⚡ Bolt: [yenc decode optimization]

.jules/bolt.md

@@ -0,0 +1,3 @@
+## 2024-06-24 - [yEnc decoding optimization]
+**Learning:** Python's manual byte-by-byte iteration is slow for yEnc decoding. We can shift the workload to C-backed functions by using `bytes.translate` to decode non-escaped segments, and `bytes.find` to quickly locate the escape character (`=`). When escaping, `(char - 64 - 42) % 256` simplifies algebraically to `(char - 106) % 256`. Also, trailing escapes at line boundaries need to be correctly identified by checking length.
+**Action:** Always prefer C-backed built-in methods like `translate()` and `find()` over manual loops when processing raw byte strings in Python.

verify_nzb.py

@@ -115,19 +115,27 @@ def _parse_yenc_attrs(line: bytes) -> dict[str, str]:
     return attrs
 
 
+_YENC_DECODE_MAP = bytes((i - 42) % 256 for i in range(256))
+
 def _decode_yenc_lines(lines: Iterable[bytes]) -> bytes:
     decoded = bytearray()
     for line in lines:
-        index = 0
-        while index < len(line):
-            byte = line[index]
-            if byte == 61:
-                index += 1
-                if index >= len(line):
-                    raise ValueError("dangling yEnc escape")
-                byte = (line[index] - 64) % 256
-            decoded.append((byte - 42) % 256)
-            index += 1
+        pos = 0
+        line_len = len(line)
+        while True:
+            escape_pos = line.find(b"=", pos)
+            if escape_pos == -1:
+                decoded.extend(line[pos:].translate(_YENC_DECODE_MAP))
+                break
+
+            decoded.extend(line[pos:escape_pos].translate(_YENC_DECODE_MAP))
+
+            if escape_pos + 1 >= line_len:
+                raise ValueError("dangling yEnc escape")
+
+            decoded.append((line[escape_pos + 1] - 106) % 256)
+            pos = escape_pos + 2
+
     return bytes(decoded)