Posts found by their words

/api/v2/search now searches statuses by words for a signed-in persona, in
Mastodon's scope: what it wrote, boosted, favourited, bookmarked or was
named in, and public posts of authors who let themselves be indexed
(indexable, off by default). Newest first, only what the persona may see,
through a text index over the posts' words in every language alike.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01LsXgEaXee4GCU1hwYgPJXw

PrivaPub/Api/Mastodon/Controllers/SearchController.cs

@@ -15,7 +15,7 @@ using PostEntity = PrivaPub.Models.Post.Post;
 
 namespace PrivaPub.Api.Mastodon.Controllers
 {
-	public class SearchController : MastodonController
+	public partial class SearchController : MastodonController
 	{
 		readonly MastodonMapper _mapper;
 		readonly DbEntities _dbEntities;
@@ -73,12 +73,52 @@ namespace PrivaPub.Api.Mastodon.Controllers
 			{
 				results.Accounts = await _search.Find(q, resolve, Limit(), Math.Max(0, offset), token);
 			}
+			if (type is null or "statuses" && MyId != default && !q.StartsWith('#'))
+				results.Statuses = await Words(q, Math.Max(0, offset), Limit(), token);
 			if (type is null or "hashtags" && q.TrimStart('#') is { Length: > 0 } tag)
 				results.Hashtags.Add(new { name = tag.ToLowerInvariant(), url = $"{_localActors.BaseAddress}/tags/{Uri.EscapeDataString(tag.ToLowerInvariant())}", history = Array.Empty<object>() });
 			return Json(results);
 		}
 	}
 
+	public partial class SearchController
+	{
+		const int WordCandidates = 400;
+
+		// Posts by their words, as Mastodon searches them: those the persona wrote, boosted, favourited, bookmarked or was
+		// named in, and public posts of authors who let themselves be indexed; newest first, what the persona may see only
+		async Task<List<Status>> Words(string q, int offset, int limit, CancellationToken token)
+		{
+			var candidates = await _dbEntities.Posts.Match(f => f.Text(q)).Match(p => !p.DeletedAt.HasValue && p.ReblogOfPostId == null && !p.AuthorGone)
+				.Sort(p => p.ID, MongoDB.Entities.Order.Descending).Limit(WordCandidates).ExecuteAsync(token);
+			if (candidates.Count == 0)
+				return new List<Status>();
+			var ids = candidates.Select(p => p.ID).ToList();
+			var mine = new HashSet<string>();
+			mine.UnionWith((await _dbEntities.Favourites.Match(f => f.AccountId == MyId && ids.Contains(f.PostId)).ExecuteAsync(token)).Select(f => f.PostId));
+			mine.UnionWith((await MongoDB.Entities.DB.Default.Find<Models.Social.Bookmark>().Match(b => b.AvatarId == MyId && ids.Contains(b.PostId)).ExecuteAsync(token))
+				.Select(b => b.PostId));
+			mine.UnionWith((await _dbEntities.Posts.Match(p => p.AuthorAccountId == MyId && ids.Contains(p.ReblogOfPostId) && !p.DeletedAt.HasValue).ExecuteAsync(token))
+				.Select(p => p.ReblogOfPostId));
+			var authors = candidates.Select(p => p.AuthorAccountId ?? p.GroupUserId).Where(a => a != default).Distinct().ToList();
+			var indexable = (await _dbEntities.Avatars.Match(a => authors.Contains(a.ID) && a.Settings.IsIndexable).ExecuteAsync(token)).Select(a => a.ID)
+				.Concat((await _dbEntities.ForeignAvatars.Match(f => authors.Contains(f.ID) && f.IsIndexable).ExecuteAsync(token)).Select(f => f.ID))
+				.ToHashSet();
+			var found = new List<PostEntity>();
+			foreach (var post in candidates)
+			{
+				var author = post.AuthorAccountId ?? post.GroupUserId;
+				var reachable = author == MyId && !post.IsFederatedCopy || mine.Contains(post.ID) || post.Mentions.Any(m => m.AccountId == MyId)
+					|| post.Visibility == PostVisibility.Public && indexable.Contains(author);
+				if (reachable && await VisibilityPolicy.CanSee(post, MyId, token))
+					found.Add(post);
+				if (found.Count >= offset + limit)
+					break;
+			}
+			return await _mapper.Statuses(found.Skip(offset).Take(limit).ToList(), MyId, token);
+		}
+	}
+
 	public class StubsController : MastodonController
 	{
 		[HttpGet("/api/v1/custom_emojis"), Microsoft.AspNetCore.Authorization.AllowAnonymous]

PrivaPub/Infrastructure/Data/Indexes.cs

@@ -110,6 +110,11 @@ namespace PrivaPub.Infrastructure.Data
 			await Plain<PersonaFilter>(token, f => f.AvatarId);
 			await Plain<FollowedTag>(token, t => t.Name);
 			await Plain<Post>(token, p => p.Tags);
+			//the words status search reads, in every language alike (no stemming, no stop words); Language is not the
+			//index's language field
+			await DB.Default.Index<Post>().Key(p => p.Text, KeyType.Text).Key(p => p.ContentHtml, KeyType.Text).Key(p => p.SpoilerText, KeyType.Text)
+				.Key(p => p.Title, KeyType.Text).Option(o => { o.DefaultLanguage = "none"; o.LanguageOverride = "textSearchLanguage"; o.Name = "words"; })
+				.CreateAsync(token);
 			await Plain<PersonaListMember>(token, m => m.AvatarId);
 			await Plain<Mute>(token, m => m.TargetActorURI);
 			await Plain<Report>(token, r => r.IsResolved, r => r.ID);