Details
Description
Summary: When importing data from CSV, it would be helpful to be able to omit certain fields from being loaded into the document. A practical application would be that you want to use a certain field to construct the document key, but don't want to include that field in the actual content.
Suggestion: Maybe another flag like --omit-fields fieldname1,fieldname2, etc? Any fields being omitted should still be able to be used with --generate-key
Attachments
Issue Links
- relates to
-
MB-27676 cbimport: should add flag of fields to ignore
-
- Closed
-
- links to
| For Gerrit Dashboard: MB-29468 | ||||||
|---|---|---|---|---|---|---|
| # | Subject | Branch | Project | Status | CR | V |
| 116066,5 | MB-29468 Added the ability to ignore multiple fields in cbimport | master | backup | Status: MERGED | +2 | +1 |
Activity
I guess the one point that is worth making is that to generated the key, the CSV has to have that data, so there is data duplication between a field in the document and the key. This is a good reason why the feature has to be in cbimport.
Need to add support to CSV and the ability to specify multiple fields.
Build couchbase-server-7.0.0-1007 contains backup commit 086ca6b with commit message:
MB-29468 Added the ability to ignore multiple fields in cbimport
Build couchbase-server-6.6.0-7625 contains backup commit 086ca6b with commit message:
MB-29468 Added the ability to ignore multiple fields in cbimport
There are already command line tools that are very good at changing CSV files, could one of those be used first before passing the file to cbimport.
On a unix system using cut
From the Windows side powershell has great support for this too
If the user is not experience with the command line then all spreadsheet programs support importing and exporting CSV files. The file can be imported, mutated and exported.
I be careful about adding a feature that can easily be achieve by use one extra simple step.