fix(together): x.com ボット拒否検出・Gemini JSONコードブロック除去

This commit is contained in:
posimai 2026-04-22 10:08:18 +09:00
parent 9f0bb8eae7
commit 324d892b28
1 changed files with 21 additions and 5 deletions

View File

@ -2619,11 +2619,30 @@ ${excerpt}
const titleMatch = fullContent.match(/^Title:\s*(.+)/m);
const jinaTitle = titleMatch ? titleMatch[1].trim().slice(0, 300) : null;
// x.com でタイトルがドメイン名のみ = ボット拒否ページ → AI 要約をスキップして skipped で保存
try {
const u = new URL(url);
const host = u.hostname.replace('www.', '');
if (/^(x\.com|twitter\.com)$/.test(host) && (!jinaTitle || jinaTitle === host)) {
await pool.query(
`UPDATE together_shares SET full_content=$1, archive_status='skipped' WHERE id=$2`,
[fullContent, shareId]
);
return;
}
} catch (_) {}
await pool.query(
`UPDATE together_shares SET full_content=$1, title=COALESCE(title, $2) WHERE id=$3`,
[fullContent, jinaTitle, shareId]
);
function parseGeminiJson(raw) {
// Gemini がコードブロックで返す場合に対応
const cleaned = raw.replace(/^```(?:json)?\s*/i, '').replace(/\s*```$/, '').trim();
return JSON.parse(cleaned);
}
let summary = null;
let tags = [];
if (togetherGenAI && fullContent) {
@ -2638,7 +2657,7 @@ ${excerpt}
const result = await Promise.race([model.generateContent(prompt), timeoutP]);
const raw = result.response.text().trim();
try {
const parsed = JSON.parse(raw);
const parsed = parseGeminiJson(raw);
summary = (parsed.summary || '').slice(0, 300);
tags = Array.isArray(parsed.tags) ? parsed.tags.slice(0, 4).map(t => String(t).slice(0, 20)) : [];
} catch {
@ -2647,9 +2666,6 @@ ${excerpt}
break;
} catch (aiErr) {
console.error(`[together archive AI] ${modelName} share=${shareId}`, aiErr.message);
if (modelName === modelsToTry[modelsToTry.length - 1]) {
// 全モデル失敗: Jina 本文は保存済みなので Reader は使える状態で done にする
}
}
}
}
@ -2697,7 +2713,7 @@ ${excerpt}
const result = await Promise.race([model.generateContent(prompt), timeoutP]);
const raw = result.response.text().trim();
let summary = null, tags = [];
try { const p = JSON.parse(raw); summary = (p.summary || '').slice(0, 300); tags = Array.isArray(p.tags) ? p.tags.slice(0, 4).map(t => String(t).slice(0, 20)) : []; }
try { const cleaned = raw.replace(/^```(?:json)?\s*/i,'').replace(/\s*```$/,'').trim(); const p = JSON.parse(cleaned); summary = (p.summary || '').slice(0, 300); tags = Array.isArray(p.tags) ? p.tags.slice(0, 4).map(t => String(t).slice(0, 20)) : []; }
catch { summary = raw.slice(0, 300); }
await pool.query(`UPDATE together_shares SET summary=$1, tags=$2, archive_status='done' WHERE id=$3`, [summary, tags, shareId]);
rearchiveDone = true;