Closed chrismattmann closed 4 years ago
Thanks for the report! Could you post an example repo that causes the issue? (replication case)
Sure as an example, I used your tool to migrate apache/drat back to its original home of chrismattmann/drat and to copy the issues over. I followed the install guide and created a token, and then just used the exporter which if you don't specify any fields you get all the fields. Then I tried to import that CSV file using the importer and it error'ed out b/c all the fields included milestones that existed in apache/drat
but not chrismattmann/drat
. By hand editing the CSV to remove the additional milestone related URI fields, then the import worked. Make sense?
Thanks for the example. I didn't actually consider the "GitHub repo to GitHub repo" copy example - I personally was more interested in exporting or importing to/from GitHub.
Given all of that, I appreciate your workflow and there is a certain logic to the fact that the default export should be able to be imported into a fresh GitHub repo. I would propose we make the following changes:
milestone.title
(with column title milestone
)--exportAll
that will export all possible columns (Similar to what is default right now, but maybe even more in the future)Would those changes solve your issue?
@gavinr completely agree these changes would solve my issue and I appreciate you considering them.
Importing milestones requires the milestone ID -- not the milestone name. So for the whole export->import process to work, we'll also need to complete #27.
This is resolved in v2.0.0. Please see the new default fields, the new --exportAll
option, and #27. Thanks!
thank you @gavinr !!!
The problem is that if your source repo has a milestone assigned to those fields, they are listed as a URI in them on the export. When you try to import that CSV into your dest repository, it complains and throws a weird Github API error that it doesn't know what that URI is.
A better approach would be to post-process the URIs and/or to create a corresponding milestone in the dest repository if not already there, and then to scrub the CSV data with the new milestone. Or simply to by default exclude any fields that only exist in the source domain unless otherwise specified.