Create a Bulk Action that imports a CSV file.
The idea is to leverage the Bulk Service to
- provide reliable import (retry, survive cluster restart ..)
- distribute the import process
- get a progress status & rest API
What needs to be done:
- Since
NXP-28334we can contribute a scroller, just create a FileSplitScroller implementation that takes a file or file path as input and split the content in chunks of lines, reporting the total number of lines - A simple processor with a single CSV import computation receive a batch of CSV lines and create documents
Now if you are able to create documents using the same CSV format that is produced by the CSV Export you have a basic document export/import for free.
- is related to
-
NXP-29038 Extend Bulk Service to accept non Document command
- Resolved