noham/N_m3u8DL-RE
1
0
Fork 0
mirror of https://github.com/NohamR/N_m3u8DL-RE.git synced 2025-05-25 06:41:57 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity

优化特殊ttml解析

Browse Source
This commit is contained in:
nilaoda 2022-08-21 15:52:09 +08:00
parent 708e90ea57
commit 01de8a53ad
1 changed files with 18 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

20
src/N_m3u8DL-RE.Parser/Mp4/MP4TtmlUtil.cs
View File

@ -32,6 +32,8 @@ namespace Mp4SubtitleParser
{
[RegexGenerator("<p.*?>(.+?)<\\/p>")]
private static partial Regex LabelFixRegex();
[RegexGenerator("\\<tt[\\s\\S]*?\\<\\/tt\\>")]
private static partial Regex MultiElementsFixRegex();
public static bool CheckInit(byte[] data)
{
@ -110,6 +112,12 @@ namespace Mp4SubtitleParser
return sb.ToString();
}
public static List<string> SplitMultipleRootElements(string xml)
{
if (!MultiElementsFixRegex().IsMatch(xml)) return new List<string>();
return MultiElementsFixRegex().Matches(xml).Select(m => m.Value).ToList();
}
public static WebVttSub ExtractFromMp4s(IEnumerable<string> items, long segTimeMs)
{
//read ttmls
@ -129,11 +137,19 @@ namespace Mp4SubtitleParser
// mdats.
if (segTimeMs != 0)
{
xmls.Add(ShiftTime(Encoding.UTF8.GetString(data), segTimeMs, segIndex));
var datas = SplitMultipleRootElements(Encoding.UTF8.GetString(data));
foreach (var item in datas)
{
xmls.Add(ShiftTime(item, segTimeMs, segIndex));
}
}
else
{
xmls.Add(Encoding.UTF8.GetString(data));
var datas = SplitMultipleRootElements(Encoding.UTF8.GetString(data));
foreach (var item in datas)
{
xmls.Add(item);
}
}
}))
.Parse(dataSeg,/* partialOkay= */ false);