Skip to main content
Skip to main content
Edit this page

system.formats

Description​

Contains a list of all the formats along with flags whether a format is suitable for input/output or whether it supports parallelization.

Columns​

  • name (String) — Format name.
  • is_input (UInt8) — Flag that indicates whether the format is suitable for data input.
  • is_output (UInt8) — Flag that indicates whether the format is suitable for data output.
  • supports_parallel_parsing (UInt8) — Flag that indicates whether the format supports parallel parsing.
  • supports_parallel_formatting (UInt8) — Flag that indicates whether the format supports parallel formatting.
  • is_tty_friendly (UInt8) — Flag that indicates whether the format usually displays fine in the terminal. For other formats, CLI will ask before output.
  • content_type (String) — HTTP Content-Type corresponding to the output format. May depend on the current format settings.
  • supports_random_access (UInt8) — Flag that indicates whether the format supports random access in the input.
  • has_schema_inference (UInt8) — The format can dynamically determine the schema from the data (either from embedded header/metadata or from the piece of data).
  • has_external_schema (UInt8) — The format either has a fixed schema or accepts a predefined schema in its own format.
  • prefers_large_blocks (UInt8) — The format will write larger blocks into output and generate larger blocks on input.
  • supports_append (UInt8) — It's possible to append into a single file with this format.
  • supports_subsets_of_columns (UInt8) — The input format can recognize when certain columns are omitted.