CSV import attachments

[Jim Winsor]

I know you can specify the creation date, author, description, and where the attachment file resides in the markup to be used by the CSV importer to import attachments.

Is there any way to include the mime/type information in this markup? If not, how does the importer determine what type of file it is dealing with?

[Fabio Racobaldo _Herzum_]

Hello Jim,

did u try this approach?

Importing attachments

You can attach files to issues created from your CSV file. To do this, specify the URL of your attachment in an 'Attachments' column within your CSV file.

Assignee, Summary, Description, Attachment, Comment

Admin, "Issue demonstrating the CSV attachment import", "Please check the attached image below.", "https://jira-server:8080/secure/attachment/image-name.png", "01/01/2012 10:10;Admin; This comment works"

Admin, "CSV attachment import with timestamp,author and filename", "Please check the attached image below.", "01/01/2012 13:10;Admin;image.png;file://image-name.png", "01/01/2012 10:10;Admin; This comment works"

URLs for attachments support the HTTP and HTTPS protocols and can be any location that your JIRA server must be able to access. You can also use the FILE protocol to access files in the import/attachments subdirectory of your JIRA Home Directory.

https://confluence.atlassian.com/display/JIRA/Importing+Data+from+CSV

Regards,

Fabio

[Jim Winsor]

Hi Fabio, Thanks for the reply.

Yes, we are using the noted markup to import the attachments. To further describe the situation, the attachments are coming from a Bugzilla database, and are extracted to files that are named the same as the attachment ID in Bugzilla. The import CSV is then built to provide the date, author, attachment name, and file name as noted. Unfortunately, since the files are stored as a numeric id for the filename, no mime/type data is implied in the markup.

We attempted to use the bugzilla importer directly, but found that it did not align with our enviroment with regards to workflows, issue types, etc., so we opted to extract the issues to CSV and import that way.