Perf: Convert only the required source region in DrawImage

[christianrondeau]

Prerequisites

• I have written a descriptive pull-request title
• I have verified that there are no overlapping pull-requests open
• I have verified that I am following matches the existing coding patterns and practice as demonstrated in the repository. These follow strict Stylecop rules 👮.
• I have provided test coverage for my change (where applicable)

Description

While testing for performance improvement opportunities, I noticed that using the Draw API with an Image onto an Image, ALL pixels of the source image were converted before drawing (https://github.com/SixLabors/ImageSharp.Drawing/blob/main/src/ImageSharp.Drawing/Processing/DrawingCanvas%7BTPixel%7D.cs#L665).

This PR crops the source first and converts the cropped region after.

Impact

Measured on a 22.9 MP image (5700×4014), drawing a 1/8-area region using a custom MemoryAllocator for measuring:

Before: 103MB, 24.4ms (4.3x less memory)
After: 24MB, 14.4 ms (1.7x faster)

NOTE: This also fixes a small leak, this clone was never disposed: https://github.com/SixLabors/ImageSharp.Drawing/blob/main/src/ImageSharp.Drawing/Processing/DrawingCanvas%7BTPixel%7D.cs#L665

[Copilot]

Pull request overview

This PR optimizes DrawingCanvas<TPixel>.DrawImage(Image, ...) when drawing from a source Image whose pixel format differs from the canvas pixel format by converting only the actually-sampled (clipped) source region instead of converting the entire source image up-front. As part of that refactor, it also ensures temporary converted images are disposed when the draw operation short-circuits early (fixing the reported leak).

Changes:

• Pre-compute the effective clipped source/destination rectangles and, for foreign pixel formats, crop in the source format before converting to TPixel.
• Centralize source/destination clipping logic into TryGetDrawImageClip(...) and reuse it from both the public and internal draw paths.
• Add tests asserting that “convert clipped region then draw” matches “convert whole image then draw” across multiple clipping/transform scenarios (including a projective transform).

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated no comments.

File	Description
tests/ImageSharp.Drawing.Tests/Processing/DrawingCanvasTests.DrawImage.cs	Adds regression tests ensuring foreign-format, clipped-region conversion produces identical output to full-image conversion across several scenarios.
src/ImageSharp.Drawing/Processing/DrawingCanvas{TPixel}.cs	Crops before conversion for foreign pixel formats, factors clipping into TryGetDrawImageClip, and disposes owned temporary images on early-exit.

---

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

[codecov]

Codecov Report

❌ Patch coverage is 58.33333% with 10 lines in your changes missing coverage. Please review.
✅ Project coverage is 86%. Comparing base (f23980d) to head (6e136c3).
⚠️ Report is 3 commits behind head on main.

Files with missing lines	Patch %	Lines
...eSharp.Drawing/Processing/DrawingCanvas{TPixel}.cs	58%	5 Missing and 5 partials ⚠️

Additional details and impacted files

@@         Coverage Diff         @@
##           main   #407   +/-   ##
===================================
- Coverage    86%    86%   -1%     
===================================
  Files       109    109           
  Lines      9059   9071   +12     
  Branches   1087   1091    +4     
===================================
+ Hits       7826   7833    +7     
- Misses      982    985    +3     
- Partials    251    253    +2     

Flag	Coverage Δ
unittests	86% <58%> (-1%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.