views:

109

answers:

4

I have the following... I wouldn't say problem, but situation.

I have some HTML with tags and everything. I want to search the HTML for every URL. I'm doing it now by checking where it says 'h' then 't' then 't' then 'p', but I don't think is a great solution

Any good ideas?


Added: I'm looking for some kind of pseudocode but, just in case, I'm using Java for this project in particular

A: 

Use a DOM parser to extract all <a href> tags, and, if desired, additionally scan the source for http:// outside of those tags.

Borealid
not all url are in tags, some are text and some are in links or other tags :S
Saikios
@Saikios : That's what I said. The part about scanning for other links outside the tags. The second half of my single-sentence answer. Was it too long?
Borealid
Hey, no it wasn't but if I do that what beneffit i'm going to have ... just a few href? :( my Idea was trying to do something like the str_replace using indexes or a reg_exp but i have it flying over my head
Saikios
Take a look at [this](http://stackoverflow.com/questions/1732348/regex-match-open-tags-except-xhtml-self-contained-tags/1732454#1732454)
Borealid
I can't parse html with regex but I can eliminate everything that isn't between . .edu . .com . .gov, etc but is a really complex regex, :S
Saikios
+4  A: 

Try using a HTML parsing library then search for <a> tags in the HTML document.

Document doc = Jsoup.parse(input, "UTF-8", "http://example.com/");
Elements links = doc.select("a[href]"); // a with href

not all url are in tags, some are text and some are in links or other tags

You shouldn't scan the HTML source to achieve this.

You will end up with link elements that are not necessarily in the 'text' of the page, i.e you could end up with 'links' of JS scripts in the page for example.

Best way is still that you use a tool made for the job.

You should grab HTML tags and cover the most likely ones to have 'links' inside them (say: <h1>, <p>, <div> etc) . HTML parsers provide regex-like functionalities to filter through the content of the tags, something similar to your logic of "starts with HTTP".

[attr^=value], [attr$=value], [attr*=value]: elements with attributes that start with, end with, or contain the value, e.g. select("[href*=/path/]")

See: jSoup.

Bakkal
not all url are in tags, some are text and some are in links or other tags :Sby the way, I add it to my bookmarks the jSoup page for future reference, but I can't use it in this project :( I need all the urls
Saikios
I will give it a try with this thanks @Bakkal
Saikios
+1  A: 

You may want to have a look at XPath or Regular Expressions.

Hamid Nazari
Hi I'm using java, but anyway like I told to the other guys I'm trying to get all the urls on the string text, a, link, etc. Thanks to everybody that answer ;)
Saikios
A: 

The best way should be to google for regexes. One example is this one:

    /^(https?):\/\/((?:[a-z0-9.\-]|%[0-9A-F]{2}){3,})(?::(\d+))?((?:\/(?:[a-z0-9\-._~!$&'()+,;=:@]|%[0-9A-F]{2})))(?:\?((?:[a-z0-9\-._~!$&'()+,;=:\/?@]|%[0-9A-F]{2})))?(?:#((?:[a-z0-9\-._~!$&'()+,;=:\/?@]|%[0-9A-F]{2})*))?$/i

found in a hacker news article. As far as I can follow it, it looks good. But there is, as far as I know, no formal regex for this problem. So the best solution is to google for some and try which one matches most of what you want.

erikb
coolregex but is for something else, this is for checking if the url is a correct url but not to obtain the url's from a big string the one that I was thinking is probably bigger :P
Saikios