Skip to content

AI-native knowledge base & Evernote/Obsidian alternative with instant sync & native MCP. Free on Cloudflare or self-hosted via Docker.

feat(import): enhance WeChat chat note formatting and image layout

tianma-if628245d2c7d8
diff --git a/apps/desktop/src/main/wechat-chat-archive.mjs b/apps/desktop/src/main/wechat-chat-archive.mjs
index 4192a2c4..dd9cdb29 100644
--- a/apps/desktop/src/main/wechat-chat-archive.mjs
+++ b/apps/desktop/src/main/wechat-chat-archive.mjs
@@ -193,6 +193,135 @@ const parseMessages = (text) => {
return sections;
};
+const WECHAT_EMOJI_MAP = {
+ "[微笑]": "😊",
+ "[撇嘴]": "🥺",
+ "[色]": "😍",
+ "[发呆]": "😶",
+ "[得意]": "😎",
+ "[流泪]": "😭",
+ "[害羞]": "😳",
+ "[闭嘴]": "🤐",
+ "[睡]": "😴",
+ "[大哭]": "😭",
+ "[尴尬]": "😅",
+ "[发怒]": "😡",
+ "[调皮]": "😜",
+ "[呲牙]": "😁",
+ "[惊讶]": "😲",
+ "[难过]": "🙁",
+ "[抓狂]": "😫",
+ "[吐]": "🤮",
+ "[偷笑]": "🤭",
+ "[愉快]": "😄",
+ "[白眼]": "🙄",
+ "[傲慢]": "😤",
+ "[困]": "🥱",
+ "[惊恐]": "😱",
+ "[憨笑]": "😄",
+ "[悠闲]": "😌",
+ "[咒骂]": "🤬",
+ "[疑问]": "❓",
+ "[嘘]": "🤫",
+ "[晕]": "😵",
+ "[衰]": "😞",
+ "[骷髅]": "💀",
+ "[敲打]": "🔨",
+ "[再见]": "👋",
+ "[擦汗]": "😓",
+ "[抠鼻]": "🤏",
+ "[鼓掌]": "👏",
+ "[坏笑]": "😏",
+ "[左哼哼]": "😤",
+ "[右哼哼]": "😤",
+ "[鄙视]": "😒",
+ "[委屈]": "🥺",
+ "[快哭了]": "😢",
+ "[阴险]": "😏",
+ "[亲亲]": "😘",
+ "[可怜]": "🥺",
+ "[笑脸]": "🙂",
+ "[生病]": "🤒",
+ "[脸红]": "😳",
+ "[破涕为笑]": "😂",
+ "[恐惧]": "😨",
+ "[失望]": "😞",
+ "[无语]": "🙄",
+ "[嘿哈]": "🤣",
+ "[捂脸]": "🤦",
+ "[奸笑]": "😏",
+ "[机智]": "🤓",
+ "[皱眉]": "😟",
+ "[耶]": "✌️",
+ "[吃瓜]": "🍉",
+ "[加油]": "💪",
+ "[汗]": "😅",
+ "[天啊]": "😱",
+ "[Emm]": "🤔",
+ "[社会社会]": "🤝",
+ "[旺柴]": "🐕",
+ "[好的]": "👌",
+ "[打脸]": "🤦",
+ "[加油加油]": "🎉",
+ "[哇]": "🤩",
+ "[翻白眼]": "🙄",
+ "[666]": "🤙",
+ "[让我看看]": "👀",
+ "[叹气]": "😮‍💨",
+ "[苦涩]": "🥲",
+ "[裂开]": "💔",
+ "[嘴唇]": "💋",
+ "[爱心]": "❤️",
+ "[心碎]": "💔",
+ "[蛋糕]": "🎂",
+ "[炸弹]": "💣",
+ "[便便]": "💩",
+ "[月亮]": "🌙",
+ "[太阳]": "☀️",
+ "[拥抱]": "🫂",
+ "[强]": "👍",
+ "[弱]": "👎",
+ "[握手]": "🤝",
+ "[胜利]": "✌️",
+ "[抱拳]": "🤛",
+ "[勾引]": "🤙",
+ "[拳头]": "✊",
+ "[OK]": "👌",
+ "[合十]": "🙏",
+ "[点赞]": "👍",
+ "[玫瑰]": "🌹",
+ "[凋谢]": "🥀",
+ "[红包]": "🧧",
+};
+
+const WECHAT_SPECIAL_TOKENS = new Set([
+ "动画表情",
+ "视频号",
+ "小程序",
+ "微信红包",
+ "位置",
+ "名片",
+ "语音通话",
+ "视频通话",
+ "微信转账",
+]);
+
+const renderTextMessage = (line) => {
+ let text = line;
+ for (const [key, emoji] of Object.entries(WECHAT_EMOJI_MAP)) {
+ text = text.replaceAll(key, emoji);
+ }
+ text = text.replace(/\[([^\]]+)\]/g, (match, inner) => {
+ if (WECHAT_SPECIAL_TOKENS.has(inner)) {
+ return `\uE000${inner}\uE001`;
+ }
+ return match;
+ });
+ let escaped = escapeMarkdownText(text);
+ escaped = escaped.replace(/\uE000(.*?)\uE001/g, "*[$1]*");
+ return escaped;
+};
+
const parseLink = (line) => {
const match = LINK_LINE.exec(line.trim());
if (!match) return null;
@@ -209,19 +338,161 @@ const mediaMarkdown = (item) => {
return isImage(item.filename) ? `![${label}](${url})` : `[附件:${label}](${url})`;
};
-const renderBody = (lines, mediaByName) => lines.map((line) => {
