Nuancery · Software practice
Nuancery combines visual judgment, interaction design, and technical research in product development. Labs publishes the parts of that work that can stand on their own.
Swift components
Each component is extracted from production work and documents its API, supported platforms, dependencies, examples, and application-owned responsibilities.
Keep rich-text documents editable without losing their authored structure.
A document codec and editor-routing layer for attributed text, checklist semantics, runtime metadata, and safe reconciliation of unsupported values.
Load media progressively while keeping gallery state predictable.
A gallery coordination layer for lazy image and video loading, cancellation, dismissal, player release, and UIKit photo-browser integration.
Run OCR stages with explicit geometry and model boundaries.
A staged OCR pipeline for image preprocessing, Vision text recognition, reading-order reconstruction, and optional PaddleOCR Core ML inference.
Add an accessible page-curl transition with a controlled fallback.
A SwiftUI transition layer that keeps page-curl presentation, accessibility behavior, and reduced-motion handling within a small integration boundary.
Use navigation-style segmented controls across native glass and material fallback paths.
SwiftUI and UIKit segmented controls with generic selection identity, reselect callbacks, drag cancellation, accessibility identifiers, and Reduce Motion handling. Native glass is used where iOS provides it.
Normalize image-derived colors before they become muddy surfaces.
A synchronous color policy that keeps a source hue when possible, raises dark or overly saturated colors into a readable range, and selects a foreground color for the resulting surface.
Derive a palette from the image region visible in a host container.
A Core Graphics palette analyzer that accounts for the host container, content mode, and crop anchor. It returns colors, ranked candidates, contrast-aware foregrounds, and a deterministic cache key.
Allocate deterministic lanes for multi-day all-day event bars.
A UI-independent layout engine for clipped calendar-grid ranges. It normalizes dates with the supplied Calendar, splits bars at row boundaries, and reports deterministic overflow and daily lane coverage.
Select an HTTP(S) URL from explicit input or shared text.
A synchronous resolver for explicit URLs, text candidates, provider domains, and legacy source markers. It performs no network requests and rejects documented local-address forms.
Create square subject or room thumbnails from a CGImage.
A synchronous thumbnail processor using Apple Vision, Core Image, and Core Graphics. It reports whether the result used a subject cutout, fallback, or room treatment.
Render grouped bar data with Swift Charts or Canvas.
Flat and dimensional SwiftUI bar charts with application-owned selection, formatting, colors, and navigation. Geometry and hit-testing helpers are available for tests.
Order OCR candidates into horizontal or vertical reading lines.
A synchronous geometry engine for horizontal, vertical, automatic, and multipage reading order. It preserves caller-owned candidate identifiers and does not perform OCR.
Maintenance
Each repository includes regression tests for the behavior exposed by its public API.
Storage, localization, navigation, and domain-specific decisions remain in the integrating application.
Ordering, fallback, geometry, failure, and concurrency behavior are described in the API documentation.
wondays: The components are maintained separately from wondays, a Nuancery application that uses the related code.