre">is_convertible&lt;Dr2,Dr1&gt;::value</span></tt></p>
<dl class="last docutils">
<dt>then</dt>
<dd><p class="first last"><tt class="docutils literal"><span class="pre">-((Dr1</span> <span class="pre">const&amp;)lhs).distance_to((Dr2</span> <span class="pre">const&amp;)rhs)</span></tt>.</p>
</dd>
<dt>Otherwise,</dt>
<dd><p class="first last"><tt class="docutils literal"><span class="pre">((Dr2</span> <span class="pre">const&amp;)rhs).distance_to((Dr1</span> <span class="pre">const&amp;)lhs)</span></tt>.</p>
</dd>
</dl>
</td>
</tr>
</tbody>
</table>
</div>
</div>
<div class="section" id="tutorial-example">
<h1><a class="toc-backref" href="#id31">Tutorial Example</a></h1>
<!-- Copyright David Abrahams 2004. Use, modification and distribution is -->
<!-- subject to the Boost Software License, Version 1.0. (See accompanying -->
<!-- file LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) -->
<p>In this section we'll walk through the implementation of a few
iterators using <tt class="docutils literal"><span class="pre">iterator_facade</span></tt>, based around the simple
example of a linked list of polymorphic objects. This example was
inspired by a <a class="reference external" href="http://thread.gmane.org/gmane.comp.lib.boost.user/5100">posting</a> by Keith Macdonald on the <a class="reference external" href="http://www.boost.org/more/mailing_lists.htm#users">Boost-Users</a>
mailing list.</p>
<div class="section" id="the-problem">
<h2><a class="toc-backref" href="#id32">The Problem</a></h2>
<p>Say we've written a polymorphic linked list node base class:</p>
<pre class="literal-block">
# include &lt;iostream&gt;
struct node_base
{
node_base() : m_next(0) {}
// Each node manages all of its tail nodes
virtual ~node_base() { delete m_next; }
// Access the rest of the list
node_base* next() const { return m_next; }
// print to the stream
virtual void print(std::ostream&amp; s) const = 0;
// double the value
virtual void double_me() = 0;
void append(node_base* p)
{
if (m_next)
m_next-&gt;append(p);
else
m_next = p;
}
private:
node_base* m_next;
};
</pre>
<p>Lists can hold objects of different types by linking together
specializations of the following template:</p>
<pre class="literal-block">
template &lt;class T&gt;
struct node : node_base
{
node(T x)
: m_value(x)
{}
void print(std::ostream&amp; s) const { s &lt;&lt; this-&gt;m_value; }
void double_me() { m_value += m_value; }
private:
T m_value;
};
</pre>
<p>And we can print any node using the following streaming operator:</p>
<pre class="literal-block">
inline std::ostream&amp; operator&lt;&lt;(std::ostream&amp; s, node_base const&amp; n)
{
n.print(s);
return s;
}
</pre>
<p>Our first challenge is to build an appropriate iterator over these
lists.</p>
</div>
<div class="section" id="a-basic-iterator-using-iterator-facade">
<h2><a class="toc-backref" href="#id33">A Basic Iterator Using <tt class="docutils literal"><span class="pre">iterator_facade</span></tt></a></h2>
<p>We will construct a <tt class="docutils literal"><span class="pre">node_iterator</span></tt> class using inheritance from
<tt class="docutils literal"><span class="pre">iterator_facade</span></tt> to implement most of the iterator's operations.</p>
<pre class="literal-block">
# include &quot;node.hpp&quot;
# include &lt;boost/iterator/iterator_facade.hpp&gt;
class node_iterator
: public boost::iterator_facade&lt;...&gt;
{
...
