chore: Replace python parquet generation script with ts

[ZhongpinWang]

Update:

Since it is at the moment not possible to test if the generated parquet works with the context registry service, I would pause this PR for a moment.

---

As in the title.

Also added a flag to the generation script to leave out items with [PREDICT] placeholder. This is a preparation step for the upcoming context registry feature support.

[hyperspace-insights]

Summary

The following content is AI-generated and provides a summary of the pull request:

---

Replace Python Parquet Generation Script with TypeScript

Chore

♻️ Replaces the existing Python script for generating Parquet files with TypeScript equivalents, aligning the tooling with the rest of the project's TypeScript-based stack.

Changes

• .gitignore: Added generated directory to the ignore list to exclude locally generated Parquet files from version control.
• package.json: Added @dsnp/parquetjs as a dev dependency to enable Parquet file generation in TypeScript.
• pnpm-lock.yaml: Updated lock file with @dsnp/parquetjs and its transitive dependencies (including AWS SDK, Smithy, Thrift, and other supporting libraries).
• scripts/generate-parquet.ts: New TypeScript script that generates a sample payments.parquet file with customer payment data, saved to a local generated/ directory.
• sample-code/resources/generate-parquet.ts: New TypeScript script that generates product data Parquet files for use with SAP RPT, replacing the removed Python script. Supports a --no-predict flag to optionally exclude rows with [PREDICT] placeholders.
• sample-code/resources/generate_parquet.py: Removed — replaced by the TypeScript script above.
• sample-code/resources/product_data_with_placeholders.parquet: New Parquet file containing product data rows including [PREDICT] placeholder entries, generated by the new script.
• sample-code/resources/product_data.parquet: Updated Parquet file (regenerated via the new TypeScript script).
• sample-code/src/rpt.ts: Updated the Parquet file reference from product_data.parquet to product_data_with_placeholders.parquet to use the new file that includes prediction placeholder rows.

---

• 🔄 Regenerate and Update Summary
• ✏️ Insert as PR Description (deletes this comment)
• 🗑️ Delete comment

PR Bot Information

Version: 1.20.51

• Event Trigger: pull_request.opened
• LLM: anthropic--claude-4.6-sonnet
• File Content Strategy: Full file content
• Output Template: Default Template
• Summary Prompt: Default Prompt
• Correlation ID: 28e0c725-1921-43d4-a694-e43ff32494ac

[davidkna-sap]

[pp] I would prefer something lighter (dependency-tree wise), e.g. hyparquet-writer, parquet-wasm or @duckdb/duckdb-wasm.

[ZhongpinWang]

I picked this one only because it is quite popular. Also different parquet library has different implementation of the protocol. For now I can't really test if the generated parquet works with our services (as they might expect parquet file exported from HANA). We can hold this PR for a moment and see if the package in the end works.

[davidkna-sap]

[nth] Consider enabling compression. This file is quite a bit larger despite similar contents.