koztv-blog-tools 1.2.9 → 1.2.11

package/dist/index.d.mts

@@ -178,6 +178,18 @@ interface TelegramExportOptions {
     /** Callback when session string is generated (save this for future use) */
     onSession?: (session: string) => void;
 }
+interface ForwardInfo {
+    /** Original channel/user ID */
+    fromId?: number;
+    /** Original channel username (if channel) */
+    fromUsername?: string;
+    /** Original channel/user name */
+    fromName?: string;
+    /** Original post ID in source channel */
+    postId?: number;
+    /** Original post date */
+    date?: Date;
+}
 interface ExportedPost {
     msgId: number;
     date: Date;
@@ -190,6 +202,10 @@ interface ExportedPost {
     channelUsername: string;
     channelTitle: string;
     replyToMsgId?: number;
+    /** Message IDs merged into this post (for grouped posts) */
+    mergedMsgIds?: number[];
+    /** Forwarding info if this post was forwarded from another source */
+    fwdFrom?: ForwardInfo;
 }
 interface ExportResult {
     channelMeta: {
@@ -280,9 +296,34 @@ interface ExportAndTranslateResult {
     /** Session string for future use */
     session: string;
 }
+/**
+ * Options for processing posts from intermediate files
+ */
+interface ProcessFromFilesOptions {
+    /** Directory with intermediate export files (.telegram-export) */
+    exportDir: string;
+    /** Output directory for posts */
+    outputDir: string;
+    /** Translation config (optional) */
+    translate?: TranslationConfig;
+    /** Progress callback */
+    onProgress?: (message: string) => void;
+    /** Only process specific message IDs (if empty, process all) */
+    msgIds?: number[];
+    /** Force reprocess even if file exists */
+    force?: boolean;
+}
+/**
+ * Process posts from intermediate files (Step 2 only, no Telegram API)
+ * Use this after running exportTelegramChannel separately
+ */
+declare function processFromFiles(options: ProcessFromFilesOptions): Promise<{
+    processed: number;
+    skipped: number;
+}>;
 /**
  * Export posts from Telegram channel with optional translation
  */
 declare function exportAndTranslate(options: ExportAndTranslateOptions): Promise<ExportAndTranslateResult>;
 
-export { type AnalyticsConfig, type ExportAndTranslateOptions, type ExportAndTranslateResult, type ExportResult, type ExportedPost, type GoalName, type GoalParams, type GroupedPost, type ParsePostOptions, type Post, type TelegramExportOptions, type TranslateOptions, type TranslationConfig, categorizePost, cleanContent, configureAnalytics, deduplicatePosts, exportAndTranslate, exportTelegramChannel, extractAttachments, extractExcerpt, extractTitle, formatPostMarkdown, generateEnglishSlug, generateSlug, groupPosts, parsePost, resumeExport, trackBookAppointment, trackGoal, trackLearnMore, trackServiceClick, trackTelegramClick, translateContent, translateTitle };
+export { type AnalyticsConfig, type ExportAndTranslateOptions, type ExportAndTranslateResult, type ExportResult, type ExportedPost, type GoalName, type GoalParams, type GroupedPost, type ParsePostOptions, type Post, type ProcessFromFilesOptions, type TelegramExportOptions, type TranslateOptions, type TranslationConfig, categorizePost, cleanContent, configureAnalytics, deduplicatePosts, exportAndTranslate, exportTelegramChannel, extractAttachments, extractExcerpt, extractTitle, formatPostMarkdown, generateEnglishSlug, generateSlug, groupPosts, parsePost, processFromFiles, resumeExport, trackBookAppointment, trackGoal, trackLearnMore, trackServiceClick, trackTelegramClick, translateContent, translateTitle };

package/dist/index.d.ts

@@ -178,6 +178,18 @@ interface TelegramExportOptions {
     /** Callback when session string is generated (save this for future use) */
     onSession?: (session: string) => void;
 }
+interface ForwardInfo {
+    /** Original channel/user ID */
+    fromId?: number;
+    /** Original channel username (if channel) */
+    fromUsername?: string;
+    /** Original channel/user name */
+    fromName?: string;
+    /** Original post ID in source channel */
+    postId?: number;
+    /** Original post date */
+    date?: Date;
+}
 interface ExportedPost {
     msgId: number;
     date: Date;
@@ -190,6 +202,10 @@ interface ExportedPost {
     channelUsername: string;
     channelTitle: string;
     replyToMsgId?: number;
+    /** Message IDs merged into this post (for grouped posts) */
+    mergedMsgIds?: number[];
+    /** Forwarding info if this post was forwarded from another source */
+    fwdFrom?: ForwardInfo;
 }
 interface ExportResult {
     channelMeta: {
@@ -280,9 +296,34 @@ interface ExportAndTranslateResult {
     /** Session string for future use */
     session: string;
 }
+/**
+ * Options for processing posts from intermediate files
+ */
+interface ProcessFromFilesOptions {
+    /** Directory with intermediate export files (.telegram-export) */
+    exportDir: string;
+    /** Output directory for posts */
+    outputDir: string;
+    /** Translation config (optional) */
+    translate?: TranslationConfig;
+    /** Progress callback */
+    onProgress?: (message: string) => void;
+    /** Only process specific message IDs (if empty, process all) */
+    msgIds?: number[];
+    /** Force reprocess even if file exists */
+    force?: boolean;
+}
+/**
+ * Process posts from intermediate files (Step 2 only, no Telegram API)
+ * Use this after running exportTelegramChannel separately
+ */
+declare function processFromFiles(options: ProcessFromFilesOptions): Promise<{
+    processed: number;
+    skipped: number;
+}>;
 /**
  * Export posts from Telegram channel with optional translation
  */
 declare function exportAndTranslate(options: ExportAndTranslateOptions): Promise<ExportAndTranslateResult>;
 
-export { type AnalyticsConfig, type ExportAndTranslateOptions, type ExportAndTranslateResult, type ExportResult, type ExportedPost, type GoalName, type GoalParams, type GroupedPost, type ParsePostOptions, type Post, type TelegramExportOptions, type TranslateOptions, type TranslationConfig, categorizePost, cleanContent, configureAnalytics, deduplicatePosts, exportAndTranslate, exportTelegramChannel, extractAttachments, extractExcerpt, extractTitle, formatPostMarkdown, generateEnglishSlug, generateSlug, groupPosts, parsePost, resumeExport, trackBookAppointment, trackGoal, trackLearnMore, trackServiceClick, trackTelegramClick, translateContent, translateTitle };
+export { type AnalyticsConfig, type ExportAndTranslateOptions, type ExportAndTranslateResult, type ExportResult, type ExportedPost, type GoalName, type GoalParams, type GroupedPost, type ParsePostOptions, type Post, type ProcessFromFilesOptions, type TelegramExportOptions, type TranslateOptions, type TranslationConfig, categorizePost, cleanContent, configureAnalytics, deduplicatePosts, exportAndTranslate, exportTelegramChannel, extractAttachments, extractExcerpt, extractTitle, formatPostMarkdown, generateEnglishSlug, generateSlug, groupPosts, parsePost, processFromFiles, resumeExport, trackBookAppointment, trackGoal, trackLearnMore, trackServiceClick, trackTelegramClick, translateContent, translateTitle };

package/dist/index.js

@@ -44,6 +44,7 @@ __export(index_exports, {
   generateSlug: () => generateSlug,
   groupPosts: () => groupPosts,
   parsePost: () => parsePost,
+  processFromFiles: () => processFromFiles,
   resumeExport: () => resumeExport,
   trackBookAppointment: () => trackBookAppointment,
   trackGoal: () => trackGoal,
@@ -333,7 +334,7 @@ Do not add any explanations or notes.`
   let lastError = null;
   for (let attempt = 0; attempt < maxRetries; attempt++) {
     if (attempt > 0) {
-      const delay = 3e3 * Math.pow(2, attempt - 1);
+      const delay = 1e3 * Math.pow(2, attempt - 1);
       await new Promise((r) => setTimeout(r, delay));
     }
     const response = await fetch(endpoint, {
@@ -370,44 +371,123 @@ function generateEnglishSlug(title) {
 
