P5: remote objects are read in every shape, and videos, events, audio and links keep their own details

- ObjectShapes reads `url`, `icon` and `image` as a value, a Link or an array; Markdown `content` (PeerTube) is
  rendered; a missing `mediaType` is inferred from the extension or the attachment type (Bridgy); alt text comes from
  `name`, `summary` or their language maps; thumbnails come from attachment `icon`, object `icon[]`, `image` or a Loops
  `preview`; durations are ISO 8601 or seconds; per-attachment `sensitive` is kept; the language falls back to
  `@context` `@language` (Pleroma); titles and excerpts are plain text; hashtags normalise with NFKC like Mastodon.
- Typed details on Post: Link (Lemmy link posts, Mbin `source` URLs, Mastodon 4.7 Link attachments with an FEP-8967
  publisher preview), Video (PeerTube files with their streams, HLS playlist, poster, captions, chapters, licence,
  channel, live state, comment policy), Audio (Funkwhale), Event (Mobilizon, Gancio, Hubzilla: times, zone, floating
  time, places and addresses, online link, capacity, status), the cover image, and the author's own source text.
- Mastodon API: a card built from those details without fetching anything, an event's "when · where" line, attachment
  thumbnails and durations, and no media file offered as a preview image.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_012CzABvBkbcFqoHdmi8b9WB

PrivaPub/Federation/Objects/NoteParser.cs

@@ -1,6 +1,7 @@
 using PrivaPub.Models.Post;
 
 using System.Globalization;
+using System.Text;
 using System.Text.Json.Nodes;
 
 using static PrivaPub.Federation.Objects.ActivityJson;
@@ -31,6 +32,12 @@ namespace PrivaPub.Federation.Objects
 		public IReadOnlyList<PostMention> Mentions { get; init; } = Array.Empty<PostMention>();
 		public IReadOnlyList<string> Tags { get; init; } = Array.Empty<string>();
 		public IReadOnlyList<PostMedia> Attachments { get; init; } = Array.Empty<PostMedia>();
+		public PostSource Source { get; init; }
+		public string CoverURL { get; init; }
+		public PostLink Link { get; init; }
+		public VideoDetails Video { get; init; }
+		public AudioDetails Audio { get; init; }
+		public EventDetails Event { get; init; }
 	}
 
 	public static class NoteParser
@@ -50,6 +57,8 @@ namespace PrivaPub.Federation.Objects
 
 			var (contentHtml, language) = Content(note);
 			var summary = Value(note, "summary");
+			var excerpt = ObjectShapes.Text(summary, 1000)
+				?? (note["preview"] is JsonObject preview && Value(preview, "type") == "Note" ? ObjectShapes.Text(Value(preview, "content"), 1000) : default);
 			var sensitive = note["sensitive"] is JsonValue flag && flag.TryGetValue<bool>(out var marked) && marked;
 			var summaryWarns = type is "Note" or "Question" || sensitive;
 			return new NoteDocument
@@ -59,9 +68,9 @@ namespace PrivaPub.Federation.Objects
 				Raw = note,
 				AttributedTo = Attribution(note["attributedTo"]),
 				ContentHtml = ContentSanitizer.Html(contentHtml),
-				Title = Plain(Value(note, "name"), 500),
+				Title = ObjectShapes.Text(Value(note, "name"), 500),
 				SpoilerText = summaryWarns ? Plain(summary, 500) : default,
-				Excerpt = summaryWarns ? default : Plain(summary, 1000),
+				Excerpt = summaryWarns ? default : excerpt,
 				Sensitive = sensitive || summaryWarns && !string.IsNullOrWhiteSpace(summary),
 				Language = language,
 				Url = UrlOf(note["url"]) ?? id,
@@ -78,12 +87,18 @@ namespace PrivaPub.Federation.Objects
 					.DistinctBy(m => m.ActorURI)
 					.Take(MaxTags)
 					.ToList(),
-				Tags = Tags(note, "Hashtag").Select(t => Value(t, "name")?.TrimStart('#').ToLowerInvariant())
+				Tags = Tags(note, "Hashtag").Select(t => Value(t, "name")?.TrimStart('#').Normalize(NormalizationForm.FormKC).ToLowerInvariant())
 					.Where(t => !string.IsNullOrEmpty(t))
 					.Distinct()
 					.Take(MaxTags)
 					.ToList(),
-				Attachments = Attachments(note)
+				Attachments = Attachments(note),
+				Source = ObjectShapes.Source(note),
+				CoverURL = ObjectShapes.Cover(note),
+				Link = ObjectShapes.Link(note),
+				Video = ObjectShapes.Video(note),
+				Audio = ObjectShapes.Audio(note),
+				Event = ObjectShapes.Event(note)
 			};
 		}
 
