From 5071e7b4b629338f84ca21ba0c9cdb12e68213d7 Mon Sep 17 00:00:00 2001
From: 雾聪 <wucong.lyb@alibaba-inc.com>
Date: 星期三, 20 十二月 2023 16:20:17 +0800
Subject: [PATCH] update TimestampSentence & TimestampIsPunctuation
---
runtime/onnxruntime/src/util.cpp | 8 +++++---
1 files changed, 5 insertions(+), 3 deletions(-)
diff --git a/runtime/onnxruntime/src/util.cpp b/runtime/onnxruntime/src/util.cpp
index ac793f5..661a50d 100644
--- a/runtime/onnxruntime/src/util.cpp
+++ b/runtime/onnxruntime/src/util.cpp
@@ -305,6 +305,10 @@
}
bool TimestampIsPunctuation(U16CHAR_T &u16) {
+ // (& ' -) in the dict
+ if (u16 == 0x26 || u16 == 0x27 || u16 == 0x2D){
+ return false;
+ }
return (u16 >= 0x21 && u16 <= 0x2F) // 鏍囧噯ASCII鏍囩偣
|| (u16 >= 0x3A && u16 <= 0x40) // 鏍囧噯ASCII鏍囩偣
|| (u16 >= 0x5B && u16 <= 0x60) // 鏍囧噯ASCII鏍囩偣
@@ -590,9 +594,7 @@
} else{
ts_sentences += ts_sent + ",";
}
-
// clear
- idx_str++;
text_seg = "";
ts_sent = "";
start = 0;
@@ -605,9 +607,9 @@
text_seg += " " + characters[idx_str];
}
ts_seg.push_back(timestamps[idx_ts]);
- idx_str++;
idx_ts++;
}
+ idx_str++;
}
// for none punc results
if(ts_seg.size() >0){
--
Gitblit v1.9.1