Skip to content

tail

Read a tabular data file and return its last N rows. The mirror of the first-N-rows behaviour you get from read_table with a limit; same response shape.

When to use

When the interesting records are at the end of a file (the newest log lines, the latest appended rows) and you don't want to pull the whole table.

Input schema

Parameter Type Required? Default Description
path string yes (no default) Absolute or working-directory-relative path to the file
limit int no server default (1000) Number of trailing rows to return. 0 = the whole loaded window
table string no (no default) Specific table to read for multi-table sources
unlimited bool no false Lift the 5,000,000-row file-loader cap so the true end is reached

Response shape

Identical to read_table: { schema, rows, row_count, truncated, total_rows_available, cell_truncated }. The rows are the last limit rows of the loaded window.

Notes

  • For streaming formats the file loads with the 5 M-row cap, so the tail reflects the end of that window. Pass unlimited: true to tail the genuine end of a very large file.
  • For multi-table sources pass table.

Example call

{
  "name": "tail",
  "arguments": { "path": "/var/log/events.csv", "limit": 20 }
}

See also