rescience / articles Goto Github PK

In publish script, article.code.doi is checked for None, where as its an empty string, even when code doi is removed from metadata.yaml. This leads to the following error:

Updating metadata to Zenodo... [{'field': 'metadata.related_identifiers.0.identifier', 'message': 'Not a valid persistent identifier. Identifier is required.'}]
Traceback (most recent call last):
  File "articles/publish.py", line 219, in <module>
    update_metadata(server, token, article_id, article)
  File "articles/publish.py", line 103, in update_metadata
    raise IOError("%s: " % response.status_code + response.json()["message"])
OSError: 400: Validation error.

Changing it to this logic will solve it:

if article.code.doi is not None and len(article.code.doi) > 0:

I am having trouble running the script from my Mac. Any ideas @rougier @khinsen?

(base) MBpro:articles olivia$ echo $ZENODO_SANDBOX_TOKEN
REDACTED
(base) MBpro:articles olivia$ python process.py --sandbox --metadata ../mahmoudian-2020-rescience/article/metadata.yaml --pdf ../mahmoudian-2020-rescience/article/Reproduction_of_Smyth_et_al__1996.pdf 
No token found (ZENODO_SANDBOX_TOKEN)
You can request one from https://sandbox.zenodo.org/account/settings/applications/tokens/new/

Hi EICs,
I just handled my first submission using these new instructions. I followed the readme closely but was confused the whole time. I get it now.

Things that are not clear:

• Do I copy the article.pdf and metadata.yaml here and then run the scripts? OR
• Do I copy the scripts over to the forked author repo and run it there?

This wasn't clear to me. I now realize I should have done the first and not the second.

Here is my first take on a revised README: https://gist.github.com/karthik/3907aae7b96a4d23ed885e78cc146804

Is this useful? Should I submit this as a pull-request?

I am trying to publish ReScience/submissions#27. I am following the instructions in https://github.com/ReScience/articles/blob/master/README.md. I can run 'process.py' just fine with the --sandbox option, but it fails when I try the --zenodo option:

$ ./process.py --zenodo --metadata metadata.yaml --pdf article.pdf
Traceback (most recent call last):
  File "./process.py", line 77, in <module>
    article_id, article_doi = reserve_doi(server, token)
  File "./process.py", line 21, in reserve_doi
    response.json()["message"])
OSError: 401: The server could not verify that you are authorized to access the URL requested. You either supplied the wrong credentials (e.g. a bad password), or your browser doesn't understand how to supply the credentials required.

Both my ZENODO_SANDBOX_TOKEN and ZENODO_TOKEN are set. I have both tried setting ZENODO_TOKEN to the same as ZENODO_SANDBOX_TOKEN as well as setting it to a separately created token. Neither works and both give me the same error message.

Hi @rougier

I notice that there is no DOI in the article's bibliography.

I usually make sure that all references have a clickable DOI, which is also useful to verify the references.

Changing "doi = false" to "doi = true" in rescience.cls 's biber settings fixes it. Should I do that?

In the publishing instructions, it should be explicitly stated that before step 4, the metadata.yaml file has to be (partially) completed by the editor, to at least include the received, accepted and published dates, so that the command in step 4 executes correctly.

In order to publish a new article (after acceptance), you'll need to have the article metadata file (YAML format) and the corresponding article (PDF format). 

Is best practice to take the YAML and PDF and drop them in the same directory as the Python scripts?