You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Currently if we are executing notebooks via binderbot, the TOC file is ignored during execution phase and we just run every *.ipynb file.
On the other hand if execution occurs via jupyter-book, then TOC is parsed within jupyter-book and only those notebooks that appear in the TOC are executed.
Ideally the list of executed notebooks would be identical regardless of where the execution takes place. Only notebooks in the TOC should be executed.
One possibility is to add some code to the binderbot-specific sections of the build-book workflow to parse the TOC and generate a list of notebooks to pass to the binderbot call.
Alternatively, it may be possible to leverage jupyter-book for this. Is is possible to get the jupyter-book CLI to just report back the list of notebooks it "would have executed" without actually executing them?
The text was updated successfully, but these errors were encountered:
This came up in discussion on ProjectPythia/cmip6-cookbook#48. It would be helpful to remove the failing notebook from the TOC until a solution to the secrets issue is found, so that the failure is not holding up other development on the CMIP6 cookbook. But as things stand right now, removing the notebook from the TOC will not prevent it from being executed.
Currently if we are executing notebooks via binderbot, the TOC file is ignored during execution phase and we just run every *.ipynb file.
On the other hand if execution occurs via jupyter-book, then TOC is parsed within jupyter-book and only those notebooks that appear in the TOC are executed.
Ideally the list of executed notebooks would be identical regardless of where the execution takes place. Only notebooks in the TOC should be executed.
One possibility is to add some code to the binderbot-specific sections of the build-book workflow to parse the TOC and generate a list of notebooks to pass to the binderbot call.
Alternatively, it may be possible to leverage jupyter-book for this. Is is possible to get the jupyter-book CLI to just report back the list of notebooks it "would have executed" without actually executing them?
The text was updated successfully, but these errors were encountered: