Regex optimization for Tag Extraction

Question

I am extracting the tags and their surrounding text in the HTML source code. This is the Regex pattern that I use in my code:

String a_pattern = "(.*?)<a (.*?)</a>(.*)";

I tried to use "?" to make the quantifier lazy. However, this pattern takes lots of time while compiling a long string.

Can you please give me some hints on optimizing this pattern?

I should note that I need all three groups of text (before, within, and after )

Thank you

This is a pretty bad idea. Use a proper HTML parser instead. See http://stackoverflow.com/questions/1732348/regex-match-open-tags-except-xhtml-self-contained-tags?rq=1 — p.s.w.g, Aug 04 '14 at 14:42
@p.s.w.g Is right, you should probably use a parser. But if you're set on using regex, tell us the language, and give us some examples of input and expected output. But a parser is probably your best bet in terms of runtime and code readability and just making it work in general — skamazin, Aug 04 '14 at 14:53
@sp00m, @p.s.w.g, @skamazin, thank you all for your comments. The point is that I don't need to only extract . I know that I can do this using some parsers like jericho. In fact, having `Stackoverflow is a good website for the beginners and and also the experts `, I expect to extract two strings: 1. Stackoverflow is a good website for the beginners and, 2. for the beginners and and also the experts . In fact, the goal is to extracted the surrounding text of tag. I am also programming in Java. Thank you. — n4z4nin, Aug 05 '14 at 07:19

score 0 · Answer 1 · answered Aug 04 '14 at 15:00

0

If you want only to get the content from those tags you can use this regex:

<a.*?>(.*?)<\/a>

You can check the performance for both regex in this link:

As you can see, the regex you are using is 97% slower.

enter image description here

This result uses the javascript regex engine, but it is really useful to know regex performance.

answered Aug 04 '14 at 15:00

Federico Piazza

Thank you Fede for your answer. What if I want to also extract the content before and also after ? – n4z4nin Aug 05 '14 at 07:06
@n4z4nin You can try using a complement regex, [check this] link(http://jsperf.com/regex-optimize/4). You can use multiple regex OR like: `(.*?)(.*)`. So, you could your two regex and even so it would be really faster than your current. Btw, if this answer your question you can consider mark it as resolved – Federico Piazza Aug 05 '14 at 13:57
Thank you Fede. However, I also need the text in between. That's why I am using three groups in my regex. – n4z4nin Aug 06 '14 at 13:33
@n4z4nin Can you please check this one: http://regex101.com/r/lO4kU8/2. By the way, here you have the performance http://jsperf.com/regex-optimize/6. This one work very well 97% faster. – Federico Piazza Aug 06 '14 at 13:44
@n4z4nin The regex is `(.*?)(.*)<|>(.*)` – Federico Piazza Aug 06 '14 at 13:51

1 Answers1