Ensure next/prev syntax works for `--startcp` option #7024

MetRonnie · 2025-10-09T15:01:02Z

E.g. cylc play --startcp 'previous(T00)'

Check List

I have read CONTRIBUTING.md and added my name as a Code Contributor.
Contains logically grouped changes (else tidy your branch by rebase).
Does not contain off-topic changes (use other PRs for other changes).
No dependency changes
Tests are included (or explain why tests are not needed).
Changelog entry included if this is a change that can affect users
Docs updated
If this is a bug fix, PR should be raised against the relevant ?.?.x branch.

oliver-sanders · 2025-10-14T09:42:57Z

Ensure next/prev syntax works for --startcp option

Is this not adding support for the next/prev syntax here?

MetRonnie · 2025-10-14T09:56:08Z

Yes, it seems to be an odd omission seeing as it works for the initial cycle point option

oliver-sanders · 2025-10-14T10:06:49Z

Er, it was only ever intended / documented to work at the ICP, but I can see why it might be useful for the start cycle (though I don't think opps would want to specify the warm cycle point implicitly like this?). Plz can you change from "fix" to "feat".

MetRonnie · 2025-10-14T10:24:04Z

though I don't think opps would want to specify the warm cycle point implicitly like this?

Warm start from 00Z, 12Z etc? This is why I implemented it!

MetRonnie · 2025-10-14T10:26:43Z

cylc/flow/scheduler_cli.py

-            "Set the start cycle point, which may be after"
-            " the initial cycle point. If the specified start point is"
-            " not in the sequence, the next on-sequence point will"
-            " be used. (Not to be confused with the initial cycle point)"),


(I notice leading spaces have crept into the codebase - not sure what the motivation is, as this is not how we write paragraphs ordinarily. Perhaps we should specify the style to follow to avoid inconsistency?)

???

Whether the space goes at the start or end of the line is arbitrary, neither is "how we write paragraphs ordinarily" since we don't write this whitespace in plain text. So neither way is right or wrong, natural or unnatural, it's just a style choice like any other.

Leading spaces are preferable because it is much easier to spot when a space has been omitted as they are all aligned in the first column (we have had several cases where tailing spaces were omitted) and with this arrangement commenting out or removing a line will never leave an erroneous space.

It's exactly the same reason why placing operators at the start rather than end of the line is considered preferable (PEP8, flake8, black, ruff, prettier).

FYI: There are no linting rules for strings, so style cannot be enforced.

I agree with @oliver-sanders

oliver-sanders

LGTM but can't test ATM.

oliver-sanders · 2025-10-23T12:08:50Z

changes.d/7024.feat.md

@@ -0,0 +1 @@
+Ensure `next()/previous()` syntax works for the `cylc play --startcp` option.


Suggested change

Ensure `next()/previous()` syntax works for the `cylc play --startcp` option.

Add support for the `next()/previous()` syntax with the `cylc play --startcp` option.

oliver-sanders · 2025-10-23T12:22:35Z

cylc/flow/scheduler_cli.py

-            "Set the start cycle point, which may be after"
-            " the initial cycle point. If the specified start point is"
-            " not in the sequence, the next on-sequence point will"
-            " be used. (Not to be confused with the initial cycle point)"),


???

Whether the space goes at the start or end of the line is arbitrary, neither is "how we write paragraphs ordinarily" since we don't write this whitespace in plain text. So neither way is right or wrong, natural or unnatural, it's just a style choice like any other.

Leading spaces are preferable because it is much easier to spot when a space has been omitted as they are all aligned in the first column (we have had several cases where tailing spaces were omitted) and with this arrangement commenting out or removing a line will never leave an erroneous space.

It's exactly the same reason why placing operators at the start rather than end of the line is considered preferable (PEP8, flake8, black, ruff, prettier).

FYI: There are no linting rules for strings, so style cannot be enforced.

oliver-sanders · 2025-10-23T12:28:26Z

cylc/flow/option_parsers.py

+        "Set the initial cycle point. "
+        "Required if not defined in flow.cylc.\n"
+        "Can be an absolute point or an offset relative to the "
+        "current time - see "
+        f"{SHORTLINK_TO_ICP_DOCS} (Cylc documentation link)."


Why change this?

For clarity

oliver-sanders · 2025-10-23T12:28:41Z

cylc/flow/scheduler_cli.py

+            "Set the start cycle point, which may be after "
+            "the initial cycle point. "
+            "If the specified start point is not in the sequence, the next "
+            "on-sequence point will be used.\n"
+            "Can be an absolute point or an offset relative to the "
+            "current time, like the --initial-cycle-point option - see "
+            f"{SHORTLINK_TO_ICP_DOCS} (Cylc documentation link)."