};
</pre>
<div class="section" id="template-arguments-for-iterator-facade">
<h3><a class="toc-backref" href="#id34">Template Arguments for <tt class="docutils literal"><span class="pre">iterator_facade</span></tt></a></h3>
<p><tt class="docutils literal"><span class="pre">iterator_facade</span></tt> has several template parameters, so we must decide
what types to use for the arguments. The parameters are <tt class="docutils literal"><span class="pre">Derived</span></tt>,
<tt class="docutils literal"><span class="pre">Value</span></tt>, <tt class="docutils literal"><span class="pre">CategoryOrTraversal</span></tt>, <tt class="docutils literal"><span class="pre">Reference</span></tt>, and <tt class="docutils literal"><span class="pre">Difference</span></tt>.</p>
<div class="section" id="derived">
<h4><a class="toc-backref" href="#id35"><tt class="docutils literal"><span class="pre">Derived</span></tt></a></h4>
<p>Because <tt class="docutils literal"><span class="pre">iterator_facade</span></tt> is meant to be used with the CRTP
<a class="citation-reference" href="#cop95" id="id10">[Cop95]</a> the first parameter is the iterator class name itself,
<tt class="docutils literal"><span class="pre">node_iterator</span></tt>.</p>
</div>
<div class="section" id="value">
<h4><a class="toc-backref" href="#id36"><tt class="docutils literal"><span class="pre">Value</span></tt></a></h4>
<p>The <tt class="docutils literal"><span class="pre">Value</span></tt> parameter determines the <tt class="docutils literal"><span class="pre">node_iterator</span></tt>'s
<tt class="docutils literal"><span class="pre">value_type</span></tt>. In this case, we are iterating over <tt class="docutils literal"><span class="pre">node_base</span></tt>
objects, so <tt class="docutils literal"><span class="pre">Value</span></tt> will be <tt class="docutils literal"><span class="pre">node_base</span></tt>.</p>
</div>
<div class="section" id="categoryortraversal">
<h4><a class="toc-backref" href="#id37"><tt class="docutils literal"><span class="pre">CategoryOrTraversal</span></tt></a></h4>
<p>Now we have to determine which <a class="reference external" href="new-iter-concepts.html#iterator-traversal-concepts-lib-iterator-traversal">iterator traversal concept</a> our
<tt class="docutils literal"><span class="pre">node_iterator</span></tt> is going to model. Singly-linked lists only have
forward links, so our iterator can't can't be a <a class="reference external" href="new-iter-concepts.html#bidirectional-traversal-iterators-lib-bidirectional-traversal-iterators">bidirectional
traversal iterator</a>. Our iterator should be able to make multiple
passes over the same linked list (unlike, say, an
<tt class="docutils literal"><span class="pre">istream_iterator</span></tt> which consumes the stream it traverses), so it
must be a <a class="reference external" href="new-iter-concepts.html#forward-traversal-iterators-lib-forward-traversal-iterators">forward traversal iterator</a>. Therefore, we'll pass
<tt class="docutils literal"><span class="pre">boost::forward_traversal_tag</span></tt> in this position<a class="footnote-reference" href="#category" id="id11"><sup>1</sup></a>.</p>
<table class="docutils footnote" frame="void" id="category" rules="none">
<colgroup><col class="label" /><col /></colgroup>
<tbody valign="top">
<tr><td class="label"><a class="fn-backref" href="#id11">[1]</a></td><td><tt class="docutils literal"><span class="pre">iterator_facade</span></tt> also supports old-style category
tags, so we could have passed <tt class="docutils literal"><span class="pre">std::forward_iterator_tag</span></tt> here;
either way, the resulting iterator's <tt class="docutils literal"><span class="pre">iterator_category</span></tt> will
end up being <tt class="docutils literal"><span class="pre">std::forward_iterator_tag</span></tt>.</td></tr>
</tbody>
</table>
</div>
<div class="section" id="id12">
<h4><a class="toc-backref" href="#id38"><tt class="docutils literal"><span class="pre">Reference</span></tt></a></h4>
<p>The <tt class="docutils literal"><span class="pre">Reference</span></tt> argument becomes the type returned by
<tt class="docutils literal"><span class="pre">node_iterator</span></tt>'s dereference operation, and will also be the
same as <tt class="docutils literal"><span class="pre">std::iterator_traits&lt;node_iterator&gt;::reference</span></tt>. The
library's default for this parameter is <tt class="docutils literal"><span class="pre">Value&amp;</span></tt>; since
<tt class="docutils literal"><span class="pre">node_base&amp;</span></tt> is a good choice for the iterator's <tt class="docutils literal"><span class="pre">reference</span></tt>
type, we can omit this argument, or pass <tt class="docutils literal"><span class="pre">use_default</span></tt>.</p>
</div>
<div class="section" id="difference">
<h4><a class="toc-backref" href="#id39"><tt class="docutils literal"><span class="pre">Difference</span></tt></a></h4>
<p>The <tt class="docutils literal"><span class="pre">Difference</span></tt> argument determines how the distance between
two <tt class="docutils literal"><span class="pre">node_iterator</span></tt>s will be measured and will also be the
same as <tt class="docutils literal"><span class="pre">std::iterator_traits&lt;node_iterator&gt;::difference_type</span></tt>.
