Hi Martin,
In the 2.23/12 release notes, we talk about an extension
to support automatically variable length arrays.
We have modified several Root class header files where
variable length arrays are defined. The length of the array
can be specified in the comment field of the data member
declaration. The version of rootcint in 2.24 (now implemented)
is able to generate automatically valid statements in the Streamer
function. We realized that this was an important feature
missing for a frequent case.
The TTree::Branch functions will also be modified to support
this case in split mode.
Sorry if the release notes were ambiguous.
Rene Brun
On Fri, 17 Mar 2000 schulte@pktw09.phy.tu-dresden.de wrote:
> hello,
>
> i tried out the rootcint-extension:
>
> Float_t *fY; //[fN] Array of Y coordinates
>
> using ULong_t's in my class.
> it compiles, an i can write to, and read from a file.
>
> but THE VALUES IN THE ARRAY ARE NOT THE VALUES i have putten in.
>
> i loked in the streamer method (...Dict.C) an it seams, as if this
> behaviour
> is correct (although not satisfying) since i find the lines:
>
> //R__b.ReadArray(farraydata);
> and
> //R__b.WriteArray(farraydata, __COUNTER__);
> commented out.
>
> am i doing something wrong (missing something in the LinkDef, missing
> -,!,?,&,$
> at some place, ...) or is this feature not jet implemented.
> the statement :
> "Thanks to this change, rootcint will be able to
> automatically generate
> a valid Streamer for the class."(ROOT Version 2.23 Release Notes)
> is somewhat ambigious!
> ('... will NOW be able to ...' or '... will IN SOME FUTURE be able to
> ...')
>
>
> i'm using :
> Version 2.23/12 1 February 2000
> on aix.
>
> thanks,
>
> martin schulte-wissermann
>
>
This archive was generated by hypermail 2b29 : Tue Jan 02 2001 - 11:50:21 MET