 // src/telegram.ts
 var import_telegram = require("telegram");
-var import_sessions = require("telegram/sessions");
+var import_sessions = require("telegram/sessions/index.js");
 var fs = __toESM(require("fs"));
 var path = __toESM(require("path"));
 var readline = __toESM(require("readline"));
 function entitiesToMarkdown(text, entities) {
   if (!entities || entities.length === 0) return text;
-  const sorted = [...entities].sort((a, b) => b.offset - a.offset);
-  let result = text;
+  const mergedEntities = [];
+  const processedIndices = /* @__PURE__ */ new Set();
+  const links = [];
+  for (const entity of entities) {
+    if (entity instanceof import_telegram.Api.MessageEntityTextUrl) {
+      links.push({
+        start: entity.offset,
+        end: entity.offset + entity.length,
+        entity
+      });
+    }
+  }
+  const formatEntities = entities.filter(
+    (e) => e instanceof import_telegram.Api.MessageEntityBold || e instanceof import_telegram.Api.MessageEntityItalic
+  );
+  for (let i = 0; i < formatEntities.length; i++) {
+    if (processedIndices.has(entities.indexOf(formatEntities[i]))) continue;
+    const current = formatEntities[i];
+    const currentEnd = current.offset + current.length;
+    const bridgingLink = links.find((l) => l.start === currentEnd);
+    if (bridgingLink) {
+      const nextEntity = formatEntities.find(
+        (e) => e !== current && e.constructor === current.constructor && e.offset === bridgingLink.end
+      );
+      if (nextEntity) {
+        const mergedLength = nextEntity.offset + nextEntity.length - current.offset;
+        if (current instanceof import_telegram.Api.MessageEntityBold) {
+          mergedEntities.push(new import_telegram.Api.MessageEntityBold({ offset: current.offset, length: mergedLength }));
+        } else if (current instanceof import_telegram.Api.MessageEntityItalic) {
+          mergedEntities.push(new import_telegram.Api.MessageEntityItalic({ offset: current.offset, length: mergedLength }));
+        }
+        processedIndices.add(entities.indexOf(current));
+        processedIndices.add(entities.indexOf(nextEntity));
+        continue;
+      }
+    }
+    if (!processedIndices.has(entities.indexOf(current))) {
+      mergedEntities.push(current);
+      processedIndices.add(entities.indexOf(current));
+    }
+  }
+  for (let i = 0; i < entities.length; i++) {
+    if (processedIndices.has(i)) continue;
+    const entity = entities[i];
+    if (!(entity instanceof import_telegram.Api.MessageEntityBold) && !(entity instanceof import_telegram.Api.MessageEntityItalic)) {
+      mergedEntities.push(entity);
+    }
+  }
+  const sorted = [...mergedEntities].sort((a, b) => {
+    if (a.offset !== b.offset) return a.offset - b.offset;
+    return b.length - a.length;
+  });
+  const markers = [];
   for (const entity of sorted) {
     const start = entity.offset;
     const end = entity.offset + entity.length;
-    const content = result.substring(start, end);
-    let replacement = content;
+    let startMark = "";
+    let endMark = "";
+    let priority = 0;
     if (entity instanceof import_telegram.Api.MessageEntityBold) {
-      replacement = `**${content}**`;
+      startMark = "**";
+      endMark = "**";
+      priority = 1;
     } else if (entity instanceof import_telegram.Api.MessageEntityItalic) {
-      replacement = `*${content}*`;
+      startMark = "*";
+      endMark = "*";
+      priority = 1;
     } else if (entity instanceof import_telegram.Api.MessageEntityCode) {
-      replacement = `\`${content}\``;
+      startMark = "`";
+      endMark = "`";
+      priority = 2;
     } else if (entity instanceof import_telegram.Api.MessageEntityPre) {
-      replacement = `\`\`\`
-${content}
-\`\`\``;
+      startMark = "```\n";
+      endMark = "\n```";
+      priority = 2;
     } else if (entity instanceof import_telegram.Api.MessageEntityStrike) {
-      replacement = `~~${content}~~`;
+      startMark = "~~";
+      endMark = "~~";
+      priority = 1;
     } else if (entity instanceof import_telegram.Api.MessageEntityUnderline) {
-      replacement = `**${content}**`;
+      startMark = "**";
+      endMark = "**";
+      priority = 1;
     } else if (entity instanceof import_telegram.Api.MessageEntityTextUrl) {
-      replacement = `[${content}](${entity.url})`;
-    } else if (entity instanceof import_telegram.Api.MessageEntityUrl) {
-      replacement = content;
-    } else if (entity instanceof import_telegram.Api.MessageEntityMention) {
-      replacement = content;
-    } else if (entity instanceof import_telegram.Api.MessageEntityHashtag) {
-      replacement = content;
+      startMark = "[";
+      endMark = `](${entity.url})`;
+      priority = 10;
+    }
+    if (startMark) {
+      markers.push({ pos: start, insert: startMark, priority });
+      markers.push({ pos: end, insert: endMark, priority: -priority });
     }
-    result = result.substring(0, start) + replacement + result.substring(end);
   }
+  markers.sort((a, b) => {
+    if (a.pos !== b.pos) return a.pos - b.pos;
+    if (a.priority < 0 && b.priority < 0) {
+      return a.priority - b.priority;
+    }
+    if (a.priority > 0 && b.priority > 0) {
+      return b.priority - a.priority;
+    }
+    return a.priority - b.priority;
+  });
+  let result = "";
+  let lastPos = 0;
+  for (const marker of markers) {
+    result += text.substring(lastPos, marker.pos) + marker.insert;
+    lastPos = marker.pos;
+  }
+  result += text.substring(lastPos);
+  result = result.replace(/(\S) \*\*/g, "$1** ").replace(/\*\* (\S)/g, " **$1").replace(/(\S) \*/g, "$1* ").replace(/\* (\S)/g, " *$1");
   return result;
 }
 async function defaultReadline(prompt) {
@@ -579,6 +659,43 @@ async function exportTelegramChannel(options) {
     if (message.replyTo && "replyToMsgId" in message.replyTo) {
       replyToMsgId = message.replyTo.replyToMsgId;
     }
+    let fwdFrom;
+    if (message.fwdFrom) {
+      fwdFrom = {};
+      if (message.fwdFrom.date) {
+        fwdFrom.date = new Date(message.fwdFrom.date * 1e3);
+      }
+      const fromId = message.fwdFrom.fromId;
+      if (fromId) {
+        if (fromId instanceof import_telegram.Api.PeerChannel) {
+          fwdFrom.fromId = fromId.channelId.toJSNumber();
+          try {
+            const channelEntity = await client.getEntity(fromId.channelId);
+            if (channelEntity instanceof import_telegram.Api.Channel) {
+              fwdFrom.fromUsername = channelEntity.username || void 0;
+              fwdFrom.fromName = channelEntity.title;
+            }
+          } catch (e) {
+          }
+        } else if (fromId instanceof import_telegram.Api.PeerUser) {
+          fwdFrom.fromId = fromId.userId.toJSNumber();
+          try {
+            const userEntity = await client.getEntity(fromId.userId);
+            if (userEntity instanceof import_telegram.Api.User) {
+              fwdFrom.fromUsername = userEntity.username || void 0;
+              fwdFrom.fromName = [userEntity.firstName, userEntity.lastName].filter(Boolean).join(" ");
+            }
+          } catch (e) {
+          }
+        }
+      }
+      if (!fwdFrom.fromName && message.fwdFrom.fromName) {
+        fwdFrom.fromName = message.fwdFrom.fromName;
+      }
+      if (message.fwdFrom.channelPost) {
+        fwdFrom.postId = message.fwdFrom.channelPost;
+      }
+    }
     const post = {
       msgId,
       date: new Date(message.date * 1e3),
@@ -590,7 +707,8 @@ async function exportTelegramChannel(options) {
       link,
       channelUsername: channelMeta.username,
       channelTitle: channelMeta.title,
-      replyToMsgId
+      replyToMsgId,
+      fwdFrom
     };
     posts.push(post);
     const markdown = formatPostMarkdown(post);
@@ -628,6 +746,29 @@ forwards: ${post.forwards}`;
     frontmatter += `
 reply_to_msg_id: ${post.replyToMsgId}`;
   }
+  if (post.fwdFrom) {
+    frontmatter += "\nfwd_from:";
+    if (post.fwdFrom.fromName) {
+      frontmatter += `
+  from_name: "${post.fwdFrom.fromName.replace(/"/g, '\\"')}"`;
+    }
+    if (post.fwdFrom.fromUsername) {
+      frontmatter += `
+  from_username: "${post.fwdFrom.fromUsername}"`;
+    }
+    if (post.fwdFrom.fromId) {
+      frontmatter += `
+  from_id: ${post.fwdFrom.fromId}`;
+    }
+    if (post.fwdFrom.postId) {
+      frontmatter += `
+  post_id: ${post.fwdFrom.postId}`;
+    }
+    if (post.fwdFrom.date) {
+      frontmatter += `
+  date: ${post.fwdFrom.date.toISOString()}`;
+    }
+  }
   frontmatter += "\n---\n\n";
   let body = post.content || "";
   if (post.mediaFiles.length > 0) {
@@ -700,8 +841,68 @@ function generateSlug2(text, lang) {
   }
   return processed.toLowerCase().replace(/[^a-z0-9\s-]/g, "").replace(/\s+/g, "-").replace(/-+/g, "-").replace(/^-|-$/g, "").substring(0, 60) || "untitled";
 }
+function groupPostsByTime(posts, timeWindowMs = 5 * 60 * 1e3) {
+  const sortedPosts = [...posts].sort(
+    (a, b) => new Date(a.date).getTime() - new Date(b.date).getTime()
+  );
+  const groups = [];
+  let currentGroup = [];
+  let lastDate = null;
+  for (const post of sortedPosts) {
+    const postDate = new Date(post.date);
+    if (lastDate && Math.abs(postDate.getTime() - lastDate.getTime()) <= timeWindowMs) {
+      currentGroup.push(post);
+    } else {
+      if (currentGroup.length > 0) {
+        groups.push(currentGroup);
+      }
+      currentGroup = [post];
+    }
+    lastDate = postDate;
+  }
+  if (currentGroup.length > 0) {
+    groups.push(currentGroup);
