tags, index, file sizes and multiple event components

[Henrik Nordberg <HNordberg@lbl.gov>]

Dear Everyone,

Scripts to generate a list of file sizes for each event component are
located in CVS in QE/index/size. This how you use them:
Check out all the files from that directory. Edit a file called
component.hpss.paths. It should contain the name and HPSS path for each
component. For example:

tracks /home/starreco/FD_123/tracks
comp1 /home/starreco/FD_123/comp1
comp0 /home/starreco/FD_123/

Then run ./get_sizes.csh giving the boot file name as the only 
argument. The script reads component.hpss.paths and creates a file
called <component name>.sizes.txt for each component. (For tracks it
would be called tracks.sizes.txt.) Those files are later used to create
an index from FID to size.

The scripts no longer gets the size for files one at the time, but now
get a listing for an entire directory at the time. Then it matches the
filenames in that directory with the ones from the output of
oodumpcatalog. This is _much_ faster than the previous method.

ascii2vert, the program that creates the full scan index from the ASCII
output from the TagDB, has been updated to output a file called
components.txt. That file contains a list of the components that the
Storage Manager will know about ones the index is built. It can be used
as a template for component.hpss.paths. (You need to add the HPSS path
for each component, even if they are all the same.)

smDefs.idl and qe.idl have been updated. I added FID_T as a typedef
for SMQ_FID_T (which is awkward) to smDefs.idl. To the QE IDL interface
I added 
unsigned long getSize(in FID_T fid);
which can be used to get the size of a file, if you know the FID.
Feel free to use this function. It was created because the QE would
use it internally anyway, and it seemed to me as a useful function
for others too.

Regards, Henrik
_________________________________________
Henrik Nordberg       <hnordberg@lbl.gov>
Scientific Data Management Research Group
Lawrence Berkeley National Laboratory