No much code, but a general idea how I usually do it:
Read all of the html content into a TStringList.
Create a function that finds the first entry, then
another similar function that find a next entry
Then run the findnext routine in a loop until it returns false.
Just look for suitable keywords inside the html and strip out the data you need.
if findfirstrecord then
GrabResults()
else
Exit;
while findnextrecord do
begin
GrabResults();
end;