[Boost-commit] svn:boost r50652 - in sandbox/filesystem-v3: boost/filesystem libs/filesystem/src libs/filesystem/test

Subject: [Boost-commit] svn:boost r50652 - in sandbox/filesystem-v3: boost/filesystem libs/filesystem/src libs/filesystem/test
From: bdawes_at_[hidden]
Date: 2009-01-17 09:36:47

Author: bemandawes
Date: 2009-01-17 09:36:46 EST (Sat, 17 Jan 2009)
New Revision: 50652
URL: http://svn.boost.org/trac/boost/changeset/50652

Log:
Filesystem.v3: path_unit_test working on Linux/GCC
Text files modified: 
   sandbox/filesystem-v3/boost/filesystem/path.hpp               |     3 -                                       
   sandbox/filesystem-v3/libs/filesystem/src/path.cpp            |    27 ++++++++++++---                         
   sandbox/filesystem-v3/libs/filesystem/test/path_unit_test.cpp |    70 +++++++++++++++++++++++++-------------- 
   3 files changed, 66 insertions(+), 34 deletions(-)

Modified: sandbox/filesystem-v3/boost/filesystem/path.hpp
==============================================================================
--- sandbox/filesystem-v3/boost/filesystem/path.hpp	(original)
+++ sandbox/filesystem-v3/boost/filesystem/path.hpp	2009-01-17 09:36:46 EST (Sat, 17 Jan 2009)
@@ -17,7 +17,6 @@
                               TO DO
 
    * Windows, POSIX, conversions for char16_t, char32_t for supporting compilers.
-   * Windows, POSIX, conversions for user-defined types.
    * Fix inserter/extractor problems; see comments in commented out code below.
    * Need an error category for codecvt errors. Seed path.cpp detail::append, etc.
    * Add Alternate Data Stream test cases. See http://en.wikipedia.org/wiki/NTFS Features.
@@ -34,7 +33,7 @@
    * path.cpp: locale and detail append/convert need error handling.
    * path_unit_test needs to probe error handling, verify exceptions are thrown when
      requested.
-   * Provide the name check functions for more character types. Templatize?
+   * Provide the name check functions for more character types? Templatize?
      
                          Design Questions
 

Modified: sandbox/filesystem-v3/libs/filesystem/src/path.cpp
==============================================================================
--- sandbox/filesystem-v3/libs/filesystem/src/path.cpp	(original)
+++ sandbox/filesystem-v3/libs/filesystem/src/path.cpp	2009-01-17 09:36:46 EST (Sat, 17 Jan 2009)
@@ -18,6 +18,11 @@
 #include <cstring>
 #include <cassert>
 
+#ifdef BOOST_FILESYSTEM_DEBUG
+# include <iostream>
+# include <iomanip>
+#endif
+
 #ifdef BOOST_WINDOWS_PATH
 #  include <windows.h>
 #endif
@@ -764,13 +769,23 @@
                    value_type * to, value_type * to_end,
                    string_type & target, error_code & ec )
   {
+    //std::cout << std::hex
+    //          << " from=" << std::size_t(from)
+    //          << " from_end=" << std::size_t(from_end)
+    //          << " to=" << std::size_t(to)
+    //          << " to_end=" << std::size_t(to_end)
+    //          << std::endl;
+
     std::mbstate_t state  = std::mbstate_t();  // perhaps unneeded, but cuts bug reports
     const interface_value_type * from_next;
     value_type * to_next;
 
-    if ( wchar_t_codecvt_facet()->APPEND_DIRECTION( state, from, from_end, from_next,
-           to, to_end, to_next ) != std::codecvt_base::ok )
+    std::codecvt_base::result res;
+
+    if ( (res=wchar_t_codecvt_facet()->APPEND_DIRECTION( state, from, from_end, from_next,
+           to, to_end, to_next )) != std::codecvt_base::ok )
     {
+      //std::cout << " result is " << static_cast<int>(res) << std::endl;
       assert( 0 && "append error handling not implemented yet" );
       throw "append error handling not implemented yet";
     }
@@ -795,7 +810,7 @@
       return;
     }
 
