Updating a Datatable multiple times a second causes hugh CPU spikes

[hellow554]

The bug

I use a datatable to display data. I have roughly 150 rows, each consiting of 5 fields.
I receive real time items every 100ms and update the datatable according. This is my update code:

def _create_or_update_row(self, key: str, data: ContentType) -> None:
    if key in self.table.rows:
        for field in fields(data):
            self.table.update_cell(
                key,
                field.name,
                _dump_field(field.name, getattr(data, field.name)),
                update_width=True,
            )
    else:
        self.table.add_row(
            key,
            *(_dump_field(field.name, getattr(data, field.name)) for field in fields(data)),
            key=key,
        )

This is triggered by a message:

@on(DataChanged)
def data_changed(self, message: DataChanged) -> None:
    self._create_or_update_row(message.key, message.ct)

This results in a huge CPU spike, roughly to 60% and higher:

This is due to a lot of ioctl calls, if I see that correctly:

and this happens a lot!!!

I can try to set up a simple project that reproduces the problem, but maybe you can tell me, what I'm doing wrong here.
Maybe this could be fixed by cachine the TIOCGWINSZ result and wait for SIGWINCH or similar?

[github-actions[bot]]

Thank you for your issue. Give us a little time to review it.

PS. You might want to check the FAQ if you haven't done so already.

This project is developed and maintained by Will McGugan. Consider sponsoring Will's work on this project (and others).

This is an automated reply, generated by FAQtory

[hellow554]

I removed the update_width=True, and the ioctl calls are gone.

Nevertheless I think it's worth investigating, so so many calls are made.
Btw. If I'm running it in server mode, the ioctl calls are still made and all responded with not supported.
Maybe there shouldn't be an ioctl call on the web driver?!

[willmcgugan]

The size is cached. Textual should only retrieve the size when you resize the window. Since I can't reproduce it, could you provide an MRE?

Updating a table is an expensive operation, especially if columns change size. Consider profiling your Python code, but the solution is typically to batch updates.