This demo does dynamic pattern matching in Snappy-compressed string data stored within an Arrow table. A Scala program is available to generate the data for this demo from Wikipedia database dumps.
The tools in this repo expect the following repositories to be checked out in the parent directory, next to this one:
The following Python 3 packages need to be installed:
data
: contains the Scala sources to do data preparation.hardware
: contains the VHDL code for the decompress + matching kernel. Some of this is pre-generated by Fletcher based on the*.as
schema definitions and byvhdmmio
based on the YAML files.alveo
: contains build scripts for Alveo. The scripts are based on the vadd example from Xilinx.