Skip to main content
IMPLEMENTED

Single/Multiple Text File data types

  • September 11, 2018
  • 4 replies
  • 16 views

Søren Sørensen
Contributor
Forum|alt.badge.img+1

I'm working with a company that regularly receives large comma separated text files from external parties ... large as in 2-300 columns. It would be nice with a couple of features for specifying column data types:

1. Based on the first X rows (where X is user specified), Discovery Hub could make an initial data type suggestion, just like when we import data from a file in SSMS.

2. Multi-selection of fields (the usual Ctrl and Shift functionality) to update the data type for all fields in one go, also like importing text file data in SSMS.

4 replies

Forum|alt.badge.img
  • Contributor
  • November 8, 2018

Having worked with a text file with many columns yesterday, I totally agree. 


Forum|alt.badge.img
  • Contributor
  • November 8, 2018

I would add: 3. A good text source sync. Sometimes csv files get additional columns which aren't appended at the end but inserted somewhere between existing columns.

So a sync feature that would not throw away existing columns but detect that an additional column has been inserted somewhere would be great.

 


Forum|alt.badge.img
  • Contributor
  • November 8, 2018

I would like to add the column "guessing" feature of  "Flat File Import" in SSMS 17.3. Although the flat file import has some significant flaws when it comes to actually performing the import, it is quite good at guessing the data types. As long as one can review and change. 

I will add the suggestion about changed structure with a new field in the middle. I think it would require that field names are in the first row. 


JTreadwell
Community Manager
Forum|alt.badge.img+5
  • Community Manager
  • February 26, 2021

Thanks for you suggestion!

This functionality has been can be utilized by using the CData CSV data source in the Operational Data Exchange. 

I will go ahead and mark this as completed.