refactor: drop the dead anchor: kwarg + gate the legacy trees as alias-only (ADR-008 step 7)

[pftg]

Part A — kill the dead anchor: kwarg

SnapDiff::Capture::Viewport.prepare!(expected_window_size, anchor: nil) accepted a kwarg no caller ever passed non-nil — both screenshot_matcher.rb call sites hardcoded anchor: nil — plus a unit test pinning its acceptance and a doc comment promising it to v3.

Adding an optional kwarg later is 100% non-breaking, so reserving it now bought nothing and shipped a test defending dead code. v3's scroll-preservation work will design the real contract; the panel found this seam's shape wrong for that anyway (no session access, no resize responsibility).

Deleted: the kwarg, both call-site args, the @param anchor doc, and the test.

Part B — the mechanical CI gate

test/unit/legacy_tree_is_alias_only_test.rb — a plain unit test, so it runs in rake test:unit and in CI with no new workflow wiring.

What it asserts. Every .rb file under lib/capybara/ and lib/capybara_screenshot_diff/ (plus both root entry files) contains nothing but:

• require / require_relative / autoload
• module / class / class << self / end / private / extend / include
• constant aliases (Foo = SnapDiff::Foo)
• def_delegator(s)
• defs whose entire body is a single line delegating into SnapDiff

That last rule is the one that matters: a one-line forwarder is not lost when the file is deleted, real logic is. This is what keeps 3.0 a git rm instead of a refactor.

Mechanism: strip comments/blanks, fold trailing-comma continuations into one logical line, then walk. Failures name the file and the offending line.

The allowlist — the honest state of the tree

One file.

File

Reason

lib/capybara/screenshot/diff/config_legacy.rb

Generates the old accessor names from SnapDiff::Config::MAPPING via define_method (ADR-008 step 1 — it is delegation, but contains def/define_method). It also still holds derived config logic that never moved: .active? precedence, .screenshot_area path assembly, and .default_options — which carries one literal default, the vips tolerance 0.001. AVAILABLE_DRIVERS also lives here on purpose (#227: test_helper reads it at boot, image_compare_test stubs it) — the gate does not push it out.

Narrowed with a second cheap assertion: config_legacy.rb's method inventory is pinned to exactly active?, screenshot_area, screenshot_area_abs, self.compare, self.configure, self.default_options. Any new method there reds the gate.

Findings (files that did NOT need allowlisting, contrary to the brief)

• lib/capybara_screenshot_diff.rb needs no exception. Its error-class eager aliases are plain constant assignments and it hosts no defs at all — the session forwarders actually live in lib/capybara_screenshot_diff/screenshot_assertion.rb.
• screenshot_assertion.rb and static.rb need no exception either. Every method in them is a one-line SnapDiff.* forwarder, so the shape rule accepts them without a per-file carve-out.
• The real finding is config_legacy.rb. ADR-008 step 1 moved config storage to SnapDiff::Config but left derived config logic behind. ADR-008's "alias-only" claim holds for the whole tree except this file. Worth amending the ADR to say so.

Gate-check evidence

Three mutation probes, each reverted with a targeted edit (no git checkout):

1. One-line def in a non-allowlisted file (lib/capybara_screenshot_diff/static.rb):

capybara_screenshot_diff/static.rb: `def probe_me; end` is not a one-line forwarder into SnapDiff
2 runs, 5 assertions, 1 failures

2. Realistic multi-line method (lib/capybara/screenshot/diff/region.rb):

capybara/screenshot/diff/region.rb: `def probe_me(other)` is not a one-line forwarder into SnapDiff
2 runs, 5 assertions, 1 failures

3. New method in the ALLOWLISTED file (config_legacy.rb) — proves the allowlist is not a blanket pass:

-["active?", "screenshot_area", "screenshot_area_abs", "self.compare", "self.configure", "self.default_options"]
+["active?", "probe_me", "screenshot_area", ...]
2 runs, 5 assertions, 1 failures

After all three reverts (git diff --stat clean of probes): 2 runs, 5 assertions, 0 failures on the untouched tree. The gate also refute_emptys the file glob, so it cannot pass vacuously.

Test numbers

	Baseline (923f588)	This branch
rake test:unit	528 runs, 1486 assertions, 0 failures	529 runs, 1490 assertions, 0 failures
rake test	—	562 runs, 1535 assertions, 0 failures, 6 skips
standardrb lib test	—	140 files, no offenses

529 = 528 − 1 (deleted anchor test) + 2 (gate tests).

No CHANGELOG or version.rb changes.

🤖 Generated with Claude Code

Summary by Sourcery

Remove the dead viewport anchor seam and enforce that legacy compatibility code remains removable alias-only plumbing.

Enhancements:

• Remove the unused viewport anchor: parameter and its associated test and call-site arguments.
• Add a unit-test gate that enforces the legacy compatibility trees remain alias-only, while explicitly tracking the remaining derived configuration logic.

Tests:

• Add coverage for legacy-tree shape enforcement and the allowed configuration method inventory.

[sourcery-ai]

Sorry @pftg, you have reached your weekly rate limit of 500000 diff characters.

Please try again later or upgrade to continue using Sourcery

[sourcery-ai]

Reviewer's Guide

Removes an unused anchor: keyword argument from the viewport preparation seam and introduces a unit test gate that enforces the legacy capybara/capybara_screenshot_diff trees remain alias-only shims around SnapDiff, with a narrowly allowlisted config file.

