touch: read -d dates in the local zone and honor their offsets by jay23606 · Pull Request #15292 · uutils/coreutils

jay23606 · 2026-10-13T00:12:27Z

touch -d read a date without an offset, such as 2026-08-01 12:00:00, as UTC. It now reads it in the system zone, as the documentation describes.

This also changes dates that have an offset. The offset used to be dropped, so 2004-01-16 12:00 +0200 was set to 12:00 UTC. It now honors the offset and sets 10:00 UTC. test_touch_set_date7 has its expected value updated to match.

Closes #15291

Integration tests set TZ to a non-UTC zone and check fixed timestamps for both forms. Run locally on Windows only. Six symlink tests fail there for a privilege reason, unrelated to this change.

Copilot

🟡 Changes recommended

The locale-format path remains UTC, and the regression test is ineffective on UTC hosts.

1 open finding

Apply system-zone conversion to POSIX_LOCALE date parsing · New

What changed in this PR

Updates touch -d to interpret offset-less numeric timestamps in the system time zone.

Changes:

Uses the local zone for numeric dates without offsets.
Adds a unit regression test.

File	Description
`src/uu/touch/src/touch.rs`	Updates date parsing and adds coverage.

🧠 Review effort: Balanced

💡 Configure MCP servers for context-aware, tailored reviews. Learn more in the docs.

Copilot

🟡 Changes recommended

Explicit offsets are discarded, and the regression tests are ineffective on UTC CI hosts.

2 open findings

1 resolved since last review

Apply system-zone conversion to POSIX_LOCALE date parsing

🧠 Review effort: Balanced

Copilot

🔵 Needs a closer look

The regression tests pass under the previous behavior on UTC-based CI runners and need an explicit non-UTC zone.

2 open findings

🧠 Review effort: Balanced

Copilot

🟡 Changes recommended

The primary regression test is nondeterministic on UTC hosts, and the description omits the explicit-offset behavior change.

2 open findings

1 resolved since last review

Preserve parsed timezone offsets when converting timestamps

🧠 Review effort: Balanced

Copilot

🔵 Needs a closer look

The description incorrectly claims explicit-offset behavior is unchanged.

1 open finding

PR description omits changed behavior for nonzero %z inputs

1 resolved since last review

Test timestamp conversion under a non-UTC system timezone

🧠 Review effort: Balanced

Copilot

🟢 Approval recommended

The focused implementation correctly addresses both reported parsing errors with adequate regression coverage.

0 open findings

1 resolved since last review

PR description omits changed behavior for nonzero %z inputs

🧠 Review effort: Balanced

Copilot

🟡 Changes recommended

Nonexistent local times during DST gaps can fall through and be silently normalized.

1 open finding

Reject nonexistent DST-gap times in -d date parsing · New

🧠 Review effort: Balanced

Copilot

🟢 Approval recommended

The focused implementation is consistent with the reported behavior and includes adequate regression coverage.

0 open findings

1 resolved since last review

Reject nonexistent DST-gap times in -d date parsing

🧠 Review effort: Balanced

Copilot

🟡 Changes recommended

DST-gap detection misses accepted formats, and existing fixed-epoch tests become host-timezone-dependent.

1 open finding

Set TZ=UTC for date and reference test comparisons · New

🧠 Review effort: Balanced

Copilot

🟡 Changes recommended

The DST-gap guard misses accepted minute-only timestamps, allowing silent normalization.

2 open findings

🧠 Review effort: Balanced

Copilot

🔵 Needs a closer look

DST-gap detection misses the supported minute-only date format.

2 open findings

🧠 Review effort: Balanced

Copilot

🔵 Needs a closer look

Minute-only timestamps bypass the new daylight-saving gap validation.

2 open findings

Previously missed (1)

In code that hasn't changed since last review

Reject nonexistent DST times consistently regardless of minute format

src/uu/touch/src/touch.rs:885

Include %Y-%m-%d %H:%M in this guard and add a regression test: with TZ=EST+5EDT,M3.2.0/2,M11.1.0/2, -d '2020-03-08 02:30' bypasses the gap check and reaches the normalizing fallback, while the equivalent input ending in :00 is rejected.

🧠 Review effort: Balanced

codspeed · 2026-10-15T09:31:54Z

Merging this PR will degrade performance by 17.48%

⚠️

120 benchmarks spent significant time in system calls

System calls cannot be consistently instrumented, so they are not included in the measure, which understates the real cost. Please switch to the Walltime instrument to accurately measure system calls.

