ab80: Adding changes of cws mib21, ab77, ab77run2, dr77i to dev300

ab77: #163789# Handle class module factories document specific
ab77: #163732# Defer removal of documents until XCloseListener::notifyClosing
ab77: #163808# make VBA symbols Application.ScreenUpdating and Application.Interactive work globally on all documents
ab77: wae
ab77: #163840# read UTF-8 BOM
ab77: #163732# VBA UserForm_Terminate triggered too late while closing document
ab77: minor correction
ab77: assertion: do not call ::rtl::OUString::copy() with negative index
ab77run2: #163869# do not call Class_Terminate VBA macros when document disposes
dr77i: #163941# do not update drawing layer when pasting from clipboard after cut
mib21: #163944# ignore trailing whitespace in Basic source lines
mib21: #163948# allow to manually pack MSVC DLLs into extensions
mib21: #163948# multiprocess build fix
Authors:
Andreas Bregas <ab@openoffice.org>
Daniel Rentz [dr] <daniel.rentz@oracle.com>
Eike Rathke [er] <eike.rathke@oracle.com>

2 files changed, 56 insertions, 7 deletions

diff --git a/tools/inc/tools/stream.hxx b/tools/inc/tools/stream.hxx
index 61af361aa5b7..440eb2283f20 100644
--- a/tools/inc/tools/stream.hxx
+++ b/tools/inc/tools/stream.hxx
@@ -460,9 +460,20 @@ public:
 
                 /// Switch to no endian swapping and write 0xfeff
     sal_Bool        StartWritingUnicodeText();
-                /// Read 16bit, if 0xfeff do nothing, if 0xfffe switch
-                /// endian swapping, if none of them put back
-    sal_Bool        StartReadingUnicodeText();
+
+                /** If eReadBomCharSet==RTL_TEXTENCODING_DONTKNOW: read 16bit,
+                    if 0xfeff do nothing (UTF-16), if 0xfffe switch endian
+                    swapping (UTF-16), if 0xefbb or 0xbbef read another byte
+                    and check for UTF-8. If no UTF-* BOM was detected put all
+                    read bytes back. This means that if 2 bytes were read it
+                    was an UTF-16 BOM, if 3 bytes were read it was an UTF-8
+                    BOM. There is no UTF-7, UTF-32 or UTF-EBCDIC BOM detection!
+
+                    If eReadBomCharSet!=RTL_TEXTENCODING_DONTKNOW: only read a
+                    BOM of that encoding and switch endian swapping if UTF-16
+                    and 0xfffe.
+                 */
+    sal_Bool        StartReadingUnicodeText( rtl_TextEncoding eReadBomCharSet );
 
                 /// Read a line of Unicode
     sal_Bool        ReadUniStringLine( String& rStr );
diff --git a/tools/source/stream/stream.cxx b/tools/source/stream/stream.cxx
index 9fdef8436f1a..92ce4214dfc4 100644
--- a/tools/source/stream/stream.cxx
+++ b/tools/source/stream/stream.cxx
@@ -1046,21 +1046,59 @@ sal_Bool SvStream::StartWritingUnicodeText()
 |*
 *************************************************************************/
 
-sal_Bool SvStream::StartReadingUnicodeText()
+sal_Bool SvStream::StartReadingUnicodeText( rtl_TextEncoding eReadBomCharSet )
 {
+    if (!(  eReadBomCharSet == RTL_TEXTENCODING_DONTKNOW ||
+            eReadBomCharSet == RTL_TEXTENCODING_UNICODE ||
+            eReadBomCharSet == RTL_TEXTENCODING_UTF8))
+        return sal_True;    // nothing to read
+
+    bool bTryUtf8 = false;
     sal_uInt16 nFlag;
+    sal_sSize nBack = sizeof(nFlag);
     *this >> nFlag;
     switch ( nFlag )
     {
         case 0xfeff :
-            // native
+            // native UTF-16
+            if (    eReadBomCharSet == RTL_TEXTENCODING_DONTKNOW ||
+                    eReadBomCharSet == RTL_TEXTENCODING_UNICODE)
+                nBack = 0;
         break;
         case 0xfffe :
-            SetEndianSwap( !bSwap );
+            // swapped UTF-16
+            if (    eReadBomCharSet == RTL_TEXTENCODING_DONTKNOW ||
+                    eReadBomCharSet == RTL_TEXTENCODING_UNICODE)
+            {
+                SetEndianSwap( !bSwap );
+                nBack = 0;
+            }
+        break;
+        case 0xefbb :
+            if (nNumberFormatInt == NUMBERFORMAT_INT_BIGENDIAN &&
+                    (eReadBomCharSet == RTL_TEXTENCODING_DONTKNOW ||
+                     eReadBomCharSet == RTL_TEXTENCODING_UTF8))
+                bTryUtf8 = true;
+        break;
+        case 0xbbef :
+            if (nNumberFormatInt == NUMBERFORMAT_INT_LITTLEENDIAN &&
+                    (eReadBomCharSet == RTL_TEXTENCODING_DONTKNOW ||
+                     eReadBomCharSet == RTL_TEXTENCODING_UTF8))
+                bTryUtf8 = true;
         break;
         default:
-            SeekRel( -((sal_sSize)sizeof(nFlag)) );      // no BOM, pure data
+            ;   // nothing
+    }
+    if (bTryUtf8)
+    {
+        sal_uChar nChar;
+        nBack += sizeof(nChar);
+        *this >> nChar;
+        if (nChar == 0xbf)
+            nBack = 0;      // it is UTF-8
     }
+    if (nBack)
+        SeekRel( -nBack );      // no BOM, pure data
     return nError == SVSTREAM_OK;
 }