LAPACK/BLAS Documentation Search

Quickly search through official LAPACK/BLAS documentation directly from Raycast.

Features

• Fast Local Search: Search through 30+ LAPACK/BLAS routines instantly
• Offline Documentation: All documentation is stored locally as markdown files - no internet connection required
• Quick Access: View function signatures, parameters, and descriptions without leaving Raycast
• Open in Browser: Jump to the official online documentation when needed

Usage

1. Open Raycast
2. Search for "LAPACK/BLAS Docs"
3. Type the name of a LAPACK or BLAS routine (e.g., "dgemm", "dgesv", "daxpy")
4. View the documentation inline or press Enter to see the full documentation
5. Use the actions menu to:
   • View full documentation
   • Open in browser
   • Copy URL
   • Copy routine name

Included Routines

This extension currently includes documentation for:

BLAS Level 1 (Vector Operations)

• daxpy, saxpy - Vector addition (y := alpha*x + y)
• ddot, sdot - Dot product
• dscal, sscal - Scale a vector
• dnrm2, snrm2 - Euclidean norm
• dcopy, scopy - Copy a vector

BLAS Level 2 (Matrix-Vector Operations)

• dgemv, sgemv - Matrix-vector multiplication
• dger, sger - Rank-1 update

BLAS Level 3 (Matrix-Matrix Operations)

• dgemm, sgemm - Matrix-matrix multiplication
• dtrsm, strsm - Triangular solve

LAPACK Linear Systems

• dgesv, sgesv - Solve Ax = b using LU factorization
• dgetrf, sgetrf - LU factorization

LAPACK Least Squares and QR

• dgeqrf, sgeqrf - QR factorization

LAPACK Symmetric/Hermitian

• dpotrf, spotrf - Cholesky factorization
• dsyev, ssyev - Eigenvalue decomposition

LAPACK SVD

• dgesvd, sgesvd - Singular value decomposition

Architecture

This extension is designed to work completely offline:

1. Inventory (assets/docs/inventory.json): Contains metadata for all routines including names, descriptions, categories, and official URLs
2. Documentation Files (assets/docs/*.md): Markdown files containing the full documentation for each routine
3. Local Loading: All files are bundled with the extension and loaded from the local filesystem

No network requests are made during normal operation, making the search instant and reliable.

Adding New Documentation

To add documentation for additional LAPACK/BLAS routines:

1. Add an entry to assets/docs/inventory.json:

{
  "id": "routine-name",
  "name": "routine-name",
  "role": "blas:routine" or "lapack:driver",
  "category": "BLAS Level 1" or "LAPACK Linear Equations",
  "url": "https://netlib.org/lapack/...",
  "docPath": "routine-name.md",
  "description": "Brief description of the routine"
}

2. Create a markdown file at assets/docs/routine-name.md with the documentation

3. Rebuild the extension: npm run build

Development

# Install dependencies
npm install

# Run in development mode
npm run dev

# Build the extension
npm run build

# Lint and format code
npm run fix-lint

Credits

Based on the NumPy Documentation Search extension pattern.

Official LAPACK/BLAS documentation is provided by Netlib.