+  }
+  return groups;
+}
+function mergePostGroup(posts) {
+  const sorted = [...posts].sort(
+    (a, b) => new Date(a.date).getTime() - new Date(b.date).getTime()
+  );
+  const mainPostIndex = sorted.findIndex((p) => !p.fwdFrom);
+  const mainPost = mainPostIndex >= 0 ? sorted[mainPostIndex] : sorted[0];
+  const contentParts = [];
+  for (const p of sorted) {
+    if (!p.content && !p.fwdFrom) continue;
+    if (p.fwdFrom && p === mainPost) {
+      contentParts.push(p.content || "");
+    } else if (p.fwdFrom) {
+      const fwd = p.fwdFrom;
+      const quotedContent = (p.content || "").split("\n").map((line) => `> ${line}`).join("\n");
+      const sourceLink = fwd.fromUsername && fwd.postId ? `https://t.me/${fwd.fromUsername}/${fwd.postId}` : fwd.fromUsername ? `https://t.me/${fwd.fromUsername}` : null;
+      const sourceName = fwd.fromName || fwd.fromUsername || "Unknown";
+      const attribution = sourceLink ? `\u2014 [${sourceName}](${sourceLink})` : `\u2014 ${sourceName}`;
+      contentParts.push(`${quotedContent}
+>
+> ${attribution}`);
+    } else {
+      contentParts.push(p.content || "");
+    }
+  }
+  const allMediaFiles = sorted.flatMap((p) => p.mediaFiles || []);
+  return {
+    ...mainPost,
+    content: contentParts.filter(Boolean).join("\n\n"),
+    hasMedia: sorted.some((p) => p.hasMedia),
+    mediaFiles: allMediaFiles,
+    // Store all msgIds for reference
+    mergedMsgIds: sorted.map((p) => p.msgId),
+    // Keep fwdFrom only if main post is forwarded
+    fwdFrom: mainPost.fwdFrom
+  };
+}
 function cleanContent2(text) {
-  return text.replace(/##\s*Attachments\n(?:- [^\n]+\n?)*/g, "").replace(/- media\/\d+\/[^\n]+/g, "").replace(/#\w+@\w+/g, "").trim();
+  return text.replace(/##\s*Attachments\n(?:- [^\n]+\n?)*/g, "").replace(/- media\/\d+\/[^\n]+/g, "").replace(/!\[\]\(\/media\/\d+\/[^)]+\)/g, "").replace(/<video src="\/media\/\d+\/[^"]+"\s*[^>]*><\/video>/g, "").replace(/#\w+@\w+/g, "").trim();
 }
 function extractTitleAndBody(content) {
   const lines = content.split("\n").filter((l) => l.trim());
@@ -716,39 +917,23 @@ async function processPost(post, options, exportDir) {
   const { title: originalTitle, body: originalBody } = extractTitleAndBody(cleanedContent);
   const date = post.date.toISOString().split("T")[0];
   const languages = [];
-  const translationExists = (lang) => {
-    const langDir = path2.join(outputDir, lang);
-    if (!fs2.existsSync(langDir)) return false;
-    try {
-      for (const slug of fs2.readdirSync(langDir)) {
-        const indexPath = path2.join(langDir, slug, "index.md");
-        if (fs2.existsSync(indexPath)) {
-          const content = fs2.readFileSync(indexPath, "utf-8");
-          if (content.includes(`original_link: "${post.link}"`)) {
-            return true;
-          }
-        }
+  const baseSlug = String(post.msgId);
+  const postDir = path2.join(outputDir, baseSlug);
+  const translationExistsInDir = (lang) => {
+    if (!fs2.existsSync(postDir)) return false;
+    const langFile = path2.join(postDir, `${lang}.md`);
+    if (fs2.existsSync(langFile)) {
+      const content = fs2.readFileSync(langFile, "utf-8");
+      if (content.includes(`original_link: "${post.link}"`)) {
+        return true;
       }
-    } catch {
     }
     return false;
   };
   if (translate && translate.targetLangs.length > 0) {
     const sourceLang = translate.sourceLang || "ru";
-    if (translate.keepOriginal && !translationExists(sourceLang)) {
-      languages.push({
-        lang: sourceLang,
-        title: originalTitle,
-        body: originalBody,
-        isOriginal: true
-      });
-    }
     for (const targetLang of translate.targetLangs) {
       if (targetLang === sourceLang) continue;
-      if (translationExists(targetLang)) {
-        onProgress?.(`  ${targetLang}: already exists, skipping`);
-        continue;
-      }
       onProgress?.(`  Translating to ${targetLang}...`);
       const translateOpts = {
         apiKey: translate.apiKey,
@@ -759,18 +944,34 @@ async function processPost(post, options, exportDir) {
       };
       try {
         const translatedTitle = await translateTitle(originalTitle, translateOpts);
+        if (originalBody) {
+          await new Promise((r) => setTimeout(r, 2e3));
+        }
         const translatedBody = originalBody ? await translateContent(originalBody, translateOpts) : "";
         languages.push({
           lang: targetLang,
           title: translatedTitle,
           body: translatedBody,
-          isOriginal: false
+          isOriginal: false,
+          slug: generateEnglishSlug(translatedTitle)
+          // slug from translated title for URL
         });
         onProgress?.(`  \u2192 ${targetLang}: "${translatedTitle.substring(0, 40)}..."`);
       } catch (error) {
         onProgress?.(`  Error translating to ${targetLang}: ${error.message}`);
       }
-      await new Promise((r) => setTimeout(r, 2e3));
+      await new Promise((r) => setTimeout(r, 5e3));
+    }
+    if (translate.keepOriginal) {
+      const russianSlug = generateSlug2(originalTitle);
+      languages.push({
+        lang: sourceLang,
+        title: originalTitle,
+        body: originalBody,
+        isOriginal: true,
+        slug: russianSlug
+        // Custom slug for Russian URL
+      });
     }
   } else {
     const defaultLang = translate?.sourceLang || "ru";
@@ -781,25 +982,21 @@ async function processPost(post, options, exportDir) {
       isOriginal: true
     });
   }
-  for (const { lang, title, body, isOriginal } of languages) {
-    const slug = generateSlug2(title, lang);
-    const postDir = path2.join(outputDir, lang, slug);
-    if (fs2.existsSync(postDir)) {
-      onProgress?.(`  Skipping existing: ${lang}/${slug}`);
+  for (const { lang, title, body, isOriginal, slug: customSlug } of languages) {
+    const langFile = path2.join(postDir, `${lang}.md`);
+    if (fs2.existsSync(langFile) || translationExistsInDir(lang)) {
+      onProgress?.(`  Skipping existing: ${baseSlug}/${lang}.md`);
       continue;
     }
     fs2.mkdirSync(postDir, { recursive: true });
     let finalBody = body;
     const mediaFiles = [];
     if (post.mediaFiles && post.mediaFiles.length > 0) {
-      const paddedId = String(post.msgId).padStart(6, "0");
-      const targetMediaDir = mediaDir ? path2.join(mediaDir, paddedId) : path2.join(outputDir, "..", "media", paddedId);
-      fs2.mkdirSync(targetMediaDir, { recursive: true });
       for (let i = 0; i < post.mediaFiles.length; i++) {
         const sourcePath = path2.join(exportDir, post.mediaFiles[i]);
         const ext = path2.extname(sourcePath).toLowerCase() || ".jpg";
         const newName = `image${i + 1}${ext}`;
-        const targetPath = path2.join(targetMediaDir, newName);
+        const targetPath = path2.join(postDir, newName);
         if (fs2.existsSync(sourcePath)) {
           if (!fs2.existsSync(targetPath)) {
             fs2.copyFileSync(sourcePath, targetPath);
@@ -810,8 +1007,8 @@ async function processPost(post, options, exportDir) {
       if (mediaFiles.length > 0) {
         const images = mediaFiles.filter((f) => !f.match(/\.(mp4|mov|webm|m4v)$/i));
         const videos = mediaFiles.filter((f) => f.match(/\.(mp4|mov|webm|m4v)$/i));
-        const imageMarkdown = images.map((f) => `![](/media/${paddedId}/${f})`).join("\n\n");
-        const videoMarkdown = videos.map((f) => `<video src="/media/${paddedId}/${f}" controls></video>`).join("\n\n");
+        const imageMarkdown = images.map((f) => `![](${f})`).join("\n\n");
+        const videoMarkdown = videos.map((f) => `<video src="${f}" controls></video>`).join("\n\n");
         const mediaMarkdown = [imageMarkdown, videoMarkdown].filter(Boolean).join("\n\n");
         if (mediaMarkdown) {
           finalBody = finalBody + "\n\n" + mediaMarkdown;
@@ -820,22 +1017,116 @@ async function processPost(post, options, exportDir) {
     }
     const replyLine = post.replyToMsgId ? `reply_to_msg_id: ${post.replyToMsgId}
 ` : "";
+    const slugLine = customSlug ? `slug: "${customSlug}"
+` : "";
+    let fwdFromLine = "";
+    let bodyWithForwardQuote = finalBody;
+    if (post.fwdFrom) {
+      const fwdFrom = post.fwdFrom;
+      fwdFromLine = "fwd_from:\n";
+      if (fwdFrom.fromName) {
+        fwdFromLine += `  from_name: "${fwdFrom.fromName.replace(/"/g, '\\"')}"
+`;
+      }
+      if (fwdFrom.fromUsername) {
+        fwdFromLine += `  from_username: "${fwdFrom.fromUsername}"
+`;
+      }
+      if (fwdFrom.postId) {
+        fwdFromLine += `  post_id: ${fwdFrom.postId}
+`;
+      }
+      const quotedBody = finalBody.split("\n").map((line) => `> ${line}`).join("\n");
+      const sourceLink = fwdFrom.fromUsername && fwdFrom.postId ? `https://t.me/${fwdFrom.fromUsername}/${fwdFrom.postId}` : fwdFrom.fromUsername ? `https://t.me/${fwdFrom.fromUsername}` : null;
+      const sourceName = fwdFrom.fromName || fwdFrom.fromUsername || "Unknown";
+      const attribution = sourceLink ? `\u2014 [${sourceName}](${sourceLink})` : `\u2014 ${sourceName}`;
+      bodyWithForwardQuote = `${quotedBody}
+>
+> ${attribution}`;
+    }
     const markdown = `---
 title: "${title.replace(/"/g, '\\"')}"
 date: ${date}
 lang: ${lang}
 original_link: "${post.link || ""}"
-${replyLine}${isOriginal ? "" : `translated_from: "${translate?.sourceLang || "ru"}"
+${slugLine}${replyLine}${fwdFromLine}${isOriginal ? "" : `translated_from: "${translate?.sourceLang || "ru"}"
 `}---
 
