Data 2025a (#266)

* Update .pre-commit-config.yaml

* Update .pre-commit-config.yaml

* update file_converter.py for h3>4

* run pre commit hooks

* version 6.5.8

* end json files with newlines

* Update poetry.lock

* update data 2025a

* refactor file converter json io

* install build dependency group

* bugfix cbuildwheel missing poetry

* remove build dependency group

.github/workflows/build.yml

@@ -71,9 +71,6 @@ jobs:
       - name: "Checkout repo"
         uses: actions/checkout@v4
 
-      - name: Install poetry
-        run: pip install poetry
-
       - name: "Build wheels"
         uses: pypa/[email protected]
         with:

.pre-commit-config.yaml

@@ -1,6 +1,6 @@
 repos:
     -   repo: https://github.com/pre-commit/pre-commit-hooks
-        rev: v4.6.0
+        rev: v5.0.0
         hooks:
             -   id: check-ast  # Is it valid Python?
             -   id: debug-statements # no debbuging statements used
@@ -19,15 +19,15 @@ repos:
 
 
     - repo: https://github.com/astral-sh/ruff-pre-commit
-      rev: v0.6.6
+      rev: v0.9.2
       hooks:
           # linter.
           - id: ruff
             args: [ --fix ]
           - id: ruff-format
 
     -   repo: https://github.com/asottile/blacken-docs
-        rev: 1.18.0
+        rev: 1.19.1
         hooks:
             -   id: blacken-docs
 
@@ -39,17 +39,17 @@ repos:
                 exclude: ^(docs/)
 
     -   repo: https://github.com/abravalheri/validate-pyproject
-        rev: v0.19
+        rev: v0.23
         hooks:
             -   id: validate-pyproject
 
     -   repo: https://github.com/asottile/pyupgrade
-        rev: v3.17.0
+        rev: v3.19.1
         hooks:
             -   id: pyupgrade
 
     -   repo: https://github.com/pre-commit/mirrors-mypy
-        rev: v1.11.2
+        rev: v1.14.1
         hooks:
             -   id: mypy
                 exclude: ^((tests|scripts)/)
@@ -62,6 +62,6 @@ repos:
     #        additional_dependencies: [ numpy, poetry==1.1.11 ]
 
     -   repo: https://github.com/pre-commit/mirrors-clang-format
-        rev: v18.1.8
+        rev: v19.1.7
         hooks:
             -   id: clang-format

CHANGELOG.rst

@@ -2,6 +2,14 @@
 Changelog
 =========
 
+
+6.5.8 (2025-01-21)
+------------------
+
+* updated the data to `2025a <https://github.com/evansiroky/timezone-boundary-builder/releases/tag/2025a>`__
+* internal: updated ``file_converter.py`` for ``h3>=4``
+
+
 6.5.7 (2024-12-02)
 ------------------
 

pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "timezonefinder"
-version = "6.5.7"
+version = "6.5.8"
 description = "python package for finding the timezone of any point on earth (coordinates) offline"
 authors = ["jannikmi <[email protected]>"]
 license = "MIT"

scripts/file_converter.py

@@ -62,9 +62,7 @@
 
 import functools
 import itertools
-import json
 from dataclasses import dataclass
-from os.path import abspath, join
 from pathlib import Path
 from typing import Dict, List, NamedTuple, Optional, Set, Tuple, Union
 
@@ -90,6 +88,7 @@
     write_boundary_data,
     write_coordinate_data,
     write_json,
+    load_json,
 )
 from timezonefinder.configs import (
     DTYPE_FORMAT_H,
@@ -147,8 +146,8 @@ def parse_polygons_from_json(input_path: Path) -> int:
     global polygons, polygon_lengths, poly_zone_ids, poly_boundaries
 
     print(f"parsing input file: {input_path}\n...\n")
-    with open(input_path) as json_file:
-        tz_list = json.loads(json_file.read()).get("features")
+    input_json = load_json(input_path)
+    tz_list = input_json["features"]
 