Why change this?

wxtim · 2025-10-24T14:34:45Z

Looks LGTM, and seemed to work when I tried it out, apart from a bug where Cylc VIP or Play seems to freeze if the startcp >> icp because checking whether it is in the sequence(s) can take a long time if there are lots of items in the sequence.

oliver-sanders · 2025-10-24T14:45:57Z

(unrelated issue)

MetRonnie · 2025-11-04T16:57:14Z

@wxtim are you going to approve?

wxtim · 2025-10-23T14:13:42Z

cylc/flow/scheduler_cli.py

-            "Set the start cycle point, which may be after"
-            " the initial cycle point. If the specified start point is"
-            " not in the sequence, the next on-sequence point will"
-            " be used. (Not to be confused with the initial cycle point)"),


I agree with @oliver-sanders

wxtim · 2025-11-05T09:07:36Z

@wxtim are you going to approve?

Sort of - I failed to spot that I had pending comments - I agree with Oliver about where spacing goes, and even if I didn't I'd rather not have changes like that in this PR.

Technically, I think it works. 👍🏼

MetRonnie · 2025-11-05T09:49:09Z

I haven't changed the whitespace just for the sake of it, it was along with a bit of a rewrite for clarity. Happy to change the whitespace back although it sounded like there was no consensus on it (and I can't say I am compelled by the argument for leading whitespace within strings like that, I find it a bit more difficult to read that way).

MetRonnie added 2 commits October 9, 2025 15:57

Type annotations

5f4c36f

Ensure next/prev syntax works for --startcp option

3af69d4

MetRonnie added this to the 8.6.1 milestone Oct 9, 2025

MetRonnie self-assigned this Oct 9, 2025

MetRonnie added the small label Oct 9, 2025

MetRonnie force-pushed the startcp branch from ea2b27d to 8b33cf4 Compare October 10, 2025 17:12

Docs & changelog

f611a0c

MetRonnie force-pushed the startcp branch from 8b33cf4 to f611a0c Compare October 14, 2025 10:23

MetRonnie requested a review from oliver-sanders October 14, 2025 10:24

MetRonnie commented Oct 14, 2025

View reviewed changes

MetRonnie mentioned this pull request Oct 17, 2025

Fix icp = now changing on reload #7048

Open

8 tasks

oliver-sanders reviewed Oct 23, 2025

View reviewed changes

oliver-sanders requested a review from wxtim October 23, 2025 12:47

wxtim approved these changes Nov 5, 2025

View reviewed changes

		@@ -0,0 +1 @@
		Ensure `next()/previous()` syntax works for the `cylc play --startcp` option.

	Ensure `next()/previous()` syntax works for the `cylc play --startcp` option.
	Add support for the `next()/previous()` syntax with the `cylc play --startcp` option.

Ensure next/prev syntax works for --startcp option #7024

Are you sure you want to change the base?

Ensure next/prev syntax works for --startcp option #7024

Conversation

MetRonnie commented Oct 9, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

oliver-sanders commented Oct 14, 2025

Uh oh!

MetRonnie commented Oct 14, 2025

Uh oh!

oliver-sanders commented Oct 14, 2025

Uh oh!

MetRonnie commented Oct 14, 2025

Uh oh!

MetRonnie Oct 14, 2025

Choose a reason for hiding this comment

Uh oh!

oliver-sanders Oct 23, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

wxtim Oct 23, 2025

Choose a reason for hiding this comment

Uh oh!

oliver-sanders left a comment

Choose a reason for hiding this comment

Uh oh!

oliver-sanders Oct 23, 2025

Choose a reason for hiding this comment

Uh oh!

oliver-sanders Oct 23, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

oliver-sanders Oct 23, 2025

Choose a reason for hiding this comment

Uh oh!

MetRonnie Oct 29, 2025

Choose a reason for hiding this comment

Uh oh!

oliver-sanders Oct 23, 2025

Choose a reason for hiding this comment

Uh oh!

wxtim commented Oct 24, 2025

Uh oh!

oliver-sanders commented Oct 24, 2025

Uh oh!

MetRonnie commented Nov 4, 2025

Uh oh!

wxtim Oct 23, 2025

Choose a reason for hiding this comment

Uh oh!

wxtim commented Nov 5, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

MetRonnie commented Nov 5, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Ensure next/prev syntax works for `--startcp` option #7024

Ensure next/prev syntax works for `--startcp` option #7024

MetRonnie commented Oct 9, 2025 •

edited

Loading

oliver-sanders Oct 23, 2025 •

edited

Loading

oliver-sanders Oct 23, 2025 •

edited

Loading

wxtim commented Nov 5, 2025 •

edited

Loading

MetRonnie commented Nov 5, 2025 •

edited

Loading