-${finalBody}
+${bodyWithForwardQuote}
 `;
-    fs2.writeFileSync(path2.join(postDir, "index.md"), markdown);
-    createdPosts.push(`${lang}/${slug}`);
-    onProgress?.(`  Created: ${lang}/${slug}`);
+    fs2.writeFileSync(langFile, markdown);
+    createdPosts.push(`${baseSlug}/${lang}.md`);
+    onProgress?.(`  Created: ${baseSlug}/${lang}.md`);
   }
   return createdPosts;
 }
+async function processFromFiles(options) {
+  const { exportDir, outputDir, translate, onProgress, msgIds, force } = options;
+  const ndjsonPath = path2.join(exportDir, "posts.ndjson");
+  if (!fs2.existsSync(ndjsonPath)) {
+    throw new Error(`posts.ndjson not found in ${exportDir}`);
+  }
+  fs2.mkdirSync(outputDir, { recursive: true });
+  const lines = fs2.readFileSync(ndjsonPath, "utf-8").split("\n").filter((l) => l.trim());
+  const posts = lines.map((line) => {
+    const data = JSON.parse(line);
+    return {
+      ...data,
+      date: new Date(data.date)
+    };
+  });
+  onProgress?.(`Found ${posts.length} posts in ${ndjsonPath}`);
+  const postGroups = groupPostsByTime(posts, 5 * 60 * 1e3);
+  onProgress?.(`Grouped into ${postGroups.length} post groups`);
+  const mergedPosts = postGroups.map(
+    (group) => group.length > 1 ? mergePostGroup(group) : group[0]
+  );
+  let processed = 0;
+  let skipped = 0;
+  for (const post of mergedPosts) {
+    const postMsgIds = post.mergedMsgIds || [post.msgId];
+    if (msgIds && msgIds.length > 0 && !postMsgIds.some((id) => msgIds.includes(id))) {
+      continue;
+    }
+    const postId = `${post.channelUsername}-${post.msgId}`;
+    if (!post.content && !post.hasMedia) {
+      skipped++;
+      continue;
+    }
+    onProgress?.(`Processing: ${postId}`);
+    const processOptions = {
+      apiId: 0,
+      apiHash: "",
+      channel: "",
+      outputDir,
+      translate,
+      onProgress
+    };
+    try {
+      if (force) {
+        const postDir = path2.join(outputDir, String(post.msgId));
+        if (fs2.existsSync(postDir)) {
+          for (const file of fs2.readdirSync(postDir)) {
+            if (file.endsWith(".md")) {
+              fs2.unlinkSync(path2.join(postDir, file));
+            }
+          }
+        }
+      }
+      const created = await processPost(post, processOptions, exportDir);
+      if (created.length > 0) {
+        processed++;
+      } else {
+        skipped++;
+      }
+    } catch (error) {
+      onProgress?.(`  Error: ${error.message}`);
+      skipped++;
+    }
+    await new Promise((r) => setTimeout(r, 50));
+  }
+  return { processed, skipped };
+}
 async function exportAndTranslate(options) {
   const {
     apiId,
@@ -871,9 +1162,14 @@ async function exportAndTranslate(options) {
   });
   onProgress?.(`  Exported ${exportResult.posts.length} posts from "${exportResult.channelMeta.title}"`);
   onProgress?.("\nStep 2: Processing posts...");
+  const postGroups = groupPostsByTime(exportResult.posts, 5 * 60 * 1e3);
+  onProgress?.(`  Grouped into ${postGroups.length} post groups`);
+  const mergedPosts = postGroups.map(
+    (group) => group.length > 1 ? mergePostGroup(group) : group[0]
+  );
   let processed = 0;
   let skipped = 0;
-  for (const post of exportResult.posts) {
+  for (const post of mergedPosts) {
     const postId = `${post.channelUsername}-${post.msgId}`;
     if (!post.content && !post.hasMedia) {
       skipped++;
@@ -925,6 +1221,7 @@ Processing: ${postId}`);
   generateSlug,
   groupPosts,
   parsePost,
+  processFromFiles,
   resumeExport,
   trackBookAppointment,
   trackGoal,