The library's default for <tt class="docutils literal"><span class="pre">Difference</span></tt> is <tt class="docutils literal"><span class="pre">std::ptrdiff_t</span></tt>, an
appropriate type for measuring the distance between any two
addresses in memory, and one that works for almost any iterator,
so we can omit this argument, too.</p>
<p>The declaration of <tt class="docutils literal"><span class="pre">node_iterator</span></tt> will therefore look something
like:</p>
<pre class="literal-block">
# include &quot;node.hpp&quot;
# include &lt;boost/iterator/iterator_facade.hpp&gt;
class node_iterator
: public boost::iterator_facade&lt;
node_iterator
, node_base
, boost::forward_traversal_tag
&gt;
{
...
};
</pre>
</div>
</div>
<div class="section" id="constructors-and-data-members">
<h3><a class="toc-backref" href="#id40">Constructors and Data Members</a></h3>
<p>Next we need to decide how to represent the iterator's position.
This representation will take the form of data members, so we'll
also need to write constructors to initialize them. The
<tt class="docutils literal"><span class="pre">node_iterator</span></tt>'s position is quite naturally represented using
a pointer to a <tt class="docutils literal"><span class="pre">node_base</span></tt>. We'll need a constructor to build an
iterator from a <tt class="docutils literal"><span class="pre">node_base*</span></tt>, and a default constructor to
satisfy the <a class="reference external" href="new-iter-concepts.html#forward-traversal-iterators-lib-forward-traversal-iterators">forward traversal iterator</a> requirements<a class="footnote-reference" href="#default" id="id13"><sup>2</sup></a>.
Our <tt class="docutils literal"><span class="pre">node_iterator</span></tt> then becomes:</p>
<pre class="literal-block">
# include &quot;node.hpp&quot;
# include &lt;boost/iterator/iterator_facade.hpp&gt;
class node_iterator
: public boost::iterator_facade&lt;
node_iterator
, node_base
, boost::forward_traversal_tag
&gt;
{
public:
node_iterator()
: m_node(0)
{}
explicit node_iterator(node_base* p)
: m_node(p)
{}
private:
...
node_base* m_node;
};
</pre>
<table class="docutils footnote" frame="void" id="default" rules="none">
<colgroup><col class="label" /><col /></colgroup>
<tbody valign="top">
<tr><td class="label"><a class="fn-backref" href="#id13">[2]</a></td><td>Technically, the C++ standard places almost no
requirements on a default-constructed iterator, so if we were
really concerned with efficiency, we could've written the
default constructor to leave <tt class="docutils literal"><span class="pre">m_node</span></tt> uninitialized.</td></tr>
</tbody>
</table>
</div>
<div class="section" id="implementing-the-core-operations">
<h3><a class="toc-backref" href="#id41">Implementing the Core Operations</a></h3>
<p>The last step is to implement the <a class="reference internal" href="#core-operations">core operations</a> required by
the concepts we want our iterator to model. Referring to the
<a class="reference internal" href="#core-operations">table</a>, we can see that the first three rows are applicable
because <tt class="docutils literal"><span class="pre">node_iterator</span></tt> needs to satisfy the requirements for
<a class="reference external" href="new-iter-concepts.html#readable-iterators-lib-readable-iterators">readable iterator</a>, <a class="reference external" href="new-iter-concepts.html#single-pass-iterators-lib-single-pass-iterators">single pass iterator</a>, and <a class="reference external" href="new-iter-concepts.html#incrementable-iterators-lib-incrementable-iterators">incrementable
iterator</a>.</p>
<p>We therefore need to supply <tt class="docutils literal"><span class="pre">dereference</span></tt>,
<tt class="docutils literal"><span class="pre">equal</span></tt>, and <tt class="docutils literal"><span class="pre">increment</span></tt> members. We don't want these members
to become part of <tt class="docutils literal"><span class="pre">node_iterator</span></tt>'s public interface, so we can
make them private and grant friendship to
<tt class="docutils literal"><span class="pre">boost::iterator_core_access</span></tt>, a &quot;back-door&quot; that
<tt class="docutils literal"><span class="pre">iterator_facade</span></tt> uses to get access to the core operations:</p>
<pre class="literal-block">
# include &quot;node.hpp&quot;
# include &lt;boost/iterator/iterator_facade.hpp&gt;
class node_iterator
: public boost::iterator_facade&lt;
node_iterator
, node_base
, boost::forward_traversal_tag
&gt;
{
public:
node_iterator()
: m_node(0) {}
explicit node_iterator(node_base* p)
: m_node(p) {}
private:
friend class boost::iterator_core_access;
void increment() { m_node = m_node-&gt;next(); }
