fix: compare review diagrams against target baseline

[ivanmilevtues]

Ivan's comment:
There was an issue with the mermaid renders, that is fixed. I stumbled across another issue which was related to forks, we used to check with the real repo instead of our fork (fixed).

Example PR:
ivanmilevtues/optimizerDuck#1

Summary

• compare review-mode diagrams against .codeboarding/analysis.json from the PR target/base commit, not PR-head history
• re-copy the committed base analysis after cache restore so cached generated base artifacts cannot replace the versioned baseline used for the diff
• fix Mermaid label escaping from invalid #amp;/#quot; forms to GitHub-compatible numeric refs (#38;, #34;, etc.)

Root cause

The deployed action searched backwards through the PR head branch for .codeboarding/analysis.json. If the target branch had a synced baseline that was not present in the PR branch history, review mode missed it and regenerated a fresh base analysis. That made the diagram compare fresh-base output against fresh-head output instead of comparing against the versioned target-branch baseline.

For CodeBoarding/MORT PR #2, the workflow log showed exactly that fallback and produced n_changed: 0, even though comparing main:.codeboarding/analysis.json against the generated PR artifact produced changes.

Validation

Local:

python3 -m unittest tests.test_diff_to_mermaid
python3 -m unittest discover -s tests

GitHub checks on this PR:

• Tests / unittest ✅
• Tests / lint ✅
• CodeBoarding review / review ✅

MORT integration test:

• Test PR: Test CodeBoarding action baseline fix MORT#3
• Test run used CodeBoarding/CodeBoarding-action@fix/review-target-baseline
• Log confirmed the fixed base selection:

Using committed baseline at PR base 7402b58c4d5713a29081f855d01ec934a3603710; analysis metadata source is 9ec8f3906a5b055fb39c9d8fa8600cff3d756e26.

That run later stopped because the hosted/free-tier quota returned 402 Resource exhausted, so it did not complete a final MORT diagram, but it did verify the critical corrected path: the action now reads the target/base branch baseline.

[codeboarding-review]

Architecture review · 14 components changed

graph LR
    n_Analysis_Orchestrator["Analysis Orchestrator"]
    n_Structural_Diffing_Engine["Structural Diffing Engine"]
    n_Mermaid_Visualization_Engine["Mermaid Visualization Engine"]
    n_UX_Integration_Layer["UX #38; Integration Layer"]
    n_Analysis_Orchestrator -- "triggers health checks and status reporting" --> n_UX_Integration_Layer
    n_Structural_Diffing_Engine -- "provides structural change data for rendering" --> n_Mermaid_Visualization_Engine
    n_Structural_Diffing_Engine -- "supplies change detection flags and diff summar…" --> n_UX_Integration_Layer
    n_Mermaid_Visualization_Engine -- "queries specific file changes for component map…" --> n_Structural_Diffing_Engine
    n_UX_Integration_Layer -- "filters visual noise for final presentation" --> n_Mermaid_Visualization_Engine
    n_Mermaid_Visualization_Engine -- "provides component change counts for summary re…" --> n_UX_Integration_Layer
    classDef added fill:#1f883d,stroke:#0b5d23,color:#ffffff;
    classDef modified fill:#bf8700,stroke:#7d4e00,color:#ffffff;
    classDef deleted fill:#cf222e,stroke:#82071e,color:#ffffff,stroke-dasharray:5 3;
    class n_Structural_Diffing_Engine,n_Mermaid_Visualization_Engine modified;
    linkStyle 1,2,3,4 stroke:#7d4e00,stroke-width:2px;
    linkStyle 5 stroke:#82071e,stroke-width:2px,stroke-dasharray:5 3;

Loading

Colors indicate component changes compared to main: 🟩 Added · 🟨 Modified · 🟥 Removed

Download the PR analysis artifacts from this workflow artifact.

Structural Diffing Engine : 1 file changed

• scripts/diff_to_mermaid.py

Mermaid Visualization Engine : 1 file changed

• scripts/diff_to_mermaid.py

---

⚠️ 1 architecture issue found — open CodeBoarding to explore them.

Explore this PR’s architecture in your browser or VS Code.

_{codeboarding-action · run 27653739150}

[ivanmilevtues]

These were causing friction and were the reason why we wound't render properly the diff.

[ivanmilevtues]

This for some really odd reason was comparing with the main of the real repo and not the main of the fork.

Hopefully this is resolved now.

This action.yml is getting to be quite big. I would flag it for refactor, slim-down, weight loss program.

[Svilen-Stefanov]

I'm wondering, how did you even verify this is working? I suppose you saw that the high level components between the action and the webview are the same?
I'm looking at 14 components changed, then opening the webview, where I see 2 high level component changed and 7 lower lvl components modified.

The webview is still quite off with the actual modifications, I'm wondering if you've seen these and if you have an assumption where these are coming from?

[ivanmilevtues]

I am merging this so I can continue with the investigation and will do a new PR for that.