dev-array-viewerprotocol v2workspace: none
How Anyfile's NumPy Array Viewer Works
The custom parser reads NumPy headers and safe primitive layouts, then exposes array metadata and paged values without Python.
Viewing architecture
- 1NPZ entries are indexed before an array is selected.
- 2Binary values are decoded in bounded pages.
Main limitations
- — Object arrays are intentionally rejected.
- — High-dimensional values use an inspection-oriented flattened view.
MANIFEST
Formats declared by the plugin
Open-source projects and licenses
This viewer's core format logic is implemented within the Anyfile project; there is no separately credited external runtime project for this page. Repository licenses and third-party notices remain authoritative.