misnamed in headers

[rburghol]

@COBrogan -- I just fixed a small bug, that is, the header info for odbc_delete was set to export odbc_post which seemed to not cause any trouble that I could discern, but I was chasing down why the WA metrics did not seem to be updating and I found something that could be related if there was indeed some real confusion as to the function in question.

Here's what I observed:

• When running a segment of the James that showed NA for the WA_90_mgd in scenario 17, I openeed the slurm log. (nano slurm-404338.out )
• The log showed that the routine called WA_export.R properly, and the routine seemingly executed correctly, printing out the valid WA results in the final line of the log as Calculating basinwide available flow as -1.55203619909502 + 0 / 90 = -1.552
• BUT, there was no property to be found with the name WA_90_mgd on the runid_17 prop, though all others were there.l
• I scrolled up in the slum log, lookoing for the output of the save command and noticed the message Final na/null check for pk vid val 9022116, so I checked the database for the pid 9022116, but found nothing.
• Upon further investigation, the message Final na/null check for pk vid val 9022116 only occurs in the odbc_delete() function.
• Now, there are probably reasons why the odbc_delete is called, but I could not find a reason inside the context of this WA script. It is the only script that I know of that has failed to save properties correctly.
• So, I found this mislabeled export and thought maybe there was a connection.
• I am doing this PR figuring it needs to happen one way or the other, and maybe it will help.

[COBrogan]

Good catch! It's weird this wasn't causing issues elsewhere. In the future, it seems we can use the generic @export rather than @export name to avoid similar issues. Roxygen should find the name of the function and insert it into the namespace easily enough
https://roxygen2.r-lib.org/articles/namespace.html