bool equal(node_iterator const&amp; other) const
{
return this-&gt;m_node == other.m_node;
}
node_base&amp; dereference() const { return *m_node; }
node_base* m_node;
};
</pre>
<p>Voilà; a complete and conforming readable, forward-traversal
iterator! For a working example of its use, see <a class="reference external" href="../example/node_iterator1.cpp">this program</a>.</p>
</div>
</div>
<div class="section" id="a-constant-node-iterator">
<h2><a class="toc-backref" href="#id42">A constant <tt class="docutils literal"><span class="pre">node_iterator</span></tt></a></h2>
<div class="sidebar">
<p class="first sidebar-title">Constant and Mutable iterators</p>
<p>The term <strong>mutable iterator</strong> means an iterator through which
the object it references (its &quot;referent&quot;) can be modified. A
<strong>constant iterator</strong> is one which doesn't allow modification of
its referent.</p>
<p>The words <em>constant</em> and <em>mutable</em> don't refer to the ability to
modify the iterator itself. For example, an <tt class="docutils literal"><span class="pre">int</span> <span class="pre">const*</span></tt> is a
non-<tt class="docutils literal"><span class="pre">const</span></tt> <em>constant iterator</em>, which can be incremented
but doesn't allow modification of its referent, and <tt class="docutils literal"><span class="pre">int*</span>
<span class="pre">const</span></tt> is a <tt class="docutils literal"><span class="pre">const</span></tt> <em>mutable iterator</em>, which cannot be
modified but which allows modification of its referent.</p>
<p class="last">Confusing? We agree, but those are the standard terms. It
probably doesn't help much that a container's constant iterator
is called <tt class="docutils literal"><span class="pre">const_iterator</span></tt>.</p>
</div>
<p>Now, our <tt class="docutils literal"><span class="pre">node_iterator</span></tt> gives clients access to both <tt class="docutils literal"><span class="pre">node</span></tt>'s <tt class="docutils literal"><span class="pre">print(std::ostream&amp;)</span> <span class="pre">const</span></tt> member function, but also its
mutating <tt class="docutils literal"><span class="pre">double_me()</span></tt> member. If we wanted to build a
<em>constant</em> <tt class="docutils literal"><span class="pre">node_iterator</span></tt>, we'd only have to make three
changes:</p>
<pre class="literal-block">
class const_node_iterator
: public boost::iterator_facade&lt;
const_node_iterator
, node_base <strong>const</strong>
, boost::forward_traversal_tag
&gt;
{
public:
const_node_iterator()
: m_node(0) {}
explicit const_node_iterator(node_base* p)
: m_node(p) {}
private:
friend class boost::iterator_core_access;
void increment() { m_node = m_node-&gt;next(); }
bool equal(const_node_iterator const&amp; other) const
{
return this-&gt;m_node == other.m_node;
}
node_base <strong>const</strong>&amp; dereference() const { return *m_node; }
node_base <strong>const</strong>* m_node;
};
</pre>
<div class="sidebar">
<p class="first sidebar-title"><tt class="docutils literal"><span class="pre">const</span></tt> and an iterator's <tt class="docutils literal"><span class="pre">value_type</span></tt></p>
<p class="last">The C++ standard requires an iterator's <tt class="docutils literal"><span class="pre">value_type</span></tt> <em>not</em> be
<tt class="docutils literal"><span class="pre">const</span></tt>-qualified, so <tt class="docutils literal"><span class="pre">iterator_facade</span></tt> strips the
<tt class="docutils literal"><span class="pre">const</span></tt> from its <tt class="docutils literal"><span class="pre">Value</span></tt> parameter in order to produce the
iterator's <tt class="docutils literal"><span class="pre">value_type</span></tt>. Making the <tt class="docutils literal"><span class="pre">Value</span></tt> argument
<tt class="docutils literal"><span class="pre">const</span></tt> provides a useful hint to <tt class="docutils literal"><span class="pre">iterator_facade</span></tt> that the
iterator is a <em>constant iterator</em>, and the default <tt class="docutils literal"><span class="pre">Reference</span></tt>
argument will be correct for all lvalue iterators.</p>
</div>
<p>As a matter of fact, <tt class="docutils literal"><span class="pre">node_iterator</span></tt> and <tt class="docutils literal"><span class="pre">const_node_iterator</span></tt>
are so similar that it makes sense to factor the common code out
into a template as follows:</p>
<pre class="literal-block">
template &lt;class Value&gt;
class node_iter
: public boost::iterator_facade&lt;
node_iter&lt;Value&gt;
, Value
, boost::forward_traversal_tag
&gt;
{
public:
node_iter()
: m_node(0) {}
