All notable changes to json2sql will be documented in this file.
- CLI integration tests with comprehensive coverage (#8, #9, #10)
- Standalone dialect tests and converter edge cases (#10)
- npm wrapper (
package.json+cli.js) for npm publishing - GitHub Actions: PyPI publish workflow (release or manual dispatch)
- GitHub Actions: GitHub Pages deployment workflow
- GitHub Actions: npm publish workflow (release or manual dispatch)
CONTRIBUTING.mdwith development setup and PR guidelinesSECURITY.mdwith security policy- Homebrew and Scoop install methods with bucket setup docs
- Directory listing badges: Open Source Alternative, LibHunt, Awesome Python
- Awesome DevOps submission badge (#433)
[project.urls]metadata inpyproject.toml(Homepage, Source, Issues, Changelog)- GitHub-based install fallback documentation
- Schema-only mode (
--schema-onlyflag for CREATE TABLE only)
- UTF-8 encoding (mojibake) in file output
- Ruff lint warnings: E501, B904, F821, X|None syntax,
datetime.UTCdeprecation - README install section formatting (broken Homebrew/Scoop code blocks)
- Broken PyPI badges replaced with GitHub release badge
- PyPI package name conflict — renamed to
json2sql-cli revenueholdings_licenseimport made optional (fixes CI failures when not installed)- Missing
ruffdev dependency inpyproject.toml - CONTRIBUTING.md typos in lint commands (
uff→ruff) npm-publish.ymlremoved (wrong-language workflow for Python repo)
- CI test matrix expanded to include Python 3.13
- CI security hardened:
persist-credentials: falseon checkout steps - CI permissions restricted to
contents: read - npm keywords optimized for discoverability (15 terms), npm version bumped to 0.1.1
- README rewritten with unified pricing tables, Revenue Holdings branding
- README expanded with CI/CD integration examples and alternatives comparison
- README tool count and DevForge URLs updated
- npm install instructions added to README
- Initial release
- JSON to SQL INSERT conversion
- Nested JSON auto-flattening into relational tables
- Multi-dialect output: PostgreSQL, MySQL, SQLite
- Array-of-objects handling as multiple INSERT rows
- Type inference for strings, numbers, booleans, nulls
- stdin pipe support
- Table name specification
- Output file support
- Python 3.10+ support, zero dependencies beyond typer