Resolves #270: add support for --time-style=relative

[syphar]

This adds showing relative timestamps for #270 .

For the first version this uses timeago from the issue.

I didn't find tests covering this part of the codebase, I'm happy to add them with a better idea on what should be covered and how.

Also I'm happy to change the format if that's something we want (the LS++ format also looks neat).

Different colouring for different ages I would add separately.

[syphar]

How it looks like on my machine:

[anthraxx]

How it looks like on my machine:

I'm wondering if it would be possible to omit the repetitive " ago", this just consumes a lot of space without adding value to the output

[syphar]

"ago" is gone now:

[ariasuni]

I didn't find tests covering this part of the codebase, I'm happy to add them with a better idea on what should be covered and how.

You can check the Testing with vagrant section of the README, and the README in the xtests folder. You can probably extend these date tests (tip: use e.g. exa -l --time-style=relative --color=always > xtests/outputs/dates_long_timestyle_relative.ansitxt to get the output to test).

[syphar]

I didn't find tests covering this part of the codebase, I'm happy to add them with a better idea on what should be covered and how.

You can check the Testing with vagrant section of the README, and the README in the xtests folder. You can probably extend these date tests (tip: use e.g. exa -l --time-style=relative --color=always > xtests/outputs/dates_long_timestyle_relative.ansitxt to get the output to test).

Thanks for the hint! Somehow I missed that part.

Will add tests

[planet36]

How are relative future dates printed?
lsd prints "in XXXX" for future dates, in contrast to "XXXX ago" for past dates.

(edit: changed underscores to Xs)

[syphar]

I didn't find tests covering this part of the codebase, I'm happy to add them with a better idea on what should be covered and how.

You can check the Testing with vagrant section of the README, and the README in the xtests folder. You can probably extend these date tests (tip: use e.g. exa -l --time-style=relative --color=always > xtests/outputs/dates_long_timestyle_relative.ansitxt to get the output to test).

@ariasuni I added tests as described. Many other tests were failing on my machine, is that intended?

[syphar]

How are relative future dates printed?
lsd prints "in " for future dates, in contrast to " ago" for past dates.

Currently it would print just now, also since timeago doesn't support future times yet. From the source, LS++ also only renders past dates correctly.

[anthraxx]

How are relative future dates printed?
lsd prints "in " for future dates, in contrast to " ago" for past dates.

Currently it would print just now, also since timeago doesn't support future times yet. From the source, LS++ also only renders past dates correctly.

I would personally consider sind more an edge case that shouldn't normally happen. printing just now on future dates sounds reasonable enough

[vi]

For the first version this uses timeago from the issue.

Note that timeago was developed specifically for #270.

If something is amiss in timeago, more features can be added (especially if it does not require translators to update their translations).

[vi]

omit the repetitive " ago"

Note that timeago crate has special simplified function format_5chars for compressed time intervals representations of predictable width. This one is English-only and is not supposed to be translated though.

[syphar]

omit the repetitive " ago"

Note that timeago crate has special simplified function format_5chars for compressed time intervals representations of predictable width. This one is English-only and is not supposed to be translated though.

This probably boils down to a design decision:

1. change to 5chars
2. add language to relative format (I see other date formats are localized, so probably we could use timeago languages too, where supported).

[vi]

change to 5chars

It may make sense to provide both options depending on listing options. Terse, but less user-friendly look may be poor default mode.

[ariasuni]

@ariasuni I added tests as described. Many other tests were failing on my machine, is that intended?

Uh yeah since they’re not on the CI, there are some that got broken and I need to merge the fixes and stuff… if your tests work then it’s fine. Also, it would be nice to have a test for future dates, see there:

exa/xtests/details-view-dates.toml

Lines 28 to 36 in dceca33

	# distant past and far future dates
	[[cmd]]
	name = "‘exa -l’ handles dates far past and future dates"
	shell = "exa -l /testcases/far-dates"
	stdout = { file = "outputs/far_dates_long.ansitxt" }
	stderr = { empty = true }
	status = 0
	tags = [ 'long', 'time' ]

[syphar]

@ariasuni I added tests as described. Many other tests were failing on my machine, is that intended?

Uh yeah since they’re not on the CI, there are some that got broken and I need to merge the fixes and stuff… if your tests work then it’s fine. Also, it would be nice to have a test for future dates, see there:

done, added the far future test.

[syphar]

@ariasuni is there anything else missing? I'm happy to change where necessary :)

[Hyoihz]

I'm having troubles displaying files properly with exa.

exa when run as is (without any flags), displays my files one per line as if I'd passed the -1 or --oneline flag.

But when I use exa that's from my package manager, it works normally as how ls and lsd was displayed.

[syphar]

just went through old open PRs and came back to this one :)

@ariasuni I would love to see this in exa, I'm happy to add anything that's needed to get this finished up

[syphar]

Hello @ariasuni I just remembered that I still had this PR open, and would love to see this feature in exa :)
Is there anything that is missing that I can do to get this to land?

To start, I just rebased on the current master :)

[varon]

This would be great to have.

[ariasuni]

Closing this, since exa is unmaintained (see #1243), and this has been done in the active fork eza. Thanks!