Escape arrow sequences when writing closed captions to an SRT file

Closes #755
Tyrrrz · Dec 27, 2023 · f08ac9d · f08ac9d
1 parent 4784fb3
commit f08ac9d
Show file tree

Hide file tree

Showing 2 changed files with 18 additions and 5 deletions.
diff --git a/YoutubeExplode.Converter.Tests/SubtitleSpecs.cs b/YoutubeExplode.Converter.Tests/SubtitleSpecs.cs
@@ -24,15 +24,15 @@ public async Task I_can_download_a_video_as_a_single_mp4_file_with_subtitles()
         using var dir = TempDir.Create();
         var filePath = Path.Combine(dir.Path, "video.mp4");
 
-        var streamManifest = await youtube.Videos.Streams.GetManifestAsync("YltHGKX80Y8");
+        var streamManifest = await youtube.Videos.Streams.GetManifestAsync("NtQkz0aRDe8");
         var streamInfos = streamManifest
             .GetVideoStreams()
             .Where(s => s.Container == Container.Mp4)
             .OrderBy(s => s.Size)
             .Take(1)
             .ToArray();
 
-        var trackManifest = await youtube.Videos.ClosedCaptions.GetManifestAsync("YltHGKX80Y8");
+        var trackManifest = await youtube.Videos.ClosedCaptions.GetManifestAsync("NtQkz0aRDe8");
         var trackInfos = trackManifest.Tracks;
 
         // Act
@@ -61,15 +61,15 @@ public async Task I_can_download_a_video_as_a_single_webm_file_with_subtitles()
         using var dir = TempDir.Create();
         var filePath = Path.Combine(dir.Path, "video.webm");
 
-        var streamManifest = await youtube.Videos.Streams.GetManifestAsync("YltHGKX80Y8");
+        var streamManifest = await youtube.Videos.Streams.GetManifestAsync("NtQkz0aRDe8");
         var streamInfos = streamManifest
             .GetVideoStreams()
             .Where(s => s.Container == Container.WebM)
             .OrderBy(s => s.Size)
             .Take(1)
             .ToArray();
 
-        var trackManifest = await youtube.Videos.ClosedCaptions.GetManifestAsync("YltHGKX80Y8");
+        var trackManifest = await youtube.Videos.ClosedCaptions.GetManifestAsync("NtQkz0aRDe8");
         var trackInfos = trackManifest.Tracks;
 
         // Act

diff --git a/YoutubeExplode/Videos/ClosedCaptions/ClosedCaptionClient.cs b/YoutubeExplode/Videos/ClosedCaptions/ClosedCaptionClient.cs
@@ -165,7 +165,20 @@ static string FormatTimestamp(TimeSpan value) =>
                 .Append(FormatTimestamp(caption.Offset + caption.Duration))
                 .AppendLine()
                 // Content
-                .AppendLine(caption.Text);
+                .AppendLine(
+                    caption
+                        .Text
+                        // Caption text may contain valid SRT-formatted data in itself.
+                        // This can happen, for example, if the subtitles for a YouTube video
+                        // were imported from an SRT file, but something went wrong in the
+                        // process, resulting in parts of the file being read as captions
+                        // rather than control sequences.
+                        // SRT file format does not provide any means of escaping special
+                        // characters, so as a workaround we just replace the dashes in the
+                        // arrow sequence with en-dashes, which look similar enough.
+                        // https://github.com/Tyrrrz/YoutubeExplode/issues/755
+                        .Replace("-->", "––>")
+                );
 
             await writer.WriteLineAsync(buffer.ToString());
             buffer.Clear();