     poly_id = 0
     zone_id = 0
@@ -184,7 +183,7 @@ def parse_polygons_from_json(input_path: Path) -> int:
             for hole_nr, hole in enumerate(poly_with_hole):
                 nr_of_holes += 1  # keep track of how many holes there are
                 print(
-                    f"\rpolygon {poly_id}, zone {tz_name}, hole number {nr_of_holes}, {hole_nr+1} in polygon",
+                    f"\rpolygon {poly_id}, zone {tz_name}, hole number {nr_of_holes}, {hole_nr + 1} in polygon",
                     end="",
                 )
                 polynrs_of_holes.append(poly_id)
@@ -205,24 +204,24 @@ def parse_polygons_from_json(input_path: Path) -> int:
     assert nr_of_polygons >= 0
     assert nr_of_polygons >= nr_of_zones
     assert zone_id == nr_of_zones - 1
-    assert (
-        poly_id == nr_of_polygons
-    ), f"polygon counter {poly_id} and entry amount in all_length {nr_of_polygons} are different."
+    assert poly_id == nr_of_polygons, (
+        f"polygon counter {poly_id} and entry amount in all_length {nr_of_polygons} are different."
+    )
 
     if 0 in polygon_lengths:
         raise ValueError()
 
     # binary file value range tests:
-    assert (
-        nr_of_polygons < THRES_DTYPE_H
-    ), f"address overflow: #{nr_of_polygons} polygon ids cannot be encoded as {DTYPE_FORMAT_H}!"
-    assert (
-        nr_of_zones < THRES_DTYPE_H
-    ), f"address overflow: #{nr_of_zones} zone ids cannot be encoded as {DTYPE_FORMAT_H}!"
+    assert nr_of_polygons < THRES_DTYPE_H, (
+        f"address overflow: #{nr_of_polygons} polygon ids cannot be encoded as {DTYPE_FORMAT_H}!"
+    )
+    assert nr_of_zones < THRES_DTYPE_H, (
+        f"address overflow: #{nr_of_zones} zone ids cannot be encoded as {DTYPE_FORMAT_H}!"
+    )
     max_poly_length = max(polygon_lengths)
-    assert (
-        max_poly_length < THRES_DTYPE_I
-    ), f"address overflow: the maximal amount of coords {max_poly_length} cannot be represented by {DTYPE_FORMAT_I}"
+    assert max_poly_length < THRES_DTYPE_I, (
+        f"address overflow: the maximal amount of coords {max_poly_length} cannot be represented by {DTYPE_FORMAT_I}"
+    )
     max_hole_poly_length = max(all_hole_lengths)
     assert max_hole_poly_length < THRES_DTYPE_H, (
         f"address overflow: the maximal amount of coords in hole polygons "
@@ -242,7 +241,7 @@ def parse_polygons_from_json(input_path: Path) -> int:
     return polygon_space
 
 
-def update_zone_names(output_path):
+def update_zone_names(output_path: Path):
     # update all the zone names and set the right ids to be written in the poly_zone_ids.bin
     global poly_zone_ids
     global list_of_pointers
@@ -252,7 +251,7 @@ def update_zone_names(output_path):
     global polynrs_of_holes
     global nr_of_zones
     global nr_of_polygons
-    file_path = abspath(join(output_path, TIMEZONE_NAMES_FILE))
+    file_path = output_path / TIMEZONE_NAMES_FILE
     print(f"updating the zone names in {file_path} now.")
     # pickle the zone names (python array)
     write_json(all_tz_names, file_path)
@@ -483,18 +482,18 @@ def zones_in_cell(self) -> Set[int]:
 
     @property
     def children(self) -> Set[int]:
-        return set(h3.h3_to_children(self.id))
+        return set(h3.cell_to_children(self.id))
 
     @property
     def outer_children(self) -> Set[int]:
         child_set = self.children
-        center_child = h3.h3_to_center_child(self.id)
+        center_child = h3.cell_to_center_child(self.id)
         child_set.remove(center_child)
         return child_set
 
     @property
     def neighbours(self) -> HexIdSet:
-        return set(h3.k_ring(self.id, k=1))
+        return set(h3.grid_ring(self.id, k=1))
 
