Skip to main content

Format conversion

Convert any readable iterabledata format to any writable one. There is no fixed pairwise matrix — inspect your install with undatum formats list --writable.

CSV to Parquet​

undatum convert people.csv people.parquet
undatum stats people.parquet

For multi-GB inputs:

undatum convert huge.jsonl.zst huge.parquet --low-memory

Filter and compute while converting​

undatum convert data.csv berlin.parquet --where "city = 'Berlin'" --add "total = price * quantity"

Partitioned output​

undatum convert data.csv by_country --partition-by country -O parquet

Convert a stream​

cat data.csv | undatum convert - data.jsonl

XML to JSON Lines​

undatum convert --tagname item data.xml data.jsonl

Excel sheet to Parquet​

undatum formats tables workbook.xlsx
undatum convert workbook.xlsx sales.parquet --table Sheet2

Bulk directory convert​

undatum convert ./raw ./processed --recursive --to-ext parquet
undatum convert ./raw ./out --recursive --to-ext jsonl --filename-pattern "{stem}.converted.jsonl"

Cloud to cloud​

undatum convert s3://bucket/input.jsonl gs://other/output.parquet

See convert, repack, and the format matrix.