ASCII2NC error

[jwraby]

I'm getting an error when I run MET V12.0.0 ASCII2NC as follows:
DEBUG 1: Start ascii2nc by jraby(921059) at 2025-02-26 22:20:42Z cmd: /usr/local/bin/ascii2nc /p/home/jraby/MET_obs/MSA/NATurE/qc_obs_used_earth_relative.2022_07_05_121500 /p/home/jraby/MET_obs/MSA/NATurE/qc_obs_used_earth_relative.2022_07_05_121500_test2.nc -log /p/home/jraby/MET/MET_ascii2nc/logs/NATurE_MATLAB_TWR12_std_obs_test2 -v 5
DEBUG 1: Config File: /usr/local/share/met/config/Ascii2NcConfig_default
ERROR :
ERROR : StringArray::operator const -> range check error!

But I ran ASCII2NC using the same input obs file using MET V11.1.0 and it runs successfully and produces the expected output. Need assistance with troubleshooting this problem. I can provide the input obs file and other data as needed by email using a file exchange server if you could provide me with an email address. Thanks.

[CPKalb]

Hi:

Would you be able to send over the input obs file so that we can test these in MET 11 versus MET 12? If it's just a single obs file and your configuration, you may be able to upload the data directly to this discussion following the information for attaching files. If it's too large, we have a google form to fill out to send us data. The instructions are here.

Thanks,
Christina

[jwraby]

Hi Christina - Thanks for reaching out on this. In the past I have been unable to upload small files into the discussion. I'll give it a try. I have also tried uploading the files using the other method and it doesn't work for me. I think the issue is that I'm on an Army network and they probably block such upload/transfers. The only methods which works reliably is to upload the file to an Army Secure Access File Exchange (SAFE) system which uses the recipients email address to send a download link email. I've used the SAFE to send files to Julie Prestopnik in the past. I'll comment on the results of my attempt to upload into this discussion.

[jwraby]

qc_obs_used_earth_relative.2022_07_05_121500.TXT

[jwraby]

Christina - It looks like my attempt was successful. Please let me know. Thanks.

[jwraby]

FYI, I did not use a config file.

[CPKalb]

I'm able to duplicate the behavior. The file runs fine in MET version 11.1 but produces an error in MET version 12. It looks like the error is occurring on line 223. I need to talk to a software engineer because it's not clear why this is error is happening. One of us will get back to you.

Christina

[jwraby]

Thanks for the update, Christina. Nice to know that it's not just an issue at my end. I have a workaround by using MET V11.1.0 which works for me. Also, good to find out that I can upload files into the discission. Don't know why I had issues before.

[JohnHalleyGotway]

@jwraby I took a look at your file and agree that it worked fine with 11.1 but is broken by 12.0.

ascii2nc -format met_point qc_obs_used_earth_relative.2022_07_05_121500.TXT ascii2nc_12.0.0.nc
...
ERROR  : StringArray::operator[](int) const -> range check error!

The "cause" of the failure is that your file is tab-delimited. When I run sed to switch it from tab to space-delimited (sed -i 's/\t/ /g' qc_obs_used_earth_relative.2022_07_05_121500.TXT) it runs fine. The error occurs when trying to determine which format parsing logic to apply.

The easiest fix is simply adding -format met_point to your ASCII2NC command line to bypass the format parsing logic. This commands runs just fine (using the original tab-delimited input):

ascii2nc -format met_point qc_obs_used_earth_relative.2022_07_05_121500.TXT ascii2nc_12.0.0.nc

Please confirm that adding -format met_point gets you past this issue. And I'll go track down the underlying parsing issue. Should be an easy fix once I find it. And I'll be sure to add a test with tab-delimited input to make sure we don't break it again in the future!

[JohnHalleyGotway]

@jwraby, FYI, I wrote this up as a bugfix issue in dtcenter/MET#3095. The fix easy but since its in already-released code, I want to document it well and work through the bugfix process.

I'd recommend that you mark this discussion as being answered and then we can close/lock it. And any further discussion could be done as comments on that bugfix issue.