The wavelength calibration files are usually delivered as an Excel spreadsheet
named 20150827CalibrationResults.xlsx.  The updates for new dates are added as
new columns to the existing spreadsheet.

To make something the pipeline understands:
1. Save the spreadsheet to a new name in this directory (data/wave_cal),
   indicating the current date.  For example, 20170210CalibrationResults.xlsx
   for a file received February 10, 2017.
2. Open the new Excel file.
3. Select all numerical fields and select Format->Cells.  Change the format
   to Number, with 30 decimal precision.
4. Delete any extra (non-date) fields in the Date(YearMonth) row.
5. For the new date, check that there are columns for:
     Red D105
     Red D130
     Blue 1st Order
     Blue 2nd Order
   in that order.  If there are not, make the appropriate columns, and
   copy the nearest available data.
6. Check that all 25 spexels have wavelength information for all dates.
   If not, copy it from the appropriate column for the previous date.
7. Save the Excel file.
8. Save a copy of the Excel file as a .csv file with the same base name
   (File->Save as...).
9. Compare the new CSV file (eg. with opendiff/FileMerge) to the most
   recent previous version and verify that the only change is the addition
   of new column(s) for additional wavelength calibration dates.
10. Copy the new CSV file to CalibrationResults.csv.

Notes by Dario Fadda:

Follow previous procedure from 1. to 5.
Then, save the excel file as csv.

At this point, use emacs to convert csv into org, by going into
org-mode (M-x org-mode).  Select all the data in the file and
convert it: C-u C-c |
Then, select the new columns and append to the end of the previous
calibration file.
Once this is done, reconvert to csv using:
Esc-x  org-table-export    name in .csv
