如何捕获Java中正则表达式匹配后的文本？

Question

I am working on an assignment in which I need to search through a web site and extract conversion rates.

If I was able to simply match the rates this would be simple to capture and extract, but I need to be able to hit an update button and have the program search for the updated conversion rates, so I am not able to simply hard code a match to search for.

Is there a way in which I can match the text that precedes the rates and capture all text that follows the match?

If there is a better way to do this I am also open to suggestions. I just need help to know how to get the rates from the website when I do not know exactly what the rates will be. I simply know the format of the rates and where the rates are located within the site.

Here is what I have so far:

String regex = "(?<=EUR'>)\\d+(?:\\.\\d*)?(?=<)";

Pattern pattern = Pattern.compile(regex);
Matcher match = pattern.matcher(?);

while (match.find()) {   
  System.out.println("Found a match: " + match.group(1).toString());  
  System.out.println("Start position: " + match.start(1)); 
  System.out.println("End position: " + match.end(1)); 
} 

I think i understand how to set up the pattern, but I am unsure as to what I should put for the match string if I only know what the beginning and end will be and not the end...

An example of what I would need to grab is the line below

<td class='rtRates'><a href='/graph/?from=USD&amp;to=EUR'>0.772000</a></td>

I need to grab the rate in this line, but it will constantly be changing

Answer 1

Do not use regex to parse html, or a velociraptor will come and eat you. Use something like jsoup and query value of an <a> element that is inside a <td> with class rtRates .

Answer 2

I am not sure what your problem is, because your expression is matching what you expect (I think). See it on Regexr .

If you want to be more flexible on what the part looks like between the tags, you can use this:

(?<=EUR'>)[^<]*

The [^<] is a negatied character class. It will match any character but < . Then you can also remove the lookahead assertion.

See it on Regexr

Answer 3

Can't you just use this?

EUR'>(\d+(?:\.\d+)?)<

The rate is captured in group #1, which is handy since you're already using group(1) to extract it. ;) But seriously, there are no capturing groups in your regex, so calling group(1) on the Matcher result an exception. What gives?

ps Notice that I changed your \\d* to \\d+ . Almost everyone who tries to match decimal numbers requires at least one digit after the decimal point. If that's not the case here, go ahead and change it back.