PHP正则去除末尾字符串遇问题，寻求技术指导

Hey there! Let’s tackle your regex problem step by step. I get why you’re avoiding DOM parsing—when your content isn’t consistently <a> or <img> tags, regex is a reasonable approach (even if HTML regex has its quirks!).

First, let’s look at the example HTML you shared:

1. Extracting URLs #

For Tag Attributes #

Use these regex patterns to capture values from common URL-bearing attributes:

• Extract href values: href=['"]([^'"]+)['"]
  • This matches href=" or href=', then captures all characters until the closing quote.
• Extract src values: src=['"]([^'"]+)['"]
  • Same logic as above, targeted at image sources.
• Extract srcset values: srcset=['"]([^'"]+)['"]
  • Since srcset often has multiple URLs separated by commas, you’ll need to split the captured value afterward to get individual URLs.

For Plain Text URLs #

If your content has raw URLs (not inside tags), use this pattern to catch them:
https?:\/\/[^\s<>"]+

• Matches http:// or https://, then captures all characters until a whitespace, <, >, or quote.

Example Code (JavaScript) #

Here’s how to put this into practice to extract all URLs from your content:

const content = '<a href="http://domain.co.uk.co.uk/wp-content/uploads/2016/06/DSCN8900.jpg"><img class="alignnone size-medium wp-image-4181" src="http://domain.co.uk/wp-content/uploads/2016/06/DSCN8900-300x225.jpg" alt="dscn8900" width="300" height="225" srcset="http://domain.co.uk/wp-content/uploads/2016/06/DSCN8900-1024x768.jpg 1024w, http://domain.co.uk/wp-content/uploads/2016/06/DSCN8900-300x225.jpg 300w"> Some plain text URL: http://domain.co.uk/another-page.html';

const extractedUrls = [];

// Extract from attributes
const attrRegex = /(href|src|srcset)=['"]([^'"]+)['"]/g;
let attrMatch;
while ((attrMatch = attrRegex.exec(content)) !== null) {
  const attrName = attrMatch[1];
  if (attrName === 'srcset') {
    // Split srcset into individual URLs
    const srcsetUrls = attrMatch[2].split(',').map(item => item.trim().split(' ')[0]);
    extractedUrls.push(...srcsetUrls);
  } else {
    extractedUrls.push(attrMatch[2]);
  }
}

// Extract plain text URLs
const plainRegex = /https?:\/\/[^\s<>"]+/g;
let plainMatch;
while ((plainMatch = plainRegex.exec(content)) !== null) {
  if (!extractedUrls.includes(plainMatch[0])) {
    extractedUrls.push(plainMatch[0]);
  }
}

console.log(extractedUrls);

2. Modifying URLs (e.g., Updating Domains) #

If you need to replace parts of URLs (like changing domain.co.uk to new-domain.com), use a regex with capture groups:

const modifiedContent = content.replace(/(href|src|srcset)=['"](https?:\/\/)domain\.co\.uk([^'"]+)['"]/g, '$1="$2new-domain.com$3"');

This preserves the attribute name, protocol, and path while swapping out the domain.

Key Notes to Avoid Headaches #

• Quotation Marks: The patterns above support both single and double quotes. If your content uses unquoted attributes (rare but possible), adjust the regex to (href|src|srcset)\s*=\s*(['"]?)([^'"\s>]+)\1 to handle that case.
• Edge Cases: Regex isn’t perfect for HTML—if your content has nested tags, escaped quotes, or malformed markup, you might need to tweak the patterns. For super messy content, a lightweight HTML parser might still be worth considering, but regex works great for structured, predictable content.

Let me know if you have a specific end goal (like replacing URLs, extracting specific types, etc.) and I can refine this further!

内容的提问来源于stack exchange，提问作者YaBCK