met file error could produce better error message to diagnose location of problem

[AllanPeake]

Hart_modified_CombGSE2.txt

The attached met file is producing the following error message.

Cannot cast DBNull.Value to type 'System.Single'. Please use a nullable type

There is clearly an error within the met file (some missing temp data represented by hyphens in 2016 day 259-267

) - but it would be nice if the error message could report the line where the problem exists.

[hol430]

Is "222022-GSE_Radn" valid data for the Code column?

[AllanPeake]

It works OK in other met files, surprisingly.

From: Drew Holzworth [mailto:notifications@github.com] Sent: Thursday, 7 February 2019 2:16 PM To: APSIMInitiative/ApsimX ApsimX@noreply.github.com Cc: Peake, Allan (A&F, Toowoomba) Allan.Peake@csiro.au; Author author@noreply.github.com Subject: Re: [APSIMInitiative/ApsimX] met file error could produce better error message to diagnose location of problem (#3488)

Is "222022-GSE_Radn" valid data for the Code column?

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHubhttps://github.com/APSIMInitiative/ApsimX/issues/3488#issuecomment-461282712, or mute the threadhttps://github.com/notifications/unsubscribe-auth/Ah1hT9JNWndWteJYqEOnfTxi4jKV1bjdks5vK6h-gaJpZM4amdVN.

[hol430]

Interesting....should it work? Or should we just throw an error in this scenario? With a decent error message it shouldn't be hard to track down the problem.

[AllanPeake]

I don’t think we want more error messages than we already have – so I think we ought to leave that be!

From: Drew Holzworth [mailto:notifications@github.com] Sent: Thursday, 7 February 2019 2:40 PM To: APSIMInitiative/ApsimX ApsimX@noreply.github.com Cc: Peake, Allan (A&F, Toowoomba) Allan.Peake@csiro.au; Author author@noreply.github.com Subject: Re: [APSIMInitiative/ApsimX] met file error could produce better error message to diagnose location of problem (#3488)

Interesting....should it work? Or should we just throw an error in this scenario? With a decent error message it shouldn't be hard to track down the problem.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHubhttps://github.com/APSIMInitiative/ApsimX/issues/3488#issuecomment-461286009, or mute the threadhttps://github.com/notifications/unsubscribe-auth/Ah1hT8PA-xZjRUZnzvsUQYZHLheFMvQEks5vK64pgaJpZM4amdVN.

[sno036]

It would be nice if it gave a date or something to locate the issue ...

[hol430]

From an implementation perspective, it will be much easier to provide a date than a line number if we want to ignore faulty data in the code column of a met file. On the other hand, is a date as useful as a line number? For a large file it could still take a few seconds for a user to find the faulty row of data.

[hol430]

@sno036 @AllanPeake is this enough information?

Invalid data in column maxt on date 15/05/2016 (day of year = 136)

If I open notepad++ and search for "2016 136" it takes me straight to the faulty data.

[AllanPeake]

Looks good!

A

From: Drew Holzworth [mailto:notifications@github.com] Sent: Friday, 8 February 2019 9:54 AM To: APSIMInitiative/ApsimX ApsimX@noreply.github.com Cc: Peake, Allan (A&F, Toowoomba) Allan.Peake@csiro.au; Mention mention@noreply.github.com Subject: Re: [APSIMInitiative/ApsimX] met file error could produce better error message to diagnose location of problem (#3488)

@sno036https://github.com/sno036 @AllanPeakehttps://github.com/AllanPeake is this enough information?

Invalid data in column maxt on date 15/05/2016 (day of year = 136)

If I open notepad++ and search for "2016 136" it takes me straight to the faulty data.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHubhttps://github.com/APSIMInitiative/ApsimX/issues/3488#issuecomment-461640221, or mute the threadhttps://github.com/notifications/unsubscribe-auth/Ah1hT01t3kZKFAizB50Ju9ICENI4vPyHks5vLLyjgaJpZM4amdVN.