From d84ecabd59ec62b01ba595676285d85509ac7568 Mon Sep 17 00:00:00 2001
From: BingLingGroup <42505588+BingLingGroup@users.noreply.github.com>
Date: Mon, 27 Apr 2020 11:21:07 +0800
Subject: [PATCH] Fix #114 Baidu's strange error code handling Add limitation
 in SplitIntoAudioPiece with an audio length of at least 4 bytes

---
 CHANGELOG.md              | 2 ++
 autosub/api_baidu.py      | 6 ++++--
 autosub/ffmpeg_utils.py   | 5 +++++
 docs/CHANGELOG.zh-Hans.md | 2 ++
 requirements.txt          | 2 +-
 setup.py                  | 2 +-
 6 files changed, 15 insertions(+), 4 deletions(-)

diff --git a/CHANGELOG.md b/CHANGELOG.md
index 8ce7b877..ccec4013 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -57,6 +57,7 @@ Click up arrow to go back to TOC.
 - Add merge_src_assfile, merge_bilingual_assfile methods.
 - Add stop words to split events in merge_src_assfile method.
 - Add punctuations split in merge_src_assfile method.
+- Add limitation in SplitIntoAudioPiece with an audio length of at least 4 bytes.
 
 #### Changed(Unreleased)
 
@@ -74,6 +75,7 @@ Click up arrow to go back to TOC.
 - Fix os.remove() PermissionError in api_xfyun.py.
 - Fix DEFAULT_AUDIO_PRCS_CMDS and DEFAULT_CHECK_CMD.
 - Fix the issue with path checking in dependency finding.
+- Fix Baidu's strange error code handling. [issue #114](https://github.com/BingLingGroup/autosub/issues/114)
 
 #### Removed(Unreleased)
 
diff --git a/autosub/api_baidu.py b/autosub/api_baidu.py
index 5e9abaec..486a2d2b 100644
--- a/autosub/api_baidu.py
+++ b/autosub/api_baidu.py
@@ -49,8 +49,10 @@ def get_baidu_transcript(
     try:
         err_no = result_dict["err_no"]
         if err_no != 0:
-            raise exceptions.SpeechToTextException(
-                json.dumps(result_dict, indent=4, ensure_ascii=False))
+            if err_no not in (3301, 3303, 3307, 3313, 3315):
+                raise exceptions.SpeechToTextException(
+                    json.dumps(result_dict, indent=4, ensure_ascii=False))
+            raise KeyError
         if delete_chars:
             result = result_dict["result"][0].translate(
                 str.maketrans(delete_chars, " " * len(delete_chars)))
diff --git a/autosub/ffmpeg_utils.py b/autosub/ffmpeg_utils.py
index 3450464b..075875ce 100644
--- a/autosub/ffmpeg_utils.py
+++ b/autosub/ffmpeg_utils.py
@@ -87,6 +87,11 @@ def __call__(self, region):
             err = prcs.communicate()[1]
             if err:
                 return None
+            audio_file = open(filename, mode="rb")
+            audio_data = audio_file.read()
+            audio_file.close()
+            if len(audio_data) <= 4:
+                return None
             return filename
 
         except KeyboardInterrupt:
diff --git a/docs/CHANGELOG.zh-Hans.md b/docs/CHANGELOG.zh-Hans.md
index fab0b5b8..34892b30 100644
--- a/docs/CHANGELOG.zh-Hans.md
+++ b/docs/CHANGELOG.zh-Hans.md
@@ -55,6 +55,7 @@
 - 添加方法merge_src_assfile，merge_bilingual_assfile。
 - 添加停用词用于merge_src_assfile方法里的断句。
 - 添加标点符号分割功能在merge_src_assfile方法里。
+- 添加音频长度至少为4字节的检测，在SplitIntoAudioPiece里。
 
 #### 改动(未发布)
 
@@ -72,6 +73,7 @@
 - 修复api_xfyun.py中的os.remove()文件占用问题。
 - 修复DEFAULT_AUDIO_PRCS_CMDS和DEFAULT_CHECK_CMD。
 - 修复依赖查找中，路径检查的问题。
+- 修复百度奇怪的错误代码处理。[issue #114](https://github.com/BingLingGroup/autosub/issues/114)
 
 #### 删除(未发布)
 
diff --git a/requirements.txt b/requirements.txt
index 6283f8b9..a6711458 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -1,7 +1,7 @@
 requests>=2.3.0
 pysubs2>=0.2.4
 progressbar2>=3.34.3
-auditok>=0.1.5
+auditok==0.1.5
 googletrans>=2.4.0
 langcodes>=1.2.0
 wcwidth>=0.1.7
diff --git a/setup.py b/setup.py
index d7203f6e..1b4b03fe 100644
--- a/setup.py
+++ b/setup.py
@@ -34,7 +34,7 @@
         'requests>=2.3.0',
         'pysubs2>=0.2.4',
         'progressbar2>=3.34.3',
-        'auditok>=0.1.5',
+        'auditok==0.1.5',
         'googletrans>=2.4.0',
         'langcodes>=1.2.0',
         'wcwidth>=0.1.7',