     @property
     def true_parents(self) -> HexIdSet:
@@ -509,7 +508,7 @@ def true_parents(self) -> HexIdSet:
             raise ValueError("not defined for resolution 0")
         lower_res = self.res - 1
         # NOTE: (lat,lng) pairs!
-        coord_pairs = h3.h3_to_geo_boundary(self.id)
+        coord_pairs = h3.cell_to_boundary(self.id)
         return {h3.latlng_to_cell(pt[0], pt[1], lower_res) for pt in coord_pairs}
 
 
@@ -582,9 +581,9 @@ def report_progress():
 def all_res_candidates(res: int) -> HexIdSet:
     print(f"compiling hex candidates for resolution {res}.")
     if res == 0:
-        return set(h3.get_res0_indexes())
+        return set(h3.get_res0_cells())
     parent_res_candidates = all_res_candidates(res - 1)
-    child_iter = (h3.h3_to_children(h) for h in parent_res_candidates)
+    child_iter = (h3.cell_to_children(h) for h in parent_res_candidates)
     return set(itertools.chain.from_iterable(child_iter))
 
 
@@ -734,8 +733,8 @@ def compile_addresses(
                 }
             )
 
-    with open(join(output_path, HOLE_REGISTRY_FILE), "w") as json_file:
-        json.dump(hole_registry, json_file, indent=4)
+    path = output_path / HOLE_REGISTRY_FILE
+    write_json(hole_registry, path)
 
     # '<H'  Y times [H unsigned short: nr of values (coordinate PAIRS! x,y in int32 int32) in this hole]
     assert len(all_hole_lengths) == nr_of_holes
@@ -775,6 +774,10 @@ def parse_data(
     input_path: Union[Path, str] = DEFAULT_INPUT_PATH,
     output_path: Union[Path, str] = DEFAULT_OUTPUT_PATH,
 ):
+    input_path = Path(input_path)
+    output_path = Path(output_path)
+    output_path.mkdir(parents=True, exist_ok=True)
+
     polygon_space = parse_polygons_from_json(input_path)
     update_zone_names(output_path)
     hole_space = compile_polygon_binaries(output_path)

scripts/utils.py

@@ -35,6 +35,8 @@ def write_json(obj, path):
     print("writing json to ", path)
     with open(abspath(path), "w") as json_file:
         json.dump(obj, json_file, indent=2)
+        # write a newline at the end of the file
+        json_file.write("\n")
 
 
 def time_execution(func):
@@ -133,12 +135,12 @@ def export_mapping(file_name: str, obj: Dict, res: int):
 
 
 def write_value(output_file, value, data_format, lower_value_limit, upper_value_limit):
-    assert (
-        value > lower_value_limit
-    ), f"trying to write value {value} subceeding lower limit {lower_value_limit} (data type {data_format})"
-    assert (
-        value < upper_value_limit
-    ), f"trying to write value {value} exceeding upper limit {upper_value_limit} (data type {data_format})"
+    assert value > lower_value_limit, (
+        f"trying to write value {value} subceeding lower limit {lower_value_limit} (data type {data_format})"
+    )
+    assert value < upper_value_limit, (
+        f"trying to write value {value} exceeding upper limit {upper_value_limit} (data type {data_format})"
+    )
     output_file.write(struct.pack(data_format, value))
 
 

tests/main_test.py

@@ -45,26 +45,26 @@ def ocean2land(test_locations):
 
 def check_geometry(geometry_obj: List):
     coords = geometry_obj[0][0]
-    assert (
-        len(coords) == 2
-    ), "the polygon does not consist of two latitude longitude lists"
+    assert len(coords) == 2, (
+        "the polygon does not consist of two latitude longitude lists"
+    )
     x_coords, y_coords = coords
     nr_x_coords = len(x_coords)
     nr_y_coords = len(y_coords)
     assert nr_x_coords > 2, "a polygon must consist of more than 2 coordinates"
-    assert (
-        nr_x_coords == nr_y_coords
-    ), "the amount of x and y coordinates (lng, lat) must be equal"
+    assert nr_x_coords == nr_y_coords, (
+        "the amount of x and y coordinates (lng, lat) must be equal"
+    )
 