- const mediaMatch = MEDIA_LINE.exec(line.trim());
- if (mediaMatch) {
- const item = mediaByName.get(posix.basename(mediaMatch[2]));
- if (item) {
- item.used = true;
- return mediaMarkdown(item);
+const QUOTE_SEPARATOR = /^[-—─\s]{5,}$/;
+
+const renderBody = (lines, mediaByName) => {
+ const rendered = [];
+ let index = 0;
+
+ while (index < lines.length) {
+ const line = lines[index];
+ const trimmed = line.trim();
+
+ // 识别微信引用回复:以「开头,以」结尾,紧随其后是分隔线
+ if (trimmed.startsWith("「")) {
+ let quoteEndIndex = -1;
+ for (let cursor = index; cursor < lines.length; cursor += 1) {
+ if (lines[cursor].includes("」")) {
+ quoteEndIndex = cursor;
+ break;
+ }
+ }
+ if (quoteEndIndex !== -1 && quoteEndIndex + 1 < lines.length && QUOTE_SEPARATOR.test(lines[quoteEndIndex + 1].trim())) {
+ const quoteLines = [];
+ for (let cursor = index; cursor <= quoteEndIndex; cursor += 1) {
+ let quoteLine = lines[cursor].trim();
+ if (cursor === index) quoteLine = quoteLine.replace(/^「/, "");
+ if (cursor === quoteEndIndex) quoteLine = quoteLine.replace(/」$/, "");
+ if (quoteLine) quoteLines.push(quoteLine);
+ }
+ if (quoteLines.length > 0) {
+ const firstQuote = quoteLines[0];
+ const colonMatch = firstQuote.match(/^([^::]+)[::](.*)$/);
+ if (colonMatch) {
+ rendered.push(`> **${escapeMarkdownText(colonMatch[1].trim())}**:${renderTextMessage(colonMatch[2].trim())}`);
+ for (let cursor = 1; cursor < quoteLines.length; cursor += 1) {
+ rendered.push(`> ${renderTextMessage(quoteLines[cursor])}`);
+ }
+ } else {
+ for (const q of quoteLines) {
+ rendered.push(`> ${renderTextMessage(q)}`);
+ }
+ }
+ }
+ index = quoteEndIndex + 2;
+ continue;
+ }
}
+
+ const mediaMatch = MEDIA_LINE.exec(trimmed);
+ if (mediaMatch) {
+ const item = mediaByName.get(posix.basename(mediaMatch[2]));
+ if (item) {
+ item.used = true;
+ rendered.push(mediaMarkdown(item));
+ index += 1;
+ continue;
+ }
+ }
+
+ const link = parseLink(trimmed);
+ if (link) {
+ rendered.push(`[${escapeMarkdownText(link.title)}](${link.url})`);
+ index += 1;
+ continue;
+ }
+
+ if (trimmed) {
+ rendered.push(renderTextMessage(trimmed));
+ }
+ index += 1;
+ }
+
+ return rendered.join("\n\n");
+};
+
+const parseTimeMs = (timeStr) => {
+ const match = /^(\d{4})年(\d{1,2})月(\d{1,2})日\s+(\d{1,2}):(\d{2})(?::(\d{2}))?$/.exec(timeStr?.trim() || "");
+ if (!match) return null;
+ const [, y, m, d, hh, mm, ss] = match;
+ return new Date(Number(y), Number(m) - 1, Number(d), Number(hh), Number(mm), ss ? Number(ss) : 0).getTime();
+};
+
+const groupMessages = (messages) => {
+ const groups = [];
+ for (const message of messages) {
+ if (message.raw) {
+ groups.push(message);
+ continue;
+ }
+ const prev = groups.at(-1);
+ if (prev && !prev.raw && prev.sender === message.sender) {
+ const prevMs = parseTimeMs(prev.time);
+ const currMs = parseTimeMs(message.time);
+ const isClose = (prevMs !== null && currMs !== null && Math.abs(currMs - prevMs) <= 3 * 60 * 1000)
+ || prev.time === message.time;
+ if (isClose) {
+ prev.body.push(...message.body);
+ continue;
+ }
+ }
+ groups.push({
+ sender: message.sender,
+ time: message.time,
+ body: [...message.body],
+ });
+ }
+ return groups;
+};
+
+const formatHeader = (sender, time, state) => {
+ const match = /^(\d{4}年\d{1,2}月\d{1,2}日)\s+(\d{1,2}:\d{2})(?::\d{2})?$/.exec(time?.trim() || "");
+ if (!match) {
+ return `**${escapeMarkdownText(sender)}** · ${escapeMarkdownText(time)}`;
}
- const link = parseLink(line);
- if (link) return `[${escapeMarkdownText(link.title)}](${link.url})`;
- return escapeMarkdownText(line);
-}).join("\n\n");
+ const [, date, clock] = match;
+ if (state.lastDate === date) {
+ return `**${escapeMarkdownText(sender)}** · ${clock}`;
+ }
+ state.lastDate = date;
+ return `**${escapeMarkdownText(sender)}** · ${date} ${clock}`;
+};
+
+const buildSummaryCard = (messages, mediaItems) => {
+ const structured = messages.filter((m) => !m.raw && m.sender && m.time);
+ if (structured.length < 2) return null;
+
+ const senders = [...new Set(structured.map((m) => m.sender))];
+ const firstTime = structured[0]?.time;
+ const lastTime = structured.at(-1)?.time;
+
+ let timeRange = firstTime;
+ if (lastTime && lastTime !== firstTime) {
+ const firstMatch = /^(\d{4}年\d{1,2}月\d{1,2}日)\s+(\d{1,2}:\d{2})/.exec(firstTime || "");
+ const lastMatch = /^(\d{4}年\d{1,2}月\d{1,2}日)\s+(\d{1,2}:\d{2})/.exec(lastTime || "");
+ if (firstMatch && lastMatch && firstMatch[1] === lastMatch[1]) {
+ timeRange = `${firstTime} ~ ${lastMatch[2]}`;
+ } else {
+ timeRange = `${firstTime} ~ ${lastTime}`;
+ }
+ }
+
+ const imageCount = mediaItems.filter((item) => isImage(item.filename)).length;
+ const otherCount = mediaItems.length - imageCount;
+ const mediaSummary = [];
+ if (imageCount > 0) mediaSummary.push(`${imageCount} 张图片`);
+ if (otherCount > 0) mediaSummary.push(`${otherCount} 个附件`);
+
+ const lines = [
+ "> 💬 **微信聊天记录**",
+ `> **参与者**:${senders.map(escapeMarkdownText).join("、")}`,
+ `> **时间**:${escapeMarkdownText(timeRange)} · 共 ${structured.length} 条消息`,
+ ];
+ if (mediaSummary.length > 0) {
+ lines.push(`> **包含媒体**:${mediaSummary.join("、")}`);
+ }
+ return lines.join("\n");
+};
const titleTimestamp = (time) => {
const match = /^(\d{4})年(\d{1,2})月(\d{1,2})日 (\d{1,2}):(\d{2})(?::\d{2})?$/.exec(time || "");
@@ -284,16 +555,19 @@ export const wechatChatNoteFromArchive = (bytes, archiveName = TRANSCRIPT_NAME)
}
const messages = parseMessages(text);
- const sections = messages.length > 0
- ? messages.map((message) => {
+ const grouped = groupMessages(messages);
+ const dateState = { lastDate: null };
+ const sections = grouped.length > 0
+ ? grouped.map((message) => {
if (message.raw) return renderBody(message.raw, mediaByName);
const body = renderBody(message.body, mediaByName);
- const header = `**${escapeMarkdownText(message.sender)}** · ${escapeMarkdownText(message.time)}`;
+ const header = formatHeader(message.sender, message.time, dateState);
return body ? `${header}\n\n${body}` : header;
})
: [escapeMarkdownText(text.trim())];
const unused = items.filter((item) => !item.used).map((item) => mediaMarkdown(item));
- const markdown = [...sections, ...unused].filter(Boolean).join("\n\n").trim();
+ const summaryCard = buildSummaryCard(messages, items);
+ const markdown = [summaryCard, ...sections, ...unused].filter(Boolean).join("\n\n").trim();
if (!markdown) throw new WeChatArchiveError("unrecognized");
return {
title: archiveTitle(archiveName, messages),
diff --git a/apps/desktop/src/main/wechat-chat-archive.test.mjs b/apps/desktop/src/main/wechat-chat-archive.test.mjs
index 5990b1e2..6818738d 100644
--- a/apps/desktop/src/main/wechat-chat-archive.test.mjs
+++ b/apps/desktop/src/main/wechat-chat-archive.test.mjs
@@ -94,11 +94,14 @@ describe("WeChat chat archive", () => {
test("turns the exported zip into messages, an inline image, and attachments", () => {
const note = wechatChatNoteFromArchive(sampleArchive(), "聊天记录_20260922_223222.zip");
expect(note.title).toBe("聊天记录_20260922_223222");
+ expect(note.markdown).toContain("> 💬 **微信聊天记录**");
+ expect(note.markdown).toContain("> **参与者**:鱼、Flechazo");
+ expect(note.markdown).toContain("> **包含媒体**:1 张图片、2 个附件");
expect(note.markdown).toContain("**鱼** · 2026年9月22日 22:15");
expect(note.markdown).toContain("[发布说明](https://example.com/post)");
expect(note.markdown).toContain("![微信图片_202609222215_1.jpg](edgeever-wechat-media://m1)");
- expect(note.markdown).toContain("\\[发呆\\]还没注册");
- expect(note.markdown).toContain("\\[动画表情\\]");
+ expect(note.markdown).toContain("😶还没注册");
+ expect(note.markdown).toContain("*[动画表情]*");
expect(note.markdown).toContain("[附件:微信视频_1.mp4](edgeever-wechat-media://m2)");
expect(note.markdown).toContain("[附件:说明.pdf](edgeever-wechat-media://m3)");
expect(note.markdown).not.toContain(".zip");
@@ -109,6 +112,45 @@ describe("WeChat chat archive", () => {
]);
});
+ test("formats quoted replies as markdown blockquotes and maps emoji", () => {
+ const quoteTranscript = [
+ "·张三",
+ "2026年9月24日 10:00",
+ "明天开会讨论方案",
+ "",
+ "·李四",
+ "2026年9月24日 10:05",
+ "「张三:明天开会讨论方案」",
+ "- - - - - - - - - - - - - - -",
+ "收到,我准备一下材料[OK]",
+ ].join("\n");
+ const archive = zipOf([{ name: "聊天记录.txt", data: quoteTranscript }]);
+ const note = wechatChatNoteFromArchive(archive, "项目组讨论.zip");
+ expect(note.markdown).toContain("> **张三**:明天开会讨论方案");
+ expect(note.markdown).toContain("收到,我准备一下材料👌");
+ });
+
+ test("simplifies times on the same day and keeps full date across days", () => {
+ const multiDayTranscript = [
+ "·张三",
+ "2026年9月24日 10:00",
+ "第一天上午",
+ "",
+ "·李四",
+ "2026年9月24日 10:05",
+ "第一天同天回复",
+ "",
+ "·张三",
+ "2026年9月25日 09:00",
+ "第二天上午",
+ ].join("\n");
+ const archive = zipOf([{ name: "聊天记录.txt", data: multiDayTranscript }]);
+ const note = wechatChatNoteFromArchive(archive, "跨天讨论.zip");
+ expect(note.markdown).toContain("**张三** · 2026年9月24日 10:00");
+ expect(note.markdown).toContain("**李四** · 10:05");
+ expect(note.markdown).toContain("**张三** · 2026年9月25日 09:00");
+ });
+
test("adds the first message time when WeChat provides only a generic archive name", () => {
expect(wechatChatNoteFromArchive(sampleArchive(), "聊天记录.zip").title).toBe("聊天记录 · 2026-09-22 22:15");
expect(wechatChatNoteFromArchive(sampleArchive()).title).toBe("聊天记录 · 2026-09-22 22:15");
diff --git a/apps/web/src/lib/wechat-chat-import.test.mjs b/apps/web/src/lib/wechat-chat-import.test.mjs
index 452382cf..b3734351 100644
--- a/apps/web/src/lib/wechat-chat-import.test.mjs
+++ b/apps/web/src/lib/wechat-chat-import.test.mjs
@@ -68,4 +68,32 @@ describe("WeChat chat note", () => {
})).rejects.toThrow("offline");
expect(deleted).toEqual(["memo_1"]);
});
+
+ test("groups consecutive images into an image gallery", async () => {
+ let updated = null;
+ await createWeChatChatMemo({
+ notebookId: "nb_inbox",
+ title: "多图聊天记录",
+ markdown: [
+ "![图1.jpg](edgeever-wechat-media://m1)",
+ "",
+ "![图2.jpg](edgeever-wechat-media://m2)",
+ ].join("\n"),
+ media: [
+ { id: "m1", filename: "图1.jpg", mimeType: "image/jpeg", byteSize: 2 },
+ { id: "m2", filename: "图2.jpg", mimeType: "image/jpeg", byteSize: 2 },
+ ],
+ createMemo: async () => ({ memo }),
+ readMedia: async (item) => new File(["a"], item.filename, { type: item.mimeType }),
+ uploadResource: async (_id, file) => ({ url: `edgeever-staged://${file.name}` }),
+ updateMemo: async (_created, content) => {
+ updated = content;
+ return { memo: { ...memo, contentMarkdown: content.contentMarkdown } };
+ },
+ });
+ expect(updated.contentJson.content.some((node) => node.type === "edgeeverImageGallery")).toBe(true);
+ const gallery = updated.contentJson.content.find((node) => node.type === "edgeeverImageGallery");
+ expect(gallery?.content?.length).toBe(2);
+ });
});
+
diff --git a/apps/web/src/lib/wechat-chat-import.ts b/apps/web/src/lib/wechat-chat-import.ts
index ca4b4a24..fb4e20e1 100644
--- a/apps/web/src/lib/wechat-chat-import.ts
+++ b/apps/web/src/lib/wechat-chat-import.ts
@@ -1,4 +1,9 @@
-import { docToMarkdown, markdownToDoc, type TiptapDoc } from "@edgeever/shared";
+import {
+ docToMarkdown,
+ groupConsecutiveImagesIntoGalleries,
+ markdownToDoc,
+ type TiptapDoc,
+} from "@edgeever/shared";
import { IMAGE_WIDTH_PRESETS } from "@edgeever/shared/image-display";
export const WECHAT_MEDIA_URL_PREFIX = "edgeever-wechat-media://";
@@ -25,9 +30,10 @@ const noteContent = (markdown: string) => {
const nodes = (parsed.content ?? []).map((node) => node.type === "image"
? { ...node, attrs: { ...node.attrs, width: WECHAT_IMAGE_WIDTH } }
: node);
+ const groupedNodes = groupConsecutiveImagesIntoGalleries(nodes);
const contentJson: TiptapDoc = {
type: "doc",
- content: nodes.at(-1)?.type === "paragraph" ? nodes : [...nodes, { type: "paragraph" }],
+ content: groupedNodes.at(-1)?.type === "paragraph" ? groupedNodes : [...groupedNodes, { type: "paragraph" }],
};
return { contentJson, contentMarkdown: docToMarkdown(contentJson) };
};