@cloud411716/fancy-webnovel 0.3.9 → 0.3.11
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
package/package.json
CHANGED
|
@@ -68,17 +68,22 @@ async function extractCategories(page, channel, type) {
|
|
|
68
68
|
/** 从 __INITIAL_STATE__ 提取当前品类页的作品列表 */
|
|
69
69
|
async function extractBookList(page) {
|
|
70
70
|
// 等 __INITIAL_STATE__ 有书单数据后再提取(防止网络慢时页面 JS 还没注数据)
|
|
71
|
-
|
|
72
|
-
(
|
|
73
|
-
|
|
74
|
-
|
|
75
|
-
|
|
76
|
-
|
|
77
|
-
|
|
78
|
-
|
|
79
|
-
|
|
80
|
-
|
|
81
|
-
|
|
71
|
+
try {
|
|
72
|
+
await page.waitForFunction(
|
|
73
|
+
() => {
|
|
74
|
+
const s = window.__INITIAL_STATE__ || {};
|
|
75
|
+
const cands = [s.rank, s.rankData, s.page];
|
|
76
|
+
for (const c of cands) {
|
|
77
|
+
if (c && (c.book_list || c.bookList || c.rankList)) return true;
|
|
78
|
+
}
|
|
79
|
+
return false;
|
|
80
|
+
},
|
|
81
|
+
{ timeout: 15000 }
|
|
82
|
+
);
|
|
83
|
+
} catch (waitErr) {
|
|
84
|
+
console.log(` ✗ 等待页面数据超时(15s),可能是网络慢或页面结构变化:${waitErr.message}`);
|
|
85
|
+
return []; // 返回空数组而非抛异常,避免中断整个采集
|
|
86
|
+
}
|
|
82
87
|
return page.evaluate(() => {
|
|
83
88
|
const s = window.__INITIAL_STATE__ || {};
|
|
84
89
|
const cands = [
|
|
@@ -299,7 +304,9 @@ async function scrapeChannel(ch, type) {
|
|
|
299
304
|
|
|
300
305
|
try {
|
|
301
306
|
// 打开入口页
|
|
302
|
-
await page.goto(initUrl, { waitUntil: "networkidle" })
|
|
307
|
+
await page.goto(initUrl, { waitUntil: "networkidle" }).catch(gotoErr => {
|
|
308
|
+
console.log(` ✗ 打开入口页失败:${gotoErr.message}`);
|
|
309
|
+
});
|
|
303
310
|
|
|
304
311
|
// 等待登录(用户可在此时手动登录)
|
|
305
312
|
if (LOGIN_WAIT > 0) {
|
|
@@ -310,7 +317,7 @@ async function scrapeChannel(ch, type) {
|
|
|
310
317
|
// 连通性自检
|
|
311
318
|
const probe = await probePage(page);
|
|
312
319
|
if (!probe.host || probe.host.indexOf("fanqie") === -1) {
|
|
313
|
-
console.
|
|
320
|
+
console.log(
|
|
314
321
|
` ✗ 当前页面非番茄(host=${probe.host}),可能被重定向到登录/验证页,已跳过。`
|
|
315
322
|
);
|
|
316
323
|
return null;
|
|
@@ -416,7 +423,9 @@ async function scrapeChannel(ch, type) {
|
|
|
416
423
|
console.log(` [${ci + 1}/${categories.length}] ${cat.name}`);
|
|
417
424
|
|
|
418
425
|
try {
|
|
419
|
-
await page.goto(`https://fanqienovel.com${cat.href}`, { waitUntil: "networkidle" })
|
|
426
|
+
await page.goto(`https://fanqienovel.com${cat.href}`, { waitUntil: "networkidle" }).catch(gotoErr => {
|
|
427
|
+
console.log(` ✗ 打开品类页失败:${gotoErr.message}`);
|
|
428
|
+
});
|
|
420
429
|
await page.waitForTimeout(2500);
|
|
421
430
|
await scrollLoad(page, 2);
|
|
422
431
|
|
|
@@ -451,7 +460,7 @@ async function scrapeChannel(ch, type) {
|
|
|
451
460
|
|
|
452
461
|
bodyLines.push("---", "");
|
|
453
462
|
} catch (catErr) {
|
|
454
|
-
console.
|
|
463
|
+
console.log(
|
|
455
464
|
` [fanqie] 品类 ${cat.name} 处理出错,跳过: ${catErr && catErr.message ? catErr.message : catErr}`
|
|
456
465
|
);
|
|
457
466
|
bodyLines.push(`## ${cat.name} — 采集失败`, "", "---", "");
|
|
@@ -491,7 +500,7 @@ async function main() {
|
|
|
491
500
|
written++;
|
|
492
501
|
console.log(` ✓ 已保存: ${filepath}`);
|
|
493
502
|
} catch (chErr) {
|
|
494
|
-
console.
|
|
503
|
+
console.log(
|
|
495
504
|
`[fanqie] ${channelLabel(ch)}${typeLabel(ty)} 采集失败,跳过: ${chErr && chErr.message ? chErr.message : chErr}`
|
|
496
505
|
);
|
|
497
506
|
}
|