refactor: add find_spans

proselint/checks/__init__.py

@@ -357,21 +357,33 @@ def check_matching_quotes(chars: tuple[str, str]) -> bool:
     )
 
 
+@functools.lru_cache
 def find_quoted_ranges(text: str) -> list[tuple[int, int]]:
     """Find the range of a quote pair."""
-    active_quotes: list[tuple[str, int]] = [
-        (_m.group(0), _m.start()) for _m in re.finditer(r"[\"'“”]", text)
+    # NOTE: regex implementation is ~3x faster than previously, but this could
+    # still be optimized further.
+    return find_spans(text, r"[\"'“”]", check_matching_quotes)
+
+
+# TODO: allow this to return unmatched spans, to provide a way to check for
+# entries that are unmatched, like in misc.braces
+def find_spans(
+    text: str, regex: str, predicate: Callable[[tuple[str, str]], bool]
+) -> list[tuple[int, int]]:
+    """Find spans of matching characters (e.g. braces, quotes)."""
+    active: list[tuple[str, int]] = [
+        (_m.group(0), _m.start()) for _m in re.finditer(regex, text)
     ]
-    prev_quotes: list[tuple[str, int]] = []
+    prev: list[tuple[str, int]] = []
     spans: list[tuple[int, int]] = []
-    for quote, span_end in active_quotes:
-        for potential, span_start in reversed(prev_quotes):
-            if check_matching_quotes((quote, potential)):
-                prev_quotes.pop()
+    for char, span_end in active:
+        for potential, span_start in reversed(prev):
+            if predicate((char, potential)):
+                prev.pop()
                 spans.append((span_start, span_end))
                 break
         else:
-            prev_quotes.append((quote, span_end))
+            prev.append((char, span_end))
     return spans