Spectra not loading anymore

[hamed-musallam]

I tested these samples using version 0.44, and the majority of the spectra loaded correctly but I still got some errors

unfortunately with current development version not working at all

https://s3.uni-jena.de/nmrxiv/production/archive/247bab07-dc3c-45cc-807d-f21bb83882f6/Genistein%20annotated%20NMR%20400%20MHz%20DMSOd6%20data.zip
https://s3.uni-jena.de/nmrxiv/production/archive/1a4d675d-f078-4628-ac78-85236e0ad8af/%20Capsaicin%20400%20MHz%20DMSOd6%20NMR%20data%20.zip
https://s3.uni-jena.de/nmrxiv/production/archive/247bab07-dc3c-45cc-807d-f21bb83882f6/Genistein%20annotated%20NMR%20400%20MHz%20DMSOd6%20data.zip
https://s3.uni-jena.de/nmrxiv/production/archive/1fd327ba-9a52-4120-88a5-b72529cb5a98/Curcumin%20annotated%20NMR%20400%20MHz%20DMSOd6%20data.zip
https://s3.uni-jena.de/nmrxiv/production/archive/5a1698da-4270-4114-b588-f258b9fd44d3/Apigenin%20annotated%20NMR%20400MHz%20DMSOd6%20data.zip
https://s3.uni-jena.de/nmrxiv/production/archive/432597cd-27c8-4130-b797-7e11a7599562/Myricetin%20annotated%20NMR%20400%20MHz%20DMSOd6%20data.zip
https://s3.uni-jena.de/nmrxiv/production/archive/99d715bc-34bb-40bc-80c0-75722a68d23f/Kaempferol%20annotated%20NMR%20400%20MHz%20DMSOd6%20data.zip
https://s3.uni-jena.de/nmrxiv/production/archive/fd127472-6962-46ad-9390-8ae9e0ad3000/%20Glycyrrhizin%20900MHz%20400MHz%20DMSOd6%20NMR%20data%20.zip
https://s3.uni-jena.de/nmrxiv/production/archive/25bc00c8-e1f2-4bde-b281-342ac7fe6abe/%20Glycyrrhetinic%20acid%3AEnoxolone%20900_400MHz%20DMSOd6%20NMR%20data%20.zip
https://s3.uni-jena.de/nmrxiv/production/archive/2cd25088-4513-40ad-8aa6-21ff1f4e3527/%20Lupeol%20900MHz%20400MHz%20CDCl3%20NMR%20data%20.zip

[jobo322]

let me check what is happening

[hamed-musallam]

it seems some of the meta does not exist in some of these files but is it mandatory?

[jobo322]

The first file that I did check, I found it

DATACLASS = NTUPLES but the data is a single xydata and the last or first LDR does not exists. That is why this jcamp is not loaded.

@lpatiny is it possible and should we add support for it or the jcamp file was wrongly generated?

[jobo322]

changing the DATACLASS to XYDATA, the spectrum is loaded. The assignment does not yet loaded but it is another issue

[lpatiny]

When I drag and drop the zip file we actually try to upload 23 files and it seems that only one is failing. If would be nice in the logs to know which one it is ! @jobo322 Could you check how to get the filename and add another issue for this ?

Practically it seems related to the 13C.jdx that is generated by JEOL and that indeed pretends it is NTUPLES

13C.jdx.zip

Based on the draft specifications: http://www.jcamp-dx.org/drafts/JCAMP6_2b%20Draft.pdf we are expected to see pages

And they are not in this JCAMP-DX.

[hamed-musallam]

some of those files and maybe all of them come from Harvard public datasets https://dataverse.harvard.edu/dataverse/cenaptnmr;jsessionid=51df1d14ff071004936a072f6bc1

[targos]

Is it a regression or it worked before?

[targos]

If it worked before, you can use old cloudflare preview links on the pull requests to identify when it started to fail.

[targos]

I did a git bisect session.
The regression happened in commit 470abab

[targos]

@jobo322 In case you suspect the regression is in nmr-load-save, here is the diff: https://github.com/cheminfo/nmr-load-save/compare/v0.12.1...v0.13.1

[jobo322]

Currently, nmr-load-save is checking that the properties for NTUPLES data-class like LAST, FIRST, and SYMBOLS exist even if the DATACLASS LDR is NTUPLES. It avoids throwing an error in the case of a wrong definition of DATACLASS to NTUPLES.

After version 0.23.3 of nmr-load-save, there are two remnants issues at the moment to load this data:

• https://github.com/cheminfo/nmr-load-save/issues/305
• https://github.com/cheminfo/nmr-load-save/issues/304

The last issue is bug related with 2D JEOL Data and only appear if experimental features is enabled, it is related with automatic processing of 2D. I will add the corresponding test cases with 2D data of each supported brand.