package/dist/index.mjs

@@ -276,7 +276,7 @@ Do not add any explanations or notes.`
   let lastError = null;
   for (let attempt = 0; attempt < maxRetries; attempt++) {
     if (attempt > 0) {
-      const delay = 3e3 * Math.pow(2, attempt - 1);
+      const delay = 1e3 * Math.pow(2, attempt - 1);
       await new Promise((r) => setTimeout(r, delay));
     }
     const response = await fetch(endpoint, {
@@ -313,44 +313,123 @@ function generateEnglishSlug(title) {
 
 // src/telegram.ts
 import { TelegramClient, Api } from "telegram";
-import { StringSession } from "telegram/sessions";
+import { StringSession } from "telegram/sessions/index.js";
 import * as fs from "fs";
 import * as path from "path";
 import * as readline from "readline";
 function entitiesToMarkdown(text, entities) {
   if (!entities || entities.length === 0) return text;
-  const sorted = [...entities].sort((a, b) => b.offset - a.offset);
-  let result = text;
+  const mergedEntities = [];
+  const processedIndices = /* @__PURE__ */ new Set();
+  const links = [];
+  for (const entity of entities) {
+    if (entity instanceof Api.MessageEntityTextUrl) {
+      links.push({
+        start: entity.offset,
+        end: entity.offset + entity.length,
+        entity
+      });
+    }
+  }
+  const formatEntities = entities.filter(
+    (e) => e instanceof Api.MessageEntityBold || e instanceof Api.MessageEntityItalic
+  );
+  for (let i = 0; i < formatEntities.length; i++) {
+    if (processedIndices.has(entities.indexOf(formatEntities[i]))) continue;
+    const current = formatEntities[i];
+    const currentEnd = current.offset + current.length;
+    const bridgingLink = links.find((l) => l.start === currentEnd);
+    if (bridgingLink) {
+      const nextEntity = formatEntities.find(
+        (e) => e !== current && e.constructor === current.constructor && e.offset === bridgingLink.end
+      );
+      if (nextEntity) {
+        const mergedLength = nextEntity.offset + nextEntity.length - current.offset;
+        if (current instanceof Api.MessageEntityBold) {
+          mergedEntities.push(new Api.MessageEntityBold({ offset: current.offset, length: mergedLength }));
+        } else if (current instanceof Api.MessageEntityItalic) {
+          mergedEntities.push(new Api.MessageEntityItalic({ offset: current.offset, length: mergedLength }));
+        }
+        processedIndices.add(entities.indexOf(current));
+        processedIndices.add(entities.indexOf(nextEntity));
+        continue;
+      }
+    }
+    if (!processedIndices.has(entities.indexOf(current))) {
+      mergedEntities.push(current);
+      processedIndices.add(entities.indexOf(current));
+    }
+  }
+  for (let i = 0; i < entities.length; i++) {
+    if (processedIndices.has(i)) continue;
+    const entity = entities[i];
+    if (!(entity instanceof Api.MessageEntityBold) && !(entity instanceof Api.MessageEntityItalic)) {
+      mergedEntities.push(entity);
+    }
+  }
+  const sorted = [...mergedEntities].sort((a, b) => {
+    if (a.offset !== b.offset) return a.offset - b.offset;
+    return b.length - a.length;
+  });
+  const markers = [];
   for (const entity of sorted) {
     const start = entity.offset;
     const end = entity.offset + entity.length;
-    const content = result.substring(start, end);
-    let replacement = content;
+    let startMark = "";
+    let endMark = "";
+    let priority = 0;
     if (entity instanceof Api.MessageEntityBold) {
-      replacement = `**${content}**`;
+      startMark = "**";
+      endMark = "**";
+      priority = 1;
     } else if (entity instanceof Api.MessageEntityItalic) {
-      replacement = `*${content}*`;
+      startMark = "*";
+      endMark = "*";
+      priority = 1;
     } else if (entity instanceof Api.MessageEntityCode) {
-      replacement = `\`${content}\``;
+      startMark = "`";
+      endMark = "`";
+      priority = 2;
     } else if (entity instanceof Api.MessageEntityPre) {
-      replacement = `\`\`\`
-${content}
-\`\`\``;
+      startMark = "```\n";
+      endMark = "\n```";
+      priority = 2;
     } else if (entity instanceof Api.MessageEntityStrike) {
-      replacement = `~~${content}~~`;
+      startMark = "~~";
+      endMark = "~~";
+      priority = 1;
     } else if (entity instanceof Api.MessageEntityUnderline) {
-      replacement = `**${content}**`;
+      startMark = "**";
+      endMark = "**";
+      priority = 1;
     } else if (entity instanceof Api.MessageEntityTextUrl) {
-      replacement = `[${content}](${entity.url})`;
-    } else if (entity instanceof Api.MessageEntityUrl) {
-      replacement = content;
-    } else if (entity instanceof Api.MessageEntityMention) {
-      replacement = content;
-    } else if (entity instanceof Api.MessageEntityHashtag) {
-      replacement = content;
+      startMark = "[";
+      endMark = `](${entity.url})`;
+      priority = 10;
+    }
+    if (startMark) {
+      markers.push({ pos: start, insert: startMark, priority });
+      markers.push({ pos: end, insert: endMark, priority: -priority });
     }
-    result = result.substring(0, start) + replacement + result.substring(end);
   }
+  markers.sort((a, b) => {
+    if (a.pos !== b.pos) return a.pos - b.pos;
+    if (a.priority < 0 && b.priority < 0) {
+      return a.priority - b.priority;
+    }
+    if (a.priority > 0 && b.priority > 0) {
+      return b.priority - a.priority;
+    }
+    return a.priority - b.priority;
+  });
+  let result = "";
+  let lastPos = 0;
+  for (const marker of markers) {
+    result += text.substring(lastPos, marker.pos) + marker.insert;
+    lastPos = marker.pos;
+  }
+  result += text.substring(lastPos);
+  result = result.replace(/(\S) \*\*/g, "$1** ").replace(/\*\* (\S)/g, " **$1").replace(/(\S) \*/g, "$1* ").replace(/\* (\S)/g, " *$1");
   return result;
 }
 async function defaultReadline(prompt) {
@@ -522,6 +601,43 @@ async function exportTelegramChannel(options) {
     if (message.replyTo && "replyToMsgId" in message.replyTo) {
       replyToMsgId = message.replyTo.replyToMsgId;
     }
+    let fwdFrom;
+    if (message.fwdFrom) {
+      fwdFrom = {};
+      if (message.fwdFrom.date) {
+        fwdFrom.date = new Date(message.fwdFrom.date * 1e3);
+      }
+      const fromId = message.fwdFrom.fromId;
+      if (fromId) {
+        if (fromId instanceof Api.PeerChannel) {
