0

This idea was prepared rather for GTK+, because KDE Frameworks allows to use normal widget like list, iconview, etc. in place of datamodel/datastorage based widget.

Idea is to add option to export data from datastorage/datamodel connected to selected widget. User must select widget and use some keyboard combination to brings smart copy window. In this window user will select output format(like csv, xml-html, etc.), what to export, what to exclude and where save results, or maybe to with program start to read this data.

Why? Imagine we need to get list of each file in directory. We ran dolphin, went into our input directory, press ctrl+w(or whatever), select csv, select file name, select enter as record separator, selects output file and press save.
Another example: Imagine we need only list of directories. We start file manager, went into our input directory, switch into details view, do anything we done in previous example, but add filter "type==directory"

Another example? We need to list each running service. We start program, which allows to manage service and take steps very similar what we done in previous example.

This could be a lot better/faster and easier than using terminal.

Whilst I do understand your intention and examples I doubt strongly that something like that is a good idea...

"Smart copy/export" certainly is a catchy name and allows for typical KDE like overwhelming feature set, but you will never be able to offer all possible transformation, selection and conversion features in a GUI that any living being is able or willing to use and understand. Data conversion is a very complex topic. To be able to support a majority of needs a GUI for that would require endless options. Only thing about the escaping, quoting, encoding and encapsulation options that are essential for CSV data handling - you want to pack all such options into a "copy" command? For many formats? Desired markup? Layouting?

Your showcase example is a typical situation that is solved with a bit of individual scripting. That is what makes unixoid systems so powerful. That is an existing, usable and proven solution to the task cited in your suggestion.

Yes, I do understand that not all users are able to implement a script just like that. But you are not going to solve this lack of ability by throwing some complex GUI engine at their feet. Those overly complex features that already exist within KDE clearly show that they are rarely accepted or even used by KDE users and I can understand that. Take the "window specific rules" for example: a mighty tool with great flexibility - but where are the users actually using it? Where are the users willing to wrap their minds around the idea? To what benefit compared to the effort?

Maybe you have right about users won't use this feature, but currently many frameworks uses data/model solutions. In my opinion, there's no many problems with conversions, because we add this feature to widgets using data/model.