 
 def check_pairwise_geometry(geometry_obj: List):
     # list of all coord pairs of the first polygon
     cord_pairs = geometry_obj[0][0]
     assert len(cord_pairs) > 2, "a polygon must consist of more than 2 coordinates"
     first_coord_pair = cord_pairs[0]
-    assert (
-        len(first_coord_pair) == 2
-    ), "the polygon does not consist of coordinate pairs as expected."
+    assert len(first_coord_pair) == 2, (
+        "the polygon does not consist of coordinate pairs as expected."
+    )
 
 
 def is_valid_lng_int(x: int) -> bool:
@@ -192,9 +192,9 @@ def test_timezone_name_attribute(self):
         timezone_names_stored = self.test_instance.timezone_names
         with open(join(abs_default_path, TIMEZONE_NAMES_FILE)) as json_file:
             timezone_names_json = json.loads(json_file.read())
-        assert (
-            timezone_names_stored == timezone_names_json
-        ), f"the content of the {TIMEZONE_NAMES_FILE} and the attribute {timezone_names_stored} are different."
+        assert timezone_names_stored == timezone_names_json, (
+            f"the content of the {TIMEZONE_NAMES_FILE} and the attribute {timezone_names_stored} are different."
+        )
 
 
 class BaseClassTestMEM(BaseTimezoneFinderClassTest):
@@ -285,9 +285,9 @@ def test_get_geometry(self):
             )
             # not necessary:
             # assert nested_list_equal(geometry_from_id, geometry_from_name), \
-            assert (
-                len(geometry_from_name) == len(geometry_from_id)
-            ), "the results for querying the geometry for a zone with zone name or zone id are NOT equal."
+            assert len(geometry_from_name) == len(geometry_from_id), (
+                "the results for querying the geometry for a zone with zone name or zone id are NOT equal."
+            )
             check_geometry(geometry_from_id)
 
             geometry_from_name = self.test_instance.get_geometry(
@@ -296,9 +296,9 @@ def test_get_geometry(self):
             geometry_from_id = self.test_instance.get_geometry(
                 tz_name=zone_name, tz_id=zone_id, use_id=False, coords_as_pairs=True
             )
-            assert (
-                len(geometry_from_name) == len(geometry_from_id)
-            ), "the results for querying the geometry for a zone with zone name or zone id are NOT equal."
+            assert len(geometry_from_name) == len(geometry_from_id), (
+                "the results for querying the geometry for a zone with zone name or zone id are NOT equal."
+            )
 
             check_pairwise_geometry(geometry_from_id)
             check_pairwise_geometry(geometry_from_name)

tests/shortcut_test.py

@@ -38,9 +38,9 @@ def test_resolutions():
     shortcut_hex_ids = shortcuts.keys()
     resolutions = [h3.get_resolution(h) for h in shortcut_hex_ids]
     res_matched = [res == configs.SHORTCUT_H3_RES for res in resolutions]
-    assert all(
-        res_matched
-    ), f"not all shortcut resolutions match the expected resolution {configs.SHORTCUT_H3_RES}"
+    assert all(res_matched), (
+        f"not all shortcut resolutions match the expected resolution {configs.SHORTCUT_H3_RES}"
+    )
 
 
 def test_empty_shortcut():
@@ -60,12 +60,12 @@ def test_unique_pole_cells():
         if hex.surr_n_pole:
             n_pole_ctr += 1
 
-    assert (
-        s_pole_ctr == 1
-    ), f"{s_pole_ctr} cells are considered to surround the south pole"
-    assert (
-        n_pole_ctr == 1
-    ), f"{n_pole_ctr} cells are considered to surround the north pole"
+    assert s_pole_ctr == 1, (
+        f"{s_pole_ctr} cells are considered to surround the south pole"
+    )
+    assert n_pole_ctr == 1, (
+        f"{n_pole_ctr} cells are considered to surround the north pole"
+    )
 
 
 def has_coherent_sequences(lst: List[int]) -> bool: