使用Java正则表达式从xml中提取作者

Question

I understand that regex is not ideal for this task. But I couldn't use parser since I need preserve the OFFSET . So I have two questions here, one is about regex and other is to extract "author". If you recommend me using any parser, please let me know if there's a parser can preserve the offset. I have xml like this:

<post author="lafeat" datetime="2014-04-03T04:26:00" id="p1">
For legions of young couples, there is no wedding venue more desirable than a barn in the country.
</post>

My code is here:

String regex = "<post\\s*?author=\"(?!\")*\"?.*?>.*?</post>";
Pattern p = Pattern.compile(regex);
Matcher m = p.matcher(text);
while (m.find()) {
    System.out.println("start from: " + m.start());
    System.out.println("end to: " + m.end());
    System.out.println("the text is: " + text.substring(m.start(), m.end()));
}

But I didn't get anything back from this regex? Any suggestion will be great thank.

Answer 1

Using a dedicated HTML parser is better than any regex you can come up with.

---

To answer your question:

A negative lookahead isn't required here. It's being used incorrectly anyway:

1. You cannot apply quantifiers on zero-width assertions, ie you can't do this: (?!\\")* . This is because the preceding token, the zero-width negative lookahead expression, is not quantifiable.

2. You're not traversing through the string. As your regex is currently written, it only checks a single position. It's important to note that lookaround assertions are zero-width — it doesn't match any characters. So in order to have all the characters from the first double-quote to the next captured, you will have to actually match the text. You can use a dot for this purpose: (?:(?!\\").)* . It will advance through the string character by character until it reaches a position that is followed by a double-quote.

This is how you should write the expression ( see demo ):

<post\\s*?author=\"((?:(?!\").)*).*?>

But it doesn't need to be that complicated. You can just use a negated character class and be done with it ( see demo ):

<post\\s*?author=\"([^\"]+)\".*?>

\\"([^\\"]+)\\" is a negated character class that matches any characters other than a double-quote, one or more times.

Answer 2

You're not getting anything back because you're using a Negative Lookahead incorrectly and no capturing group. If you want to extract author , use a capturing group.

String regex = "<post\\s*author=\"([^\"]+)\"[^>]+>[^><]+</post>";

And then return the matched group here:

while (m.find()) {
    System.out.println("start from: " + m.start());
    System.out.println("end to: " + m.end());
    System.out.println("the text is: " + m.group(1));
}