This page describes how to add fits.gz files in /archive/new_release to the public EtaCar database/website.

You want to complete the following tasks on Amonra, as that's where the database is.

The website uses two types of files it needs to know about and have. 1) The actual fits files, at /archive/new_release 2) The preview images, at /archive/new_release_pics

It is assumed that the files in 1 already exist, and the files in 2 are generated from the files in 1. Furthermore, the database needs information from both files.

So, first, to add the actual fits files to the database, preview images ignored for the moment:

In /home/jtolds/fitsfiles, there is 4 significant perl scripts: insertline-hstimages.pl, insertline-spectra.pl, updateline-hstimages.pl, and updateline-spectra.pl

In theory, the difference between the insert and update scripts should be just the style of generated SQL. As of 5/28/8, the hstimage scripts are in sync in this regard, while the spectra ones are not.

insertline-spectra.pl takes a list of fits.gz files as arguments and outputs SQL. you should just be able to dump this into the amonra EtaCar MySQL database.

If Roberta or Kris or John want to make a note on an observation, /software/noteupdater is where to start

Preview images need to be cropped, and John Martin wrote a routine to determine where. '/home/jtolds/archivefiles/imagegeneration/image_rowdb/getimgarea' is a binary John made does this. create_row_info.pl in the same directory is a script that I wrote to wrap getimgarea. It takes a list of fits.gz files as args, and outputs SQL on STDOUT and tabular data on STDERR. The SQL is to be dumped into the EtaCar db, and the tabular data is used by the preview image creation program.

In /home/jtolds/archivefiles/imagegeneration/make_pics', there are more important scripts. create-imagebounds will regenerate the tabular data from stderr of the previous program from the database in case you lost it. create-preview4 takes the tabular data file as an argument and creates preview images in the appropriate places. create-preview4-hacknstack is the same, except it first uses the 'hacknstack' binary that John wrote. This is useful for very wide and short observations. Instead of having to scroll right-to-left, the image is cut off at a certain amount of observation pixels and wrapped.

As of 5/28/8, there is also a create-preview3 script, but I believe this is just an outdated version.