22. Metadata and EXIF Inspector

The metadata inspector is opened by a compact information icon in the active document bar or the platform menu. Desktop uses a dismissible right-side overlay up to 360 pixels wide. Android uses a bottom sheet on phones and a side sheet on tablets. Closing it restores the full document surface.

Facts are grouped as Source, Content, Embedded, Derived, and Renderer. Every fact includes a typed value, source, availability, unit, sensitivity, and copy policy. The interface shows Not provided, Unsupported, Redacted, Pending, or a stable error instead of inventing a value.

Source facts include display name, source kind, size, observed modification time, creation/birth time when the host supplies it, permissions or provider write state, identity confidence, and current external revision. Text facts include encoding, BOM, newline pattern, line and character counts, language mode, and round-trip safety. Mermaid facts include detected diagram type, parser version and status, current and preview revisions, stale-preview state, authored accessibility annotation presence, source bytes, edge count, output bytes, timings, and active limit result when available. Image facts include dimensions, frames, color model, orientation, EXIF, IPTC, XMP, and ICO entries. PDF and office facts include title, author, subject, keywords, creator, producer, page or structural counts, and format version when parsable.

EXIF GPS facts are collapsed and marked sensitive by default. Copying coordinates requires an explicit action. Checksums are absent until requested; SHA-256 calculation runs cancellably and reports the external revision it represents.

Metadata extraction obeys the same parser and resource limits as rendering. A metadata failure cannot prevent content viewing unless it proves the source is malformed for the selected renderer.