[Boost-commit] svn:boost r68991 - in trunk: boost/property_tree/detail libs/property_tree/test

Subject: [Boost-commit] svn:boost r68991 - in trunk: boost/property_tree/detail libs/property_tree/test
From: sebastian.redl_at_[hidden]
Date: 2011-02-18 11:16:26

Author: cornedbee
Date: 2011-02-18 11:16:25 EST (Fri, 18 Feb 2011)
New Revision: 68991
URL: http://svn.boost.org/trac/boost/changeset/68991

Log:
Apply patch from bug 1678 with slight modification: allow BOM in XML parsing for wide streams. Add test case.
Text files modified: 
   trunk/boost/property_tree/detail/rapidxml.hpp            |    18 ++++++++++++++----                      
   trunk/libs/property_tree/test/test_xml_parser_common.hpp |     6 ++++++                                  
   trunk/libs/property_tree/test/xml_parser_test_data.hpp   |     6 +++++-                                  
   3 files changed, 25 insertions(+), 5 deletions(-)

Modified: trunk/boost/property_tree/detail/rapidxml.hpp
==============================================================================
--- trunk/boost/property_tree/detail/rapidxml.hpp	(original)
+++ trunk/boost/property_tree/detail/rapidxml.hpp	2011-02-18 11:16:25 EST (Fri, 18 Feb 2011)
@@ -1701,16 +1701,26 @@
         ///////////////////////////////////////////////////////////////////////
         // Internal parsing functions
         
-        // Parse BOM, if any
+        // Parse UTF-8 BOM, if any
         template<int Flags>
-        void parse_bom(Ch *&text)
+        void parse_bom(char *&text)
         {
-            // UTF-8?
             if (static_cast<unsigned char>(text[0]) == 0xEF && 
                 static_cast<unsigned char>(text[1]) == 0xBB && 
                 static_cast<unsigned char>(text[2]) == 0xBF)
             {
-                text += 3;      // Skip utf-8 bom
+                text += 3;
+            }
+        }
+        
+        // Parse UTF-16/32 BOM, if any
+        template<int Flags>
+        void parse_bom(wchar_t *&text)
+        {
+            const wchar_t bom = 0xFEFF;
+            if (text[0] == bom)
+            {
+                ++text;
             }
         }
 

Modified: trunk/libs/property_tree/test/test_xml_parser_common.hpp
==============================================================================
--- trunk/libs/property_tree/test/test_xml_parser_common.hpp	(original)
+++ trunk/libs/property_tree/test/test_xml_parser_common.hpp	2011-02-18 11:16:25 EST (Fri, 18 Feb 2011)
@@ -124,6 +124,12 @@
         ReadFuncWS(), WriteFuncWS(), bug_data_pr2855, NULL,
         "testpr2855.xml", NULL, "testpr2855out.xml", 3, 7, 14
     );
+    
+    generic_parser_test_ok<Ptree, ReadFuncWS, WriteFuncWS>
+    (
+        ReadFuncWS(), WriteFuncWS(), bug_data_pr1678, NULL,
+        "testpr1678.xml", NULL, "testpr1678out.xml", 2, 0, 4
+    );
 
 }
 

Modified: trunk/libs/property_tree/test/xml_parser_test_data.hpp
==============================================================================
--- trunk/libs/property_tree/test/xml_parser_test_data.hpp	(original)
+++ trunk/libs/property_tree/test/xml_parser_test_data.hpp	2011-02-18 11:16:25 EST (Fri, 18 Feb 2011)
@@ -761,7 +761,11 @@
     "<start>";      // XML tag not closed
 
 const char *bug_data_pr2855 =
-    "<?xml version=\"1.0\" encoding\"utf-8\"?>\n"
+    "<?xml version=\"1.0\" encoding=\"utf-8\"?>\n"
     "<start> notrim</start>";
 
+// byte order mark
+const char *bug_data_pr1678 =
+    "\xEF\xBB\xBF<?xml version=\"1.0\" encoding=\"utf-8\"?><root/>";
+
 #endif