test: test coverage for core transformation functions

[jamesbhobbs]

Add comprehensive test coverage for core transformation functions

Summary

Added test suites for three critical transformation functions in the jupyterlab-deepnote extension, improving overall test coverage from 26.57% to 55.94%. The newly tested functions handle the core logic for converting Deepnote files to Jupyter notebook format.

Files added:

• convert-deepnote-block-type-to-jupyter.spec.ts - 30 tests covering all block type mappings (code, markdown, SQL, visualizations, input types, etc.)
• convert-deepnote-block-to-jupyter-cell.spec.ts - 13 tests for block-to-cell conversion including output handling and metadata preservation
• transform-deepnote-yaml-to-notebook-content.spec.ts - 11 tests for the main YAML transformation pipeline

All three files now have 100% line coverage.

Review & Testing Checklist for Human

• Verify mock accuracy: The tests heavily mock @deepnote/blocks functions (createPythonCode, createMarkdown, deserializeDeepnoteFile). Ensure the mocked behavior matches the actual library functions, especially for edge cases.
• Manual end-to-end test: Open a real .deepnote file in JupyterLab locally to verify the transformation still works correctly with actual data. The tests are unit-level only and don't exercise the full integration.
• Review type assertions: Several as any[] and ! operators were added to handle TypeScript strict checking in tests. While necessary for the test infrastructure, verify these don't indicate actual type safety issues in the production code.

Test Plan

1. Clone the branch and run jlpm test to verify all 54 tests pass
2. Start JupyterLab locally with jupyter lab --debug
3. Open the example.deepnote file (or create a test .deepnote file)
4. Verify it opens correctly and cells display as expected
5. Test the NotebookPicker dropdown to switch between notebooks within the file

Notes

• All tests pass locally (54 tests)
• Lint and typecheck pass
• No changes to production code, only test additions
• The mocking approach was necessary since @deepnote/blocks is a proprietary package with limited testability

Requested by: James Hobbs (james@deepnote.com) (@jamesbhobbs)
Session: https://app.devin.ai/sessions/13b72387222a4a1a9322d6dbd266d01c

Summary by CodeRabbit

• Tests
  • Added extensive unit and UI tests covering Deepnote→Jupyter conversions, YAML→notebook transformations, notebook picker behavior, and block-type mappings.
• Chores
  • Updated test runner configuration and stricter linting rules.
• Bug Fixes
  • Improved initialization and error handling in the notebook picker and request handling to surface clearer failures.

[devin-ai-integration]

🤖 Devin AI Engineer

I'll be helping with this pull request! Here's what you should know:

✅ I will automatically:

• Address comments on this PR that start with 'DevinAI' or '@devin'.
• Look at CI failures and help fix them

⚙️ Control Options:

• Disable automatic comment and CI monitoring

[coderabbitai]

Caution

Review failed

The head commit changed during the review from 4a5bd7e to 915f9fd.

📝 Walkthrough

Walkthrough

This PR adds and updates test suites and adjusts configs/source handling: new unit tests for convertDeepnoteBlockToJupyterCell, convertDeepnoteBlockTypeToJupyter, transformDeepnoteYamlToNotebookContent, and a rewritten NotebookPicker test file. Tests cover many Deepnote block types and mappings (code/sql/visualization/input/markdown/text-cell variants), output handling (multiple outputs, truncated flags), nbformat/kernel metadata population, multi-notebook and empty-notebook cases, deserialization error propagation, and non-mutation of inputs. Code changes include improved promise/error handling and null-checks in NotebookPicker, stricter error and JSON handling in requestAPI, jest transform allowlist additions, and stricter TypeScript ESLint rules in package.json. No exported/public API signatures were changed.

Sequence Diagram(s)

sequenceDiagram
  autonumber
  participant Test
  participant Transformer as transformDeepnoteYamlToNotebookContent
  participant Deserializer as deserializeDeepnoteFile
  participant Converter as convertDeepnoteBlockToJupyterCell
  Note right of Transformer #f0f8ff: Top-level YAML -> NotebookContent

  Test->>Transformer: call(YAML)
  Transformer->>Deserializer: deserializeDeepnoteFile(YAML)
  alt deserialization fails
    Deserializer-->>Transformer: throw Error
    Transformer-->>Test: propagate Error
  else deserialization succeeds
    Deserializer-->>Transformer: DeepnoteFile(notebooks[])
    loop for each notebook
      loop for each block
        Transformer->>Converter: convertDeepnoteBlockToJupyterCell(block)
        Converter-->>Transformer: Jupyter cell(s) (type, source, outputs, metadata)
      end
      Transformer->>Transformer: assemble cells + metadata (kernelspec, language_info, deepnote.notebooks)
    end
    Transformer-->>Test: NotebookContent(nbformat, nbformat_minor, cells, metadata)
  end

Loading

sequenceDiagram
  autonumber
  participant Test
  participant Mapper as convertDeepnoteBlockTypeToJupyter
  Note right of Mapper #e6f7ff: Maps Deepnote block.type -> "code" | "markdown"

  Test->>Mapper: call(block.type)
  alt type ∈ {code, sql, visualization, notebook-function, big-number}
    Mapper-->>Test: "code"
  else type ∈ {input-text, input-checkbox, input-textarea, input-file, input-select, input-date-range, input-date, input-slider}
    Mapper-->>Test: "code"
  else type ∈ {markdown, image, button, separator, text-cell-*}
    Mapper-->>Test: "markdown"
  else unknown or empty
    Mapper-->>Test: "markdown"
  end

Loading

Possibly related PRs

• fix: Add ESLint rules for type safety and fix violations #46 — overlapping edits to ESLint rules, NotebookPicker promise/null-check changes, handler.ts parsing/error behavior, and corresponding tests.
• feat(charts): Add visualization blocks support vscode-deepnote#66 — implements visualization block conversion paths referenced by the visualization-to-code tests.

Suggested reviewers

• Artmann
• saltenasl

Pre-merge checks

✅ Passed checks (3 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title Check	✅ Passed	The title "test: test coverage for core transformation functions" directly describes the primary objective of this PR. The changeset's dominant focus is adding three comprehensive test files (54 tests total) covering core transformation functions like convertDeepnoteBlockTypeToJupyter, convertDeepnoteBlockToJupyterCell, and transformDeepnoteYamlToNotebookContent. While the PR includes secondary changes to jest.config.js, package.json, NotebookPicker.tsx, and handler.ts, these are supporting changes rather than the main point. The title is clear, concise, and would help a teammate scanning history understand the primary change at a glance.
Docstring Coverage	✅ Passed	Docstring coverage is 100.00% which is sufficient. The required threshold is 80.00%.

---

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

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 69.56%. Comparing base (48898e4) to head (915f9fd).
⚠️ Report is 1 commits behind head on main.
✅ All tests successful. No failed tests found.

Additional details and impacted files

@@             Coverage Diff             @@
##             main      #45       +/-   ##
===========================================
+ Coverage   52.56%   69.56%   +16.99%     
===========================================
  Files          13       13               
  Lines         253      253               
  Branches       28       28               
===========================================
+ Hits          133      176       +43     
+ Misses        116       73       -43     
  Partials        4        4               

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

🚀 New features to boost your workflow:

• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[coderabbitai]

Actionable comments posted: 3

📜 Review details

Configuration used: CodeRabbit UI

Review profile: ASSERTIVE

Plan: Pro

📥 Commits

Reviewing files that changed from the base of the PR and between ef0931b and 0105761.

📒 Files selected for processing (3)

• src/__tests__/convert-deepnote-block-to-jupyter-cell.spec.ts (1 hunks)
• src/__tests__/convert-deepnote-block-type-to-jupyter.spec.ts (1 hunks)
• src/__tests__/transform-deepnote-yaml-to-notebook-content.spec.ts (1 hunks)

🧰 Additional context used

🧠 Learnings (1)

📚 Learning: 2025-10-14T13:43:36.364Z

Learnt from: CR
PR: deepnote/deepnote#0
File: CLAUDE.md:0-0
Timestamp: 2025-10-14T13:43:36.364Z
Learning: Applies to packages/*/src/**/*.test.{ts,tsx} : Follow existing test patterns (see packages/blocks/src/blocks/*.test.ts); write clear names, group with describe(), cover edge cases, errors, special characters, and verify exact output for code-generation functions

