Crate oxicode_snapcompact 

Snapcompact — bitmap-frame context compression for vision-capable LLMs.

Ported from omp (MIT) — packages/snapcompact/src/snapcompact.ts. See docs/ref-porter/xai-org-grok-build.md section E2 for design rationale.

How it works

Instead of asking an LLM to summarize discarded history, snapcompact renders discarded text as dense bitmap frames that vision-capable models read back directly. Local and deterministic — no LLM call, no API key, no latency beyond rendering.

Scope of this crate

Shape system, conversation serialization, shape selection, and the full text→PNG rasterizer (ported from omp’s pi-natives/src/snapcompact.rs — see the renderer submodule). The renderer uses bundled BDF/HEX bitmap fonts plus the bundled Silver TrueType font for non-Latin glyphs, all under permissive licenses (see NOTICE.md).

compact() always returns real PNG-encoded bytes — there is no “no-op renderer” path. When a frame fails to render, the error is logged via tracing and that frame’s bytes are empty; the caller can decide whether to drop or surface partial results.

Modules

renderer

Renderer submodule — pi-natives ported text→PNG rasterizer.

Structs

BoundedSource

The minimal envelope returned by compact when no renderer is available — shape chosen, text chunked into per-frame source ranges, but frame bytes are empty.

CompactOptions

Options for compact.

CompactPreparation

Inputs to a compaction pass.

CompactResult

Outcome of a compaction pass.

FrameRange

Per-frame source range.

FrameRef

A reference to a rendered frame.

Shape

One eval-validated frame shape.

Enums

Font

Bundled font name.

Variant

Ink variant.

Constants

SHAPES

Eval-validated shape table. Names match omp’s SHAPE_VARIANTS keys exactly so the Rust and TS registries stay in sync.

Functions

compact

Run a compaction pass — render each frame slice through the pi-natives ported renderer (render_snapcompact_png) so the returned bytes are real PNG-encoded image data.

normalize

Normalize conversation text for rasterization.

prepare

Render the conversation text into a serializable compact envelope.

resolve_shape

Pick the eval-validated shape for model_id.

serialize_conversation

Compact conversation text to one line per turn.