refactor: extract download logic and clean up caption handling

* feat: enhance download feature with caption support for Instagram

* feat: improve message handling with enhanced caching and error management

* feat: add hashtag removal utility and integrate it into caption formatting

* fix: bot sent caption without content from cache

* feat: refactor Instagram download URL handling to simplify video response structure

apps/bot/src/bot/features/download.ts

@@ -4,17 +4,157 @@ import { logHandle } from '../helpers/logging';
 import { TTL_URLS } from '@/config/redis';
 import { getInstagramDownloadUrl } from '@/utils/instagram';
 import { getRedisInstance } from '@/utils/redis';
+import { removeHashtags } from '@/utils/text';
 import { getTiktokDownloadUrl } from '@/utils/tiktok';
 import { validateInstagramUrl, validateTikTokUrl, validateYoutubeUrl } from '@/utils/urls';
 import { getYoutubeDownloadUrl } from '@/utils/youtube';
+import { code, expandableBlockquote, fmt } from '@grammyjs/parse-mode';
 import { Composer, InputFile } from 'grammy';
 import { cluster } from 'radashi';
 
 const composer = new Composer<Context>();
 const feature = composer.chatType('private');
-
 const redis = getRedisInstance();
 
+type DownloadResult = {
+  caption?: string;
+  imagesUrls?: string[];
+  videoUrl?: string;
+};
+
+async function checkCacheAndReply(context: Context, url: string) {
+  let contentMessageId: number | undefined;
+
+  const cachedVideoId = await redis.get(url);
+  if (cachedVideoId) {
+    const cachedMessage = await context.replyWithVideo(cachedVideoId);
+    contentMessageId = cachedMessage.message_id;
+  }
+
+  if (contentMessageId) {
+    const cachedCaption = await redis.get(`caption:${url}`);
+    if (cachedCaption) {
+      const { entities, text } = formatCaption(cachedCaption);
+
+      if (text.trim().length) {
+        await context.reply(text, {
+          entities,
+          reply_parameters: { message_id: contentMessageId },
+        });
+      }
+
+      return { captionSent: true, contentMessageId };
+    }
+  }
+
+  return { contentMessageId };
+}
+
+function formatCaption(caption: string) {
+  const cleanCaption = removeHashtags(caption);
+  return fmt`${expandableBlockquote} ${code} ${cleanCaption} ${code} ${expandableBlockquote}`;
+}
+
+async function getDownloadData(
+  url: string,
+  opts: {
+    isInstagram: boolean;
+    isTikTok: boolean;
+    isYoutube: boolean;
+  },
+): Promise<DownloadResult> {
+  const { isInstagram, isTikTok, isYoutube } = opts;
+
+  if (isTikTok) {
+    const result = await getTiktokDownloadUrl(url);
+    return {
+      caption: result.title,
+      imagesUrls: result.images,
+      videoUrl: result.play,
+    };
+  }
+
+  if (isInstagram) {
+    const result = await getInstagramDownloadUrl(url);
+    return {
+      caption: result.caption,
+      imagesUrls: result.images,
+      videoUrl: result.play,
+    };
+  }
+
+  if (isYoutube) {
+    const result = await getYoutubeDownloadUrl(url);
+    return {
+      videoUrl: result.play,
+    };
+  }
+
+  return {};
+}
+
+async function sendCaptionAndCache(
+  context: Context,
+  caption: string | undefined,
+  url: string,
+  contentMessageId?: number,
+) {
+  if (!caption) return;
+
+  const { entities, text } = formatCaption(caption);
+  await redis.set(`caption:${url}`, caption, 'EX', TTL_URLS);
+
+  if (text.trim().length) {
+    await context.reply(text, {
+      entities,
+      reply_parameters: contentMessageId ? { message_id: contentMessageId } : undefined,
+    });
+  }
+}
+
+async function sendImages(
+  context: Context,
+  imagesUrls: string[],
+  existingContentMessageId?: number,
+) {
+  if (!imagesUrls.length) return existingContentMessageId;
+
+  const chunks = cluster(imagesUrls, 10);
+  let contentMessageId = existingContentMessageId;
+
+  for (const chunk of chunks) {
+    const imageMessages = await context.replyWithMediaGroup(
+      chunk.map((imageUrl) => ({ media: imageUrl, type: 'photo' })),
+    );
+
+    if (!contentMessageId && imageMessages.length) {
+      contentMessageId = imageMessages.at(0)?.message_id;
+    }
+  }
+
+  return contentMessageId;
+}
+
+async function sendVideoAndCache(
+  context: Context,
+  videoUrl: string | undefined,
+  url: string,
+  existingContentMessageId?: number,
+) {
+  let contentMessageId = existingContentMessageId;
+
+  if (videoUrl && !contentMessageId) {
+    const { video, ...videoMessage } = await context.replyWithVideo(
+      new InputFile({ url: videoUrl }),
+    );
+
+    contentMessageId = videoMessage.message_id;
+    await redis.set(url, video.file_id, 'EX', TTL_URLS);
+  }
+
+  return contentMessageId;
+}
+
 feature.on('message:text', logHandle('download-message'), async (context) => {
   const url = context.message.text.trim();
 
