使用browserless.js获取的URL中Unicode转义无法解码的技术求助

I’ve dealt with similar weird encoding glitches when scraping with headless browsers before, so let’s unpack what’s going on here and get those images loading properly.

The Root Cause #

Your hint about matching u003d (not \u003d) working tells me everything: the URLs you’re getting don’t contain actual Unicode-escaped = characters (where \u003d resolves to =), but instead literal text sequences of u003d. Browserless must be transforming original = characters into this literal string format during DOM serialization—something that doesn’t happen when you copy the URL directly from the console (your browser probably auto-converts the literal u003d back to = when you paste it elsewhere).

Quick Fixes #

1. Target the Literal u003d Directly

Since you confirmed matching u003d works, just replace those sequences with actual = signs:

// Replace all literal u003d instances with =
const fixedImageUrl = scrapedUrl.replace(/u003d/g, '=');

This should immediately fix the broken image issue for your example URL.

2. Handle All Literal u00XX Escapes (For Future-Proofing)

If you run into other similar escapes (like u0026 for &), use a more flexible regex to convert any literal u00XX sequence to its corresponding ASCII character:

const fixedImageUrl = scrapedUrl.replace(/u00([0-9A-Fa-f]{2})/g, (match, hexCode) => {
  // Convert the 2-digit hex code to its ASCII character
  return String.fromCharCode(parseInt(hexCode, 16));
});

This will handle any standard ASCII character that’s been converted to a literal u00XX string.

Why Your Previous Methods Didn’t Work #

Let’s clear up why the approaches you tried failed:

• JSON.parse(JSON.stringify(...)): This only resolves actual JSON escape sequences (like \u003d inside a JSON string). Since your string has literal u003d (not escaped), JSON doesn’t recognize it as a special sequence.
• String.prototype.normalize(): This is for Unicode normalization (e.g., combining accent characters), which has nothing to do with ASCII escape strings.
• decodeURIComponent(): This decodes URI-encoded characters like %3D, not literal u003d text.

Verify the Issue (Optional) #

To confirm this is indeed the problem, run a quick check on your scraped URL string:

console.log(scrapedUrl.includes('u003d')); // Should return true
console.log(scrapedUrl.includes('\u003d')); // Should return false (since \u003d is just =)

If those results match, you know you’re dealing with literal text sequences rather than actual Unicode escapes.

内容的提问来源于stack exchange，提问作者user11909477