DECODR will build a database-friendly csv using a cruise's Dickson-Cervantes pH-DIC-TA csv (spreadsheet export) and sta.csvs. This version adds two pH columns (pH1, pH2) left of the DIC & TA columns (DIC1, DIC2, TA1, TA2).

A slight modification to the DIC record parser - no "COFI" is needed in front of the DIC bottle number in column 1. But, there should be no description lines above the data table - the first line should be the data column headers; lines 2+ are data.

The line.sta calculator uses station position GPS latitude & longitude to calculate the actual CalCOFI station Line.Sta. There was a problem when a negative (W) longitude was used from the YYMMcasts.csv, saved by CESL. A negative longitude causes the decimal longitude parsing back to degrees to be negative, so the longitude minutes were miscalculated. This resulted in an incorrect 'actual line' (Ac_Line) & 'actual sta' Ac_Sta calculations during the csv>database conversion. 

DECODR1410 now checks for a longitude < 0 and multiplies it by -1 to insure a positive longitude is parsed. The decimal longitude (Lon_Dec) and longitude degrees are still reported as a negative. So this change does not affect the outputted data signs, only the internal calculations.

Added Report Order Line.Sta filter for Lines 78.0 55.0, 79.0 55.0, 81.0 55.0, 82.0 55.0 to handle CalCOFI 1311NH's extra DIC stations. Also included additional DIC stations 80.0 53.0 & 80.0 57.5.

Also added a second cast option for sta 90.0 90.0 3500m cast in case there is an extra prodo cast (CalCOFI 1311NH cast 021 & 521). This feature may need to be expanded for other duplicate stations but for now it only works for 90.90.

LD recommended I update our ERDDAP server from 1.46 to 1.52. It took a couple tries but it is online. Hydrographic database is still not visible - continuing work in progress.

The extrapolation routine in DECODR was miscalculating the 200m dynamic height for CalCOFI 1311NH. The problem was the result of having dynamic height listed shallow to deep in the 1311_DHtD.csv file. Inverting the depths from shallow>deep to deep>shallow solved the problem. An inverter utility was added to DECODR's CSV tab; it reads in the YYMM_DHt?.csv and inverts the data for each cast, deep>shallow. The sta.csv>csv module which generates this file will be corrected to output deep>shallow so the next cruises will not need to be inverted.