0.1: finish all three kinds of time related types by al8n · Pull Request #3 · Findit-AI/mediatime

al8n · 2026-04-16T13:15:57Z

No description provided.

Copilot

Pull request overview

This PR converts the repository from a template crate into mediatime, implementing exact rational media-time primitives (Timebase, Timestamp, TimeRange) with no_std-first APIs, plus accompanying docs, tests, benchmarks, and CI/config updates.

Changes:

Implements Timebase, Timestamp, and TimeRange (with hashing/ordering semantics) and adds an extensive in-crate test suite.
Updates crate metadata and README to reflect the new crate name, API, and MSRV.
Adds a Criterion benchmark and Codecov configuration; tweaks formatting/CI workflow settings.

Reviewed changes

Copilot reviewed 15 out of 15 changed files in this pull request and generated 7 comments.

Show a summary per file

File	Description
`src/lib.rs`	Adds the full public API for timebase/timestamp/range types plus tests.
`Cargo.toml`	Renames crate to `mediatime`, sets MSRV/edition/metadata, bench config.
`README.md`	Rewrites documentation to describe the new crate and usage examples.
`CHANGELOG.md`	Adds an initial changelog for 0.1.0.
`benches/gcd.rs`	Adds Criterion benchmarks for GCD helpers used in hashing.
`.codecov.yml`	Adds Codecov configuration.
`.github/workflows/ci.yml`	Updates CI configuration (including schedule formatting and tool versions).
`.github/workflows/loc.yml`	Updates LoC gist upload key/name.
`rustfmt.toml`	Adjusts rustfmt configuration (imports granularity).
`tests/foo.rs`	Removes the placeholder test file.
`README-zh_CN.md`	Removes the Chinese template README.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2026-04-16T13:22:58Z

+    if diff < 0 {
+      return None;
+    }
+    let secs = (diff / NS_PER_SEC) as u64;


duration_since computes secs/nanos from an i128 diff, but then casts to u64/u32 without any overflow handling. For large pts/timebases this will wrap (not saturate), which contradicts the doc comment about pathological inputs saturating. Consider clamping to Duration::MAX (or returning None/an error) when diff / NS_PER_SEC exceeds u64::MAX.

Suggested change

let secs = (diff / NS_PER_SEC) as u64;

let secs_i128 = diff / NS_PER_SEC;

if secs_i128 > u64::MAX as i128 {

return Some(Duration::MAX);

}

let secs = secs_i128 as u64;

Copilot · 2026-04-16T13:22:58Z

+    // = pts * from.num * to.den / (from.den * to.num)
+    let numerator = (pts as i128) * (from.num as i128) * (to.den.get() as i128);
+    let denominator = (from.den.get() as i128) * (to.num as i128);
+    (numerator / denominator) as i64


rescale_pts divides using an i128 intermediate but then casts the quotient to i64. If the rescaled value exceeds the i64 range, this will wrap/truncate silently via as i64. Consider defining and enforcing overflow behavior (e.g., saturate to i64::{MIN,MAX} or return an Option/Result) to avoid surprising incorrect timestamps on large inputs.

Copilot · 2026-04-16T13:23:00Z

+  - **benches/*
+  - **examples/*
+  - **tests/*


These ignore entries start with * (e.g. **benches/*), which YAML parsers commonly interpret as an alias token; unquoted, this is likely invalid YAML. Quote the glob patterns (and consider using **/benches/* etc.) so .codecov.yml parses reliably.

Suggested change

- **benches/*

- **examples/*

- **tests/*

- "**benches/*"

- "**examples/*"

- "**tests/*"

codecov · 2026-04-16T13:29:12Z

Welcome to Codecov 🎉

Once you merge this PR into your default branch, you're all set! Codecov will compare coverage reports and display results in all future pull requests.

ℹ️ You can also turn on project coverage checks and project coverage reporting on Pull Request comment

Thanks for integrating Codecov - We've got you covered ☂️

Copilot

Pull request overview

Copilot reviewed 15 out of 15 changed files in this pull request and generated 4 comments.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2026-04-16T23:59:20Z

+    let numerator = (pts as i128) * (from.num as i128) * (to.den.get() as i128);
+    let denominator = (from.den.get() as i128) * (to.num as i128);
+    let q = numerator / denominator;


rescale_pts documents that it panics when to.num() == 0, but the implementation relies on an implicit division-by-zero panic (numerator / denominator). Consider adding an explicit assert!(to.num != 0, ...) (or similar) before computing q so callers get a clear panic message and the behavior is intentional/documented in-code.

