> add a statement like:
> tree->SearchActiveBranches("file.C");
> ( ... )
> SearchActiveBranches will read the ascii file file.C scanning
> ( ... )
> For each referenced variable, I will automatically call
> tree->SetBranchStatus. That's all !!
> You can imagine (and I trust you to propose additions/substractions)
> a more elaborated mechanism to take into account the case where
> the analysis code is spread over several files.
The problem does not begin first when the source is "spread".
Imagine you have leaf names in commentars, or in conditional code '#if ...
#endif'. The 'scanning' routine will see them all.
And it a leaf name is computed 'on the fly' it won't see it at all.
I hoped you are building some kind of "watchpoints" into the "new
algorithm" to generate traps when a particular leaf is accessed.
> My guess is that this algorithm will be more performant than you algorithm
> in most cases.
This migh be. But I accept this price for the total freedom.
> One could also imagine to combine this algorithm with your proposed
> algorithm to take into account variables used only conditionally.
Note - I don't ( DO NOT ) propose to replace the TTree:MakeClass with my
MakeClass.cxx macro. Most probably, they both have their place in life.
But if you find it useful, feel free to do it.
Jacek.
This archive was generated by hypermail 2b29 : Tue Jan 04 2000 - 00:43:32 MET