explicit node_iter(Value* p)
: m_node(p) {}
private:
friend class boost::iterator_core_access;
bool equal(node_iter&lt;Value&gt; const&amp; other) const
{
return this-&gt;m_node == other.m_node;
}
void increment()
{ m_node = m_node-&gt;next(); }
Value&amp; dereference() const
{ return *m_node; }
Value* m_node;
};
typedef node_iter&lt;node_base&gt; node_iterator;
typedef node_iter&lt;node_base const&gt; node_const_iterator;
</pre>
</div>
<div class="section" id="interoperability">
<h2><a class="toc-backref" href="#id43">Interoperability</a></h2>
<p>Our <tt class="docutils literal"><span class="pre">const_node_iterator</span></tt> works perfectly well on its own, but
taken together with <tt class="docutils literal"><span class="pre">node_iterator</span></tt> it doesn't quite meet
expectations. For example, we'd like to be able to pass a
<tt class="docutils literal"><span class="pre">node_iterator</span></tt> where a <tt class="docutils literal"><span class="pre">node_const_iterator</span></tt> was expected,
just as you can with <tt class="docutils literal"><span class="pre">std::list&lt;int&gt;</span></tt>'s <tt class="docutils literal"><span class="pre">iterator</span></tt> and
<tt class="docutils literal"><span class="pre">const_iterator</span></tt>. Furthermore, given a <tt class="docutils literal"><span class="pre">node_iterator</span></tt> and a
<tt class="docutils literal"><span class="pre">node_const_iterator</span></tt> into the same list, we should be able to
compare them for equality.</p>
<p>This expected ability to use two different iterator types together
is known as <a class="reference external" href="new-iter-concepts.html#interoperable-iterators-lib-interoperable-iterators"><strong>interoperability</strong></a>. Achieving interoperability in
our case is as simple as templatizing the <tt class="docutils literal"><span class="pre">equal</span></tt> function and
adding a templatized converting constructor<a class="footnote-reference" href="#broken" id="id16"><sup>3</sup></a><a class="footnote-reference" href="#random" id="id17"><sup>4</sup></a>:</p>
<pre class="literal-block">
template &lt;class Value&gt;
class node_iter
: public boost::iterator_facade&lt;
node_iter&lt;Value&gt;
, Value
, boost::forward_traversal_tag
&gt;
{
public:
node_iter()
: m_node(0) {}
explicit node_iter(Value* p)
: m_node(p) {}
template &lt;class OtherValue&gt;
node_iter(node_iter&lt;OtherValue&gt; const&amp; other)
: m_node(other.m_node) {}
private:
friend class boost::iterator_core_access;
template &lt;class&gt; friend class node_iter;
template &lt;class OtherValue&gt;
bool equal(node_iter&lt;OtherValue&gt; const&amp; other) const
{
return this-&gt;m_node == other.m_node;
}
void increment()
{ m_node = m_node-&gt;next(); }
Value&amp; dereference() const
{ return *m_node; }
Value* m_node;
};
typedef impl::node_iterator&lt;node_base&gt; node_iterator;
typedef impl::node_iterator&lt;node_base const&gt; node_const_iterator;
</pre>
<table class="docutils footnote" frame="void" id="broken" rules="none">
<colgroup><col class="label" /><col /></colgroup>
<tbody valign="top">
<tr><td class="label"><a class="fn-backref" href="#id16">[3]</a></td><td>If you're using an older compiler and it can't handle
this example, see the <a class="reference external" href="../example/node_iterator2.hpp">example code</a> for workarounds.</td></tr>
</tbody>
</table>
<table class="docutils footnote" frame="void" id="random" rules="none">
<colgroup><col class="label" /><col /></colgroup>
<tbody valign="top">
<tr><td class="label"><a class="fn-backref" href="#id17">[4]</a></td><td>If <tt class="docutils literal"><span class="pre">node_iterator</span></tt> had been a <a class="reference external" href="new-iter-concepts.html#random-access-traversal-iterators-lib-random-access-traversal-iterators">random access
traversal iterator</a>, we'd have had to templatize its
<tt class="docutils literal"><span class="pre">distance_to</span></tt> function as well.</td></tr>
</tbody>
</table>
<p>You can see an example program which exercises our interoperable
iterators <a class="reference external" href="../example/node_iterator2.cpp">here</a>.</p>
</div>
<div class="section" id="telling-the-truth">
<h2><a class="toc-backref" href="#id44">Telling the Truth</a></h2>
<p>Now <tt class="docutils literal"><span class="pre">node_iterator</span></tt> and <tt class="docutils literal"><span class="pre">node_const_iterator</span></tt> behave exactly as
you'd expect... almost. We can compare them and we can convert in
one direction: from <tt class="docutils literal"><span class="pre">node_iterator</span></tt> to <tt class="docutils literal"><span class="pre">node_const_iterator</span></tt>.