-    std::size_t buf_size = end - begin;  // perhaps too large, but that's OK
+    std::size_t buf_size = (end - begin) * 3;  // perhaps too large, but that's OK
 
     //  dynamically allocate a buffer only if source is unusually large
     if ( buf_size > default_codecvt_buf_size )
@@ -806,7 +821,7 @@
     else
     {
       value_type buf[default_codecvt_buf_size];
-      do_append( begin, end, buf, buf+buf_size, target, ec );
+      do_append( begin, end, buf, buf+default_codecvt_buf_size, target, ec );
     }
   }
 
@@ -844,8 +859,8 @@
       return interface_string_type();
     }
 
-    //  The codecvt length functions may not be implemented, and I don't reall
-    //  understand them either. Thus this code is really just a guess; it it turns
+    //  The codecvt length functions may not be implemented, and I don't really
+    //  understand them either. Thus this code is just a guess; if it turns
     //  out the buffer is too small then an error will be reported and the code
     //  will have to be fixed.
     std::size_t buf_size = src.size() * 4;

Modified: sandbox/filesystem-v3/libs/filesystem/test/path_unit_test.cpp
==============================================================================
--- sandbox/filesystem-v3/libs/filesystem/test/path_unit_test.cpp	(original)
+++ sandbox/filesystem-v3/libs/filesystem/test/path_unit_test.cpp	2009-01-17 09:36:46 EST (Sat, 17 Jan 2009)
@@ -49,14 +49,6 @@
 #define PATH_IS( a, b ) check_path( a, b, __FILE__, __LINE__ )
 #define IS( a,b ) check_equal( a, b, __FILE__, __LINE__ )
 
-//  It is convenient for editing to have these ready to copy and paste
-# ifdef BOOST_WINDOWS_PATH
-# else  // BOOST_POSIX_PATH
-# endif
-
-# ifndef BOOST_FILESYSTEM_NARROW_ONLY
-# endif  // narrow only
-
 namespace
 {
   int errors;
@@ -138,28 +130,30 @@
     path x2(x1);                                       // #2
     PATH_IS(x2, BOO_LIT("path x1"));
 
-    path x4(string("std::string::iterator").begin());  // #3
-    PATH_IS(x4, BOO_LIT("std::string::iterator"));
-
-    path x7(s.begin(), s.end());                       // #4
-    PATH_IS(x7, BOO_LIT("string iterators"));
-
-    path x10(string("std::string"));                   // #5
-    PATH_IS(x10, BOO_LIT("std::string"));
-
-# ifndef BOOST_FILESYSTEM_NARROW_ONLY
     path x3(L"const wchar_t *");                       // #3
     PATH_IS(x3, L"const wchar_t *");
 
+    string s3a( "s3a.c_str()" );
+    path x3a( s3a.c_str() );                           // #3
+    PATH_IS(x3a, BOO_LIT("s3a.c_str()"));
+
+    path x4(string("std::string::iterator").begin());  // #3
+    PATH_IS(x4, BOO_LIT("std::string::iterator"));
+
     path x5(wstring(L"std::wstring::iterator").begin()); // #3
     PATH_IS(x5, L"std::wstring::iterator");
 
+    path x7(s.begin(), s.end());                       // #4
+    PATH_IS(x7, BOO_LIT("string iterators"));
+
     path x8(ws.begin(), ws.end());                     // #4
     PATH_IS(x8, L"wstring iterators");
 
+    path x10(string("std::string"));                   // #5
+    PATH_IS(x10, BOO_LIT("std::string"));
+
     path x11(wstring(L"std::wstring"));                // #5
     PATH_IS(x11, L"std::wstring");
-# endif
   }
 
