Implemented image fetching from RSS Item

modules/RichSiteSummary/Service/RssFetch.cs

@@ -1,8 +1,13 @@
 using System;
 using System.Collections.Concurrent;
 using System.Collections.Generic;
+using System.Net;
+using System.Net.Http;
 using System.Threading;
 using System.Threading.Tasks;
+using System.Xml.Linq;
+using System.Text.RegularExpressions;
+using System.Linq;
 using CodeHollow.FeedReader;
 using Kruzya.TelegramBot.Core.Extensions;
 using Kruzya.TelegramBot.RichSiteSummary.Data;
@@ -14,6 +19,10 @@ using Telegram.Bot.Types;
 using Telegram.Bot.Types.Enums;
 using Feed = Kruzya.TelegramBot.RichSiteSummary.Data.Feed;
 
+using CodeHollow.FeedReader.Feeds;
+using CodeHollow.FeedReader.Feeds.Itunes;
+using CodeHollow.FeedReader.Feeds.MediaRSS;
+
 namespace Kruzya.TelegramBot.RichSiteSummary.Service
 {
     /// <summary>
@@ -123,6 +132,7 @@ namespace Kruzya.TelegramBot.RichSiteSummary.Service
             
             Post feedPost;
             var newPosts = new List<Post>();
+            var postsImages = new Dictionary<Post, string>();
             foreach (var post in parsedFeed.Items)
             {
                 feedPost = await dbContext.Posts.ByFeedAndUrl(post.Link, feed);
@@ -137,7 +147,8 @@ namespace Kruzya.TelegramBot.RichSiteSummary.Service
                 feedPost.Title = post.Title;
                 feedPost.Url = post.Link;
                 feedPost.PostedAt = post.PublishingDate.GetValueOrDefault(DateTime.Now);
-                
+
+                postsImages.Add(feedPost, await FetchImageUrl(post));
                 newPosts.Add(feedPost);
             }
 
@@ -152,7 +163,7 @@ namespace Kruzya.TelegramBot.RichSiteSummary.Service
                         var message = new UserMessage()
                         {
                             ChatId = new ChatId(subscriber.SubscriberId), DisableWebPagePreview = true,
-                            ParseMode = ParseMode.Html, Text = text
+                            ParseMode = ParseMode.Html, Text = text, ImageUrl = postsImages.GetValueOrDefault(post)
                         };
 
                         _queue.Enqueue(message);
@@ -174,7 +185,80 @@ namespace Kruzya.TelegramBot.RichSiteSummary.Service
                 return null;
             }
         }
-        
+
+        #endregion
+
+        #region Image parsing
+
+        private async Task<string> FetchImageUrl(FeedItem item)
+        {
+            var feedItem = item.SpecificItem;
+            if (feedItem is Rss20FeedItem)
+            {
+                var enclosure = ((Rss20FeedItem)feedItem).Enclosure;
+                if (enclosure != null && IsValidImage(enclosure.MediaType))
+                {
+                    return enclosure.Url;
+                }
+            }
+
+            return await FetchImageUrl(item.Description);
+        }
+
+        private async Task<string> FetchImageUrl(string description)
+        {
+            if (string.IsNullOrWhiteSpace(description))
+            {
+                return "";
+            }
+
+            using var httpClientHandler = new HttpClientHandler() { AllowAutoRedirect = true };
+            using var httpClient = new HttpClient(httpClientHandler);
+
+            var imgRegEx = new Regex(@"<img[^>]+>");
+            var srcRegEx = new Regex("src=\"([^\"]+)\"");
+            var matches = imgRegEx.Matches(description);
+
+            var img = "";
+            foreach (Match match in matches)
+            {
+                try
+                {
+                    var srcData = srcRegEx.Match(match.Value);
+                    var content = srcData.Groups[1].Value;
+
+                    var response = await httpClient.SendAsync(new HttpRequestMessage(HttpMethod.Head, content));
+                    if (response.StatusCode == HttpStatusCode.OK &&
+                        IsValidImage(response.Content.Headers.ContentType.MediaType))
+                    {
+                        img = content;
+                        break;
+                    }
+                }
+                catch (Exception e)
+                {
+                    _logger.LogError($"Caused error when fetching image for RSS post: {e.Message}");
+                }
+            }
+
+            return img;
+        }
+
+        /// <summary>
+        /// Checks if image is valid for Telegram Bot API.
+        /// </summary>
+        /// <param name="contentType">Content-type for received content</param>
+        /// <returns>True, if Telegram maybe can "use" this image, false if not.</returns>
+        private static bool IsValidImage(string contentType)
+        {
+            return new string[]
+            {
+                "image/jpeg",
+                "image/bmp",
+                "image/png"
+            }.Contains(contentType);
+        }
+
         #endregion
     }
 }