+          fwdFrom.fromId = fromId.channelId.toJSNumber();
+          try {
+            const channelEntity = await client.getEntity(fromId.channelId);
+            if (channelEntity instanceof Api.Channel) {
+              fwdFrom.fromUsername = channelEntity.username || void 0;
+              fwdFrom.fromName = channelEntity.title;
+            }
+          } catch (e) {
+          }
+        } else if (fromId instanceof Api.PeerUser) {
+          fwdFrom.fromId = fromId.userId.toJSNumber();
+          try {
+            const userEntity = await client.getEntity(fromId.userId);
+            if (userEntity instanceof Api.User) {
+              fwdFrom.fromUsername = userEntity.username || void 0;
+              fwdFrom.fromName = [userEntity.firstName, userEntity.lastName].filter(Boolean).join(" ");
+            }
+          } catch (e) {
+          }
+        }
+      }
+      if (!fwdFrom.fromName && message.fwdFrom.fromName) {
+        fwdFrom.fromName = message.fwdFrom.fromName;
+      }
+      if (message.fwdFrom.channelPost) {
+        fwdFrom.postId = message.fwdFrom.channelPost;
+      }
+    }
     const post = {
       msgId,
       date: new Date(message.date * 1e3),
@@ -533,7 +649,8 @@ async function exportTelegramChannel(options) {
       link,
       channelUsername: channelMeta.username,
       channelTitle: channelMeta.title,
-      replyToMsgId
+      replyToMsgId,
+      fwdFrom
     };
     posts.push(post);
     const markdown = formatPostMarkdown(post);
@@ -571,6 +688,29 @@ forwards: ${post.forwards}`;
     frontmatter += `
 reply_to_msg_id: ${post.replyToMsgId}`;
   }
+  if (post.fwdFrom) {
+    frontmatter += "\nfwd_from:";
+    if (post.fwdFrom.fromName) {
+      frontmatter += `
+  from_name: "${post.fwdFrom.fromName.replace(/"/g, '\\"')}"`;
+    }
+    if (post.fwdFrom.fromUsername) {
+      frontmatter += `
+  from_username: "${post.fwdFrom.fromUsername}"`;
+    }
+    if (post.fwdFrom.fromId) {
+      frontmatter += `
+  from_id: ${post.fwdFrom.fromId}`;
+    }
+    if (post.fwdFrom.postId) {
+      frontmatter += `
+  post_id: ${post.fwdFrom.postId}`;
+    }
+    if (post.fwdFrom.date) {
+      frontmatter += `
+  date: ${post.fwdFrom.date.toISOString()}`;
+    }
+  }
   frontmatter += "\n---\n\n";
   let body = post.content || "";
   if (post.mediaFiles.length > 0) {
@@ -643,8 +783,68 @@ function generateSlug2(text, lang) {
   }
   return processed.toLowerCase().replace(/[^a-z0-9\s-]/g, "").replace(/\s+/g, "-").replace(/-+/g, "-").replace(/^-|-$/g, "").substring(0, 60) || "untitled";
 }
+function groupPostsByTime(posts, timeWindowMs = 5 * 60 * 1e3) {
+  const sortedPosts = [...posts].sort(
+    (a, b) => new Date(a.date).getTime() - new Date(b.date).getTime()
+  );
+  const groups = [];
+  let currentGroup = [];
+  let lastDate = null;
+  for (const post of sortedPosts) {
+    const postDate = new Date(post.date);
+    if (lastDate && Math.abs(postDate.getTime() - lastDate.getTime()) <= timeWindowMs) {
+      currentGroup.push(post);
+    } else {
+      if (currentGroup.length > 0) {
+        groups.push(currentGroup);
+      }
+      currentGroup = [post];
+    }
+    lastDate = postDate;
+  }
+  if (currentGroup.length > 0) {
+    groups.push(currentGroup);
+  }
+  return groups;
+}
+function mergePostGroup(posts) {
+  const sorted = [...posts].sort(
+    (a, b) => new Date(a.date).getTime() - new Date(b.date).getTime()
+  );
+  const mainPostIndex = sorted.findIndex((p) => !p.fwdFrom);
+  const mainPost = mainPostIndex >= 0 ? sorted[mainPostIndex] : sorted[0];
+  const contentParts = [];
+  for (const p of sorted) {
+    if (!p.content && !p.fwdFrom) continue;
+    if (p.fwdFrom && p === mainPost) {
+      contentParts.push(p.content || "");
+    } else if (p.fwdFrom) {
+      const fwd = p.fwdFrom;
+      const quotedContent = (p.content || "").split("\n").map((line) => `> ${line}`).join("\n");
+      const sourceLink = fwd.fromUsername && fwd.postId ? `https://t.me/${fwd.fromUsername}/${fwd.postId}` : fwd.fromUsername ? `https://t.me/${fwd.fromUsername}` : null;
+      const sourceName = fwd.fromName || fwd.fromUsername || "Unknown";
+      const attribution = sourceLink ? `\u2014 [${sourceName}](${sourceLink})` : `\u2014 ${sourceName}`;
+      contentParts.push(`${quotedContent}
+>
+> ${attribution}`);
+    } else {
+      contentParts.push(p.content || "");
+    }
+  }
+  const allMediaFiles = sorted.flatMap((p) => p.mediaFiles || []);
+  return {
+    ...mainPost,
+    content: contentParts.filter(Boolean).join("\n\n"),
+    hasMedia: sorted.some((p) => p.hasMedia),
+    mediaFiles: allMediaFiles,
+    // Store all msgIds for reference
+    mergedMsgIds: sorted.map((p) => p.msgId),
+    // Keep fwdFrom only if main post is forwarded
+    fwdFrom: mainPost.fwdFrom
+  };
+}
 function cleanContent2(text) {
-  return text.replace(/##\s*Attachments\n(?:- [^\n]+\n?)*/g, "").replace(/- media\/\d+\/[^\n]+/g, "").replace(/#\w+@\w+/g, "").trim();
+  return text.replace(/##\s*Attachments\n(?:- [^\n]+\n?)*/g, "").replace(/- media\/\d+\/[^\n]+/g, "").replace(/!\[\]\(\/media\/\d+\/[^)]+\)/g, "").replace(/<video src="\/media\/\d+\/[^"]+"\s*[^>]*><\/video>/g, "").replace(/#\w+@\w+/g, "").trim();
 }
 function extractTitleAndBody(content) {
   const lines = content.split("\n").filter((l) => l.trim());
@@ -659,39 +859,23 @@ async function processPost(post, options, exportDir) {
   const { title: originalTitle, body: originalBody } = extractTitleAndBody(cleanedContent);
   const date = post.date.toISOString().split("T")[0];
   const languages = [];
-  const translationExists = (lang) => {
-    const langDir = path2.join(outputDir, lang);
-    if (!fs2.existsSync(langDir)) return false;
-    try {
-      for (const slug of fs2.readdirSync(langDir)) {
-        const indexPath = path2.join(langDir, slug, "index.md");
-        if (fs2.existsSync(indexPath)) {
-          const content = fs2.readFileSync(indexPath, "utf-8");
-          if (content.includes(`original_link: "${post.link}"`)) {
-            return true;
-          }
-        }
+  const baseSlug = String(post.msgId);
+  const postDir = path2.join(outputDir, baseSlug);
+  const translationExistsInDir = (lang) => {
+    if (!fs2.existsSync(postDir)) return false;
+    const langFile = path2.join(postDir, `${lang}.md`);
+    if (fs2.existsSync(langFile)) {
