[Libreoffice-commits] core.git: 3 commits - filter/source sal/rtl

[Michael Stahl]

 filter/source/xmlfilterdetect/filterdetect.cxx |    7 ++++++-
 sal/rtl/ustring.cxx                            |   24 +++++++++++++-----------
 2 files changed, 19 insertions(+), 12 deletions(-)

New commits:
commit ba08fe174660dc4abdda0aec02dc52b9c7fdb7b5
Author: Michael Stahl <mstahl at redhat.com>
Date:   Thu Nov 6 18:32:35 2014 +0100

    filter: use more appropriate conversion from ASCII for arbitrary input
    
    ... with unknown encoding, mapping non-ASCII chars to something that
    doesn't cause trouble (in FilterDetect::detect()).
    
    Change-Id: Ibf2a2e2fd7c0c00e55042d2ccad173fab7a1b0bd

diff --git a/filter/source/xmlfilterdetect/filterdetect.cxx b/filter/source/xmlfilterdetect/filterdetect.cxx
index a6f0089..b4d614f 100644
--- a/filter/source/xmlfilterdetect/filterdetect.cxx
+++ b/filter/source/xmlfilterdetect/filterdetect.cxx
@@ -115,7 +115,12 @@ OUString SAL_CALL FilterDetect::detect( com::sun::star::uno::Sequence< com::sun:
         }
 
         if ( nUniPos == 3 || ( nUniPos == 0 && !bTryUtf16 ) ) // UTF-8 or non-Unicode
-            resultString = OStringToOUString( read_uInt8s_ToOString( *pInStream, nSize ), RTL_TEXTENCODING_UTF8 );
+        {
+            OString const str(read_uInt8s_ToOString(*pInStream, nSize));
+            resultString = OUString(str.getStr(), str.getLength(),
+                RTL_TEXTENCODING_ASCII_US,
+                RTL_TEXTTOUNICODE_FLAGS_UNDEFINED_DEFAULT|RTL_TEXTTOUNICODE_FLAGS_MBUNDEFINED_DEFAULT|RTL_TEXTTOUNICODE_FLAGS_INVALID_DEFAULT);
+        }
         else if ( nUniPos == 2 || bTryUtf16 ) // UTF-16
             resultString = read_uInt16s_ToOUString( *pInStream, nSize );
 
commit aaccd8560e555d7e3cf062ca3672b77702397a70
Author: Michael Stahl <mstahl at redhat.com>
Date:   Thu Nov 6 18:31:24 2014 +0100

    sal: it is impossible to have non-ASCII char here
    
    ...because rtl_ImplGetFastUTF8UnicodeLen has already checked that.
    
    Change-Id: I17f2b80f374073934a8f0b1a97099d4dec89ce4e

diff --git a/sal/rtl/ustring.cxx b/sal/rtl/ustring.cxx
index 1d30dcc..3e4f5c3 100644
--- a/sal/rtl/ustring.cxx
+++ b/sal/rtl/ustring.cxx
@@ -696,10 +696,7 @@ retry:
                     pBuffer = (*ppThis)->buffer;
                     do
                     {
-                        /* Check ASCII range */
-                        SAL_WARN_IF( ((unsigned char)*pStr) > 127, "rtl.string",
-                                    "rtl_string2UString_status() - UTF8 test encoding is wrong" );
-
+                        assert(((unsigned char)*pStr) <= 127);
                         *pBuffer = *pStr;
                         pBuffer++;
                         pStr++;
commit 5936a64c2aa275992ce231183b35711d8da74ace
Author: Michael Stahl <mstahl at redhat.com>
Date:   Thu Nov 6 18:27:25 2014 +0100

    sal: do not ignore conversion flags for RTL_TEXTENCODING_ASCII_US
    
    Keep the fast path fast, fall back to the text encoder in case there's a
    fly in the ointment.
    
    Change-Id: I94507856a7f3170f770adb741aa1e282d0d2400c

diff --git a/sal/rtl/ustring.cxx b/sal/rtl/ustring.cxx
index c7622bd..1d30dcc 100644
--- a/sal/rtl/ustring.cxx
+++ b/sal/rtl/ustring.cxx
@@ -641,23 +641,28 @@ static void rtl_string2UString_status( rtl_uString** ppThis,
                 return;
             }
             pBuffer = (*ppThis)->buffer;
+            sal_Int32 nLenCopy(nLen);
+            const sal_Char *pStrCopy(pStr);
             do
             {
                 /* Check ASCII range */
-                SAL_WARN_IF( ((unsigned char)*pStr) > 127, "rtl.string",
-                            "rtl_string2UString_status() - Found char > 127 and RTL_TEXTENCODING_ASCII_US is specified" );
+                if (static_cast<unsigned char>(*pStrCopy) > 127)
+                {
+                    rtl_uString_release(*ppThis);
+                    goto retry; // cancel loop - try again with the converter
+                }
 
-                *pBuffer = *pStr;
+                *pBuffer = *pStrCopy;
                 pBuffer++;
-                pStr++;
-                nLen--;
+                pStrCopy++;
+                nLenCopy--;
             }
-            while ( nLen );
+            while (nLenCopy);
             if (pInfo != NULL) {
                 *pInfo = 0;
             }
         }
-        else
+retry:
         {
             rtl_uString*                pTemp;
             rtl_uString*                pTemp2 = NULL;