Understanding focus logic in ContentSwitcher

[robvanderleek]

I'm using ContentSwitcher to collect user input in a number of steps.
For example, a minimal example with two Input widgets looks like this:

#!/usr/bin/env -S uv run --script

# /// script
# dependencies = [
#   "textual>=6.1.0",
# ]
# ///

from textual.app import App
from textual.widgets import ContentSwitcher
from textual.widgets import Input
from textual.widgets import Static
from textual.containers import Container

class TestApp(App):

    def on_input_submitted(self):
        cs = self.query_one(ContentSwitcher)
        if cs.current == 'input-0':
            cs.current = 'input-1'
        else:
            self.app.exit()

    def compose(self):
        with ContentSwitcher(initial='input-0'):
            yield Input(id='input-0')
            yield Input(id='input-1')

if __name__ == '__main__':
    TestApp().run(inline=True)

This works fine, as the user fills in the first input and submits, the second input will automatically receive the focus.

However, if one of the ContentSwitcher widgets is a composite, the second input will not automatically be focussed:

#!/usr/bin/env -S uv run --script

# /// script
# dependencies = [
#   "textual>=6.1.0",
# ]
# ///

from textual.app import App
from textual.widgets import ContentSwitcher
from textual.widgets import Input
from textual.widgets import Static
from textual.containers import Container

class TestApp(App):

    def on_input_submitted(self):
        cs = self.query_one(ContentSwitcher)
        if cs.current == 'input-0':
            cs.current = 'input-1'
        else:
            self.app.exit()

    def compose(self):
        with ContentSwitcher(initial='input-0'):
            with Container(id='input-0'):
                yield Input()
            yield Input(id='input-1')

if __name__ == '__main__':
    TestApp().run(inline=True)

I know I can manually focus the second input, but I was wondering why the second input does not receive the focus automatically in the second example.

[agammann]

The automatic focus move in the first example comes from hiding the focused
Input, rather than ContentSwitcher selecting the next form field.

ContentSwitcher.current changes which immediate child is displayed. When the
old input becomes hidden, Textual resets its focus. In the reset path for a
widget that's no longer in the focus chain, it looks for a visible, focusable
sibling. In your first example, input-1 is a sibling of input-0. In the
second, the focused input is inside the container, so input-1 is not its
sibling. That path leaves the app with no focused widget.

For a form with a known next step, explicitly focus the next input after
switching:

def on_input_submitted(self):
    cs = self.query_one(ContentSwitcher)
    if cs.current == "input-0":
        cs.current = "input-1"
        self.query_one("#input-1", Input).focus()
    else:
        self.exit()

I reproduced both examples with Textual 6.1.0 and 8.2.8 using
App.run_test(): the direct-child version moved to input-1, the composite
version ended with no focus, and the explicit focus() call worked in both.

The relevant implementation is
ContentSwitcher.watch_current
and Screen._reset_focus;
Widget.focus()
is the public API for choosing the next target.

[robvanderleek]

Hi @agammann

Thanks for your answer!
Apologies for the late reply but your suggestion was indeed what I used [1] to get the behaviour I wanted.

KR, Rob

[1] https://github.com/robvanderleek/inquirer-textual/blob/847543e95c8bd27b86c8abce0a82637a284abbd5/inquirer_textual/widgets/InquirerMulti.py#L36