+      const content = fs2.readFileSync(langFile, "utf-8");
+      if (content.includes(`original_link: "${post.link}"`)) {
+        return true;
       }
-    } catch {
     }
     return false;
   };
   if (translate && translate.targetLangs.length > 0) {
     const sourceLang = translate.sourceLang || "ru";
-    if (translate.keepOriginal && !translationExists(sourceLang)) {
-      languages.push({
-        lang: sourceLang,
-        title: originalTitle,
-        body: originalBody,
-        isOriginal: true
-      });
-    }
     for (const targetLang of translate.targetLangs) {
       if (targetLang === sourceLang) continue;
-      if (translationExists(targetLang)) {
-        onProgress?.(`  ${targetLang}: already exists, skipping`);
-        continue;
-      }
       onProgress?.(`  Translating to ${targetLang}...`);
       const translateOpts = {
         apiKey: translate.apiKey,
@@ -702,18 +886,34 @@ async function processPost(post, options, exportDir) {
       };
       try {
         const translatedTitle = await translateTitle(originalTitle, translateOpts);
+        if (originalBody) {
+          await new Promise((r) => setTimeout(r, 2e3));
+        }
         const translatedBody = originalBody ? await translateContent(originalBody, translateOpts) : "";
         languages.push({
           lang: targetLang,
           title: translatedTitle,
           body: translatedBody,
-          isOriginal: false
+          isOriginal: false,
+          slug: generateEnglishSlug(translatedTitle)
+          // slug from translated title for URL
         });
         onProgress?.(`  \u2192 ${targetLang}: "${translatedTitle.substring(0, 40)}..."`);
       } catch (error) {
         onProgress?.(`  Error translating to ${targetLang}: ${error.message}`);
       }
-      await new Promise((r) => setTimeout(r, 2e3));
+      await new Promise((r) => setTimeout(r, 5e3));
+    }
+    if (translate.keepOriginal) {
+      const russianSlug = generateSlug2(originalTitle);
+      languages.push({
+        lang: sourceLang,
+        title: originalTitle,
+        body: originalBody,
+        isOriginal: true,
+        slug: russianSlug
+        // Custom slug for Russian URL
+      });
     }
   } else {
     const defaultLang = translate?.sourceLang || "ru";
@@ -724,25 +924,21 @@ async function processPost(post, options, exportDir) {
       isOriginal: true
     });
   }
-  for (const { lang, title, body, isOriginal } of languages) {
-    const slug = generateSlug2(title, lang);
-    const postDir = path2.join(outputDir, lang, slug);
-    if (fs2.existsSync(postDir)) {
-      onProgress?.(`  Skipping existing: ${lang}/${slug}`);
+  for (const { lang, title, body, isOriginal, slug: customSlug } of languages) {
+    const langFile = path2.join(postDir, `${lang}.md`);
+    if (fs2.existsSync(langFile) || translationExistsInDir(lang)) {
+      onProgress?.(`  Skipping existing: ${baseSlug}/${lang}.md`);
       continue;
     }
     fs2.mkdirSync(postDir, { recursive: true });
     let finalBody = body;
     const mediaFiles = [];
     if (post.mediaFiles && post.mediaFiles.length > 0) {
-      const paddedId = String(post.msgId).padStart(6, "0");
-      const targetMediaDir = mediaDir ? path2.join(mediaDir, paddedId) : path2.join(outputDir, "..", "media", paddedId);
-      fs2.mkdirSync(targetMediaDir, { recursive: true });
       for (let i = 0; i < post.mediaFiles.length; i++) {
         const sourcePath = path2.join(exportDir, post.mediaFiles[i]);
         const ext = path2.extname(sourcePath).toLowerCase() || ".jpg";
         const newName = `image${i + 1}${ext}`;
-        const targetPath = path2.join(targetMediaDir, newName);
+        const targetPath = path2.join(postDir, newName);
         if (fs2.existsSync(sourcePath)) {
           if (!fs2.existsSync(targetPath)) {
             fs2.copyFileSync(sourcePath, targetPath);
@@ -753,8 +949,8 @@ async function processPost(post, options, exportDir) {
       if (mediaFiles.length > 0) {
         const images = mediaFiles.filter((f) => !f.match(/\.(mp4|mov|webm|m4v)$/i));
         const videos = mediaFiles.filter((f) => f.match(/\.(mp4|mov|webm|m4v)$/i));
-        const imageMarkdown = images.map((f) => `![](/media/${paddedId}/${f})`).join("\n\n");
-        const videoMarkdown = videos.map((f) => `<video src="/media/${paddedId}/${f}" controls></video>`).join("\n\n");
+        const imageMarkdown = images.map((f) => `![](${f})`).join("\n\n");
+        const videoMarkdown = videos.map((f) => `<video src="${f}" controls></video>`).join("\n\n");
         const mediaMarkdown = [imageMarkdown, videoMarkdown].filter(Boolean).join("\n\n");
         if (mediaMarkdown) {
           finalBody = finalBody + "\n\n" + mediaMarkdown;
@@ -763,22 +959,116 @@ async function processPost(post, options, exportDir) {
     }
     const replyLine = post.replyToMsgId ? `reply_to_msg_id: ${post.replyToMsgId}
 ` : "";
+    const slugLine = customSlug ? `slug: "${customSlug}"
+` : "";
+    let fwdFromLine = "";
+    let bodyWithForwardQuote = finalBody;
+    if (post.fwdFrom) {
+      const fwdFrom = post.fwdFrom;
+      fwdFromLine = "fwd_from:\n";
+      if (fwdFrom.fromName) {
+        fwdFromLine += `  from_name: "${fwdFrom.fromName.replace(/"/g, '\\"')}"
+`;
+      }
+      if (fwdFrom.fromUsername) {
+        fwdFromLine += `  from_username: "${fwdFrom.fromUsername}"
+`;
+      }
+      if (fwdFrom.postId) {
+        fwdFromLine += `  post_id: ${fwdFrom.postId}
+`;
+      }
+      const quotedBody = finalBody.split("\n").map((line) => `> ${line}`).join("\n");
+      const sourceLink = fwdFrom.fromUsername && fwdFrom.postId ? `https://t.me/${fwdFrom.fromUsername}/${fwdFrom.postId}` : fwdFrom.fromUsername ? `https://t.me/${fwdFrom.fromUsername}` : null;
+      const sourceName = fwdFrom.fromName || fwdFrom.fromUsername || "Unknown";
+      const attribution = sourceLink ? `\u2014 [${sourceName}](${sourceLink})` : `\u2014 ${sourceName}`;
+      bodyWithForwardQuote = `${quotedBody}
+>
+> ${attribution}`;
+    }
     const markdown = `---
 title: "${title.replace(/"/g, '\\"')}"
 date: ${date}
 lang: ${lang}
 original_link: "${post.link || ""}"
-${replyLine}${isOriginal ? "" : `translated_from: "${translate?.sourceLang || "ru"}"
+${slugLine}${replyLine}${fwdFromLine}${isOriginal ? "" : `translated_from: "${translate?.sourceLang || "ru"}"
 `}---
 
-${finalBody}
+${bodyWithForwardQuote}
 `;
-    fs2.writeFileSync(path2.join(postDir, "index.md"), markdown);
-    createdPosts.push(`${lang}/${slug}`);
-    onProgress?.(`  Created: ${lang}/${slug}`);
+    fs2.writeFileSync(langFile, markdown);
+    createdPosts.push(`${baseSlug}/${lang}.md`);
+    onProgress?.(`  Created: ${baseSlug}/${lang}.md`);
   }
   return createdPosts;
 }
+async function processFromFiles(options) {
+  const { exportDir, outputDir, translate, onProgress, msgIds, force } = options;
+  const ndjsonPath = path2.join(exportDir, "posts.ndjson");
+  if (!fs2.existsSync(ndjsonPath)) {
+    throw new Error(`posts.ndjson not found in ${exportDir}`);
+  }
+  fs2.mkdirSync(outputDir, { recursive: true });
+  const lines = fs2.readFileSync(ndjsonPath, "utf-8").split("\n").filter((l) => l.trim());
+  const posts = lines.map((line) => {
+    const data = JSON.parse(line);
+    return {
+      ...data,
+      date: new Date(data.date)
+    };
+  });
+  onProgress?.(`Found ${posts.length} posts in ${ndjsonPath}`);
+  const postGroups = groupPostsByTime(posts, 5 * 60 * 1e3);
+  onProgress?.(`Grouped into ${postGroups.length} post groups`);
+  const mergedPosts = postGroups.map(
+    (group) => group.length > 1 ? mergePostGroup(group) : group[0]
+  );
+  let processed = 0;
+  let skipped = 0;
+  for (const post of mergedPosts) {
+    const postMsgIds = post.mergedMsgIds || [post.msgId];
+    if (msgIds && msgIds.length > 0 && !postMsgIds.some((id) => msgIds.includes(id))) {
+      continue;
+    }
+    const postId = `${post.channelUsername}-${post.msgId}`;
+    if (!post.content && !post.hasMedia) {
+      skipped++;
+      continue;
+    }
+    onProgress?.(`Processing: ${postId}`);
+    const processOptions = {
+      apiId: 0,
+      apiHash: "",
+      channel: "",
+      outputDir,
+      translate,
+      onProgress
+    };
+    try {
+      if (force) {
+        const postDir = path2.join(outputDir, String(post.msgId));
+        if (fs2.existsSync(postDir)) {
+          for (const file of fs2.readdirSync(postDir)) {
+            if (file.endsWith(".md")) {
+              fs2.unlinkSync(path2.join(postDir, file));
+            }
+          }
+        }
+      }
+      const created = await processPost(post, processOptions, exportDir);
+      if (created.length > 0) {
+        processed++;
+      } else {
+        skipped++;
+      }
+    } catch (error) {
+      onProgress?.(`  Error: ${error.message}`);
+      skipped++;
+    }
+    await new Promise((r) => setTimeout(r, 50));
+  }
+  return { processed, skipped };
+}
 async function exportAndTranslate(options) {
   const {
     apiId,
@@ -814,9 +1104,14 @@ async function exportAndTranslate(options) {
   });
   onProgress?.(`  Exported ${exportResult.posts.length} posts from "${exportResult.channelMeta.title}"`);
   onProgress?.("\nStep 2: Processing posts...");
+  const postGroups = groupPostsByTime(exportResult.posts, 5 * 60 * 1e3);
+  onProgress?.(`  Grouped into ${postGroups.length} post groups`);
+  const mergedPosts = postGroups.map(
+    (group) => group.length > 1 ? mergePostGroup(group) : group[0]
+  );
   let processed = 0;
   let skipped = 0;
-  for (const post of exportResult.posts) {
+  for (const post of mergedPosts) {
     const postId = `${post.channelUsername}-${post.msgId}`;
     if (!post.content && !post.hasMedia) {
       skipped++;
@@ -867,6 +1162,7 @@ export {
   generateSlug,
   groupPosts,
   parsePost,
+  processFromFiles,
   resumeExport,
   trackBookAppointment,
   trackGoal,

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "koztv-blog-tools",
-  "version": "1.2.9",
+  "version": "1.2.11",
   "description": "Shared utilities for Telegram-based blog sites",
   "main": "dist/index.js",
   "module": "dist/index.mjs",