README

Try it out

$ uvx --from git+https://github.com/GIScience/ohsome-filter-to-sql.git ohsome-filter-to-sql
natural = tree and leaftype = broadleaf  # type in ohsome filter and hit enter
tags @> '{"natural": "tree"}' AND tags @> '{"leaftype": "broadleaf"}'  # result

Installation

uv add git+https://github.com/GIScience/ohsome-filter-to-sql.git

Usage as Python library

from ohsome_filter_to_sql.main import ohsome_filter_to_sql

sql_query = ohsome_filter_to_sql("natural = tree")

Development Setup

uv run pytest

To install pre-commit hooks run:

uv run pre-commit install

To develop new features you will need a local instance of the ohsomeDB.

How to play around with the grammar?

Execute antlr4-parse, type in an ohsome filter and press ctlr+d.

antlr4-parse OFL.g4 root -tree
buildings=yes
(root:1 (expression:8 (tagMatch:1 (string:1 buildings) = (string:1 yes))) <EOF>)

ANTLR Lab can also be used to try out the grammar.

How to generating parser code?

When the grammar file has change generate new Python code with antlr4 and move genrated files to ohsome_filter_to_sql/.

uv run antlr4 -Dlanguage=Python3 OFL.g4 && mv *.py ohsome_filter_to_sql/

Release

This project uses CalVer.

Format is: YYYY.MM.DD (E.g 2025.9.2).

In case of releasing twice on one day add a micro number starting with 1: YYYY.MM.DD.micro (E.g. 2025.9.2.1).

To make a new release run ./scripts/release.sh YYYY.MM.DD

Resources

• ohsome filter documentation and oshdb-filter
• ANTLR with Python - Introduction
• ANTLR Listeners
• ohsomeDB schema