Skip to main content
A table engine which provides a table-like interface to SELECT from and INSERT into files, similar to the s3 table function. Use file when working with local files, and s3 when working with buckets in object storage such as S3, GCS, or MinIO. The file function can be used in SELECT and INSERT queries to read from or write to files.

​
Syntax

For SELECT queries, path can also be an expression that returns an Array(String):

​
Arguments

When the structure argument is omitted, ClickHouse infers the schema from the format itself. Different formats produce different default column names and types. To see the schema for a specific format, use DESC with the format table function.For example:

​
Returned value

A table for reading or writing data in a file.

​
Examples for Writing to a File

​
Write to a TSV file

As a result, the data is written into the file test.tsv:

​
Partitioned write to multiple TSV files

If you specify a PARTITION BY expression when inserting data into a table function of type file, then a separate file is created for each partition. Splitting the data into separate files helps to improve performance of read operations.
As a result, the data is written into three files: test_1.tsv, test_2.tsv, and test_3.tsv.

​
Examples for Reading from a File

​
SELECT from a CSV file

First, set user_files_path in the server configuration and prepare a file test.csv:
Then, read data from test.csv into a table and select its first two rows:

​
Inserting data from a file into a table

Reading data from table.csv, located in archive1.zip or/and archive2.zip:

​
Globs in path

Paths may use globbing. Files must match the whole path pattern, not only the suffix or prefix. There is one exception that if the path refers to an existing directory and does not use globs, a * will be implicitly added to the path so all the files in the directory are selected.
  • * — Represents arbitrarily many characters except / but including the empty string.
  • ? — Represents an arbitrary single character.
  • {some_string,another_string,yet_another_one} — Substitutes any of strings 'some_string', 'another_string', 'yet_another_one'. The strings can contain the / symbol.
  • {N..M} — Represents any number >= N and <= M.
  • ** - Represents all files inside a folder recursively.
Constructions with {} are similar to the remote and hdfs table functions.

​
Examples

Example Suppose there are these files with the following relative paths:
  • some_dir/some_file_1
  • some_dir/some_file_2
  • some_dir/some_file_3
  • another_dir/some_file_1
  • another_dir/some_file_2
  • another_dir/some_file_3
Query the total number of rows in all files:
An alternative path expression which achieves the same:
Query the total number of rows in some_dir using the implicit *:
If your listing of files contains number ranges with leading zeros, use the construction with braces for each digit separately or use ?.
Example Query the total number of rows in files named file000, file001, … , file999:
Example Query the total number of rows from all files inside directory big_dir/ recursively:
Example Query the total number of rows from all files file002 inside any folder in directory big_dir/ recursively:

​
Virtual Columns

  • _path — Path to the file. Type: LowCardinality(String).
  • _file — Name of the file. Type: LowCardinality(String).
  • _size — Size of the file in bytes. Type: Nullable(UInt64). If the file size is unknown, the value is NULL.
  • _time — Last modified time of the file. Type: Nullable(DateTime). If the time is unknown, the value is NULL.

​
use_hive_partitioning setting

When setting use_hive_partitioning is set to 1, ClickHouse will detect Hive-style partitioning in the path (/name=value/) and will allow to use partition columns as virtual columns in the query. These virtual columns will have the same names as in the partitioned path. Example Use virtual column, created with Hive-style partitioning

​
Settings

Last modified on July 1, 2026