Applied to files:

• src/__tests__/convert-deepnote-block-type-to-jupyter.spec.ts
• src/__tests__/convert-deepnote-block-to-jupyter-cell.spec.ts

🧬 Code graph analysis (2)

src/__tests__/convert-deepnote-block-type-to-jupyter.spec.ts (1)

src/convert-deepnote-block-type-to-jupyter.ts (1)

• convertDeepnoteBlockTypeToJupyter (1-32)

src/__tests__/convert-deepnote-block-to-jupyter-cell.spec.ts (1)

src/convert-deepnote-block-to-jupyter-cell.ts (1)

• convertDeepnoteBlockToJupyterCell (10-44)

🔇 Additional comments (6)

src/__tests__/convert-deepnote-block-type-to-jupyter.spec.ts (2)

6-144: Good coverage and structure.

Covers all current mappings, unknowns, and empty input. Nice separation of suites.

---

8-29: Reduce duplication with table-driven tests.

Consider test.each to compress repetitive cases and ease future additions.

Example:

+describe('maps to code', () => {
+  test.each([
+    'code','sql','notebook-function','big-number','visualization',
+    'input-text','input-checkbox','input-textarea','input-file',
+    'input-select','input-date-range','input-date','input-slider'
+  ])('%s -> code', (t) => {
+    expect(convertDeepnoteBlockTypeToJupyter(t)).toBe('code');
+  });
+});

Based on learnings

Also applies to: 30-68, 72-131, 133-143

<details>
<summary>⛔ Skipped due to learnings</summary>

Learnt from: CR
PR: deepnote/deepnote#0
File: CLAUDE.md:0-0
Timestamp: 2025-10-14T13:43:36.364Z
Learning: Applies to packages//src/**/.test.{ts,tsx} : Follow existing test patterns (see packages/blocks/src/blocks/*.test.ts); write clear names, group with describe(), cover edge cases, errors, special characters, and verify exact output for code-generation functions

Learnt from: CR
PR: deepnote/deepnote-internal#0
File: CONVENTIONS.md:0-0
Timestamp: 2025-10-22T11:56:07.963Z
Learning: Applies to **/*.{spec,test}.{ts,tsx,js,jsx} : Avoid defining test-specific values outside individual tests; define within each test for clarity

Learnt from: CR
PR: deepnote/deepnote#0
File: CONVENTIONS.md:0-0
Timestamp: 2025-08-18T23:41:22.727Z
Learning: Applies to **/*.{spec,e2e}.{ts,tsx} : Avoid defining test-specific values outside individual test cases

Learnt from: CR
PR: deepnote/deepnote-internal#0
File: CONVENTIONS.md:0-0
Timestamp: 2025-10-22T11:56:07.963Z
Learning: Applies to **/*.{spec,test}.{ts,tsx,js,jsx} : Only specify data relevant to the tested requirement inside a single test; move general setup to factories/beforeEach

</details>

</blockquote></details>
<details>
<summary>src/__tests__/transform-deepnote-yaml-to-notebook-content.spec.ts (2)</summary><blockquote>

`1-310`: **Overall: solid, thorough tests.**

Great coverage across success and failure paths; metadata and structure checks are on point.

---

`7-9`: **Remove the ESM mock suggestion—project uses CommonJS, not ESM.**

The codebase doesn't use ESM (package.json lacks `"type": "module"` and jest.config.js has no ESM extensions). Jest.mock() doesn't need `__esModule: true` for CommonJS module mocks with named exports. The suggestion defends against a non-existent runtime.



> Likely an incorrect or invalid review comment.

</blockquote></details>
<details>
<summary>src/__tests__/convert-deepnote-block-to-jupyter-cell.spec.ts (2)</summary><blockquote>

`7-11`: **ESM-safe mocks.**

If the repo runs Jest in ESM mode, include `__esModule: true` in mock factories to avoid interop quirks.


```diff
-jest.mock('@deepnote/blocks', () => ({
+jest.mock('@deepnote/blocks', () => ({
+  __esModule: true,
   createPythonCode: jest.fn((block: any) => block.source || 'print("test")'),
   createMarkdown: jest.fn((block: any) => block.source || '# Test')
 }));
 
-jest.mock('../convert-deepnote-block-type-to-jupyter', () => ({
+jest.mock('../convert-deepnote-block-type-to-jupyter', () => ({
+  __esModule: true,
   convertDeepnoteBlockTypeToJupyter: jest.fn((type: string) => {
     // ...
   })
 }));

Also applies to: 12-35

⛔ Skipped due to learnings

Learnt from: CR
PR: deepnote/deepnote-internal#0
File: .cursor/rules/tests.mdc:0-0
Timestamp: 2025-10-22T11:56:53.582Z
Learning: Applies to {**/*.{test,spec}.{jsx,tsx},**/__tests__/**/*.{jsx,tsx}} : Only use jest.mock() for external modules in React tests; do not mock in-repo React hooks, context providers, or GraphQL operations

Learnt from: CR
PR: deepnote/deepnote-internal#0
File: CONVENTIONS.md:0-0
Timestamp: 2025-10-22T11:56:07.963Z
Learning: Applies to **/*.{spec,test}.{tsx,jsx} : Avoid excessive mocking in React component tests; use TestProvider for context; prefer apolloMocks and flags props over jest.mock for internal hooks/contexts

Learnt from: CR
PR: deepnote/deepnote-internal#0
File: .cursor/rules/tests.mdc:0-0
Timestamp: 2025-10-22T11:56:53.582Z
Learning: Applies to {**/*.{test,spec}.{jsx,tsx},**/__tests__/**/*.{jsx,tsx}} : Avoid excessive mocking in React component tests

---

5-5: False alarm—no runtime collision risk.

DeepnoteBlock is used only as a type annotation (never as a value). TypeScript's default compiler settings automatically elide unused imports. The import will not reach runtime, so jest.mock() presents no collision. The code is correct.

Likely an incorrect or invalid review comment.

[coderabbitai]

Actionable comments posted: 1

📜 Review details

Configuration used: CodeRabbit UI

Review profile: ASSERTIVE

Plan: Pro

📥 Commits

Reviewing files that changed from the base of the PR and between 0105761 and 67b5160.

📒 Files selected for processing (2)

• src/__tests__/convert-deepnote-block-to-jupyter-cell.spec.ts (1 hunks)
• src/__tests__/transform-deepnote-yaml-to-notebook-content.spec.ts (1 hunks)

🧰 Additional context used

🧠 Learnings (1)

📚 Learning: 2025-10-14T13:43:36.364Z

Learnt from: CR
PR: deepnote/deepnote#0
File: CLAUDE.md:0-0
Timestamp: 2025-10-14T13:43:36.364Z
Learning: Applies to packages/*/src/**/*.test.{ts,tsx} : Follow existing test patterns (see packages/blocks/src/blocks/*.test.ts); write clear names, group with describe(), cover edge cases, errors, special characters, and verify exact output for code-generation functions

Applied to files:

• src/__tests__/convert-deepnote-block-to-jupyter-cell.spec.ts

🧬 Code graph analysis (2)

src/__tests__/transform-deepnote-yaml-to-notebook-content.spec.ts (2)

src/transform-deepnote-yaml-to-notebook-content.ts (1)

• transformDeepnoteYamlToNotebookContent (7-58)

src/convert-deepnote-block-to-jupyter-cell.ts (1)

• convertDeepnoteBlockToJupyterCell (10-44)

src/__tests__/convert-deepnote-block-to-jupyter-cell.spec.ts (1)

src/convert-deepnote-block-to-jupyter-cell.ts (1)

• convertDeepnoteBlockToJupyterCell (10-44)

⏰ Context from checks skipped due to timeout of 90000ms. You can increase the timeout in your CodeRabbit configuration to a maximum of 15 minutes (900000ms). (4)

• GitHub Check: check_release
• GitHub Check: build
• GitHub Check: Audit - Production
• GitHub Check: Lint & Format

🔇 Additional comments (11)

src/__tests__/convert-deepnote-block-to-jupyter-cell.spec.ts (3)

189-250: Solid markdown cell tests.

Covers basic markdown, text-cell variants, image blocks, and metadata handling. Creator assertions added per past review (lines 204-208).

---

252-304: Good coverage of special block types.

Efficiently tests SQL, visualization, and all 8 input type variants converting to code cells.

---

12-35: Mock type list is currently synchronized with the implementation.

