Stream HTML test report processing to fix OOM on large reports

[FinlayRJW]

Before this PR

HtmlReportPostProcessor loads entire HTML test report files into memory with Files.readString() and processes them with a DOTALL regex over the full content. For tests that produce enormous stdout, this causes OOM in the Gradle daemon.

See memory usage before:

After this PR

==COMMIT_MSG==
HTML test report post-processing now streams line-by-line using a StreamState state machine, bounding memory to O(single line) regardless of file size.
==COMMIT_MSG==

HtmlReportPostProcessor reads and writes one line at a time via BufferedReader/BufferedWriter, writing to a temp file and atomically replacing the original. A StreamState state machine tracks position within the HTML structure:

• DEFAULT — passes lines through verbatim
• AWAITING_PRE — after seeing a <h2>standard output/error</h2> header, waits for the <pre> tag
• IN_PRE — inside a <pre> block, applies log formatting/filtering to each line

See memory usage after:

Also adds an integration test that prints 500K JSON log lines and verifies the post-processor handles it with a constrained 256MB Gradle daemon heap.

The unused Writable interface has been removed.

Possible downsides?

[changelog-app]

Generate changelog in changelog/@unreleased

Type (Select exactly one)

• Feature (Adding new functionality)
• Improvement (Improving existing functionality)
• Fix (Fixing an issue with existing functionality)
• Break (Creating a new major version by breaking public APIs)
• Deprecation (Removing functionality in a non-breaking way)
• Migration (Automatically moving data/functionality to a new system)

Description

HTML test report post-processing now streams line-by-line using a StreamState state machine, bounding memory to O(single line) regardless of file size.

Check the box to generate changelog(s)

• Generate changelog entry

[changelog-app]

✅ Successfully generated changelog entry!

Need to regenerate?

Simply interact with the changelog bot comment again to regenerate these entries.

---

📋Changelog Preview

💡 Improvements

• HTML test report post-processing now streams line-by-line using a StreamState state machine, bounding memory to O(single line) regardless of file size. (#922)

[Copilot]

Pull request overview

This PR refactors HTML test report post-processing to stream line-by-line instead of loading entire files into memory, preventing OOM errors when processing large test reports with substantial stdout/stderr output.

Changes:

• Replaced in-memory regex processing with a streaming state machine (StreamState) that processes HTML files line-by-line
• Added integration test to verify the streaming approach handles 500K log lines within a 256MB heap constraint
• Removed the unused Writable interface

Reviewed changes

Copilot reviewed 4 out of 4 changed files in this pull request and generated 1 comment.

File	Description
TestReportFormattingPluginIntegrationTest.java	Adds integration test for large report handling with constrained heap; updates copyright year
Writable.java	Removes unused interface
StreamState.java	Introduces streaming state machine for line-by-line HTML processing
HtmlReportPostProcessor.java	Refactors from in-memory regex to streaming with temp file approach

---

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

[autorelease3]

Released 2.28.0