ty: un-ignore invalid-return-type rule

computer_vision/mosaic_augmentation.py

@@ -91,7 +91,7 @@ def update_image_and_anno(
     output_size: tuple[int, int],
     scale_range: tuple[float, float],
     filter_scale: float = 0.0,
-) -> tuple[list, list, str]:
+) -> tuple[np.ndarray, list, str]:
     """
     - all_img_list <type: list>: list of all images
     - all_annos <type: list>: list of all annotations of specific image
@@ -111,7 +111,7 @@ def update_image_and_anno(
     divid_point_y = int(scale_y * output_size[0])
 
     new_anno = []
-    path_list = []
+    path_list: list[str] = []
     for i, index in enumerate(idxs):
         path = all_img_list[index]
         path_list.append(path)

data_structures/arrays/dynamic_array.py

@@ -2,7 +2,8 @@ class DynamicArray:
     def __init__(self) -> None:
         self.size = 0  # Number of elements in the array
         self.capacity = 1  # Initial capacity of the array
-        self.array = [None] * self.capacity  # Create an array with initial capacity
+        # Create an array with initial capacity
+        self.array: list[int | None] = [None] * self.capacity
 
     def append(self, item: int) -> None:
         """
@@ -77,7 +78,9 @@ def get(self, index: int) -> int:
         """
         if index < 0 or index >= self.size:
             raise IndexError("index out of range")
-        return self.array[index]
+        value = self.array[index]
+        assert value is not None  # every filled slot holds an int
+        return value
 
     def __setitem__(self, index: int, value: int) -> None:
         if index < 0 or index >= self.size:

graphs/bi_directional_dijkstra.py

@@ -46,7 +46,7 @@ def pass_and_relaxation(
 
 def bidirectional_dij(
     source: str, destination: str, graph_forward: dict, graph_backward: dict
-) -> int:
+) -> float:
     """
     Bi-directional Dijkstra's algorithm.
 

graphs/hopcroft_karp.py

@@ -49,6 +49,7 @@
 
 import math
 from collections import deque
+from typing import cast
 
 _NIL = object()
 
@@ -228,11 +229,14 @@ def maximum_matching(self) -> dict[T, T]:
                 if self.pair_left[left_vertex] is _NIL:
                     self.depth_first_search(left_vertex)
 
-        return {
-            left_vertex: matched_right  # type: ignore[misc]
-            for left_vertex, matched_right in self.pair_left.items()
-            if matched_right is not _NIL
-        }
+        return cast(
+            "dict[T, T]",
+            {
+                left_vertex: matched_right
+                for left_vertex, matched_right in self.pair_left.items()
+                if matched_right is not _NIL
+            },
+        )
 
 
 def hopcroft_karp[T](graph: dict[T, list[T]]) -> dict[T, T]:

machine_learning/mfcc.py

@@ -58,6 +58,7 @@
 """
 
 import logging
+from typing import overload
 
 import numpy as np
 import scipy.fftpack as fft
@@ -289,7 +290,11 @@ def freq_to_mel(freq: float) -> float:
     return 2595.0 * np.log10(1.0 + freq / 700.0)
 
 
-def mel_to_freq(mels: float) -> float:
+@overload
+def mel_to_freq(mels: float) -> float: ...
+@overload
+def mel_to_freq(mels: np.ndarray) -> np.ndarray: ...
+def mel_to_freq(mels: float | np.ndarray) -> float | np.ndarray:
     """
     Convert a frequency in the mel scale to Hertz.
 

networking_flow/ford_fulkerson.py

@@ -58,7 +58,7 @@ def breadth_first_search(graph: list, source: int, sink: int, parents: list) ->
     return visited[sink]
 
 
-def ford_fulkerson(graph: list, source: int, sink: int) -> int:
+def ford_fulkerson(graph: list, source: int, sink: int) -> int | float:
     """
     This function returns the maximum flow from source to sink in the given graph.
 

other/dancing_links.py

@@ -65,8 +65,8 @@
                 # Insert node into column
                 node.down = col
                 node.up = col.up
                 col.up.down = node
                 col.up = node
                 col.size += 1
 
                 # Link nodes in the same row
@@ -106,7 +106,7 @@
         col.right.left = col
         col.left.right = col
 
-    def _choose_column(self) -> ColumnNode:
+    def _choose_column(self) -> ColumnNode | None:
         """Select the column with the smallest size (heuristic)."""
         min_size = float("inf")
         chosen = None

pyproject.toml

@@ -318,7 +318,6 @@ skip = """\
 environment.python-version = "3.15"
 rules.call-non-callable = "ignore"
 rules.invalid-argument-type = "ignore"
-rules.invalid-return-type = "ignore"
 rules.no-matching-overload = "ignore"
 rules.not-iterable = "ignore"
 rules.not-subscriptable = "ignore"

web_programming/random_anime_character.py

@@ -34,6 +34,7 @@ def random_anime_character() -> tuple[str, str, str]:
         httpx2.get(URL, headers=headers, timeout=10).text, "html.parser"
     )
     title = soup.find("meta", attrs={"property": "og:title"}).attrs["content"]
+    assert isinstance(title, str)
     image_url = soup.find("meta", attrs={"property": "og:image"}).attrs["content"]
     description = soup.find("p", id="description").get_text()
     _, image_extension = os.path.splitext(os.path.basename(image_url))