@@ -100,7 +115,9 @@ namespace PrivaPub.Federation.Objects
 			}
 			if (string.IsNullOrEmpty(content) && Value(note, "_misskey_content") is { } misskey)
 				content = System.Net.WebUtility.HtmlEncode(misskey).Replace("\n", "<br>");
-			return (content, language);
+			if (!string.IsNullOrEmpty(content) && Value(note, "mediaType")?.StartsWith("text/markdown", StringComparison.OrdinalIgnoreCase) == true)
+				content = Rendering.ActivityPubRenderer.Html(content);
+			return (content, language ?? ObjectShapes.Language(note));
 		}
 
 		static string Attribution(JsonNode node) => node switch
@@ -120,33 +137,37 @@ namespace PrivaPub.Federation.Objects
 				_ => Enumerable.Empty<JsonObject>()
 			}).Where(t => Value(t, "type") == type);
 
-		static List<PostMedia> Attachments(JsonObject note)
-		{
-			var items = note["attachment"] switch
-			{
-				JsonArray array => array.OfType<JsonObject>(),
-				JsonObject single => new[] { single },
-				_ => Enumerable.Empty<JsonObject>()
-			};
-			return items
+		static List<PostMedia> Attachments(JsonObject note) =>
+			ObjectShapes.Objects(note["attachment"])
 				.Where(a => Value(a, "type") is "Document" or "Image" or "Video" or "Audio")
-				.Select(a => new PostMedia
-				{
-					RemoteURL = UrlOf(a["url"]),
-					ContentType = Value(a, "mediaType"),
-					Description = Plain(Value(a, "name") ?? Value(a, "summary"), 1500),
-					Blurhash = Value(a, "blurhash"),
-					Width = Int(a["width"]),
-					Height = Int(a["height"]),
-					Focus = a["focalPoint"] is JsonArray focus && focus.Count == 2
-						? focus.Select(f => f is JsonValue v && v.TryGetValue<double>(out var d) ? (float)Math.Clamp(d, -1, 1) : 0f).ToArray()
-						: default
-				})
+				.Select(Attachment)
 				.Where(m => Origin.Of(m.RemoteURL) != default)
 				.Take(MaxAttachments)
 				.ToList();
+
+		static PostMedia Attachment(JsonObject attachment)
+		{
+			var (url, linkMediaType) = ObjectShapes.MediaLink(attachment["url"] ?? attachment["href"]);
+			return new PostMedia
+			{
+				RemoteURL = url,
+				ContentType = ObjectShapes.InferMediaType(Value(attachment, "mediaType") ?? linkMediaType, url, Value(attachment, "type")),
+				RemotePreviewURL = ObjectShapes.Image(attachment["icon"]),
+				DurationSeconds = ObjectShapes.Seconds(attachment["duration"]),
+				Sensitive = attachment["sensitive"] is JsonValue flag && flag.TryGetValue<bool>(out var marked) ? marked : default(bool?),
+				Description = Plain(Value(attachment, "name") ?? Value(attachment, "summary") ?? FirstOf(attachment["nameMap"]) ?? FirstOf(attachment["summaryMap"]), 1500),
+				Blurhash = Value(attachment, "blurhash"),
+				Width = Int(attachment["width"]),
+				Height = Int(attachment["height"]),
+				Focus = attachment["focalPoint"] is JsonArray focus && focus.Count == 2
+					? focus.Select(f => f is JsonValue v && v.TryGetValue<double>(out var d) ? (float)Math.Clamp(d, -1, 1) : 0f).ToArray()
+					: default
+			};
 		}
 
+		static string FirstOf(JsonNode map) =>
+			map is JsonObject languages ? languages.Select(l => l.Value is JsonValue v && v.TryGetValue<string>(out var text) ? text : default).FirstOrDefault(t => t != default) : default;
+
 		static string UrlOf(JsonNode node) => node switch
 		{
 			JsonValue value when value.TryGetValue<string>(out var text) => text,

PrivaPub/Federation/Objects/ObjectShapes.cs

@@ -0,0 +1,293 @@
+using PrivaPub.Models.Post;
+
+using System.Globalization;
+using System.Net;
+using System.Text.Json.Nodes;
+using System.Text.RegularExpressions;
+using System.Xml;
+
+using static PrivaPub.Federation.Objects.ActivityJson;
+
+namespace PrivaPub.Federation.Objects
+{
+	public static partial class ObjectShapes
+	{
+		const int MaxVariants = 20;
+		const int MaxCaptions = 20;
+		const int MaxPlaces = 4;
+		const int CoverMaxWidth = 1280;
+
+		static readonly Dictionary<string, string> MediaTypesByExtension = new(StringComparer.OrdinalIgnoreCase)
+		{
+			[".jpg"] = "image/jpeg", [".jpeg"] = "image/jpeg", [".png"] = "image/png", [".gif"] = "image/gif", [".webp"] = "image/webp",
+			[".avif"] = "image/avif", [".heic"] = "image/heic", [".svg"] = "image/svg+xml",
+			[".mp4"] = "video/mp4", [".m4v"] = "video/mp4", [".webm"] = "video/webm", [".mov"] = "video/quicktime", [".mkv"] = "video/x-matroska",
+			[".mp3"] = "audio/mpeg", [".ogg"] = "audio/ogg", [".oga"] = "audio/ogg", [".opus"] = "audio/opus", [".m4a"] = "audio/mp4",
+			[".flac"] = "audio/flac", [".wav"] = "audio/wav", [".m3u8"] = "application/x-mpegURL"
+		};
+
+		public static IEnumerable<JsonObject> Objects(JsonNode node) => node switch
+		{
+			JsonArray array => array.OfType<JsonObject>(),
+			JsonObject single => new[] { single },
+			_ => Enumerable.Empty<JsonObject>()
+		};
+
+		public static (string Url, string MediaType) MediaLink(JsonNode node)
+		{
+			switch (node)
+			{
+				case JsonValue value when value.TryGetValue<string>(out var text):
+					return (text, default);
+				case JsonObject link when Value(link, "href") is { } href:
+					return (href, Value(link, "mediaType"));
+				case JsonObject wrapper when wrapper["url"] is { } inner:
+					var (url, mediaType) = MediaLink(inner);
+					return (url, mediaType ?? Value(wrapper, "mediaType"));
+				case JsonArray array:
+					var links = array.Select(MediaLink).Where(l => l.Url != default).ToList();
+					return links.FirstOrDefault(l => l.MediaType?.StartsWith("text/html", StringComparison.OrdinalIgnoreCase) != true);
+				default:
+					return default;
+			}
+		}
+
+		public static string InferMediaType(string declared, string url, string objectType)
+		{
+			if (!string.IsNullOrWhiteSpace(declared) && declared.Contains('/'))
+				return declared.Trim();
+			if (Uri.TryCreate(url, UriKind.Absolute, out var uri) && MediaTypesByExtension.TryGetValue(Path.GetExtension(uri.AbsolutePath), out var byExtension))
+				return byExtension;
+			return objectType switch
+			{
+				"Image" => "image/*",
+				"Video" => "video/*",
+				"Audio" => "audio/*",
+				_ => default
+			};
+		}
+
+		public static string Image(JsonNode node)
+		{
+			var candidates = node switch
+			{
+				JsonValue value when value.TryGetValue<string>(out var text) => new[] { (Url: text, Width: (int?)default) },
+				JsonArray array => array.Select(item => (Url: MediaLink(item).Url, Width: Int(item as JsonObject, "width"))).ToArray(),
+				JsonObject single => new[] { (Url: MediaLink(single).Url, Width: Int(single, "width")) },
+				_ => Array.Empty<(string Url, int? Width)>()
+			};
+			var usable = candidates.Where(c => Origin.Of(c.Url) != default).ToList();
+			return usable.Where(c => c.Width <= CoverMaxWidth).OrderByDescending(c => c.Width).Select(c => c.Url).FirstOrDefault()
+				?? usable.Select(c => c.Url).FirstOrDefault();
+		}
+
+		public static string Cover(JsonObject note) =>
+			Image(note["image"]) ?? Image(note["icon"]) ?? (note["preview"] is JsonObject preview && Value(preview, "type") == "Image" ? Image(preview) : default);
+
+		public static double? Seconds(JsonNode node)
+		{
+			switch (node)
+			{
+				case JsonValue value when value.TryGetValue<double>(out var number):
+					return number >= 0 && number < 1e7 ? number : default;
+				case JsonValue value when value.TryGetValue<string>(out var text) && !string.IsNullOrWhiteSpace(text):
+					try
+					{
+						var span = XmlConvert.ToTimeSpan(text.Trim());
+						return span >= TimeSpan.Zero ? span.TotalSeconds : default;
+					}
+					catch (FormatException)
+					{