Measurement and system calls

ls_recursive_long_all_balanced_tree[(6, 4, 15)]

du_wide_tree[(5000, 500)]

cksum_crc

cksum_default

cksum_raw_output

du_deep_tree[(100, 3)]

sort_unique_utf8_locale

cksum_crc32b

ls_recursive_balanced_tree[(6, 4, 15)]

ls_recursive_mixed_tree

…and 110 more

ℹ️ 1 benchmark has no flame graph

The flamegraph could not be generated because too much time was spent in runtime internals, making the visualization not useful for performance analysis.

sort_dictionary_order[500000]

⚠️

Different runtime environments detected

Some benchmarks with significant performance changes were compared across different runtime environments,
which may affect the accuracy of the results.

Open the report in CodSpeed to investigate

❌ 3 regressed benchmarks
✅ 398 untouched benchmarks
⏩ 70 skipped benchmarks¹

Warning

Please fix the performance issues or acknowledge them on CodSpeed.

Performance Changes

	Mode	Benchmark	`BASE`	`HEAD`	Efficiency
❌	Simulation	`complex_relative_date`	277.7 µs	386.1 µs	-28.08%
❌	Simulation	`true_consecutive_calls`	294.7 ns	350.2 ns	-15.86%
❌	Memory	`du_deep_tree[(100, 3)]`	76.5 KB	82.4 KB	-7.14%

Tip

Investigate this regression by commenting @codspeedbot fix this regression on this PR, or directly use the CodSpeed MCP with your agent.

_{Comparing jay23606:fix/15291-touch-local-time (d3a9360) with main (5fda18e)²}

70 benchmarks were skipped, so the baseline results were used instead. If they were deleted from the codebase, click here and archive them to remove them from the performance reports. ↩
No successful run was found on main (b77b664) during the generation of this report, so 5fda18e was used instead as the comparison base. There might be some changes unrelated to this pull request in this report. ↩

Copilot

🔵 Needs a closer look

Cross-platform timezone behavior still needs confirmation from Linux, macOS, and Windows CI.

0 open findings

2 resolved since last review

🧠 Review effort: Balanced

touch -d read "YYYY-MM-DD HH:MM:SS" as UTC. GNU touch reads it as local time. Use the system zone for the formats that carry no offset. Adds a unit test that fails on the UTC behaviour.

The unit tests compared against the system zone, which is UTC on UTC CI hosts, so they passed without the local-zone behavior. Move the coverage to integration tests that set TZ to a non-UTC zone and check fixed epochs. Co-Authored-By: Claude Haiku 5.5 <noreply@anthropic.com>

parse_datetime normalizes a wall-clock time the zone skipped, so -d read 02:30 on a spring-forward day as a real time. Refuse it before the fallback, as the date utility does, and cover it with a -d test. Co-Authored-By: Claude Haiku 5.5 <noreply@anthropic.com>

Co-Authored-By: Claude Haiku 5.5 <noreply@anthropic.com>

parse_datetime accepts 'YYYY-MM-DD HH:MM' and normalizes a time the zone skipped, while the seconds forms were already refused. Check the minute-only form too, with a regression test. Co-Authored-By: Claude Haiku 5.5 <noreply@anthropic.com>

Copilot

Warning

Copilot couldn't run its full agentic review because it didn't start before the timeout. Make sure your repository has a runner available, or add a copilot-code-review.yml file specifying one with the runs-on attribute. See the docs for more details.

Copilot review overview

4 open issues

unambiguous() will reject not only DST gaps (which you explicitly want to reject) but also DST… · New
unambiguous() will reject not only DST gaps (which you explicitly want to reject) but also DST… · New
TimeZone::system() is recomputed for every candidate dt (and potentially does tzdb/env… · New
The comment says “Same gap as above”, but the immediately preceding test is about UTC-8 parsing,… · New

_{Review effort: Lite}

A wall-clock time that happens twice, such as 01:30 on the autumn fall-back day, was rejected by every local-time conversion because `unambiguous()` fails for folds as well as gaps. `touch -t 202011010130` therefore reported an invalid date format. Add zoned_earlier, which takes the earlier instant for a fold and still rejects a gap. Use it at each local-time conversion, and compute the system time zone once in is_system_zone_gap. Co-Authored-By: Claude Haiku 5.5 <noreply@anthropic.com> Signed-off-by: jay23606 <4602056+jay23606@users.noreply.github.com>

Copilot

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.

An organization owner can choose the repository organization as the billing source for reviews requested by members with a paid Copilot license. Learn about Copilot code review billing. Review the Copilot code review billing settings.

Copilot AI balanced review requested due to automatic review settings October 13, 2026 00:12

Copilot started reviewing on behalf of jay23606 October 13, 2026 00:12

View session