[Libreoffice-commits] update string copy semantics to be undefined in a non-crashing way
Stephan Bergmann
sbergman at redhat.com
Thu Oct 4 02:34:54 PDT 2012
On 10/02/2012 07:40 PM, Libreoffice Gerrit user wrote:
> commit 595ba03fd4769c069bbe0431ab878956ef1c37ad
> Author: Michael Meeks <michael.meeks at suse.com>
> Date: Tue Oct 2 18:39:30 2012 +0100
>
> update string copy semantics to be undefined in a non-crashing way.
>
> Change-Id: I03bb4db5931932280e368012cbaee6bef2854dd6
>
> diff --git a/sal/inc/rtl/string.hxx b/sal/inc/rtl/string.hxx
> index 84ab48e..d58999b 100644
> --- a/sal/inc/rtl/string.hxx
> +++ b/sal/inc/rtl/string.hxx
> @@ -1024,31 +1024,27 @@ public:
> /**
> Returns a new string that is a substring of this string.
>
> - The substring begins at the specified beginIndex. It is an error for
> - beginIndex to be negative or to be greater than the length of this string.
> + The substring begins at the specified beginIndex. If
> + beginIndex is negative or be greater than the length of
> + this string, behaviour is undefined.
Given that "it is an error for X to happen" and "if X happens, behaviour
is undefined" have exactly the same meaning (at least in my
understanding of computing), I wonder whether this is just a harmless
rephrasing, or whether there is a deeper misunderstanding lurking there.
> diff --git a/sal/inc/rtl/ustring.hxx b/sal/inc/rtl/ustring.hxx
> index b008054..e12a4b0 100644
> --- a/sal/inc/rtl/ustring.hxx
> +++ b/sal/inc/rtl/ustring.hxx
> @@ -1372,17 +1368,9 @@ public:
> */
> OUString copy( sal_Int32 beginIndex, sal_Int32 count ) const SAL_THROW(())
> {
> - assert(beginIndex >= 0 && beginIndex <= getLength() && count >= 0
> - && sal::static_int_cast< sal_uInt32 >(count) <=
> - sal::static_int_cast< sal_uInt32 >(getLength() - beginIndex));
> - if ( (beginIndex == 0) && (count == getLength()) )
> - return *this;
> - else
> - {
> - rtl_uString* pNew = 0;
> - rtl_uString_newFromStr_WithLength( &pNew, pData->buffer+beginIndex, count );
> - return OUString( pNew, (DO_NOT_ACQUIRE*)0 );
> - }
> + rtl_uString *pNew = 0;
> + rtl_uString_newFromSubString( &pNew, pData, beginIndex, count );
> + return OUString( pNew, (DO_NOT_ACQUIRE*)0 );
> }
>
> /**
> diff --git a/sal/rtl/source/strtmpl.cxx b/sal/rtl/source/strtmpl.cxx
> index ebb0da1..6069349 100644
> --- a/sal/rtl/source/strtmpl.cxx
> +++ b/sal/rtl/source/strtmpl.cxx
> @@ -1194,6 +1194,29 @@ void SAL_CALL IMPL_RTL_STRINGNAME( newFromStr_WithLength )( IMPL_RTL_STRINGDATA*
>
> /* ----------------------------------------------------------------------- */
>
> +void SAL_CALL IMPL_RTL_STRINGNAME( newFromSubString )( IMPL_RTL_STRINGDATA** ppThis,
> + const IMPL_RTL_STRINGDATA* pFrom,
> + sal_Int32 beginIndex,
> + sal_Int32 count )
> + SAL_THROW_EXTERN_C()
> +{
> + if ( beginIndex == 0 && count == pFrom->length )
> + {
> + IMPL_RTL_STRINGNAME( assign )( ppThis, const_cast< IMPL_RTL_STRINGDATA * >( pFrom ) );
> + return;
> + }
> + if ( count < 0 || beginIndex < 0 || beginIndex + count > pFrom->length )
Note how the original code above prevented problems with overflowing
beginIndex + count.
> + {
> + OSL_FAIL( "Out of bounds substring access" );
New code should use SAL_WARN, please. (Arguably, it should rather use
abort. Which would also make the following code moot.)
> + IMPL_RTL_STRINGNAME( newFromLiteral )( ppThis, "!!br0ken!!", 10, 0 );
> + return;
> + }
> +
> + IMPL_RTL_STRINGNAME( newFromStr_WithLength )( ppThis, pFrom->buffer + beginIndex, count );
> +}
> +
> +/* ----------------------------------------------------------------------- */
> +
> // Used when creating from string literals.
> void SAL_CALL IMPL_RTL_STRINGNAME( newFromLiteral )( IMPL_RTL_STRINGDATA** ppThis,
> const sal_Char* pCharStr,
> diff --git a/sal/util/sal.map b/sal/util/sal.map
> index 7a431c3..9efed6a 100644
> --- a/sal/util/sal.map
> +++ b/sal/util/sal.map
> @@ -627,6 +627,12 @@ LIBO_UDK_3.6 { # symbols available in >= LibO 3.6
> rtl_uStringBuffer_makeStringAndClear;
> } UDK_3.10;
>
> +LIBO_UDK_3.7 { # symbols available in >= LibO 3.7
> + global:
> + rtl_string_newFromSubString;
> + rtl_uString_newFromSubString;
> +} UDK_3.10;
Oh, what should have been a linear list of versions turned into a
multi-pronged fork starting with LIBO_UDK_3.5. Fixed that now for
LIBO_UDK_3.7 at least.
Stephan
More information about the LibreOffice
mailing list