Importing records using the merge function deletes fields
If you have a phone column in your CSV and it's empty and if the record in vtgier already has a phone, it gets deleted. What you would expect from a merge function is that it would keep the non-empty field instead of keeping the empty one.
For more context, see: http://vtiger-crm.2324883.n4.nabble.com/Vtigercrm-developers-How-to-import-and-overwrite-data-without-changing-record-IDs-td23209.html
Activity
Ideas on improvizing the import functionality.
-
Uma : I think we should enhance the import functionality, So that systems should first analyze the user's need for import on either to update the existing records or to replace the record (i.e., delete existing and creating new one). Based on selected choice consequent steps should be taken on records.
-
Ruben : I agree, there should be an "update" option which only updates fields in existing records but that doesn't create new records. And the merge option that currently exists should create records if they do not exist and update them if they do (here we would just need to fix the existing bugs). Saludos
-
-
- Martin : When the record already exists maybe, allowing the User to specify per field (in the Field Mapping section) whether to always use the supplied value (overwrite), or whichever one is not blank (merge), and if both not blank then whether to use the supplied value or the existing value?
Adding the ability to select autoincrement fields as the duplicate check (being the only true Unique ID field) is key to making this work properly though. Being the only non-editable field, it means you can then update ANY existing field even if it's a field you would previously have used as a duplicate check field! (Which you can currently do!)
This is what I think would be ideal:
- In any case the duplicate records in the CRM should not be merged
- If the operation is merge then update all the matching records fields, with the field values in the CSV file if the values in the CRM record is empty or not equal to default value for that field
- If the operation is overwrite then update all the matching records fields with the field values in CSV blindly
Since we are not merging duplicate records in CRM into one single record anymore it would also be possible to Map the auto-generated fields.
Our team has started work on this and will share the progress soon. Please feel free to share your thoughts.
Related issues:
@nilay.automatesmb I think that it would be nice to be able to specify whether to overwrite or merge at the field level as suggested by Martin as quoted above by Uma.
Added enhancement label
Reassigned to @vtiger
-
Today I faced a case where I had to update existing accounts. I figured out I would use the overwrite function. However you need to map all mandatory fields in order to be able to proceed.
If we implement the mentioned update functionality (with overwrite or merge), I think there should be no check on mandatory fields. It's better to have a CSV that only has the fields needed for update in order to avoid accidents. Right now this doesn't work if there are mandatory fields different from the ones you want to update. This update function should also only update existing records (not create new ones).
mentioned in merge request !803 (merged)
-
@uma.s I think there are various unattended problems mentioned in this issue. We should either open new issues for each one or reopen this issue.
- We need a new option in the "duplicate handling" section which gives the ability to only update existing records (not create new ones):
- When using the option in 1. you should not be asked to map mandatory fields (e.g. I would like to only update the email or phone number or something else, but not touch the already existing mandatory fields like lastname):
With the ability to use the Recrod Number as field we are closer to have this new "Update" feature, but to complete it, I feel we need the above two points.