Verification confirms the mock contains all 13 types that map to 'code' in the actual implementation. However, the hardcoded duplication creates a maintenance risk: any future changes to the actual type mappings require manual updates here. Consider extracting the type list to a shared constant or document this synchronization requirement in a code comment.

src/__tests__/transform-deepnote-yaml-to-notebook-content.spec.ts (8)

7-21: Mock setup is reasonable for unit tests.

The simplified mock for convertDeepnoteBlockToJupyterCell handles code/markdown distinction appropriately. Per PR objectives, verify against actual behavior during manual testing.

---

28-65: Solid foundation test for basic transformation.

Verifies cell conversion, metadata structure, and nbformat version.

---

67-113: Verifies all notebooks in metadata.

Correctly ensures both notebooks are indexed by name with proper IDs.

---

115-172: Correctly verifies first-notebook-only for primary cells.

Lines 155-171 properly assert converter called 3 times: twice for metadata (both notebooks), once more for primary cells (first notebook). Addresses past review concern.

---

174-210: Good edge case coverage for empty scenarios.

Correctly handles empty notebooks (0 cells) and missing notebooks (fallback message).

---

212-233: Verifies kernel metadata presence.

Ensures Python kernel metadata included in output.

---

235-243: Error handling verified.

Ensures deserialization failures are properly caught and re-thrown with context.

---

245-327: Structure and ordering tests thorough.

Verifies notebook metadata includes cell details (245-277) and that mixed block types maintain order (279-327).

[coderabbitai]

Actionable comments posted: 1

📜 Review details

Configuration used: CodeRabbit UI

Review profile: ASSERTIVE

Plan: Pro

📥 Commits

Reviewing files that changed from the base of the PR and between 67b5160 and bdab09a.

📒 Files selected for processing (1)

• src/__tests__/convert-deepnote-block-to-jupyter-cell.spec.ts (1 hunks)

🧰 Additional context used

🧬 Code graph analysis (1)

src/__tests__/convert-deepnote-block-to-jupyter-cell.spec.ts (1)

src/convert-deepnote-block-to-jupyter-cell.ts (1)

• convertDeepnoteBlockToJupyterCell (10-44)

⏰ Context from checks skipped due to timeout of 90000ms. You can increase the timeout in your CodeRabbit configuration to a maximum of 15 minutes (900000ms). (3)

• GitHub Check: check_release
• GitHub Check: build
• GitHub Check: Lint & Format

🔇 Additional comments (3)

src/__tests__/convert-deepnote-block-to-jupyter-cell.spec.ts (3)

1-7: Verify mock setup is configured.

No jest.mock() for @deepnote/blocks is visible in this file. Past review comments request assertions about createPythonCode and createMarkdown invocation, but the current tests don't contain such assertions. Either mocks are configured in a shared test setup file, or creator invocation verification needs to be added.

Can you confirm whether jest.mock('@deepnote/blocks') is configured in a test setup file (e.g., jest.config.ts, setupTests.ts, or similar)? If mocks are defined there, please share that context so I can verify they align with how the tests use them.

---

167-226: Inconsistency: Past review requested createMarkdown invocation assertions.

Past review comment (on markdown cells) explicitly requested: "Mirror the code-cell expectation for createMarkdown" and proposed adding mock assertions like expect(createMarkdown).toHaveBeenCalledTimes(1). The current tests do not include these assertions.

Clarify: Were creator invocation assertions intentionally deferred, or should they be added here? If they should be added, I can generate the required test updates.

---

9-25: Verify mock behavior aligns with implementation.

The test expects result.source to equal 'print("hello")', implying createPythonCode(block) returns the block's content directly. Past review comments requested asserting that createPythonCode is called with a cloned (not original) block.

If mocks are set up elsewhere, ensure they reflect the actual behavior: that the function receives a deep clone and that the returned source string matches expectations across all code-cell tests.

[dinohamzic]

@Artmann @saltenasl this is now ready for review.

[dinohamzic]

@coderabbitai approve

[coderabbitai]

✅ Actions performed

Comments resolved and changes approved.