Sequence diagram for viewport preparation without the dead anchor keyword

sequenceDiagram
    participant Matcher as SnapDiff::ScreenshotMatcher
    participant Viewport as SnapDiff::Capture::Viewport
    participant Browser as BrowserHelpers

    Matcher->>Viewport: prepare!(expected_window_size)
    Viewport->>Browser: window_size_is_wrong?(expected_window_size)
    Browser-->>Viewport: size mismatch or valid
    Viewport-->>Matcher: return or raise WindowSizeMismatchError

Loading

Flow diagram for the legacy tree alias-only CI gate

flowchart TD
    Start["rake test:unit"] --> Discover["Discover legacy .rb files and root entry files"]
    Discover --> Parse["Strip comments/blanks and fold continuations"]
    Parse --> Validate["Validate allowed declarations and SnapDiff forwarders"]
    Validate --> Inventory["Check config_legacy.rb method inventory"]
    Inventory --> Result{Pass?}
    Result -->|Yes| Green["Unit suite passes"]
    Result -->|No| Failure["Report file and offending line"]

Loading

File-Level Changes

Change	Details	Files
Remove the dead anchor: keyword argument from viewport preparation and its supporting test coverage.	Change Viewport.prepare! to accept only expected_window_size and drop the anchor: keyword parameter. Update both ScreenshotMatcher call sites to stop passing anchor: nil when preparing the viewport. Delete the unit test that asserted prepare! accepts the anchor parameter without using it. Simplify the viewport doc comment to describe only the window-size guard behavior.	lib/snap_diff/capture/viewport.rb lib/snap_diff/screenshot_matcher.rb test/unit/capture/viewport_test.rb
Add a CI-enforced unit test gate ensuring legacy capybara trees contain only aliasing and forwarding code, with a narrowly scoped allowlist for known remaining logic.	Add a unit test that glob-loads all legacy lib/capybara/*.rb and lib/capybara_screenshot_diff/**/*.rb files and asserts they contain only requires, namespace declarations, constant aliases, delegators, and one-line SnapDiff forwarder methods. Implement a line-shape analyzer that strips comments/blank lines, folds trailing-comma continuations, and flags any non-allowed constructs or multi-line method bodies as offences. Introduce an allowlist for config_legacy.rb with a documented reason, while separately pinning its exact set of defined methods so new logic there fails the gate. Ensure the gate test fails loudly with file/line details when real logic appears in legacy trees and cannot pass vacuously by asserting the glob is non-empty.	test/unit/legacy_tree_is_alias_only_test.rb

---

Tips and commands

Interacting with Sourcery

• Trigger a new review: Comment @sourcery-ai review on the pull request.
• Continue discussions: Reply directly to Sourcery's review comments.
• Generate a GitHub issue from a review comment: Ask Sourcery to create an
  issue from a review comment by replying to it. You can also reply to a
  review comment with @sourcery-ai issue to create an issue from it.
• Generate a pull request title: Write @sourcery-ai anywhere in the pull
  request title to generate a title at any time. You can also comment
  @sourcery-ai title on the pull request to (re-)generate the title at any time.
• Generate a pull request summary: Write @sourcery-ai summary anywhere in
  the pull request body to generate a PR summary at any time exactly where you
  want it. You can also comment @sourcery-ai summary on the pull request to
  (re-)generate the summary at any time.
• Generate reviewer's guide: Comment @sourcery-ai guide on the pull
  request to (re-)generate the reviewer's guide at any time.
• Resolve all Sourcery comments: Comment @sourcery-ai resolve on the
  pull request to resolve all Sourcery comments. Useful if you've already
  addressed all the comments and don't want to see them anymore.
• Dismiss all Sourcery reviews: Comment @sourcery-ai dismiss on the pull
  request to dismiss all existing Sourcery reviews. Especially useful if you
  want to start fresh with a new review - don't forget to comment
  @sourcery-ai review to trigger a new review!

Customizing Your Experience

Access your dashboard to:

• Enable or disable review features such as the Sourcery-generated pull request
  summary, the reviewer's guide, and others.
• Change the review language.
• Add, remove or edit custom review instructions.
• Adjust other review settings.

Getting Help

• Contact our support team for questions or feedback.
• Visit our documentation for detailed guides and information.
• Keep in touch with the Sourcery team by following us on X/Twitter, LinkedIn or GitHub.

[coderabbitai]

Warning

Review limit reached

@pftg, you've reached your PR review limit, so we couldn't start this review.

Next review available in: 3 minutes

Limit details: You’ve used the included review currently available.

You've used all free OSS reviews for now. Wait for the free limit to reset to keep reviewing this public repository.

How can I continue?

Wait for the limit to reset, then comment @coderabbitai review or push new commits to the PR.

An organization admin can change what happens after included review limits in Billing.

How do review limits work?

CodeRabbit enforces per-developer PR review limits within each organization.

For paid Pro and Pro+ reviews, CodeRabbit uses a developer's included PR review attempts over the past 7 days to set the current hourly allowance. At typical activity levels, the full plan allowance applies. Higher sustained activity can lower the allowance until earlier attempts leave the 7-day window.

Please refer docs for additional details.

Review details

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro Plus

Run ID: fca45cb8-34d3-497b-9bb4-0b69875f86db

📥 Commits

Reviewing files that changed from the base of the PR and between 923f588 and da62601.

📒 Files selected for processing (4)

• lib/snap_diff/capture/viewport.rb
• lib/snap_diff/screenshot_matcher.rb
• test/unit/capture/viewport_test.rb
• test/unit/legacy_tree_is_alias_only_test.rb

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}