Checks
System
- Device OS: [please fill in, e.g. macOS 15.0.1]
- NotePlan Version: [please fill in]
- Plugin Name & Version: Tidy Up v1.20.0, Repeat Extensions v1.2.0
Describe the bug
Since Repeat Extensions v1.2.0 (#760) added support for a concrete base date in @repeat(interval, <date>), running Repeat Extensions' own /generate repeats command correctly recognizes and processes these tags. However, running Tidy Up's Generate @repeats in recent notes command on the same note does not recognize them at all — it behaves as if no repeat is present. Plain @repeat(interval) tags (no concrete date) are still handled correctly by both commands.
Likely cause
If I'm reading it correctly, np.Tidy/src/tidyRepeats.js imports generateRepeats directly from jgclark.RepeatExtensions/src/repeatMain as a relative source-file import, so Tidy Up bundles a compiled snapshot of that code (and its transitive imports, including the repeat-detection regex in helpers/NPExtendedRepeat.js) into its own script.js at build time, rather than depending on Repeat Extensions as a live plugin. Tidy Up's last release (v1.20.0) predates the Repeat Extensions v1.2.0 merge, That would exactly explain why plain repeats still work (unaffected old code) while concrete-date repeats are silently ignored (new syntax the old bundled regex doesn't match).
Testing for the #760 update was always performed on a full build on a local fork, so this would not have arisen. Probably should have been flagged in the PR as a dependency.
If that's right, the fix would be a rebuild + version bump of Tidy Up to pick up the current helpers/NPExtendedRepeat.js, rather than a code change.
To Reproduce
- Add
* task @repeat(1m, 2026-05-12) to a note and mark it complete.
- Run Repeat Extensions' /generate repeats on that note — it correctly generates the next repeat.
- Add another
* task @repeat(1m, 2026-05-12) to a (different) note and mark it complete.
- Run Tidy Up's /Generate @repeats in recent notes — no repeat is generated for that task, even though the note was changed today (well within the default 7-day "recent" window).
- For comparison, a plain
* task @repeat(1m) completed the same way is picked up correctly by Tidy Up.
Additional context
Related PR: #760
NB*
I think I also have a bug/regression with handling concrete dates in cancelled tasks, which I'll look at separately.
Checks
System
Describe the bug
Since Repeat Extensions v1.2.0 (#760) added support for a concrete base date in
@repeat(interval, <date>), running Repeat Extensions' own /generate repeats command correctly recognizes and processes these tags. However, running Tidy Up's Generate @repeats in recent notes command on the same note does not recognize them at all — it behaves as if no repeat is present. Plain@repeat(interval)tags (no concrete date) are still handled correctly by both commands.Likely cause
If I'm reading it correctly,
np.Tidy/src/tidyRepeats.jsimportsgenerateRepeatsdirectly fromjgclark.RepeatExtensions/src/repeatMainas a relative source-file import, so Tidy Up bundles a compiled snapshot of that code (and its transitive imports, including the repeat-detection regex inhelpers/NPExtendedRepeat.js) into its ownscript.jsat build time, rather than depending on Repeat Extensions as a live plugin. Tidy Up's last release (v1.20.0) predates the Repeat Extensions v1.2.0 merge, That would exactly explain why plain repeats still work (unaffected old code) while concrete-date repeats are silently ignored (new syntax the old bundled regex doesn't match).Testing for the #760 update was always performed on a full build on a local fork, so this would not have arisen. Probably should have been flagged in the PR as a dependency.
If that's right, the fix would be a rebuild + version bump of Tidy Up to pick up the current
helpers/NPExtendedRepeat.js, rather than a code change.To Reproduce
* task @repeat(1m, 2026-05-12)to a note and mark it complete.* task @repeat(1m, 2026-05-12)to a (different) note and mark it complete.* task @repeat(1m)completed the same way is picked up correctly by Tidy Up.Additional context
Related PR: #760
NB*
I think I also have a bug/regression with handling concrete dates in cancelled tasks, which I'll look at separately.