+						return double.TryParse(text, NumberStyles.Float, CultureInfo.InvariantCulture, out var seconds) && seconds >= 0 ? seconds : default;
+					}
+				default:
+					return default;
+			}
+		}
+
+		public static string Text(string html, int max)
+		{
+			if (string.IsNullOrWhiteSpace(html))
+				return default;
+			var text = WebUtility.HtmlDecode(Tags().Replace(html, " "));
+			text = Spaces().Replace(text, " ").Trim();
+			if (text.Length == 0)
+				return default;
+			return text.Length <= max ? text : text[..max];
+		}
+
+		public static string Language(JsonObject note) =>
+			note["@context"] is JsonArray context
+				? context.OfType<JsonObject>().Select(entry => Value(entry, "@language")).FirstOrDefault(language => !string.IsNullOrWhiteSpace(language))
+				: default;
+
+		public static PostSource Source(JsonObject note) => note["source"] is JsonObject source && Value(source, "content") is { } content
+			? new PostSource { Content = content.Length <= 64 * 1024 ? content : content[..(64 * 1024)], MediaType = Value(source, "mediaType") }
+			: default;
+
+		public static PostLink Link(JsonObject note)
+		{
+			var attached = Objects(note["attachment"]).FirstOrDefault(a => Value(a, "type") == "Link" && IsWeb(Value(a, "href")));
+			if (attached != default)
+			{
+				var preview = attached["preview"] as JsonObject;
+				return new PostLink
+				{
+					Href = Value(attached, "href"),
+					MediaType = Value(attached, "mediaType"),
+					Title = Text(Value(preview, "name") ?? Value(attached, "name"), 300),
+					Description = Text(Value(preview, "summary") ?? Value(preview, "content"), 1000),
+					ImageURL = preview == default ? Image(note["image"]) : Image(preview["image"]) ?? Image(preview["icon"]),
+					FromPublisher = preview != default
+				};
+			}
+			if (note["source"] is JsonValue source && source.TryGetValue<string>(out var sourceUrl) && IsWeb(sourceUrl))
+				return new PostLink { Href = sourceUrl, ImageURL = Image(note["image"]) };
+			return default;
+		}
+
+		public static VideoDetails Video(JsonObject note)
+		{
+			if (Value(note, "type") != "Video")
+				return default;
+			var links = Objects(note["url"]).ToList();
+			return new VideoDetails
+			{
+				DurationSeconds = Seconds(note["duration"]),
+				Variants = Variants(links, "video/", "audio/"),
+				PlaylistURL = links.Where(l => Value(l, "mediaType")?.Equals("application/x-mpegURL", StringComparison.OrdinalIgnoreCase) == true)
+					.Select(l => Value(l, "href")).FirstOrDefault(IsWeb),
+				IsLive = Bool(note, "isLiveBroadcast") == true,
+				State = Int(note, "state"),
+				Views = Long(note, "views"),
+				CommentsPolicy = Int(note, "commentsPolicy"),
+				DownloadEnabled = Bool(note, "downloadEnabled"),
+				Licence = Named(note["licence"] ?? note["license"]),
+				Category = Named(note["category"]),
+				ChannelURI = Objects(note["attributedTo"]).Where(a => Value(a, "type") == "Group").Select(a => Id(a)).FirstOrDefault(),
+				EmbedURL = Value(note, "embedUrl") is { } embed && IsWeb(embed) ? embed : default,
+				ChaptersURL = Value(note, "hasParts") is { } chapters && IsWeb(chapters) ? chapters : default,
+				Captions = Objects(note["subtitleLanguage"])
+					.Select(c => new Caption
+					{
+						Language = Value(c, "identifier"),
+						Url = Objects(c["url"]).Where(u => Value(u, "mediaType") == "text/vtt").Select(u => Value(u, "href")).FirstOrDefault(IsWeb)
+							?? (Value(c, "url") is { } plain && IsWeb(plain) ? plain : default)
+					})
+					.Where(c => c.Url != default)
+					.Take(MaxCaptions)
+					.ToList()
+			};
+		}
+
+		public static AudioDetails Audio(JsonObject note)
+		{
+			if (Value(note, "type") != "Audio")
+				return default;
+			return new AudioDetails
+			{
+				DurationSeconds = Seconds(note["duration"]),
+				Variants = Variants(Objects(note["url"]).ToList(), "audio/"),
+				Album = Named(note["album"]),
+				Position = Int(note, "position"),
+				Licence = Named(note["license"] ?? note["licence"])
+			};
+		}
+
+		public static EventDetails Event(JsonObject note)
+		{
+			if (Value(note, "type") != "Event")
+				return default;
+			var start = Value(note, "startTime");