Copilot · 2026-04-16T23:59:20Z

+  /// Works across different timebases. Computes the difference in nanoseconds
+  /// via 128-bit intermediates; for realistic video PTS ranges this is exact.
+  /// If the result would exceed `Duration::MAX` (pathological: seconds don't
+  /// fit in `u64`), saturates to `Duration::MAX` rather than wrapping.
+  #[cfg_attr(not(tarpaulin), inline(always))]
+  pub const fn duration_since(&self, earlier: &Self) -> Option<Duration> {
+    // nanos = pts * tb.num * 1_000_000_000 / tb.den
+    const NS_PER_SEC: i128 = 1_000_000_000;
+    let self_ns = (self.pts as i128) * (self.timebase.num as i128) * NS_PER_SEC
+      / (self.timebase.den.get() as i128);
+    let earlier_ns = (earlier.pts as i128) * (earlier.timebase.num as i128) * NS_PER_SEC
+      / (earlier.timebase.den.get() as i128);
+    let diff = self_ns - earlier_ns;
+    if diff < 0 {
+      return None;
+    }
+    let secs_i128 = diff / NS_PER_SEC;
+    if secs_i128 > u64::MAX as i128 {
+      return Some(Duration::MAX);
+    }
+    let secs = secs_i128 as u64;
+    let nanos = (diff % NS_PER_SEC) as u32;
+    Some(Duration::new(secs, nanos))


duration_since converts each timestamp to nanoseconds via integer division and then subtracts. For mixed timebases this can differ (by up to 1 ns) from rounding the exact rational difference once (e.g., 1 tick @ 1/3 minus 1 tick @ 1/6). If the intent is “round exact difference toward zero”, compute the diff using a single cross-multiply to a common denominator and do one final division/rounding; otherwise update the doc comment to reflect per-endpoint truncation.

Suggested change

/// Works across different timebases. Computes the difference in nanoseconds

/// via 128-bit intermediates; for realistic video PTS ranges this is exact.

/// If the result would exceed `Duration::MAX` (pathological: seconds don't

/// fit in `u64`), saturates to `Duration::MAX` rather than wrapping.

#[cfg_attr(not(tarpaulin), inline(always))]

pub const fn duration_since(&self, earlier: &Self) -> Option<Duration> {

// nanos = pts * tb.num * 1_000_000_000 / tb.den

const NS_PER_SEC: i128 = 1_000_000_000;

let self_ns = (self.pts as i128) * (self.timebase.num as i128) * NS_PER_SEC

/ (self.timebase.den.get() as i128);

let earlier_ns = (earlier.pts as i128) * (earlier.timebase.num as i128) * NS_PER_SEC

/ (earlier.timebase.den.get() as i128);

let diff = self_ns - earlier_ns;

if diff < 0 {

return None;

}

let secs_i128 = diff / NS_PER_SEC;

if secs_i128 > u64::MAX as i128 {

return Some(Duration::MAX);

}

let secs = secs_i128 as u64;

let nanos = (diff % NS_PER_SEC) as u32;

Some(Duration::new(secs, nanos))

/// Works across different timebases. Computes the exact rational difference

/// first using a common denominator, then truncates once when converting to

/// nanoseconds for the returned [`Duration`].

/// If the result would exceed `Duration::MAX` (pathological: seconds don't

/// fit in `u64`), saturates to `Duration::MAX` rather than wrapping.

