Filter for display columns

tihoho's profile image tihoho posted 3 years ago in Feature discussion Permalink

Hey! It's me again :)

could you add a filter for the columns you want to display?

Like as data filter (bottom of table content).


this would speed up the browsing of data in tables that have many columns. And it's faster than selecting the displayed columns at the top.

Example: I put in filter "arrive" and realtime view only columns "dtarrive" and "is_arrived".

ansgar's profile image ansgar posted 3 years ago Permalink

That's a screen of a custom query, or? Don't you think you could just modify the query to filter columns away?

tihoho's profile image tihoho posted 3 years ago Permalink

Ansgar, yep, it's custom query.

actually changing the query is not easier. often you have to work with tables with more than 50 columns. and you need to look at a specific column for a moment. and then return it as it was. it would be very convenient to find the required column through the filter, and return the full display back. this is 2 seconds * the number of hours of work with tables. savings hours per year. :)

I have not seen this feature in any GUI program for working with tables. Everyone filters only data, not columns.

It will be faster because no need to waste time on the network (while the request is sent, while it is processed, while it returns). because the filter will work with data that has already been received.

If you decide to implement this functionality, then here is my advice for which all data analysts will be grateful to you:


all table columns: id, dtsent, dtarrive, dtpaid, firstname, status, sum, service_id, metrica_id


filter data: "dt" displayed columns: dtsent, dtarrive, dtpaid

filter data: "id" displayed columns: id, dtpaid, service_id

filter data: "_id dt" displayed columns: service_id, metrica_id, dtsent, dtarrive, dtpaid

ansgar's profile image ansgar posted 3 years ago Permalink

Related: there's a context menu on the column headers, offering to hide/show single columns: Description

Also related, but for data filtering: click on Edit > Filter panel, to show an input box below the grid, with regular expression support.

tihoho's profile image tihoho posted 3 years ago Permalink

I know the functionality you are talking about. in my screenshot in the first post it is.

content filter with regular expressions and gave me the idea that it would be great to have the same filter next to it only for column names.

Abdull's profile image Abdull posted 2 years ago Permalink

I'm working with a table that has many columns (125 columns). As you can see in my first attached screenshot, it is cumbersome to quickly select a single / deselect all but one column.

Therefore, I'm scrolling horizontally for miles back and forth, pixel-hunting for the column I'm interested in, until I spot it.

It'd be great if HeidiSQL had a "quick fuzzy column filter" input box, where one could write the column name(s) to be filtered to, fuzzy-style. It would be such a productivity boost and make HeidiSQL even more awesome.

E.g., writing "ver" in such "quick fuzzy column filter" input box would update the view to show from all query results only the columns "t3ver_oid", "t3ver_id", ""t3ver_oid" ... .

(Related, the "table definition overview" already has a column filter (my second screenshot))

2 attachment(s):
  • many-columns
  • table-columns-filter
ansgar's profile image ansgar posted 2 years ago Permalink

The filter at the bottom is always row based. What you ask is column based, so the filter is not suitable for that.

You could at first write the wanted column names in the query, instead of SELECT *.

nhantam's profile image nhantam posted 1 year ago Permalink

@9756 post Yes thank you so much

tihoho's profile image tihoho posted 1 year ago Permalink

I think this is a very simple fix. 2 years have passed, and apparently I'm not the only one who needs it. Yes, it would be another plus in the convenience of your program. but of course it's up to you to decide whether to listen to your community or not :)

Please login to leave a reply, or register at first.