* fix the value of thread_info.reply_num might be -1: lumina37/ai…

…otieba#64 (comment) @ `ThreadParser.Convert()`

* replace all `response(s) with` to `respond with`
@ crawler

crawler/src/Tieba/Crawl/Facade/ThreadArchiveCrawlFacade.cs

@@ -10,7 +10,7 @@ public ThreadArchiveCrawlFacade(ILogger<ThreadArchiveCrawlFacade> logger, TbmDbC
     ) : base(logger, dbContextFactory, crawler.Invoke, parser, saver, users, requesterTcs, locks, fid, forumName) { }
 
     protected override void PostParseHook(ThreadResponse response, CrawlRequestFlag flag, Dictionary<PostId, ThreadPost> parsedPostsInResponse)
-    { // the second response with flag is as same as the first one so just skip it
+    { // the second respond with flag is as same as the first one so just skip it
         if (flag == CrawlRequestFlag.ThreadClientVersion602) return;
         var data = response.Data;
         Users.ParseUsers(data.ThreadList.Select(t => t.Author));

crawler/src/Tieba/Crawl/Parser/ReplyParser.cs

@@ -46,7 +46,7 @@ protected override ReplyPost Convert(Reply inPost)
             o.Content = Helper.SerializedProtoBufWrapperOrNullIfEmpty(inPost.Content,
                 () => Helper.WrapPostContent(inPost.Content));
             o.OriginalContents = inPost.Content;
-            // AuthorId rarely response with 0, Author should always be null but we can guarantee
+            // AuthorId rarely respond with 0, Author should always be null but we can guarantee
             o.AuthorUid = inPost.AuthorId.NullIfZero() ?? inPost.Author?.Uid ?? 0;
             // value of AuthorExpGrade will be write back in ReplyCrawlFacade.FillAuthorInfoBackToReply()
             o.SubReplyCount = inPost.SubPostNumber.NullIfZero();

crawler/src/Tieba/Crawl/Parser/ThreadParser.cs

@@ -26,7 +26,7 @@ protected override ThreadPost Convert(Thread inPost)
             o.LatestReplyPostedAt = (uint)inPost.LastTimeInt;
             // LastReplyer will be null when LivePostType != "", but LastTimeInt will have expected timestamp value
             o.LatestReplierUid = inPost.LastReplyer?.Uid;
-            o.ReplyCount = (uint?)inPost.ReplyNum.NullIfZero();
+            o.ReplyCount = inPost.ReplyNum < 0 ? 0 : (uint?)inPost.ReplyNum.NullIfZero(); // rarely respond with -1
             o.ViewCount = (uint?)inPost.ViewNum.NullIfZero();
             o.ShareCount = (uint?)inPost.ShareNum.NullIfZero();
             // when the thread is livepost or Thread.AgreeNum == 0, the agree field will not exists

crawler/src/Tieba/Crawl/Saver/ReplySaver.cs

@@ -147,14 +147,15 @@ where imagesKeyByUrlFilename.Keys.Contains(e.UrlFilename)
         existingImages.Values.Where(e => e.ByteSize == 0)
             .Join(imagesKeyByUrlFilename.Values, e => e.UrlFilename, i => i.UrlFilename,
                 (existing, newInContent) => (existing, newInContent))
-            .ForEach(t => t.existing.ByteSize = t.newInContent.ByteSize); // randomly response with 0
+            .ForEach(t => t.existing.ByteSize = t.newInContent.ByteSize); // randomly respond with 0
         db.ReplyContentImages.AddRange(pidAndImageList.Select(t => new ReplyContentImage
         {
             Pid = t.Pid,
             // no need to manually invoke DbContent.AddRange(images) since EF Core will do these chore
             // https://stackoverflow.com/questions/5212751/how-can-i-retrieve-id-of-inserted-entity-using-entity-framework/41146434#41146434
             // reuse the same instance from imagesKeyByUrlFilename will prevent assigning multiple different instances with the same key
             // which will cause EF Core to insert identify entry more than one time leading to duplicated entry error
+            // https://github.com/dotnet/efcore/issues/30236
             Image = existingImages.TryGetValue(t.Image.UrlFilename, out var e) ? e : imagesKeyByUrlFilename[t.Image.UrlFilename]
         }));
     }

crawler/src/Tieba/Crawl/Saver/StaticCommonInSavers.cs

@@ -15,7 +15,7 @@ public record FieldChangeIgnoranceCallbackRecord(FieldChangeIgnoranceCallback Up
             if (whichPostType == typeof(TiebaUser))
             {
                 switch (propName)
-                { // possible randomly response with null
+                { // possible randomly respond with null
                     case nameof(TiebaUser.IpGeolocation) when newValue is null:
                     // possible clock drift across multiple response from tieba api, they should sync their servers with NTP
                     /* following sql can track these drift
@@ -41,7 +41,7 @@ when Math.Abs((newValue as int? ?? 0) - (oldValue as int? ?? 0)) <= 10:
                     // prevent overwrite existing value of field liker_id which is saved by legacy crawler
                     // and Zan itself is deprecated by tieba so it shouldn't get updated
                     case nameof(ThreadPost.Zan):
-                    // possible randomly response with null
+                    // possible randomly respond with null
                     case nameof(ThreadPost.Geolocation) when newValue is null:
                     // empty string means the author had not write a title
                     // its value generated from the first reply within response of reply crawler will be later set by ReplyCrawlFacade.SaveParentThreadTitle()
@@ -50,19 +50,19 @@ when newValue is ""
                              // prevent repeatedly update with different title due to the thread is a multi forum topic thread
                              // thus its title can be vary within the forum and within the thread
                              || (newValue is not "" && oldValue is not ""):
-                    // possible randomly response with 0.NullIfZero()
+                    // possible randomly respond with 0.NullIfZero()
                     case nameof(ThreadPost.DisagreeCount) when newValue is null && oldValue is not null:
                     // when the latest reply post is deleted and there's no new reply after delete
                     // this field but not LatestReplyPostedAt will be null
                     case nameof(ThreadPost.LatestReplierUid) when newValue is null:
                         return true;
                 }
             }
-            // possible randomly response with null
+            // possible randomly respond with null
             if (whichPostType == typeof(ReplyPost)
                 && propName == nameof(ReplyPost.SignatureId)
                 && newValue is null && oldValue is not null) return true;
-            // possible rarely response with the protoBuf default value 0
+            // possible rarely respond with the protoBuf default value 0
             if (propName == nameof(IPost.AuthorUid)
                 && newValue is (long)0 && oldValue is not null) return true;
             return false;

crawler/src/Worker/ArchiveCrawlWorker.cs

@@ -8,7 +8,7 @@ namespace tbm.Crawler.Worker;
 public class ArchiveCrawlWorker : BackgroundService
 {
     // as of March 2019, tieba had restrict the max accepted value for page param of forum's threads api
-    // any request with page offset that larger than 10k threads will be response with results from the first page
+    // any request with page offset that larger than 10k threads will be respond with results from the first page
     private const int MaxCrawlablePage = 334; // 10k threads / 30 per request (from Rn param) = 333.3...
     private readonly ILogger<ArchiveCrawlWorker> _logger;
     private readonly ILifetimeScope _scope0;