fix: render cells that has no content but styles

[kxxt]

Fix #181

I am not confident about this fix since I only briefly looked at the code base. So this PR should be reviewed carefully to avoid breaking other things.

[a-kenji]

Thank you for taking the time, creating this pr!
And sorry for the ci problems :).

On the surface this does make sense.
I will try it out for a bit to check for regressions.

[kxxt]

On the surface this does make sense. I will try it out for a bit to check for regressions.

I think the test failure is indeed caused by this PR but I don't know if my PR is wrong or the test snapshot is wrong.

[a-kenji]

I believe the snapshot needs to be updated.

The vttest script snapshots usually test very constrained functionality.
But I also added few snapshots that test a little more interaction, to check if general behavior changes.
So the test failing here is not a bad thing.

[kxxt]

I believe the snapshot needs to be updated.

The vttest script snapshots usually test very constrained functionality. But I also added few snapshots that test a little more interaction, to check if general behavior changes. So the test failing here is not a bad thing.

Thanks for your explanation, I have updated the failing snapshot.

[a-kenji]

LGMT, thank you for the contribution 👍 .

[kxxt]

Hi, @a-kenji. Could you please publish a new release(or prerelease) for this fix? Thanks!

[a-kenji]

@kxxt,
Not sure if you got pinged on the release, but this fix is now in v0.11.0,

[kxxt]

@kxxt, Not sure if you got pinged on the release, but this fix is now in v0.11.0,

Thanks a lot for your new release! I didn't get pinged but several days ago I found the bug magically disappeared, then I figured out that cargo update command I ran updated tui-term to 0.1.11: kxxt/tracexec@52a26c4

BTW there are some other issues regarding input with htop inside tui-term, which I am going to try to fix when I got more free time: kxxt/tracexec#11