Re: [Libreoffice-commits] core.git: Move string hash function into String class.

Hi Muthu,

tried to reach you on IRC yesterday:

> Feb 13 11:03:42 <sberg> muthusuba, 8f8bc0dcf3bc253ae49159d52db049767f476ced is broken: (a) the 
> general O[U]String::hashCode does sampling-only by design; if that's not considered good these days we 
> should change the implementation rather than add another function (is 64-bit vs. 32-bit hash code of any 
> added value? doubt it given most hash maps will not be too big, anyway); (b) adding OString::hashCode64 
> but not OUString::hashCode64? (c) rtl_str_hashCode64_WithLength must
> Feb 13 11:03:42 <sberg> go into a LIBO_UDK_4.3 section in sal.map, not UDK_3_0_0; (d) 
> rtl_ustr_hashCode64_WithLength missing from sal.map; (e) missing @since tags

Stephan


On 02/13/2014 08:41 AM, Muthu Subramanian wrote:
>   include/rtl/string.h       |   18 ++++++++++++++++++
>   include/rtl/string.hxx     |   13 +++++++++++++
>   include/rtl/ustring.h      |   18 ++++++++++++++++++
>   sal/rtl/strtmpl.cxx        |   13 +++++++++++++
>   sal/util/sal.map           |    1 +
>   sd/source/core/sdpage2.cxx |   12 +-----------
>   6 files changed, 64 insertions(+), 11 deletions(-)
>
> New commits:
> commit 8f8bc0dcf3bc253ae49159d52db049767f476ced
> Author: Muthu Subramanian <sumuthu@collabora.com>
> Date:   Thu Feb 13 13:13:53 2014 +0530
>
>      Move string hash function into String class.
>
>      hashCode() seems to do sampling while creating the hash.
>      hashCode64() will not.
>
>      Change-Id: Id30f5a2a774cf5244dbc00da9649e95a532484be
>
> diff --git a/include/rtl/string.h b/include/rtl/string.h
> index 32344bf..71e0955 100644
> --- a/include/rtl/string.h
> +++ b/include/rtl/string.h
> @@ -277,6 +277,24 @@ SAL_DLLPUBLIC sal_Int32 SAL_CALL rtl_str_hashCode(
>   SAL_DLLPUBLIC sal_Int32 SAL_CALL rtl_str_hashCode_WithLength(
>           const sal_Char * str, sal_Int32 len ) SAL_THROW_EXTERN_C();
>
> +/** Return a hash code (64bit) for a string.
> +
> +    It is not allowed to store the hash code persistently, because later
> +    versions could return other hash codes.
> +
> +    @param str
> +    a string.  Need not be null-terminated, but must be at least as long as
> +    the specified len.
> +
> +    @param len
> +    the length of the string.
> +
> +    @return
> +    a hash code for the given string.
> + */
> +SAL_DLLPUBLIC sal_uInt64 SAL_CALL rtl_str_hashCode64_WithLength(
> +        const sal_Char * str, sal_Int32 len ) SAL_THROW_EXTERN_C();
> +
>   /** Search for the first occurrence of a character within a string.
>
>       The string must be null-terminated.
> diff --git a/include/rtl/string.hxx b/include/rtl/string.hxx
> index fb7283b..24bb980 100644
> --- a/include/rtl/string.hxx
> +++ b/include/rtl/string.hxx
> @@ -892,6 +892,19 @@ public:
>       }
>
>       /**
> +      Returns a 64bit hash of the string data.
> +      This hashes the entire data, while hashCode would do sampling for larger string sizes.
> +
> +      @return   a hash code value of the string data
> +
> +      @see hashCode() for simple hashes
> +    */
> +    sal_uInt64 hashCode64() const SAL_THROW(())
> +    {
> +        return rtl_str_hashCode64_WithLength( pData->buffer, pData->length );
> +    }
> +
> +    /**
>         Returns a hashcode for this string.
>
>         @return   a hash code value for this object.
> diff --git a/include/rtl/ustring.h b/include/rtl/ustring.h
