当翻译大量“あ”等拟声词时，GPT可能会出现bug，输出大量的“啊”而不进行后续翻译

[Immortalyzy]

翻译之前手动去除这些语句，跳过翻译，翻译结束后再添加回去（语气词可能保留日文效果更好）

[Immortalyzy]

Should not just skip the block, there could be more content in the sentence. Best method would be take a list of surrounding characters and match these "cry" sentences. Remove them before translation and put them back after translation.

[Immortalyzy]

Temporary solution, don't translate blocks with more than 3 "あああ"s.

[Immortalyzy]

也可以使用专门的函数单独翻译拟声词，配备专门的额prompt让gpt翻译拟声词。

[Immortalyzy]

似乎还是会导致问题，这个很难取舍，因为很多拟声词中间的文字不应该被去除。