Full application

[tmosleyIII]

This pull request introduces the initial implementation of the nebu-sql project, which enables querying installed nebu processors as DuckDB table sources. It provides a standalone binary, integrates with DuckDB, and includes a custom nebu() table function for SQL queries. The PR also sets up CI, release automation, and a build/test framework. The most important changes are grouped below.

---

Core Functionality and Features:

• Implements the main application logic in internal/app/run.go, supporting SQL query execution, file input, JSON output, and version printing.
• Adds a custom DuckDB table function (nebu()) in internal/duck/tablefunc.go to expose nebu processor outputs as SQL tables, including schema discovery and row streaming.
• Documents project features, usage, and development workflow in a comprehensive README.md.

Build, Test, and Release Infrastructure:

• Adds a Makefile for building, testing, formatting, and running smoke tests.
• Introduces GitHub Actions workflows for CI (ci.yml), release automation (release.yml), and SBOM generation (sbom.yml). [1] [2] [3]
• Provides a GoReleaser configuration for multi-platform builds and version injection. (.goreleaser.yaml)

Testing:

• Includes integration and unit tests for the table function and application logic, ensuring correct behavior and error handling. [1] [2] [3]

Dependencies and Module Setup:

• Establishes the Go module and specifies required dependencies, primarily duckdb-go.