mirror
/
bazarr
mirror of https://github.com/morpheus65535/bazarr
1
0
Fork 0
Code Issues Releases Wiki Activity

improve character encoding detection

This commit is contained in:
josdion 2020-03-26 20:10:58 +02:00
parent 92497a8822
commit be0411d50a
1 changed files with 5 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
bazarr/list_subtitles.py
View File

@ -379,10 +379,12 @@ def guess_external_subtitles(dest_folder, subtitles):
try:
# to improve performance, use only the first 32K to detect encoding
if len(text) > 32768: guess = chardet.detect(text[:32768])
else: guess = chardet.detect(text)
if guess["confidence"] < 0.8:
guess = chardet.detect(text[:32768])
logging.debug('BAZARR detected encoding %r', guess)
if guess["confidence"] < 0.6:
raise UnicodeError
if guess["confidence"] < 0.8 or guess["encoding"] == "ascii":
guess["encoding"] = "utf-8"
text = text.decode(guess["encoding"])
detected_language = guess_language(text)
except UnicodeError: