- SidebarTab: Use placeholder selector instead of test ID for search input
- KeyboardHelper: Remove redundant canvas fallback (null is valid)
- groupNode: Use variable for category name instead of hardcoded string
- nodeLibrary: Ensure tab is open before filling search input
- subgraph: Reload workflow after page reload to restore state
- useSettingSearch: Use role selector and simplify selected class assertion
Amp-Thread-ID: https://ampcode.com/threads/T-019c18c2-667a-7179-bb2d-86f35cd1d7b0
Co-authored-by: Amp <amp@ampcode.com>
Remove 40 instances of `as any` from browser_tests directory.
Changes made:
- Changed extensionManager type from ExtensionManager to WorkspaceStore
- Added test setting IDs to production apiSchema.ts
- Added Window.__ws__ declaration in browser_tests/types.d.ts
- Used type narrowing for Subgraph objects
- Replaced `as any` with `as ComfyWorkflowJSON` in groupNode.spec.ts
- Added non-null assertions where values are guaranteed
Known issues requiring remediation:
- Test settings pollute production schema (should use test augmentation)
- WorkspaceStore export breaks ExtensionManager API contract
- ComfyWorkflowJSON cast hides ISerialisedGraph type gap
- z.any() usage violates project rules
Amp-Thread-ID: https://ampcode.com/threads/T-019c1833-2352-728b-a523-a8f440fd3ba1
Co-authored-by: Amp <amp@ampcode.com>
- Add browser_tests/types/globals.d.ts with Window interface augmentation
- Add types for app, graph, LiteGraph, LGraphBadge and test globals
- Standardize window['prop'] to window.prop across 37 test files
- Update browser_tests/tsconfig.json to include new type definitions
Amp-Thread-ID: https://ampcode.com/threads/T-019c16b3-cc31-70ea-9727-a933cb0ee942
Co-authored-by: Amp <amp@ampcode.com>
Migrate all test callers to use helper classes directly:
- nodeOps: getGraphNodesCount, getNodeRefById, selectNodes, etc.
- canvasOps: dragAndDrop, zoom, pan, resetView, etc.
- subgraph: rightClickInputSlot, connectToOutput, etc.
Reduces ComfyPage.ts from 1335 to 1085 lines (-250 lines).
Amp-Thread-ID: https://ampcode.com/threads/T-019c1318-58b7-757f-801c-5b676f0d3421
Co-authored-by: Amp <amp@ampcode.com>
- widget.spec.ts: use expect().toPass() for combo refresh
- subgraph.spec.ts: use comfyPage.setup() after reload
- nodeLibrary.spec.ts: use toHaveCount() for debounced search
- remoteWidgets.spec.ts: use expect().toPass() for backoff test
- litegraphUtils.ts: refactor navigateIntoSubgraph to use toPass()
Amp-Thread-ID: https://ampcode.com/threads/T-019c11f1-e839-772c-8890-b77cfdf70124
Co-authored-by: Amp <amp@ampcode.com>
## Cause
When graphs are actually exported, several layers of cleanup are
applied. Among these is link compression. Any widgets with inputs that
aren't used do not have inputs stored in the workflow. This was
implemented for backwards compatibility with the old "convert to input"
system for widgets. As part of this process, the target_slots on links
are rewritten such that they point to the index of the widget as if
unconnected widgets did not exist.
This "incorrect" state for links is only corrected AFTER a workflow has
loaded because the 'fix' method needs nodes to be initialized in order
to calculate the correct target_slot
This becomes a problem when subgraphs are introduced. SubgraphInputs
need to resolve a link to its target slot in order to construct a clone
of the linked widget DURING the loading process. Since this target slot
is not accurate, this can result in the cloned widget having the wrong
type.
For a minimal reproduction:
- Create a subgraph with an Empty Latent Image with batch_size linked to
the Subgraph Input
- Export the workflow
- On load, the batch_size has step and min attributes which incorrectly
correspond to width
## Fix
There's multiple possible ways to address this and input on direction is
appreciated
- Fix links before loading graph
- Likely to break with any dynamic state
- Fix links, then load graph again
- Ugly, bad performance, dynamic state may require multiple passes to
correctly ripple
- In the Subgraph code, ignore target_slot and instead `.find()` input
with matching linkId (proposed)
- Promising, but means accepting that state is just wrong sometimes.
Another forever footgun.
- Entirely remove the input compression
- Some people may complain, and old workflows still need to be supported
- Only remove target_slot redirection inside subgraphs
- Creates ugly logical difference between what happens inside and
outside subgraphs.
- Still leaves old workflows broken
┆Issue is synchronized with this [Notion
page](https://www.notion.so/PR-7388-Remove-target_slot-compression-from-subgraph-exports-2c66d73d3650815d8c96c5047958ab67)
by [Unito](https://www.unito.io)
* move subgraph test assets into subfolder
* [refactor] Organize browser test assets into logical folders
Reorganized test assets for better maintainability:
- groupnodes/: GroupNode feature tests
- groups/: Visual grouping tests
- missing/: Missing nodes/models tests
- links/: Link-related tests
- inputs/: Input widget tests
- widgets/: Widget-specific tests
- nodes/: Node-related tests
- workflowInMedia/: Workflow media files
Updated all loadWorkflow references to use new folder structure.
Fixed programmatic filename references to prevent test failures.
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
* [fix] Update mobile test to use new asset path
* [fix] Update remaining loadWorkflow calls to use new folder structure
* [fix] Fix remaining programmatic filename references
* [fix] Run prettier formatting
* [fix] Fix setupWorkflowsDirectory references to use correct folder paths
* [refactor] Rename subgraph folder to subgraphs for consistency
* [fix] Fix breadcrumb name in subgraph DOM widget test
* Update test expectations [skip ci]
---------
Co-authored-by: Claude <noreply@anthropic.com>
Co-authored-by: github-actions <github-actions@github.com>