> index 80c6bcc..2069899 100644
> --- a/include/rtl/ustring.h
> +++ b/include/rtl/ustring.h
> @@ -551,6 +551,24 @@ SAL_DLLPUBLIC sal_Int32 SAL_CALL rtl_ustr_hashCode(
>   SAL_DLLPUBLIC sal_Int32 SAL_CALL rtl_ustr_hashCode_WithLength(
>           const sal_Unicode * str, sal_Int32 len ) SAL_THROW_EXTERN_C();
>
> +/** Return a hash code (64bit) for a string.
> +
> +    It is not allowed to store the hash code persistently, because later
> +    versions could return other hash codes.
> +
> +    @param str
> +    a string.  Need not be null-terminated, but must be at least as long as
> +    the specified len.
> +
> +    @param len
> +    the length of the string.
> +
> +    @return
> +    a hash code for the given string.
> + */
> +SAL_DLLPUBLIC sal_uInt64 SAL_CALL rtl_ustr_hashCode64_WithLength(
> +        const sal_Unicode * str, sal_Int32 len ) SAL_THROW_EXTERN_C();
> +
>   /** Search for the first occurrence of a character within a string.
>
>       The string must be null-terminated.
> diff --git a/sal/rtl/strtmpl.cxx b/sal/rtl/strtmpl.cxx
> index fc618d3..f2eac10 100644
> --- a/sal/rtl/strtmpl.cxx
> +++ b/sal/rtl/strtmpl.cxx
> @@ -252,6 +252,19 @@ sal_Int32 SAL_CALL IMPL_RTL_STRNAME( hashCode )( const IMPL_RTL_STRCODE* pStr )
>
>   /* ----------------------------------------------------------------------- */
>
> +sal_uInt64 SAL_CALL IMPL_RTL_STRNAME( hashCode64_WithLength )( const IMPL_RTL_STRCODE* pStr,
> +                                                               sal_Int32 nLen )
> +    SAL_THROW_EXTERN_C()
> +{
> +    sal_uInt64 nHash = 0;
> +
> +    for( sal_Int32 i = 0; i < nLen; i++ )
> +        nHash = (nHash << 5) - nHash + *pStr++;
> +    return nHash;
> +}
> +
> +/* ----------------------------------------------------------------------- */
> +
>   sal_Int32 SAL_CALL IMPL_RTL_STRNAME( hashCode_WithLength )( const IMPL_RTL_STRCODE* pStr,
>                                                               sal_Int32 nLen )
>       SAL_THROW_EXTERN_C()
> diff --git a/sal/util/sal.map b/sal/util/sal.map
> index 1456d6d..470da31 100644
> --- a/sal/util/sal.map
> +++ b/sal/util/sal.map
> @@ -221,6 +221,7 @@ UDK_3_0_0 {
>           rtl_str_shortenedCompareIgnoreAsciiCase_WithLength;
>           rtl_str_hashCode;
>           rtl_str_hashCode_WithLength;
> +        rtl_str_hashCode64_WithLength;
>           rtl_str_indexOfChar;
>           rtl_str_indexOfChar_WithLength;
>           rtl_str_indexOfStr;
> diff --git a/sd/source/core/sdpage2.cxx b/sd/source/core/sdpage2.cxx
> index 0f7f7cf..a870895 100644
> --- a/sd/source/core/sdpage2.cxx
> +++ b/sd/source/core/sdpage2.cxx
> @@ -54,16 +54,6 @@ using namespace ::com::sun::star::office;
>
>   extern void NotifyDocumentEvent( SdDrawDocument* pDocument, const OUString& rEventName, const Reference< 
> XInterface >& xSource );
>
> -static sal_uInt64 lcl_getHash( OString aString )
> -{
> -    sal_Int32  len = aString.getLength();
> -    sal_uInt64 nHash = 0;
> -
> -    for( sal_Int32 i = 0; i < len; i++ )
> -        nHash = (nHash << 5) - nHash + aString[i];
> -    return nHash;
> -}
> -
>   /*************************************************************************
>   |*
>   |* Sets: names of layout, master page links and templates for presentation
> @@ -615,7 +605,7 @@ OString SdPage::stringify() const
>
>   sal_uInt64 SdPage::getHash() const
>   {
> -    return lcl_getHash( stringify() );
> +    return stringify().hashCode64();
>   }