{
  "$schema": "https://json-schema.org/draft/2020-12/schema",
  "$id": "https://datenoio.github.io/undatum/schemas/undatum.sniff.v1.json",
  "title": "Detected file properties",
  "description": "Printed by `undatum sniff --json`.",
  "type": "object",
  "properties": {
    "schema": {
      "const": "undatum.sniff/1",
      "description": "Result layout"
    },
    "file": {
      "type": "string",
      "description": "Input path"
    },
    "filetype": {
      "type": "string",
      "description": "Detected format id (csv, jsonl, parquet, ...)"
    },
    "compression": {
      "type": [
        "string",
        "null"
      ],
      "description": "Compression codec (gz, zst, ...), null if none"
    },
    "encoding": {
      "type": [
        "string",
        "null"
      ],
      "description": "Text encoding, null for binary formats"
    },
    "delimiter": {
      "type": [
        "string",
        "null"
      ],
      "description": "Field delimiter of delimited text, else null"
    },
    "has_header": {
      "type": [
        "boolean",
        "null"
      ],
      "description": "Whether the first line is a header (CSV/TSV)"
    },
    "record_count": {
      "type": "integer",
      "description": "Number of records"
    },
    "sample_size": {
      "type": "integer",
      "description": "Records sampled for field types"
    },
    "fields": {
      "type": "object",
      "description": "Field name -> type and up to 3 examples"
    }
  },
  "required": [
    "schema",
    "file",
    "filetype",
    "record_count",
    "fields"
  ]
}