Visit Apps Ved →

parquet-tools

inspects Apache Parquet files

TLDR

Show schema
$ parquet-tools schema [file.parquet]
View data
$ parquet-tools cat [file.parquet]
Show metadata
$ parquet-tools meta [file.parquet]
View first N rows
$ parquet-tools head -n [10] [file.parquet]
Show row count
$ parquet-tools rowcount [file.parquet]
Convert to JSON
$ parquet-tools cat --json [file.parquet]
Show column index info
$ parquet-tools column-index [file.parquet]
Dump specific columns
$ parquet-tools cat --columns [col1,col2] [file.parquet]

SYNOPSIS

parquet-tools command [options] file

DESCRIPTION

parquet-tools inspects Apache Parquet files. Parquet is a columnar storage format used in big data systems.Schema inspection shows column names, types, and nesting. This helps understand data structure without reading contents.Cat and head commands display actual data. JSON output integrates with other tools.Metadata shows compression, encoding, and statistics. Row groups and column chunks reveal physical layout.Parquet files from Spark, Hive, and other systems can be examined. Useful for debugging data pipelines.

PARAMETERS

cat

Print file contents.
head
Print first rows.
schema
Show schema.
meta
Show file metadata.
rowcount
Count rows.
column-index
Show column index.
-n N
Number of rows.
--json
JSON output format.
--columns COLS
Specific columns.

INSTALL

nix profile install nixpkgs#parquet-tools

CAVEATS

Large files may be slow to fully read. Some complex types display differently. Requires Java runtime.

HISTORY

Parquet format was developed by Twitter and Cloudera around 2013. parquet-tools provides command-line inspection for the widely-adopted columnar format.

SEE ALSO

parquet(1), avro-tools(1), orc-tools(1)