How to Decode Pix2PGP Data with the Parser

The software/scripts/axiDataParser.py script decodes AXI-Stream data generated by the VHDL testbenches. It expects a text file containing the AXI-Stream in hexadecimal format — one byte per line — which is the format produced by the VCS testbenches described in How to Simulate Using VCS.

Usage

$ cd pix2pgp/software/scripts
$ python axiDataParser.py \
     --asicType SparkPixS \
     --verbose \
     --input <path/to/testbench_dump.txt>

  • --asicType — selects the ASIC-specific decoder class (must match one of the entries in asicTypeDict: SparkPixS, SparkPixT, Thriglav, SparkPixSv2, or the user’s own custom ASIC)

  • --verbose — enables per-event printouts

  • --rawData — decode raw data words without unpacking them into ADC / row / ToA / ToT (useful when developing a new ASIC integration)

Under the Hood

The parser is a thin wrapper around the AsicData class exported by the pix2pgp Python package. The class hierarchy is:

  • AsicParameters — per-ASIC constants (numOfLanes, numOfCols, wordLen)

  • Pix2PgpHeaderFormat — decodes the ASIC lane frame header

  • Pix2PgpColMetadataFormat — decodes the per-column metadata word

  • SparseDataFormat — decodes individual hit words (ADC / row / ToA / ToT depending on the ASIC)

  • Pix2PgpFpgaRxDataFormat — decodes the consolidated FPGA-side frame (preamble, header, per-lane lengths, trailer)

  • LaneData / AsicData — event containers

All classes are defined under firmware/python/pix2pgp/.