openlr-decoder

Fast OpenLR location reference decoder for Python.

Decode HERE-encoded OpenLR location references onto OSM-based road networks. Built in Rust with Python bindings via PyO3 for high performance.

Features

• Fast: Rust core with parallel batch decoding via Rayon
• Zero-copy data loading: Pass Arrow tables, Polars DataFrames, or PyArrow RecordBatches directly — no serialization overhead
• Cross-provider: Tuned for decoding HERE-encoded OpenLR onto OSM road networks
• Configurable: 14 parameters for scoring weights, tolerances, and search behavior
• Batch-friendly: decode_batch() returns an Arrow RecordBatch for seamless integration with Polars, Pandas, and DuckDB

Quick Start

pip install openlr-decoder

from openlr_decoder import RoadNetwork, Decoder

network = RoadNetwork.from_parquet("network.parquet")
decoder = Decoder(network)

result = decoder.decode("CwRbWyNG9RpsCQCb/jsboAD/6/+E")
print(result.edge_ids)    # [12345, 12346, 12347]
print(result.length)       # 542.3

Next Steps

• Getting Started — installation, loading data, single and batch decoding
• OpenLR Concepts — background on the OpenLR standard and cross-provider challenges
• Preparing Network Data — parquet schema and data sourcing
• API Reference — complete reference for all classes and functions
• Configuration Guide — tuning the decoder for your use case
• Troubleshooting — common errors and debugging workflow