On Fri, Nov 07, 2003 at 06:55:21PM +0100, Perl Authors Upload Server wrote:
>
> The following module was proposed for inclusion in the Module List:
>
> modid: Data::Stream
> DSLIP: bdhhp
> description: Stream a data dump to a filehandle
> userid: YVES (Yves)
> chapterid: 6 (Data_Type_Utilities)
> communities:
> www.perlmonks.org
>
> similar:
> Data::BFDump Data::Dumper Data::Dump
>
> rationale:
>
> Data::Stream is a total rewrite of Data::BFDump. The name change is
> due to a completely new interface, and the addition of functionality
> which streams the output to a filehandle instead of constructing it
> in memory and then returning the total dump, (and user complaints
> that BFDump is annoying to type :-). This means that the memory
> overhead of dumping using Data::Stream is signifigantly lower than
> any of the other perl serializers.
>
> Data::Stream does a breadth first traversal of the object being
> dumped, but this pass is for analysis only. A second depth first
> pass is required to actually output the structure. This two pass
> approach means that Data::Stream can correctly handle a number of
> edge cases that none of the other existing serialization modules
> currently can. Extra effort was put into ensuring that structures
> with readonly and aliased elements are dumped correctly.
>
> Other module names I considered were Data::Streamer
> Data::Dumper::Streamer and Data::Serialize and also preserving the
> BFDump name. After discussions with various people from Perlmonks
> the consensus was that Data::Stream was the preferred choice.
A "Data::Stream" could be just about anything. Doesn't say what it does.
Although "stream" is _how_ it does it, "dump" is what it actually does.
So it should have dump in the name along with stream.
I'd suggest:
Data::StreamDump
Tim.
