v0.1.0 – SQL → NoSQL analysis & migration scripts

Highlights

• 🔍 PostgreSQL schema analysis

  • Introspects tables, columns, primary keys, uniques, and foreign keys via information_schema.
  • Builds a typed internal SqlSchema model.

• 📦 Deterministic NoSQL schema proposals

  • Generates NoSqlSchema with collections, fields, and references.
  • Writes per-table JSON files under output/analyze/ and LLM-optimized versions under output/recommend/.

• 🧠 Optional LLM recommendations (OpenAI)

  • @s2n/llm package with OpenAI provider.
  • Suggests embedding/denormalization strategies and insights.
  • Integrated into CLI via --llm, --llm-api-key, --llm-model.

• 🌐 HTML reports

  • output/view/index.html – main landing page.
  • Per-table HTML pages showing SQL vs NoSQL side-by-side.
  • Includes LLM recommendations when enabled.

• 🧩 Migration script generation

  • For each table, generates output/scripts/<table>.migrate.js with a build<Table>Doc(row, related) mapper.
  • Generates run-all.migrate.js to list all per-table scripts in dependency order.
  • Scripts are designed to be wired into your own Postgres → MongoDB runner.

---

Quickstart

# Clone the repo
git clone https://github.com/data-migration/sql2nosql.git
cd sql2nosql

# Install dependencies
yarn install

# Build core + CLI
yarn build

# Create config from template and fill in values
cp sql2nosql.config.example.json sql2nosql.config.json
# Edit sql2nosql.config.json with your Postgres + MongoDB + (optional) OpenAI details

# Run analysis (with optional LLM)
yarn analyze --llm --llm-model gpt-4.1-mini

Next steps

• Open output/view/index.html in your browser to explore the analysis.
• Inspect output/scripts/*.migrate.js to review or customize per-table migration mappers.

---