Help with regexp/regsub

[Metuant]

Hi,

I'm trying to parse some information from a website using regsub and regexp, but i'm completely useless at regexp so now that they've updated their website the regexp no longer works.

The block of information I'm trying to parse (which is sometimes repeated multiple times - hence the while in the code) is:

<tr>
<td class='tablebottom'><img src="/img/member.gif" alt="[M]"/></td>
<!--name--><td class='tablebottom'>Abyssal whip</td>
<td class="tablebottom" title="Former average price: 1,650,000gp [decreased by 50,000gp]"><img src="/img/market/p_d.gif" alt="This price has decreased" /></td>
<!--price--><td class="tablebottom">1,550,000gp - 1,650,000gp</td>
<td class="tablebottom" width="20"><a href="/priceguide.php?report=45&par=" title="Report Incorrect Price"><img src="/img/!.gif" alt="[!]" border="0" /></a></td>
<td class="tablebottom"><a href="/priceguide.php?category=45">Obsidian & Abyssal</a></td>
</tr>

</table></form><br />

I'm trying to grab the item name (Abyssal whip) and its price (1,550,000gp - 1,650,000gp)
Using...

Code: Select all

	                
		while {[regexp "<!--name--><td class=\'tablebottom\'>(.*?)</td>\n\n<!--price--><td class=\"tablebottom\">(.*?)</td>\n<td class=\"tablebottom\" width=\"20\">" $data junk tname tprice]} {
			
			regsub "<!--name--><td class=\'tablebottom\'>[addslashes $tname]</td>\n\n<!--price--><td class=\"tablebottom\">[addslashes $tprice]</td>\n<td class=\"tablebottom\" width=\"20\">" $data - data
			if {$i == 0 || ([string match [string tolower [string range $item 0 1]] [string tolower [string range $tname 0 1]]] && [string length $tname] < [string length $name])} {
				set name $tname
				set price $tprice

I'm assuming that you can't just use \n\n to skip the line of useless data as I'd hoped..

Any help is appreciated

[speechles]

Perhaps sanitize the data before you attempt to parse it. So newlines, carriage returns, tabs, etc.. get eliminated before you get to that step.

Code: Select all

regsub -all "\t" $data "" data
regsub -all "\n" $data "" data
regsub -all "\r" $data "" data
regsub -all "\v" $data "" data

You can use a quantifier to express a range. This snippet should work:

Code: Select all

      while {[regexp "<!--name--><td class=\'tablebottom\'>(.*?)</td>.*?<!--price--><td class=\"tablebottom\">(.*?)</td>.*?<td class=\"tablebottom\" width=\"20\">" $data junk tname tprice]} {

         regsub "<!--name--><td class=\'tablebottom\'>[addslashes $tname]</td>.*?<!--price--><td class=\"tablebottom\">[addslashes $tprice]</td>.*?<td class=\"tablebottom\" width=\"20\">" $data - data
         if {$i == 0 || ([string match [string tolower [string range $item 0 1]] [string tolower [string range $tname 0 1]]] && [string length $tname] < [string length $name])} {
            set name $tname
            set price $tprice

[Metuant]

Thanks for the help - it works great :]