feat: change geocoding logic (#136)

* feat: changed geocoding logic

* chore: add geocoding tests

* fix: change pandas dataframe doctests after update to version 2+

* fix: change crs setting because of geopandas bug

.pre-commit-config.yaml

@@ -24,7 +24,7 @@ repos:
       - id: mypy
         additional_dependencies: ['types-requests', 'types-six']
   - repo: https://github.com/pdm-project/pdm
-    rev: 2.16.1
+    rev: 2.17.1
     hooks:
       - id: pdm-lock-check
       - id: pdm-export

CHANGELOG.md

@@ -7,6 +7,10 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## [Unreleased]
 
+### Changed
+
+- Replaced `OSMnx` dependency with `GeoPy` for geometry geocoding [#135](https://github.com/kraina-ai/quackosm/issues/135)
+
 ## [0.8.2] - 2024-06-04
 
 ### Added

README.md

@@ -105,7 +105,7 @@ Optional:
 
 - `typer[all] (>=0.9.0)` (click, colorama, rich, shellingham): Required in CLI
 
-- `osmnx (>=1.3.0)`: For geocoding of strings. Required in CLI
+- `geopy (>=1.3.0)`: For geocoding of strings. Required in CLI
 
 - `h3 (>=4.0.0b1)`: For reading H3 strings. Required in CLI
 
@@ -238,8 +238,7 @@ files/monaco_nofilter_noclip_compact.parquet
 
 ```python
 >>> import quackosm as qosm
->>> import osmnx as ox
->>> geometry = ox.geocode_to_gdf("Vatican City").unary_union
+>>> geometry = qosm.geocode_to_geometry("Vatican City")
 >>> qosm.convert_geometry_to_geodataframe(geometry)
                                               tags                      geometry
 feature_id

examples/pbf_file_reader.ipynb

@@ -31,9 +31,9 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "from quackosm import PbfFileReader\n",
     "import urllib.request\n",
-    "import osmnx as ox"
+    "\n",
+    "from quackosm import PbfFileReader, geocode_to_geometry"
    ]
   },
   {
@@ -53,8 +53,8 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "reykjavik_gdf = ox.geocode_to_gdf(\"Reykjavík, IS\")\n",
-    "reykjavik_gdf"
+    "reykjavik_geometry = geocode_to_geometry(\"Reykjavík, IS\")\n",
+    "reykjavik_geometry"
    ]
   },
   {
@@ -75,7 +75,7 @@
    "outputs": [],
    "source": [
     "reader = PbfFileReader(\n",
-    "    geometry_filter=reykjavik_gdf.geometry.iloc[0], tags_filter={\"building\": True}\n",
+    "    geometry_filter=reykjavik_geometry, tags_filter={\"building\": True}\n",
     ")\n",
     "\n",
     "reykjavik_buildings_gpq = reader.convert_pbf_to_parquet(\"iceland.osm.pbf\")\n",

pyproject.toml

@@ -50,10 +50,10 @@ Changelog = "https://github.com/kraina-ai/quackosm/blob/main/CHANGELOG.md"
 [project.optional-dependencies]
 cli = [
     "typer[all]>=0.9.0",
-    "osmnx>=1.3.0",
     "h3>=4.0.0b1",
     "s2>=0.1.9",
     "python-geohash>=0.8",
+    "geopy>=1.3.0",
 ]
 
 [project.scripts]
@@ -102,7 +102,6 @@ docs = [
     "nbconvert",
     "nbformat",
     "notebook",
-    "osmnx",
     "matplotlib",
     "seaborn>=0.10.0",
     "py-cpuinfo",

quackosm/__init__.py

@@ -11,6 +11,7 @@
     convert_pbf_to_geodataframe,
     convert_pbf_to_parquet,
 )
+from quackosm.geocode import geocode_to_geometry
 from quackosm.pbf_file_reader import PbfFileReader
 
 __app_name__ = "QuackOSM"
@@ -22,4 +23,5 @@
     "convert_geometry_to_parquet",
     "convert_pbf_to_geodataframe",
     "convert_geometry_to_geodataframe",
+    "geocode_to_geometry",
 ]

quackosm/_exceptions.py

@@ -11,3 +11,6 @@ class InvalidGeometryFilter(Exception): ...
 
 
 class MultiprocessingRuntimeError(RuntimeError): ...
+
+
+class QueryNotGeocodedError(ValueError): ...

quackosm/cli.py

@@ -9,7 +9,6 @@
 import click
 import geopandas as gpd
 import h3
-import osmnx as ox
 import typer
 from click import Argument
 from click.exceptions import MissingParameter
@@ -22,6 +21,7 @@
 from quackosm._osm_tags_filters import GroupedOsmTagsFilter, OsmTagsFilter
 from quackosm._typing import is_expected_type
 from quackosm.functions import convert_geometry_to_parquet, convert_pbf_to_parquet
+from quackosm.geocode import geocode_to_geometry
 from quackosm.osm_extracts import OsmExtractSource
 from quackosm.pbf_file_reader import _is_url_path
 
@@ -106,8 +106,7 @@ def convert(self, value, param=None, ctx=None):  # type: ignore
             return None
 
         try:
-            gdf = ox.geocode_to_gdf(query=value, which_result=None)
-            return gdf.unary_union
+            return geocode_to_geometry(value)
         except Exception:
             raise typer.BadParameter("Cannot geocode provided Nominatim query") from None
 
@@ -313,7 +312,7 @@ def main(
             help=(
                 "Geometry to use as a filter in the"
                 " [bold dark_orange]string to geocode[/bold dark_orange] format - it will be"
-                " geocoded to the geometry using Nominatim API (OSMnx library)."
+                " geocoded to the geometry using Nominatim API (GeoPy library)."
                 " Cannot be used together with"
                 " [bold bright_cyan]geom-filter-file[/bold bright_cyan] or"
                 " [bold bright_cyan]geom-filter-geojson[/bold bright_cyan] or"

quackosm/conftest.py

@@ -6,6 +6,7 @@
 from doctest import OutputChecker
 from pathlib import Path
 
+import pandas
 import pytest
 
 IGNORE_RESULT = doctest.register_optionflag("IGNORE_RESULT")
@@ -40,3 +41,10 @@ def add_pbf_files(doctest_namespace):  # type: ignore
         urllib.request.urlretrieve(pbf_file_download_url, pbf_file_path)
         doctest_namespace[f"{extract_name}_pbf_path"] = pbf_file_path
         shutil.copy(pbf_file_path, geofabrik_pbf_file_path)
+
+
+@pytest.fixture(autouse=True, scope="session")  # type: ignore
+def pandas_terminal_width() -> None:
+    """Change pandas dataframe display options."""
+    pandas.set_option("display.width", 90)
+    pandas.set_option("display.max_colwidth", 35)