# ┌──────────────────────────────────────────────────────────────────────┐
# │  Market-Data Observatory — Data Source Registry                     │
# │                                                                     │
# │  To add a new data source, simply append a new entry to the list.  │
# │  The init script reads this file at boot and handles everything:   │
# │    • git clone / pull                                               │
# │    • ClickHouse VIEW creation (CREATE OR REPLACE)                  │
# │    • background auto-refresh                                        │
# │                                                                     │
# │  On-demand refresh: curl https://<space>/api/refresh               │
# │  No code changes needed — just edit this YAML file.                │
# └──────────────────────────────────────────────────────────────────────┘

# Global settings
refresh_interval_seconds: 300   # background git-pull interval (safety net)

# Data sources — each entry = git clone + ClickHouse view(s)
sources:

  - name: hf_ohlc_data
    description: "HuggingFace OHLCV Dataset: altool/ohlcdata"
    repo_url: "https://huggingface.co/datasets/altool/ohlcdata"
    branch: "main"
    auth_env_var: "HF_TOKEN" # Allows pulling from private HF dataset natively
    clone_depth: 1
    local_dir: "hf_ohlc_data"                   # → /app/ch/user_files/hf_ohlc_data/
    views:
      - view_name: hf_ohlc
        file_glob: "**/*.parquet"
        format: Parquet
        description: "Full OHLCV data from HF dataset, selects all 150+ columns automatically"

  # ── EXAMPLE: Adding a second data source ──────────────────────────────
  # Just uncomment and edit:
  #
  # - name: options_chain
  #   description: "NSE options chain snapshots"
  #   repo_url: "https://github.com/youruser/options_chain_data.git"
  #   branch: "main"
  #   clone_depth: 1
  #   local_dir: "options_chain"
  #   views:
  #     - view_name: options
  #       file_glob: "**/*.parquet"
  #       format: Parquet
  #       description: "Options chain snapshots"
  #       columns:                           # omit 'columns' to use SELECT *
  #         - strike
  #         - expiry
  #         - option_type
  #         - open_interest
  #         - volume
  #         - ltp
  #         - iv
  #
  # ── EXAMPLE: Multiple views from the same repo ───────────────────────
  #
  # - name: market_indices
  #   description: "Broad market index data"
  #   repo_url: "https://github.com/youruser/index_data.git"
  #   branch: "main"
  #   clone_depth: 1
  #   local_dir: "indices"
  #   views:
  #     - view_name: nifty50
  #       file_glob: "nifty50/**/*.parquet"
  #       format: Parquet
  #       description: "Nifty 50 index"
  #     - view_name: banknifty
  #       file_glob: "banknifty/**/*.parquet"
  #       format: Parquet
  #       description: "Bank Nifty index"