♻️ refactor: rustify filter_bad_chars (#309)

Author: SigureMo

packages/biliass/rust/src/reader/xml.rs

@@ -145,8 +145,26 @@ fn parse_comment(
     Ok(parsed_p)
 }
 
+fn filter_bad_chars(string: &str) -> String {
+    string
+        .chars()
+        .map(|c| {
+            if ('\u{00}'..='\u{08}').contains(&c)
+                || c == '\u{0b}'
+                || c == '\u{0c}'
+                || ('\u{0e}'..='\u{1f}').contains(&c)
+            {
+                '\u{fffd}'
+            } else {
+                c
+            }
+        })
+        .collect()
+}
+
 pub fn read_comments_from_xml(text: &str, fontsize: f32) -> Result<Vec<Comment>, BiliassError> {
-    let mut reader = Reader::from_str(text);
+    let filtered_text = filter_bad_chars(text);
+    let mut reader = Reader::from_str(&filtered_text);
 
     let mut buf = Vec::new();
     let mut comments: Vec<Comment> = Vec::new();

packages/biliass/src/biliass/biliass.py

@@ -7,7 +7,6 @@
 import math
 import random
 import re
-import xml.dom.minidom
 from typing import TYPE_CHECKING, NamedTuple, TypeVar
 
 from biliass._core import CommentPosition, DmSegMobileReply, read_comments_from_xml
@@ -49,7 +48,6 @@ class Comment(NamedTuple):
 def read_comments_bilibili_xml(text: str | bytes, fontsize: float) -> Generator[Comment, None, None]:
     if isinstance(text, bytes):
         text = text.decode()
-    text = filter_bad_chars(text)
     res_rs = read_comments_from_xml(text, fontsize)
     return (
         Comment(
@@ -588,10 +586,6 @@ def convert_type2(row, height, bottom_reserved):
     return height - bottom_reserved - row
 
 
-def filter_bad_chars(string: str) -> str:
-    return re.sub("[\\x00-\\x08\\x0b\\x0c\\x0e-\\x1f]", "\ufffd", string)
-
-
 class safe_list(list):
     def get(self, index, default=None):
         try: