title: Delphi HTML Parser description: Very small and fast module for parsing HTML pages.
author: sandbil tags: Delphi

Delphi HTML Parser

This module allows you to work with HTML documents as DOM tree and to use XPath for searching tags.
It is a very simple way to parse HTML.

It's tested with Delphi XE5, XE6 and 10.3

Requirements

OpenSSL library (libeay32.dll, ssleay32.dll)
Current versions of OpenSSL can be downloaded at https://indy.fulgan.com/SSL/

Usage

1. Add parser.pas; to uses list.
2. Example usage

  {...}
  var
    DomTree: TDomTree;
    DomTreeNode: TDomTreeNode;
    HtmlTxt: string;
    NodeList: TNodeList;
    ValueList: TStringList;
  begin
    HtmlTxt := NodeList:= TNodeList.Create;
    ValueList := TStringList.Create;
    DomTree := TDomTree.Create;
    DomTreeNode := DomTree.RootNode;
    If DomTreeNode.RunParse(HtmlTxt) then
    begin
      // short example code:
      DomTreeNode.FindXPath('//*[@id="TopBox"]/div[1]/div[@class="draw default"]', NodeList, ValueList)

      {...}
    end;
    {...}
  end;

3. enjoy!!!

Available properties and methods:

• TDomTree

  • property Count - count of nodes
  • property RootNode - root node (TDomTreeNode)
  • property ParseErr - Tstringlist contains all parsing errors and warnings

• TDomTreeNode

  • property Tag - name of tag
  • property AttributesTxt - string with all attributtes
  • property Attributes - parsed attributes (TDictionary<string, string>)
  • property Text - text
  • property TypeTag -
  • property Child - contains child's nodes (TChildList of TDomTreeNode)
  • property Parent - contains parent's node
  • property Owner - contains pointer to owner TDomTree
  • function FindNode - boolean function, if true then TNodeList contains found nodes
  • function FindTagOfIndex - boolean function, if true then TNodeList contains founded nodes
  • function GetAttrValue - returns value of attribute of current node
  • function GetComment - returns the set index comment in the current container node
  • function GetTagName - return name of tag + AttributesTxt
  • function GetText - returns the set index text in the current container node
  • function GetXPath - returns Xpath for current node
  • function RunParse - if parse is successfully then CHild property contains HTML DOM tree
  • function FindXPath - boolean function, if true then TNodeList contains found nodes
  • and TStringList contains found values of attribute, comment, text

• Xpath support:

  • attributes - //*[@id="TopBox"]/div/@class

  • comment - //*[@id="TopBox"]/div/comment()[3]

  • text - //*[@id="TopBox"]/div/text()[2]

  • previous level - /../div[@class="draw default"]/img[2]/@alt

  • partial coincidence by search in value of attribute:
    /div[@class="draw] returned nodes [[div class="draw default"],[div class="draw"], [div class="draw any"]..] like Xpath's function contains.
    Note: Xpath always starts search from current node. If you want to do global search, you must start from root's node.

demo project