This module provides an IndexedZstdFile class, which can be used to seek inside zstd files without having to decompress them first.
It's shamelessy based on indexed_bzip2, which was refactored to support zstd instead of bzip2 using libzstd-seek.
Kudos to the author for its work.
Seeking inside a block is only emulated, so IndexedZstdFile will only speed up seeking when there are more than one block.
To create multi-block archives you may use t2sz.
from indexed_zstd import IndexedZstdFile
file = IndexedZstdFile( "example.zst" )
# You can now use it like a normal file
file.seek( 123 )
data = file.read( 100 )