I want to split an HTML anchor tag into its constituent parts. I have a regular expression pattern that works with .NET's Regex class, but not with std.regexp - it errors out with "*+? not allowed in atom". I think this means something in the pattern is non-standard.

Here's my code:

if (auto m = std.regexp.search(
  "<a href=\"www.google.com\">Google</a>",
  r"<a.*?href=[""'](?<url>.*?)[""'].*?>(?<name>.*?)</a>")) {
  string url = m.match(1);
  string name = m.match(2);
}

The problematic parts are "?<url>" and "?<name>" - but not being a whiz with regular expressions, I don't know what to use instead.

Perhaps someone's got a better pattern they could post?

John.

> Perhaps someone's got a better pattern they could post?

this works for me (dmd 1.029)

import std.regexp;

void main()
{
 if (auto m = std.regexp.search(
              "<a href=\"www.google.com:8080/dfs?a1=1&a2=2\">This is Google link</a>",
              "<a[^>]+href=(['\"]?)(.*?)\\1.*?>(.*)</a>"))
 {
   for(int i=0; i<10; i++)
   {
     printf("%d=\"%.*s\"\n", i, m.match(i));
   }
 }
}

novice2 Wrote:

> > Perhaps someone's got a better pattern they could post?
> 
> this works for me (dmd 1.029)
> 
> import std.regexp;
> 
> void main()
> {
>  if (auto m = std.regexp.search(
>               "<a href=\"www.google.com:8080/dfs?a1=1&a2=2\">This is Google link</a>",
>               "<a[^>]+href=(['\"]?)(.*?)\\1.*?>(.*)</a>"))
>  {
>    for(int i=0; i<10; i++)
>    {
>      printf("%d=\"%.*s\"\n", i, m.match(i));
>    }
>  }
> }
> 

Thanks - that seems to extract the href and text. What about getting other attributes like name and title, as in this link:

<a href=\"www.google.com\" name=\"googleLink\" title=\"Click Me\">Google Link</a>

John C Wrote:

> Thanks - that seems to extract the href and text.

what you asked - that you got :)
"question is half of  answer" :)

>What about getting other attributes like name and title, as in this link:
> 
> <a href=\"www.google.com\" name=\"googleLink\" title=\"Click Me\">Google Link</a>

imho, it can't be done by one regexp match. because of random sequense of attributes. imho, you should get whole <a> tag attributes string, then iterate attributes in it.
something like this below.
but sorry, it can't catch attributes without quotes.
may be, std.strings non-regexp will be better when parsing attributes.

//////
import std.regexp;
import std.stdio;

void main()
{
 if (auto m = std.regexp.search("<anothertag><a href=\"www.google.com:8080/dfs?a1=1&a2=2\" name='google Link' color=red title=\"Click Me\"\">This is Google link</a></anothertag>",
              "<a(\\s.*?)>(.*?)</a>"))
 {
   writefln("tag attributes: \"%s\"", m.match(1));
   writefln("tag content: \"%s\"", m.match(2));
   foreach(s; RegExp("(\\S+?)=(['\"]?)(.*?)\\2").search(m.match(1)))
   {
     writefln("found attribute: name=\"%s\", value=\"%s\"", s.match(1), s.match(3));
   }
 }
}