@@ -22,36 +162,31 @@ feature.on('message:text', logHandle('download-message'), async (context) => {
   const isInstagram = validateInstagramUrl(url);
   const isYoutube = validateYoutubeUrl(url);
 
-  const isSupportedService = isTikTok || isInstagram || isYoutube;
-
-  if (!isSupportedService) {
+  if (!isTikTok && !isInstagram && !isYoutube) {
     return context.reply(context.t('err-invalid-url'));
   }
 
-  const cachedFileId = await redis.get(url);
-  if (cachedFileId) {
-    return context.replyWithVideo(cachedFileId);
-  }
+  const cacheResult = await checkCacheAndReply(context, url);
+  if (cacheResult.captionSent) return;
+
+  let contentMessageId = cacheResult.contentMessageId;
 
   let imagesUrls: string[] | undefined;
   let videoUrl: string | undefined;
+  let caption: string | undefined;
 
   try {
-    if (isTikTok) {
-      const result = await getTiktokDownloadUrl(url);
-      imagesUrls = result.images;
-      videoUrl = result.play;
-    } else if (isInstagram) {
-      const result = await getInstagramDownloadUrl(url);
-      imagesUrls = result.images;
-      videoUrl = result.play;
-    } else if (isYoutube) {
-      const result = await getYoutubeDownloadUrl(url);
-      videoUrl = result.play;
-    }
-  } catch (error_: unknown) {
-    const error = error_ as Error;
-    const message = error?.message ?? String(error);
+    const result = await getDownloadData(url, {
+      isInstagram,
+      isTikTok,
+      isYoutube,
+    });
+
+    imagesUrls = result.imagesUrls;
+    videoUrl = result.videoUrl;
+    caption = result.caption;
+  } catch (error: unknown) {
+    const message = (error as Error)?.message ?? String(error);
     if (typeof message === 'string' && message.startsWith('err-')) {
       return context.reply(context.t(message));
     }
@@ -63,21 +198,9 @@ feature.on('message:text', logHandle('download-message'), async (context) => {
     return context.reply(context.t('err-invalid-download-urls'));
   }
 
-  if (imagesUrls?.length) {
-    const chunks = cluster(imagesUrls, 10);
-    for (const chunk of chunks) {
-      await context.replyWithMediaGroup(
-        chunk.map((imageUrl) => ({ media: imageUrl, type: 'photo' })),
-      );
-    }
-
-    return;
-  }
-
-  if (videoUrl) {
-    const { video } = await context.replyWithVideo(new InputFile({ url: videoUrl }));
-    await redis.set(url, video.file_id, 'EX', TTL_URLS);
-  }
+  contentMessageId = await sendImages(context, imagesUrls ?? [], contentMessageId);
+  contentMessageId = await sendVideoAndCache(context, videoUrl, url, contentMessageId);
+  await sendCaptionAndCache(context, caption, url, contentMessageId);
 });
 
 export { composer as download };

apps/bot/src/constants/regex.ts

@@ -7,3 +7,5 @@ export const INSTAGRAM_URL_REGEX =
 
 export const YOUTUBE_URL_REGEX =
   /(?:https?:\/\/)?(?:www\.)?(?:youtu\.be\/|youtube\.com\/(?:watch\?v=|embed\/|v\/|shorts\/))([\w-]{11})(?:[?&]\S*)?/u;
+
+export const TAGS = /#[\p{L}\p{N}_]+/gu;

apps/bot/src/utils/instagram.ts

@@ -14,6 +14,7 @@ export type CarouselItem = {
 
 export type Root = {
   authorInfo: AuthorInfo;
+  caption: string;
   carouselItems: CarouselItem[];
   id: number;
   mediaUrls: string[];
@@ -33,12 +34,14 @@ export async function getInstagramDownloadUrl(url: string) {
 
   if (isVideo) {
     return {
+      caption: data.caption,
       images: [],
       play: data.mediaUrls.at(0),
     };
   }
 
   return {
+    caption: data.caption,
     images: data.mediaUrls,
     play: undefined,
   };

apps/bot/src/utils/text.ts

@@ -0,0 +1,3 @@
+export function removeHashtags(caption: string): string {
+  return caption.replaceAll(/#[\p{L}\p{N}_-]+/gu, '').trim();
+}