csv-import template
===================

Load CSV or Parquet files from a local directory into DuckDB.

Quick start
-----------

1. Copy your CSV files into the data/input/ directory inside your project:

       data/input/sales.csv
       data/input/customers.csv

2. Run the ingestion pipeline:

       tycoon data sources run files

3. The parsed rows are written to data/files_raw.duckdb under the
   raw_files schema.  Each file becomes its own table named after the
   file (without extension).

Loading Parquet files instead
------------------------------

Change the file_glob value in tycoon.yml from "*.csv" to "*.parquet":

    sources:
      files:
        type: filesystem
        schema: raw_files
        config:
          path: data/input
          file_glob: "*.parquet"

Then re-run:

    tycoon data sources run files

Notes
-----

- Subdirectory globs (e.g. "**/*.csv") are also supported but will load
  raw file metadata rather than parsed rows.  Use a flat glob ("*.csv"
  or "*.parquet") to get row-level data.
- The path value can be an S3 or GCS URL (e.g. s3://my-bucket/prefix)
  if the appropriate credentials are configured in your environment.
