Skip to content

About

No description, website, or topics provided.

Resources

Stars

0 stars

Watchers

0 watching

Forks

Latest commit

 

History

5 Commits

Folders and files

NameName
Last commit message
Last commit date
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Semantic Search for Nextpad++ (macOS)

Describe what you're looking for in plain language — every sentence in the document is tinted red → grey → green by semantic similarity to your query. Entirely on-device: Apple NaturalLanguage embeddings scored on the GPU via Metal Performance Shaders. Nothing leaves your Mac.

Credit: based on nextpad-plus-plus-macos PR #346 by Kristian Rickert (@krickert) — the embedding pipeline, Metal scoring engine, spillable vector index, color ramp and tests are his work, adapted from an in-host feature to this standalone plugin.

Usage

  • Plugins → Semantic Search → Semantic Heatmap Search opens a docked panel (bottom of the window on Nextpad++ 1.1.1+; it can be moved with the panel's dock buttons and reopens where you left it).
  • Type a query — the heatmap updates live (debounced). Enter forces an immediate run; clearing the field clears the heatmap.
  • Strict / Standard / Broad shifts how much similarity is needed to show green. Colors are relative to the model and sensitivity — they are not confidence percentages.
  • The status line reports the index state ("214 of 230 sentences indexed · Metal/MPS") and any errors. Sentences that cannot be embedded stay uncolored.

Requirements & limits

  • Nextpad++ for macOS 1.0.3+ (1.1.1+ recommended: bottom-dock default and panel restore); macOS 14+ at runtime for the embedding models — on older systems the menu item explains and does nothing else.
  • A Metal/MPS-capable GPU (no CPU fallback — the plugin reports "Metal GPU unavailable" rather than silently degrading).
  • An Apple NaturalLanguage embedding model for the document's language. The contextual model may be downloaded on demand by macOS (an OS-managed asset, like a dictation language); the static sentence model is used until it arrives.
  • Documents over 2 MiB or 4,096 sentences are refused rather than partially searched. Non-UTF-8 documents are refused.

Building

cmake -B build -DCMAKE_BUILD_TYPE=Release
cmake --build build -j
ctest --test-dir build --output-on-failure   # unit tests (colors + index)
cmake --install build                        # → ~/Library/…/Nextpad++/plugins/SemanticSearch

Requires the nextpad-plus-plus-macos repo checked out as a sibling directory (plugin interface + Scintilla headers).

License

GPL-3.0, matching the host and the original contribution.

About

No description, website, or topics provided.

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages