gzipi

Tools for indexing compressed files (currently supporting gzip and zstandard) to support random-like access.

Installing

To install library from the source code, run the following coomand:

$ python setup.py install

To install from pypi, run:

$ pip install gzipi

Testing

$ make test
$ make lint

Repacking existing archives

If your archive was not converted before, you need to repack it:

$ gzipi repack -f profiles.json.gz -i index.gzi -o repacked_profiles.json.gz --format json --field domain

This command produces the repacked archive and the index file.

Retrieving data

To quickly retrieve data, you need a repacked archive and the index file.

Retrieving multiple keys provided via stdin:

$ cat domains_to_retrieve.txt | gzipi retrieve -f repacked_profiles.json.gz -i index.gzi --format json --field domain

Retrieving a single key:

$ gzipi search --input-file profiles.json.gz --index-file index.gzi --key google.com

Using local and S3 paths:

$ gzipi retrieve -k domains.txt -f s3://logs/2019.json.gz -i index.json.gz --format json -o data.json --field domain

Indexing a file

If you gzip archive is already chunked, you can index it without repacking.

Indexing a file from stdin:

$ cat profiles.json.gz | gzipi index --format json --field id > index.json.gz

Indexing a local file:

$ gzipi profiles.json.bz -i profiles.json.gz -o index.json.gz --format csv --column 0 --delimiter ','

Help

To get more information, run the following command:

$ gzipi --help