+			var places = Objects(note["location"]).Select(Place).Where(p => p != default).Take(MaxPlaces).ToList();
+			return new EventDetails
+			{
+				Start = Moment(start),
+				End = Moment(Value(note, "endTime")),
+				TimeZone = Value(note, "timezone"),
+				FloatingTime = start?.EndsWith("-00:00", StringComparison.Ordinal) == true,
+				Places = places,
+				OnlineURL = places.Select(p => p.Url).FirstOrDefault(u => u != default)
+					?? Objects(note["attachment"]).Where(a => Value(a, "type") == "Link" && Value(a, "name") == "Website").Select(a => Value(a, "href")).FirstOrDefault(IsWeb),
+				JoinMode = Value(note, "joinMode"),
+				ParticipantCount = Int(note, "participantCount"),
+				Capacity = Int(note, "maximumAttendeeCapacity"),
+				Status = Value(note, "ical:status") ?? Value(note, "eventStatus") ?? Value(note, "status"),
+				Category = Named(note["category"])
+			};
+		}
+
+		static EventPlace Place(JsonObject location)
+		{
+			var type = Value(location, "type");
+			if (type == "VirtualLocation")
+				return Value(location, "url") is { } url && IsWeb(url) ? new EventPlace { Name = Text(Value(location, "name"), 200), Url = url } : default;
+			var address = location["address"] switch
+			{
+				JsonValue value when value.TryGetValue<string>(out var text) => text,
+				JsonObject postal => string.Join(", ", new[] { "streetAddress", "postalCode", "addressLocality", "addressRegion", "addressCountry" }
+					.Select(field => Value(postal, field)).Where(part => !string.IsNullOrWhiteSpace(part))),
+				_ => default
+			};
+			var place = new EventPlace
+			{
+				Name = Text(Value(location, "name") ?? Value(location, "content"), 200),
+				Address = Text(address, 300),
+				Latitude = Double(location, "latitude"),
+				Longitude = Double(location, "longitude")
+			};
+			return place.Name == default && place.Address == default && place.Latitude == default ? default : place;
+		}
+
+		static List<MediaVariant> Variants(List<JsonObject> links, params string[] prefixes) =>
+			links.Where(l => Value(l, "mediaType") is { } type && prefixes.Any(p => type.StartsWith(p, StringComparison.OrdinalIgnoreCase)) && IsWeb(Value(l, "href")))
+				.Select(l => new MediaVariant
+				{
+					Url = Value(l, "href"),
+					MediaType = Value(l, "mediaType"),
+					Width = Int(l, "width"),
+					Height = Int(l, "height"),
+					Fps = Double(l, "fps"),
+					Size = Long(l, "size"),
+					Streams = Objects(l["attachment"])
+						.Where(p => Value(p, "name") == "ffprobe_codec_type")
+						.Select(p => Value(p, "value"))
+						.Where(v => v is "audio" or "video")
+						.Distinct()
+						.ToList()
+				})
+				.Take(MaxVariants)
+				.ToList();
+
+		static string Named(JsonNode node) => node switch
+		{
+			JsonValue value when value.TryGetValue<string>(out var text) => Text(text, 200),
+			JsonObject named => Text(Value(named, "name") ?? Value(named, "identifier"), 200),
+			_ => default
+		};
+
+		static DateTime? Moment(string value) =>
+			DateTimeOffset.TryParse(value, CultureInfo.InvariantCulture, DateTimeStyles.AssumeUniversal, out var parsed) ? parsed.UtcDateTime : (DateTime?)null;
+
+		static bool IsWeb(string url) =>
+			Uri.TryCreate(url, UriKind.Absolute, out var uri) && (uri.Scheme == Uri.UriSchemeHttps || uri.Scheme == Uri.UriSchemeHttp);
+
+		static int? Int(JsonObject node, string property) =>
+			node?[property] is JsonValue value && value.TryGetValue<double>(out var number) && number is >= 0 and < int.MaxValue ? (int)number : (int?)null;
+
+		static long? Long(JsonObject node, string property) =>
+			node?[property] is JsonValue value && value.TryGetValue<double>(out var number) && number is >= 0 and < long.MaxValue ? (long)number : (long?)null;
+
+		static double? Double(JsonObject node, string property) =>
+			node?[property] is JsonValue value && value.TryGetValue<double>(out var number) && double.IsFinite(number) ? number : (double?)null;
+
+		static bool? Bool(JsonObject node, string property) =>
+			node?[property] is JsonValue value && value.TryGetValue<bool>(out var flag) ? flag : (bool?)null;
+
+		[GeneratedRegex("<[^>]*>")]
+		private static partial Regex Tags();
+
+		[GeneratedRegex(@"\s+")]
+		private static partial Regex Spaces();
+	}
+}