Kludge for '92 Data

The '92 data was not skimmed off in the same fashion as post-'92. It was done in a standalone skim (by Tony Johnson). This was done to standardize on the so-called EIT Z-filter which became the standard for post-'92 running. The data is represented by a datacat in the VMS SLDPM area, called KLU92.DATACAT. There is no Oracle run-by-run information and the files are not organized by run. Furthermore, the version of DERAW.TEMPLATE used in the skim is now obsolete, so a copy is kept in $usr:[sldpm.recon.klu92].

A kludged set of scripts was devised to deal with '92 data. KLU92.COM emulates the HADCOPY jobs in that it reads a tape at a time and splits the data up into files by run onto disk$sld_rec_stg:[sld_data] and submits a super job per run.

KLU92 initially writes out files with no 'HAD' in the name, due to restrictions in IDAL's INTERP syntax. As each run boundary is reached, it does a SPAWN to execute SUB92KLU.COM which renames the file to put the 'HAD' prefix back in the filename and submits the SUPER job to analyze the run.

KLU92.COM takes a modified tape name, as laid out in the .DATACAT. Submission is manual, on a per tape basis. Once this job runs, the process is just like the regular scheme.

Tape jobs can be submitted using the SUB_HAD_KLU92.COM file which takes the tape name as input (not the dsn in the DATACAT: it will prepend the 'H'). Patches of individual runs can be performed using RESUB92KLU.COM, which takes the run number as input and assumes the input Jazzdata file is still on the input disk.

Note that these .COM files all have the task's name, and relevent related files hardwired into them!