Faster to split by literal string

py/aoc2024/day3.py

@@ -11,24 +11,26 @@
 xmul(2,4)&mul[3,7]!^don't()_mul(5,5)+mul(32,64](mul(11,8)undo()?mul(8,5))
 """
 
-_pattern1 = re.compile(r"mul\((\d+),(\d+)\)")
-_pattern2 = re.compile(r"(?:\A|do\(\))(.*?)(?:\Z|don't\(\))", re.S)
+_pattern = re.compile(r"mul\((\d+),(\d+)\)")
 
 
 def part1(data: str) -> int:
     """
     >>> part1(SAMPLE_INPUT_1)
     161
     """
-    return sum(int(m.group(1)) * int(m.group(2)) for m in _pattern1.finditer(data))
+    return sum(int(m.group(1)) * int(m.group(2)) for m in _pattern.finditer(data))
 
 
 def part2(data: str) -> int:
     """
     >>> part2(SAMPLE_INPUT_2)
     48
     """
-    return sum(part1(m.group(1)) for m in _pattern2.finditer(data))
+    return sum(
+        part1(subdata[: subdata.index("don't()")] if "don't()" in subdata else subdata)
+        for subdata in data.split("do()")
+    )
 
 
 parts = (part1, part2)