Hello JP,
I updated the expression and tested it. I'm not sure what to capture
where. Maybe if you can describe the table row to me I could do a
better expression.
Right now it captures as follows: team, quarter and the score (4x).
<tr[^>]*>\s*<td[^>]*>\s*<a[^>]*>(?<team>((?!</a).)*)</a[^>]*>\s*</td[
^>]*>\s*<td[^>]*>(?<quarter>((?!</td).)*)</td[^>]*>(\s*<td[^>]*>(?<sc
ore>((?!</td).)*)</td[^>]*>)+\s*</tr[^>]*>
it results in a match that is built up as follows:
Match
+- Groups
+- Team.Value "New Orleans"
+- Quarter.Value "0"
+- Score.Captures[0].Value = 10
+- Score.Captures[1].Value = 0
+- Score.Captures[2].Value = 0
+- Score.Captures[3].Value = 10
My question now is... did I interpret it right?
Is the row built up of a team name, quarter number and 4 scores?
To explain the expression:
<tr[^>*> Find a start of a row
\s* allow whitespace between <tr> tag and the first <td>
<td[^>]*> Find the first table cell.
\s* allow whitespace between <td> and <a>
<a[^>]*> Find the opening a href tag
(?team((?!</a).)*) capture everything you find from there to the </a
tag
in a group named team.
</a[^>]*> Find the closing a tag
\s* allow whitespace between </a> and </td>
</td[^>]*> Find the closing td tag
\s* allow whitespace between </td> and <td>
<td[^>]*> Find opening <td>
(?<quarter>((?!</td).)*) capture everything from there to the </ta
tag into
a group named quarter
</td[^>]*> Find the closing td tag
( start repeating group
\s* allow whitespace between </td> and <td>
<td[^>]*> Find opening <td>
(?<score>((?!</td).)*) capture everything from there to the </ta tag
into
a group named score
</td[^>]*> Find the closing td tag
)+ end repeating group. If the named group in this repeating group
captured
multiple times. The value scan be found in the Group.Captures
collection.
\s*
</tr[^>]*> Finally make sure we've captured a complete row by finding
it's
end tag.
Jesse
Hi Jesse,
Thanks again for the assistance.
I have taken what you have posted and put it into a console app but
it doesn't seem to pick up anything using the expression. If I
uncomment the shortpat and the first foreach, I get data. Am I
missing something?
static void Main(string[] args)
{
//string shortpat = @">([\w\s]+)<";
string pat =
@"<tr[^>]*><td[^>]*>(?<team>((?!</td).)*)</td[^>]*>\s*<td[^>]*>(?<qu
ar
ter>((?!</td).)*)</td[^>]*>\s*<td[^>]*>(?<score>((?!</td).)*)</td[^>
]*
\s*</tr[^>]>";
string html = @"<table border=""0"" width=""100%""><tr><td
width=""40%"">Team</td><td width=""10%""
align=""center"">1</td>
<td
width=""10%"" align=""center"">2</td><td width=""10%""
align=""center"">3</td> <td width=""10%""
align=""center"">4</td><td
width=""20%"" align=""center"">Score</td></tr><tr><td
width=""40%""><A
href=""/default.asp?c=sportsnetwork&page=nfl/teams/078.htm"">New
Orleans</A></td><td width=""10%"" align=""center""
0</td><td
width=""10%"" align=""center"" >10</td><td width=""10%""
align=""center"" >0</td><td width=""10%"" align=""center""
0</td><td
width=""20%"" align=""center"" >10</td></tr><tr><td width=""40%""><A
href=""/default.asp?c=sportsnetwork&page=nfl/teams/
071.htm"">Indianapolis</A>-</td><td width=""10%"" align=""center""
7</
td>> <td width=""10%"" align=""center"" >3</td><td width=""10%"" td>>
align=""center"" >14</td><td width=""10%""
align=""center"" >17</
td>> <td width=""20%"" align=""center"" >41</td></tr></table>"; td>>
Regex r = new Regex(pat);
//foreach (Match m in r.Matches(html))
//{
// if (m.Groups[1].Value.Trim() == "")
// // ignore these
// continue;
// else
// Console.WriteLine(m.Groups[1].Value);
// Console.ReadLine();
//}
foreach (Match m in r.Matches(html))
{
string team = m.Groups["team"].Value;
string quarter = m.Groups["quarter"].Value;
string score = m.Groups["score"].Value;
Console.WriteLine(team);
Console.WriteLine(quarter);
Console.WriteLine(score);
}
Console.ReadLine();
}
:
Hello JP,
Jesse,
That's exactly what I am trying to do with the data within the
HTML. The expressions and code you listed don't apply to the HTML
I posted correct?
They do. by just saying <td[^>]+> you say find a <td tag. Then find
everything up to the closing > and then match the closing >. It
ignores all the border, widtch, height and other stuff in there.
For now I ignored the <a href> around the team name, but other than
that, the expression should work.
I am not sure how I could group the teams and quarters if they are
not labeled. I probably don't understand how regex works...
Well if they're always in the xth cell as in you example, you can
use their position (as I've done in the expression). The
(?<name>...) construct in the expression then labels them.
Jesse
:
Hello JP,
I am creating a screen scraping app that will extract data from
a website. The screen scraping is pretty straightforward using
.NET 2.0, but stripping out all extraneous characters is proving
to be more difficult. I am basically trying to extract the
team, quarter, score for the quarter, and score for the entire
game from this html. (This html is a subset of the entire
page)
<table border="0" width="100%"><tr><td width="40%">Team</td><td
width="10%" align="center">1</td> <td width="10%"
align="center">2</td><td width="10%" align="center">3</td> <td
width="10%" align="center">4</td><td width="20%"
align="center">Score</td></tr><tr><td width="40%"><A
href="/default.asp?c=sportsnetwork&page=nfl/teams/078.htm">New
Orleans</A></td><td width="10%" align="center" >0</td><td
width="10%" align="center" >10</td><td width="10%"
align="center"
0</td><td width="10%" align="center" >0</td><td width="20%"
align="center"
10</td></tr><tr><td width="40%"><A
href="/default.asp?c=sportsnetwork&page=nfl/teams/071.htm">India
na po li s</A></td><td width="10%" align="center" >7</td><td
width="10%" align="center" >3</td><td width="10%" align="center"
14</td><td width="10%" align="center" >17</td><td width="20%"
align="center"
41</td></tr></table>
In essance I want to be able to put the names and scores into an
array so I can add to a database. From what I read regular
expressions should be able to do this but I am a complete
beginner using regex. Could someone assist in getting me
started? Many thanks.
I posted a regex a while back that did almost this.
<tr[^>]*><td[^>]*>(?<team>((?!</td).)*)</td[^>]*>\s*<td[^>]*>(?<q
ua rt
er>((?!</td).)*)</td[^>]*>\s*<td[^>]*>(?<score>((?!</td).)*)</td[
^> ]*
\s*</tr[^>]>