   //  test_use_cases  ------------------------------------------------------------------//
@@ -201,7 +195,6 @@
     s2 = p;
     CHECK( s2 == "foo" );
 
-# ifndef BOOST_FILESYSTEM_NARROW_ONLY
     wstring ws1( p );
     CHECK( ws1 == L"foo" );
 
@@ -211,7 +204,6 @@
     ws2.clear();
     ws2 = p;
     CHECK( ws2 == L"foo" );
-# endif
 
   }
 
@@ -520,30 +512,44 @@
 
     //  \u2722 and \xE2\x9C\xA2 are UTF-16 and UTF-8 FOUR TEARDROP-SPOKED ASTERISK
 
+    std::cout << "  testing p0 ..." << std::endl;
     path p0( L"\u2722" );  // for tests that depend on detail::convert_to_string
+#   ifdef BOOST_WINDOWS_PATH
     CHECK( p0.string() != "\xE2\x9C\xA2" );
+#   endif
     string p0_string( p0.string() );
 
+    std::cout << "  testing p1 ..." << std::endl;
     path p1( "\xE2\x9C\xA2" );
+#   ifdef BOOST_WINDOWS_PATH
     CHECK( p1 != L"\u2722" );
+#   endif
     wstring p1_wstring( p1.wstring() );
 
     // So that tests are run with known encoding, use Boost UTF-8 codecvt
     std::locale global_loc = std::locale();
     std::locale loc( global_loc, new fs::detail::utf8_codecvt_facet );
+    std::cout << "  imbuing locale ..." << std::endl;
     std::locale old_loc = path::imbue( loc );
 
+    std::cout << "  testing with the imbued locale ..." << std::endl;
     CHECK( p0.string() == "\xE2\x9C\xA2" );
     path p2( "\xE2\x9C\xA2" );
     CHECK( p2 == L"\u2722" );
     CHECK( p2.wstring() == L"\u2722" );
 
+    std::cout << "  imbuing the original locale ..." << std::endl;
     path::imbue( old_loc );
 
+    std::cout << "  testing with the original locale ..." << std::endl;
     CHECK( p0.string() == p0_string );
     path p3( "\xE2\x9C\xA2" );
+#   ifdef BOOST_WINDOWS_PATH
     CHECK( p3 != L"\u2722" );
+#   endif
     CHECK( p3.wstring() == p1_wstring );
+
+    std::cout << "  locale testing complete" << std::endl;
   }
 
 //  //  test_locales  --------------------------------------------------------------------//
@@ -580,6 +586,18 @@
         target += *begin + 1;  // change so that results distinguishable from char cvts
     }
 
+#  ifdef __GNUC__
+    //  This specialization shouldn't be needed, and VC++, Intel, and others work
+    //  fine without it. But gcc 4.3.2, and presumably other versions, need it.
+    template<>
+    void append<user_string::value_type>( const user_string::value_type * begin,
+      string_type & target, system::error_code & ec )
+    {
+      path_traits::append<user_string::value_type>( begin,
+        static_cast<const user_string::value_type *>(0), target, ec );
+    }
+#  endif
+
     template<>
     user_string convert<user_string>( const string_type & source,
       system::error_code & ec )
@@ -604,11 +622,11 @@
     user_string::value_type usr_c_str[] = { 'a', 'b', 'c', 0 };
     user_string usr( usr_c_str );
 
-    //path p1( usr_c_str );
-    //CHECK( p1 == path("bcd") );
-    //CHECK( p1 == "bcd" );
-    //user_string s1( p1.string<user_string>() );
-    //CHECK( s1 == usr );
+    path p1( usr.c_str() );
+    CHECK( p1 == path("bcd") );
+    CHECK( p1 == "bcd" );
+    user_string s1( p1.string<user_string>() );
+    CHECK( s1 == usr );
   }
 
 }  // unnamed namespace