citation.cff files discovered but not displayed in the provenance?

[jameshowison]

Repositories that have a citation.cff file appear to be found, but the provenance chain doesn't make that clear.

See for example, https://citeas.org/cite/ruby-cff

I'm guessing that is the same for CITATION files found in repositories.

I think expected behavior would be to have the provenance chain include steps for:

1. how the text was used to find the repo
2. which files were looked for in the repo
3. What was done with the file (ie in this case we ignore the citation.cff itself, but identify a DOI and use crossref to get the data).

[caseydm]

As a progress update - I refactored the code quite a bit and separated the steps into separate files. This made it easier to add a separate google step which is working properly. You can see it in action here: http://citeas.org/cite/Qv2ray

I still have some work to do in showing all provenance but we are getting closer.

In the case of ruby-cff, the reason you are not seeing the citation file step is because data was found before it go to that step. The order of steps starts with user input here. So the software gets to the github step then goes into that step here and tries a crossref step. It finds a DOI in the crossref step so stops going further. If I change the steps I can cause it to find the citation file, but then it finds the DOI similar to what happened in the github readme:

Maybe it would help to show steps that are skipped? So after finding the DOI in the readme the provenance says that it skipped trying to find a citation file, description file, etc?

[jameshowison]

This is great. Wonderful. I think a tree structure using indents would be cool, so if steps were skipped they would be indented and the lighter font. Actually the lighter font is for skips and a red/black cross but in regular font is for tried-but-found-nothing? At the moment we find the citation.cff file but we don't parse it, right? So that's why it finds the DOI then uses that?

…

On Thu, Aug 19, 2021 at 7:05 AM caseydm ***@***.***> wrote: As a progress update - I refactored the code quite a bit and separated the steps into separate files. This made it easier to add a separate google step which is working properly. You can see it in action here: http://citeas.org/cite/Qv2ray I still have some work to do in showing all provenance but we are getting closer. In the case of ruby-cff, the reason you are not seeing the citation file step is because data was found before it go to that step. The order of steps starts with user input here <https://github.com/ourresearch/citeas-api/blob/master/steps/user_input.py#L18>. So the software gets to the github step then goes into that step here <https://github.com/ourresearch/citeas-api/blob/7c971113b9cc1a43459a2b40dd859e5ff1b7ac63/steps/github.py#L28> and tries a crossref step. It finds a DOI in the crossref step so stops going further. If I change the steps I can cause it to find the citation file, but then it finds the DOI similar to what happened in the github readme: [image: Screen Shot 2021-08-18 at 3 59 00 PM] <https://user-images.githubusercontent.com/3642984/130065150-c68235f4-7225-4884-8172-e3fde7b5d75f.png> Maybe it would help to show steps that are skipped? So after finding the DOI in the readme the provenance says that it skipped trying to find a citation file, description file, etc? — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#67 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAAWOUUYXGBNIXW4QORQJF3T5TXPTANCNFSM5BPUE25A> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&utm_campaign=notification-email> .

[caseydm]

Glad you like it! Yes I agree a lighter font for skips would be good, and maybe we can make it collapsible so it does not clutter the provenance chain too much.

That is correct, we find citation.cff but do not parse it. I have a plan to add that though. Right now we look for the DOI first, but that can be adjusted by changing the order of steps.

[jameshowison]

Great stuff.

…

On Thu, Aug 19, 2021 at 12:59 PM caseydm ***@***.***> wrote: Glad you like it! Yes I agree a lighter font for skips would be good, and maybe we can make it collapsible so it does not clutter the provenance chain too much. That is correct, we find citation.cff but do not parse it. I have a plan to add that though. Right now we look for the DOI first, but that can be adjusted by changing the order of steps. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#67 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAAWOUSRSCPE72JYVSWQYBTT5VBA3ANCNFSM5BPUE25A> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&utm_campaign=notification-email> .