=================
Repacking Status:
-----------------

John has graciously proposed that Az does the repacking of the SSC post-FTZ 
files into Az extensions.  IST will continue our FTZ role and will have scripts 
to back up the AZ repacking process if needed (more manually intensive).

What John/Az has requested (Aug2002):
-------------------------------------


>For the reformatting into extension files issue, there are 3 basic
>categories: Scan, Phot, IERs.
>
>
>SCAN:
>	One fits file per array per scan leg. I think
>	that means equivalently one file per array per
>	exposure.
>
>PHOT, SED, TPM:
>	One fits file per array per primary band per AOR.
>	By primary band I mean the band was selected for
>	the AOR. If the AOR calls for 24 and 70microns,
>	then there are 2 primary bands. We'd want two sets
>	of fits files, each set containing 3 files (one for
>	each array), the first set containing the data from 
>	when the 24um array was pointed at the target, the 
>	second set from when the 70um array was pointed. In 
>	this case each file could include data from more than 
>	one exposure command, e.g. the pre-first-stim exposure, 
>	the first dither pattern, the next pre-first-stim 
>	exposure, and the second dither pattern.
>
>	CLUSTER / RASTER-MAP: 
>		Include DCEs from all cluster/raster pointings
>		in a single file, subject to provisions above.
>IERs:
>	These will usually just need to be packaged into files
>	containing single exposures, but we need to actually go
>	through the IERs we have and decide what is needed for
>	each of them, and then capture that information someplace
>	where it can be used during IOC. 
>


================================================================
How to do this (assuming SSC pointing+FPG+uplink/SODB works...):
----------------------------------------------------------------
Quick summary (doesn't cover all cases):
For each AOR read EXPTYPE of DCENUM=0 and EXPID=0 for one array.
If EXPTYPE='scn', then packup by EXPID and CHNLNUM. 
(where EXPID and CHNLNUM can be parsed from the filename).

If EXPTYPE='pht', then need to pack up by FOVID, CHNLNUM, and CLPOSNUM 
Just need to check the DCENUM=0 files for all EXPIDs for one array.

IF EXPTYPE='d*', then this is a dark-ier and need to pack by EXPTYPE (or 
CSM_PIDX), EXPTIME, RDOUTMOD, and CHNLNUM.
Just need to check the DCENUM=0 files for all EXPIDs for one array.

If pre-FPG file, then FOVID-->APERTURE  (yes I know this is stupid).

If no SODB, then use IST backup plans.

--------

Current Archive naming convention (can be completely changed if desired):
The archive structure is the following:
on SSCIST21:

/mipsdata
  CAMPAIGN
     |
  --------------
   AORKEY    EXT
     |
---------------
MIPL TRAN  BCD


This data set includes 
CAMPAIGN=MIPS000400  and MIPS000500
MIPL=raw JPL files
TRAN=post tranhead (no pointing, no FPG)
BCD=bcd_fp files BCD+pointing+FPG
EXT=repacked extension files for Az

Example file names are:
MIPS.1.0123456789.1234.4321.01.mipl.fits
MIPS.1.0123456789.1234.4321.tran.fits
MIPS.1.0123456789.1234.4321.bcd.fits
mips.1.0123456789_pht_24_3sec.ext.fits

Should use EXPTYPE and whatever is needed for repacking in the ext.fits 
filename.


------

Current code attached:

appex.c does the work on the fits file (Makefile may help with libs ect.)
REPACK.pl runs appex for an input list of file names.
exprepack.csh packs up a user specified set range of EXPIDs (first-last) for 
each channel (loops through channels, expids for an AOR and gives output file 
names with a user specified extension).  exprepack calls run_appex.pl which 
calls appex (standard ssc way of doing things....lots of wrapper scripts....).


-----------

Well this should be enough to start things off......

-Dave