locations for the binding types, and two lint traps the wrapper walked into

[aleksisch]

Supersedes #44, which GitHub auto-closed: a bad squash on a shallow clone produced a parentless commit, so the head briefly shared no ancestor with master. The branch is rebuilt on 15eed84 and the diff is 3 files, +11/-11.

1. The binding types carry a location

daslang removed TypeDecl(Type) — the constructor that leaves a type with no source location — so NodeId / LinkId / PinId stamp cppBindingLineInfo(), the call daslang's own typeFactory templates use.

2. NODEEDITOR001 fired on its own wrapper

Fixing the build surfaced why every tutorial-driving integration test was red, and it was not the constructor. The lint told a raw call from a wrapper body by comparing fileInfo — but a block-taking wrapper like begin_delete(ctx, blk) is inlined during inference and its body carries the call site's location, so both files matched and the lint accused begin_delete itself.

The inliner marks spliced nodes generated; a hand-written call is not. Measured on the failing app:

call	genFlags.generated
SetNodePosition (the user's own)	false
BeginDelete / EndDelete (spliced from begin_delete)	true

With that flag in the guard the app compiles again. It had been dying at initial compile FAILED, which left the harness polling a snapshot with no widget registry until it timed out — that is the whole wait_for_widget MAIN_WIN/graph failure family.

3. The clipboard chord double-swapped on macOS

test_clipboard_tutorial fed Super under ConfigMacOSXBehaviors, but dasImgui already pre-swaps Ctrl/Super for synth keys to cancel ImGui's AddKeyEvent swap (imgui_boost_runtime.das: "a synth Ctrl lands as io.KeyCtrl on every platform"). Feeding Super double-swapped and no chord fired — on the macOS runner Ctrl+D left the node count at 2 where 3 was expected. The chord is Ctrl on every platform now, matching the tutorial it drives.

Plus the three style findings on the touched file: comment to one line (STYLE015), the emitted message loses its em dash (STYLE039), unused require strings dropped (STYLE030).

Verified

• integration dir, headless, Linux: 23 tests, 23 passed, 0 failed (from 8 failing)
• failed_node_editor_lint_raw.das still reports, so a genuinely raw call is still caught
• repo lint on both touched .das: 0 issue(s), 0 error(s)
• the module builds against current daslang master (the one without the old constructor)

Not verified locally: the macOS chord path — no macOS box here. The macos-latest lane is its proof.

🤖 Generated with Claude Code