If we try to convert from <tt class="docutils literal"><span class="pre">node_const_iterator</span></tt> to
<tt class="docutils literal"><span class="pre">node_iterator</span></tt>, we'll get an error when the converting
constructor tries to initialize <tt class="docutils literal"><span class="pre">node_iterator</span></tt>'s <tt class="docutils literal"><span class="pre">m_node</span></tt>, a
<tt class="docutils literal"><span class="pre">node*</span></tt> with a <tt class="docutils literal"><span class="pre">node</span> <span class="pre">const*</span></tt>. So what's the problem?</p>
<p>The problem is that
<tt class="docutils literal"><span class="pre">boost::</span></tt><a class="reference external" href="../../type_traits/index.html#relationships"><tt class="docutils literal"><span class="pre">is_convertible</span></tt></a><tt class="docutils literal"><span class="pre">&lt;node_const_iterator,node_iterator&gt;::value</span></tt>
will be <tt class="docutils literal"><span class="pre">true</span></tt>, but it should be <tt class="docutils literal"><span class="pre">false</span></tt>. <a class="reference external" href="../../type_traits/index.html#relationships"><tt class="docutils literal"><span class="pre">is_convertible</span></tt></a>
lies because it can only see as far as the <em>declaration</em> of
<tt class="docutils literal"><span class="pre">node_iter</span></tt>'s converting constructor, but can't look inside at
the <em>definition</em> to make sure it will compile. A perfect solution
would make <tt class="docutils literal"><span class="pre">node_iter</span></tt>'s converting constructor disappear when
the <tt class="docutils literal"><span class="pre">m_node</span></tt> conversion would fail.</p>
<p>In fact, that sort of magic is possible using
<a class="reference external" href="../../utility/enable_if.html"><tt class="docutils literal"><span class="pre">boost::enable_if</span></tt></a>. By rewriting the converting constructor as
follows, we can remove it from the overload set when it's not
appropriate:</p>
<pre class="literal-block">
#include &lt;boost/type_traits/is_convertible.hpp&gt;
#include &lt;boost/utility/enable_if.hpp&gt;
...
private:
struct enabler {};
public:
template &lt;class OtherValue&gt;
node_iter(
node_iter&lt;OtherValue&gt; const&amp; other
, typename boost::enable_if&lt;
boost::is_convertible&lt;OtherValue*,Value*&gt;
, enabler
&gt;::type = enabler()
)
: m_node(other.m_node) {}
</pre>
</div>
<div class="section" id="wrap-up">
<h2><a class="toc-backref" href="#id45">Wrap Up</a></h2>
<p>This concludes our <tt class="docutils literal"><span class="pre">iterator_facade</span></tt> tutorial, but before you
stop reading we urge you to take a look at <a class="reference external" href="iterator_adaptor.html"><tt class="docutils literal"><span class="pre">iterator_adaptor</span></tt></a>.
There's another way to approach writing these iterators which might
even be superior.</p>
</div>
</div>
</div>
<div class="footer">
<hr class="footer" />
<a class="reference external" href="iterator_facade.rst">View document source</a>.
Generated by <a class="reference external" href="http://docutils.sourceforge.net/">Docutils</a> from <a class="reference external" href="http://docutils.sourceforge.net/rst.html">reStructuredText</a> source.
</div>
</body>
</html>