Procedure IsolateText( Const S: String; Tag1, Tag2: String; list:TStrings );
 Var
  pScan, pEnd, pTag1, pTag2: PChar;
  foundText: String;
  searchtext: String;
 Begin
  { Set up pointers we need for the search. HTML is not case sensitive, so
  we need to perform the search on a uppercased copy of S.}
  searchtext := Uppercase(S);
  Tag1:= Uppercase( Tag1 );
  Tag2:= Uppercase( Tag2 );
  pTag1:= PChar(Tag1);
  pTag2:= PChar(Tag2);
  pScan:= PChar(searchtext);
  Repeat
  { Search for next occurence of Tag1. }
  pScan:= StrPos( pScan, pTag1 );
  If pScan <> Nil Then Begin
  { Found one, hop over it, then search from that position
  forward for the next occurence of Tag2. }
  Inc(pScan, Length( Tag1 ));
  pEnd := StrPos( pScan, pTag2 );
  If pEnd <> Nil Then Begin
  { Found start and end tag, isolate text between,
  add it to the list. We need to get the text from
  the original S, however, since we want the un-uppercased
  version! So we calculate the address pScan would hold if
  the search had been performed on S instead of searchtext. }
  SetString( foundText,
  Pchar(S) + (pScan- PChar(searchtext) ),
  pEnd - pScan );
  list.Add( foundText );
  { Continue next search after the found end tag. }
  pScan := pEnd + Length(tag2);
  End { If }
  Else { Error, no end tag found for start tag, abort. }
  pScan := Nil;
  End; { If }
  Until pScan = Nil;
 End;