DIHtmlParser v6.5.0 XE2/XE3 32/64

Парсер и анализатор HTML, XHTML и XML.
Полная поддержка Unicode (UnicodeString или WideString в зависимости от версии Delphi).
Чтение и запись более 70 наборов символов (не зависит от ОС). Более 150 поддерживаются с помощью DIConverters.

DIHtmlParser is a fully Unicode aware HTML parser for Borland Delphi. DIHtmlParser gives you the power and flexibility to process HTML documents fast. It fully supports Unicode / WideStrings for both reading and writing and can convert all commonly used character sets natively, i.e. without the help of the operating system.

The core TDIHtmlParser component reports all types of HTML pieces plus a few HTML-related pieces to the application, which include HTML-Tags, text, comments, scripts, styles, titles, Custom-Tags, SSI, and more. All tags are parsed into a TDITag object which can be easily queried for attribute names and values. Filtering is available for all HTML pieces as well as individual tags.

The TDIHtmlParser component also accepts plugins, which the parser feeds with information durint the parsing. Any number of plugins can be assigned to a parser, and each TDIHtmlParserPlugin has the same processing and filtering options as TDIHtmlParser. Besides this, the plugins' and the parser's settings are completely independent of each other.

This plugin architecture allows to easily extend TDIHtmlParser. A number of ready-made plugins are included and cover frequent HTML tasks: extract e-mail addresses, links and titles, automatically write HTML, monitor character set changes, track HTML tables, implement HTML events, and more. Full source code is provided for all plugins as a starting point for writing new ones.

The HTMLParser reads HTML from memory or from any TStream object, using an internal buffer to speed up the reading. It automatically decodes entities and correctly handles preformatted text. "Smart parsing" skips unused HTML pieces and results in an amazingly fast parsing performance.