#[cfg_attr(not(tarpaulin), inline(always))]

pub const fn duration_since(&self, earlier: &Self) -> Option<Duration> {

const NS_PER_SEC: i128 = 1_000_000_000;

let self_den = self.timebase.den.get();

let earlier_den = earlier.timebase.den.get();

let mut a = self_den;

let mut b = earlier_den;

while b != 0 {

let r = a % b;

a = b;

b = r;

}

let gcd = a as i128;

let self_scale = (earlier_den as i128) / gcd;

let earlier_scale = (self_den as i128) / gcd;

let common_den = (self_den as i128) * self_scale;

let diff_num = (self.pts as i128) * (self.timebase.num as i128) * self_scale

- (earlier.pts as i128) * (earlier.timebase.num as i128) * earlier_scale;

if diff_num < 0 {

return None;

}

let secs_i128 = diff_num / common_den;

if secs_i128 > u64::MAX as i128 {

return Some(Duration::MAX);

}

let rem = diff_num % common_den;

let nanos = (rem * NS_PER_SEC / common_den) as u32;

Some(Duration::new(secs_i128 as u64, nanos))

Copilot

Pull request overview

Copilot reviewed 15 out of 15 changed files in this pull request and generated 2 comments.

Comments suppressed due to low confidence (1)

.github/workflows/loc.yml:50

The workflow reads tokeit.json, but the preceding step runs tokeit --lang rust without writing JSON output to that file. As written, the Upload total loc to GitHub Gist step will fail with a missing file; emit JSON (e.g., tokeit --output json > tokeit.json) or adjust the script to read from stdout/artifacts.

      - name: Count total lines of code
        run: |
          tokeit --lang rust
      - name: Upload total loc to GitHub Gist
        uses: actions/github-script@v9
        with:
          github-token: ${{ secrets.GIST_PAT }}
          script: |
            const fs = require('fs');
            const output = fs.readFileSync('tokeit.json', 'utf8');
            const gistId = '327b2a8aef9003246e45c6e47fe63937';

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2026-04-17T00:12:34Z

+  accessors, `with_*`/`set_*` setters for both endpoints, `is_instant`.
+- `const fn` across the whole public surface — every constructor, accessor,
+  and setter can be evaluated in a `const` context.
+- `#![no_std]` always, zero dependencies. No allocation anywhere — every


Changelog claims "#![no_std] always", but the crate uses #![cfg_attr(not(test), no_std)] (tests/doctests build with std). Consider rewording to something like "no_std by default" or "no_std in non-test builds" to match the actual crate configuration.

Suggested change

- `#![no_std]` always, zero dependencies. No allocation anywhere — every

- `#![no_std]` in non-test builds, zero dependencies. No allocation anywhere — every

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

al8n added 2 commits April 17, 2026 01:08

finish 0.1.0

a580960

finish 0.1.0

a3f1a17

al8n requested a review from Copilot April 16, 2026 13:15

Copilot started reviewing on behalf of al8n April 16, 2026 13:16 View session

finish 0.1.0

1401a5d

Copilot AI reviewed Apr 16, 2026

View reviewed changes

finish 0.1.0

0408882

small optimization

b156502

al8n force-pushed the 0.1.0 branch from 5d6702c to b156502 Compare April 16, 2026 13:46

al8n requested a review from Copilot April 16, 2026 23:53

Copilot started reviewing on behalf of al8n April 16, 2026 23:54 View session

Copilot AI reviewed Apr 16, 2026

View reviewed changes

small optimization

d81b03d

al8n force-pushed the 0.1.0 branch from 1badbb1 to d81b03d Compare April 17, 2026 00:07

al8n requested a review from Copilot April 17, 2026 00:07

Copilot started reviewing on behalf of al8n April 17, 2026 00:08 View session

small optimization

4327a30

Copilot AI reviewed Apr 17, 2026

View reviewed changes

Update benches/gcd.rs

c212a82

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

al8n merged commit c1c9fc1 into main Apr 17, 2026
26 checks passed

al8n deleted the 0.1.0 branch April 17, 2026 00:14

	- `#![no_std]` always, zero dependencies. No allocation anywhere — every
	- `#![no_std]` in non-test builds, zero dependencies. No allocation anywhere — every

Conversation

al8n commented Apr 16, 2026

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Uh oh!

Copilot AI Apr 16, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Apr 16, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Copilot AI Apr 16, 2026

Choose a reason for hiding this comment

Uh oh!

codecov bot commented Apr 16, 2026

Welcome to Codecov 🎉

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Copilot AI Apr 16, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Apr 16, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Uh oh!

Copilot AI Apr 17, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants