Skip to content

fix(doubao): refresh history and detail extraction #1

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
lmmsoft wants to merge 2 commits into from
Closed

fix(doubao): refresh history and detail extraction #1

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
44acf91
fix(doubao): refresh history and detail extraction
lmmsoft Jun 5, 2026
02dd1ce
fix(doubao): carry pin query type across history pages
lmmsoft Jun 6, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion clis/doubao/history.js
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -15,7 +15,7 @@ export const historyCommand = cli({
columns: ['Index', 'Id', 'Title', 'Url'],
func: async (page, kwargs) => {
const limit = parseInt(kwargs.limit, 10) || 50;
const conversations = await getDoubaoConversationList(page);
const conversations = await getDoubaoConversationList(page, { limit });
if (conversations.length === 0) {
return [{ Index: 0, Id: '', Title: 'No conversation history found. Make sure you are logged in.', Url: '' }];
}
Expand Down
238 changes: 228 additions & 10 deletions clis/doubao/utils.js
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -795,22 +795,107 @@ export function collectDoubaoTranscriptAdditions(beforeLines, currentLines, prom
.map(({ sanitized }) => sanitized)
.join('\n');
}
function getRecentConversationsScript(limit) {
return `
(async () => {
const clean = (value) => String(value || '').replace(/\\u00a0/g, ' ').replace(/\\s+/g, ' ').trim();
const requestedLimit = Math.max(1, Math.min(Number(${JSON.stringify(limit)}) || 50, 1000));
const resources = performance.getEntriesByType('resource')
.map((entry) => entry.name)
.filter((name) => typeof name === 'string');
const recentUrl = [...resources].reverse().find((name) => name.includes('/im/chain/recent_conv'));
if (!recentUrl) return { ok: false, reason: 'recent_conv resource not found', conversations: [] };

const conversations = [];
const seen = new Set();
let convVersion = 0;
let hasMore = true;
let pcPinQueryType = 0;

for (let pageIndex = 0; pageIndex < 60 && hasMore && conversations.length < requestedLimit; pageIndex += 1) {
const batchLimit = Math.max(1, Math.min(50, requestedLimit - conversations.length));
const body = {
cmd: 3200,
sequence_id: String(Date.now()) + '_' + pageIndex,
channel: 2,
version: '1',
uplink_body: {
pull_recent_conv_chain_uplink_body: {
api_version: 1,
conv_version: Number(convVersion) || 0,
direction: Number(convVersion) === 0 ? 3 : 1,
limit: batchLimit,
message_count_per_conv: 10,
option: {
not_need_message: true,
need_complete_conversation: true,
need_coco_conversation: Number(convVersion) === 0,
need_coco_bot: Number(convVersion) === 0,
need_pc_pin_chain: true,
pc_pin_query_type: pcPinQueryType,
},
},
},
};
const response = await fetch(recentUrl, {
method: 'POST',
headers: { 'Content-Type': 'application/json; encoding=utf-8' },
body: JSON.stringify(body),
});
const json = await response.json().catch(() => ({}));
if (json.status_code !== 0) {
return {
ok: false,
reason: json.status_desc || json.message || 'recent_conv request failed',
conversations,
};
}

const downlink = json.downlink_body?.pull_recent_conv_chain_downlink_body || {};
for (const cell of downlink.cells || []) {
const conversation = cell?.conversation || {};
const id = clean(conversation.conversation_id || cell?.id || '');
if (!id || seen.has(id)) continue;
seen.add(id);
conversations.push({
id,
title: clean(conversation.name || conversation.title || 'Untitled conversation').slice(0, 200),
href: '/chat/' + id,
updateTime: clean(conversation.update_time || ''),
latestIndex: clean(conversation.latest_index || ''),
botId: clean(conversation.bot_id || ''),
botType: conversation.bot_type ?? '',
});
if (conversations.length >= requestedLimit) break;
}

hasMore = Boolean(downlink.has_more);
convVersion = downlink.next_conv_version || 0;
pcPinQueryType = downlink.extra?.pc_pin_query_type ?? pcPinQueryType;
if (!convVersion || !(downlink.cells || []).length) break;
}

return { ok: true, conversations };
})()
`;
}
function getConversationListScript() {
return `
(() => {
const sidebar = document.querySelector('[data-testid="flow_chat_sidebar"]');
const sidebar = document.querySelector('[data-testid="flow_chat_sidebar"], #flow_chat_sidebar');
if (!sidebar) return [];

const items = Array.from(
sidebar.querySelectorAll('a[data-testid="chat_list_thread_item"]')
sidebar.querySelectorAll('a[data-testid="chat_list_thread_item"], a[id^="conversation_"], a[href*="/chat/"]')
);

return items
.map(a => {
const href = a.getAttribute('href') || '';
const match = href.match(/\\/chat\\/(\\d{10,})/);
if (!match) return null;
const id = match[1];
const idFromAttr = (a.getAttribute('id') || '').match(/^conversation_(\\d{10,})$/)?.[1] || '';
const idFromHref = href.match(/\\/chat\\/(?:bot\\/chat\\/)?(\\d{10,})/)?.[1] || '';
const id = idFromAttr || idFromHref;
if (!id) return null;
const textContent = (a.textContent || a.innerText || '').trim();
const title = textContent
.replace(/\\s+/g, ' ')
Expand All @@ -821,8 +906,21 @@ function getConversationListScript() {
})()
`;
}
export async function getDoubaoConversationList(page) {
export async function getDoubaoConversationList(page, options = {}) {
await ensureDoubaoChatPage(page);
const requestedLimit = Math.max(1, parseInt(String(options.limit || '50'), 10) || 50);
const apiResult = await page.evaluate(getRecentConversationsScript(requestedLimit)).catch(() => null);
if (apiResult?.ok && Array.isArray(apiResult.conversations) && apiResult.conversations.length > 0) {
return apiResult.conversations.map((item) => ({
Id: item.id,
Title: item.title || 'Untitled conversation',
Url: `${DOUBAO_CHAT_URL}/${item.id}`,
UpdateTime: item.updateTime,
LatestIndex: item.latestIndex,
BotId: item.botId,
BotType: item.botType,
}));
}
const raw = await page.evaluate(getConversationListScript());
if (!Array.isArray(raw))
return [];
Expand All @@ -839,9 +937,22 @@ export function parseDoubaoConversationId(input) {
function getConversationDetailScript() {
return `
(() => {
const clean = (v) => (v || '').replace(/\\u00a0/g, ' ').replace(/\\n{3,}/g, '\\n\\n').trim();
const clean = (v) => (v || '')
.replace(/\\u00a0/g, ' ')
.replace(/\\n{3,}/g, '\\n\\n')
.trim();

const isVisible = (el) => {
if (!(el instanceof HTMLElement)) return false;
const style = window.getComputedStyle(el);
if (style.display === 'none' || style.visibility === 'hidden') return false;
const rect = el.getBoundingClientRect();
return rect.width > 0 && rect.height > 0;
};

const messageList = document.querySelector('[data-testid="message-list"]');
const messageList = document.querySelector(
'[data-testid="message-list"], .conversation-page-message-host, [class*="message-list-"]'
);
if (!messageList) return { messages: [], meeting: null };

const meetingCard = messageList.querySelector('[data-testid="meeting-minutes-card"]');
Expand All @@ -855,8 +966,114 @@ function getConversationDetailScript() {
};
}

const roleFor = (root) => {
if (
root.matches('[data-testid="send_message"], [class*="send-message"], [class*="justify-end"]')
|| root.querySelector('[data-testid="send_message"], [class*="send-message"], [class*="bg-g-send-msg-bubble"]')
|| root.querySelector('[data-foundation-type="send-message-action-bar"]')
) {
return 'User';
}
if (
root.matches('[data-testid="receive_message"], [data-testid*="receive_message"], [class*="receive-message"]')
|| root.querySelector('[data-testid="receive_message"], [data-testid*="receive_message"], [class*="receive-message"]')
|| root.querySelector('[data-foundation-type="receive-message-action-bar"]')
|| root.querySelector('.md-box-root, [class*="md-box-root"], .flow-markdown-body, [class*="markdown"]')
) {
return 'Assistant';
}
return '';
};

const textSelectors = [
'[data-testid="message_text_content"]',
'[data-testid="message_content"]',
'[data-testid*="message_text"]',
'[data-testid*="message_content"]',
'[class*="bg-g-send-msg-bubble"]',
'[class*="bg-g-receive-msg-bubble"]',
'.md-box-root',
'[class*="md-box-root"]',
'.flow-markdown-body',
'[class*="message-content"]',
];

const extractImageLines = (root) => Array.from(root.querySelectorAll('img'))
.filter((img) => img instanceof HTMLImageElement && isVisible(img))
.map((img) => {
const width = img.naturalWidth || img.width || 0;
const height = img.naturalHeight || img.height || 0;
if (width > 0 && height > 0 && width <= 48 && height <= 48) return '';
const url = clean(img.currentSrc || img.src || '');
return /^https?:\\/\\//i.test(url) ? 'Image: ' + url : '';
})
.filter((line, index, lines) => line && lines.indexOf(line) === index);

const extractText = (root) => {
const chunks = [];
const seenText = new Set();
for (const selector of textSelectors) {
const nodes = Array.from(root.querySelectorAll(selector)).filter(isVisible);
for (const node of nodes) {
const text = clean(node.innerText || node.textContent || '');
if (!text || seenText.has(text)) continue;
seenText.add(text);
chunks.push(text);
}
if (chunks.length > 0) break;
}
const text = chunks.length > 0 ? clean(chunks.join('\\n')) : clean(root.innerText || root.textContent || '');
const imageLines = extractImageLines(root);
return clean([text, ...imageLines].filter(Boolean).join('\\n'));
};

const roots = [];
const seenNodes = new Set();
const selectors = [
'[data-testid="union_message"]',
'[data-testid="message-block-container"]',
'.v_list_row [data-message-id]',
'[data-message-id]',
];
for (const selector of selectors) {
messageList.querySelectorAll(selector).forEach((node) => {
if (!(node instanceof HTMLElement) || seenNodes.has(node)) return;
seenNodes.add(node);
roots.push(node);
});
}

const filteredRoots = roots
.filter((node) => isVisible(node) && !node.closest('script, style, noscript'))
.filter((node, index, nodes) => !nodes.some((other, otherIndex) => otherIndex !== index && other.contains(node)));

filteredRoots.sort((a, b) => {
if (a === b) return 0;
const pos = a.compareDocumentPosition(b);
return pos & Node.DOCUMENT_POSITION_FOLLOWING ? -1 : 1;
});

const deduped = [];
const seenMessages = new Set();
for (const root of filteredRoots) {
const role = roleFor(root) || 'System';
const text = extractText(root);
if (!text) continue;
const key = role + '::' + text;
if (seenMessages.has(key)) continue;
seenMessages.add(key);
deduped.push({
role,
text,
hasMeetingCard: !!root.querySelector('[data-testid="meeting-minutes-card"]'),
});
}

const messages = deduped.filter((message) => message.text);
if (messages.length > 0) return { messages, meeting };

const unions = Array.from(messageList.querySelectorAll('[data-testid="union_message"]'));
const messages = unions.map(u => {
const legacyMessages = unions.map(u => {
const isSend = !!u.querySelector('[data-testid="send_message"]');
const isReceive = !!u.querySelector('[data-testid="receive_message"]');
const textEl = u.querySelector('[data-testid="message_text_content"]');
Expand All @@ -868,7 +1085,7 @@ function getConversationDetailScript() {
};
}).filter(m => m.text);

return { messages, meeting };
return { messages: legacyMessages, meeting };
})()
`;
}
Expand Down Expand Up @@ -1116,6 +1333,7 @@ export const __test__ = {
clickSendButtonScript,
composerStateScript,
detectDoubaoVerificationScript,
getRecentConversationsScript,
getTurnsScript,
getTranscriptLinesScript,
};
Expand Down
7 changes: 7 additions & 0 deletions clis/doubao/utils.test.js
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -162,6 +162,13 @@ describe('doubao receive strategy', () => {
expect(transcriptScript).toContain('请仔细甄别');
expect(transcriptScript).toContain('下载电脑版');
});

it('carries the server pc_pin_query_type across recent-conversation pages', () => {
const recentScript = __test__.getRecentConversationsScript(100);
expect(recentScript).toContain('let pcPinQueryType = 0');
expect(recentScript).toContain('pc_pin_query_type: pcPinQueryType');
expect(recentScript).toContain('pcPinQueryType = downlink.extra?.pc_pin_query_type ?? pcPinQueryType');
});
});
describe('collectDoubaoTranscriptAdditions', () => {
it('ignores landing-page capability chips that are not assistant content', () => {
Expand Down
Loading