Notice: The ATagParser will no longer be available for sale as of 07/31/2007 but will still be actively developed and supported by its author.

Are you tired of trying to figure out overly complex tag-based parsing engines?

Then look no further!

ATagParser is a non-visual tag based parsing component for Delphi versions 3, 4, 5, 6 & 7 and is source compilable with Delphi 8, Delphi 2005, 2006 and C++ Builder 6. ATagParser is the ideal component for parsing ASP, ASPX, HTML, XHTML, XML and RSS documents as well as any other file type based on a tag format. ATagParser comes with a full compliment of events, methods and properties and is very simple to use.

New!

The long promised cascading style sheet parser is now included with ATagParser. It allows you to parse style sheets from files as well as inline styles associated with HTML tags!

• Updated TagExplorer demo allows you to easily create ATagParser desencents
• Recognizes PHP code blocks with new PHP element type
• Parse Delphi WideStrings via the new ParseWideString method
• Parses UTF-8 and UTF-16 via the new ParseStream and ParseFile methods
• Attribute and tag integer constants for easier case statement building and readability
• New component for saving and loading ATagParser settings to a stream or file
• Ability to load and parse very large files via buffered I/O or file mapping
• New InLink property allows you to easily extract HREF text descriptions
• Fully object oriented - all data and element related events return assignable objects
• Robust code with silent exception logging capabilities
• All data and error events can be made active / inactive at design time and run time
• OnEnter and OnExit events for all attributes, tags, text and data events
• Can be loaded from a file, stream, string or a buffer.
• Can parse arbitrary memory positions via the ParsePointer method - ideal for memos
• Parsing can be aborted at any time for spot parsing needs
• Parsing can be limited to a single tag type, tag name, attribute, text, etc..
• Can selectively parse attributes on a tag-by-tag basis
• Position information for all data types including attribute names and values
• Look back features for tags and text (PreviousTag, PreviousText)
• One instance can handle multiple parsing scenarios with the ParserID property
• Recognizes PRE tags and honors their formatting
• Automatically corrects malformed tags and attribute name=value pairs
• Optionally normalizes all text and removes deadspace
• Optionally Translates ISO8859-1 [Latin-1], hex and numeric entities
• Optionally returns an extensive range of error information
• OnPercent event allows for easy updating of progress bars
• Tag and attribute name determination is completely case independent
• Returns hash values for all tags and attribute names
• Returns integer indexes for all attributes, tags, data and text
• Includes extensive help and documentation
• Small foot print
• Thread safe

ATagParser was designed to allow you to interact with the parsing engine via events and persistent objects.

All events that return information contain an Abort variable that can be set to instruct the parsing engine to exit immediately once a condition has been met. In the following screenshot, the first attribute value that equals 10 will terminate the parsing process.

Of course, you can turn element types on and off from within any of the events as well as filter tags names via the OnTagName event. This event actually allows you to allow or disallow the parsing of attributes for any tag.

A lot of time and effort was spent making ATagParser as fast and accurate as possible. As with any speed-vs-accuracy scenario, there were tradeoffs in the design of ATagParser. The parsing engine is not burdened with any unnecessary actions. Its sole purpose is to dissect the document as quickly and as accurately as possible and to return the information in a form that can be further manipulated.

ATagParser moves forward through the document examining each character to determine how to ultimately apply it to the various events. Each movement takes an average of .12µs per character on a 500Mhz CPU and .05µs on an 800Mhz CPU (that's the equivalent of 20 characters per millionth of a second!)

ATagParser includes a property called VirtualMode which can be used to make ATagParser even faster! In VirtualMode, ATagParser is unburdened with having to deal with anything other than tag names, hash values, location and type.

The speed increase is simply phenomenal. We acheived the following results while testing.

By any standard, that's fast!

* µs = microseconds or millionths of a second
* Your results may vary and are dependent on several factors including hardware configuration

Right out of the box, ATagParser supports the following tag types.

• HTML tags
• XHTML tags
• XML tags
• ASP code blocks
• ASP.NET tags
• CDATA sections
• PHP code blocks
• JSP tags
• WML tags
• SiteSpring site tags
• Your own perhaps?

Which make it perfect for the following uses:

• RSS feed parsing
• HTML import for applications
• HTML compression
• Website information aggregation
• Website content management
• HTML or custom tag based rendering engines
• Custom report generators based on a tag format
• Custom tag based scripting engines
• Custom tag based file formats
• ASP, HTML and XHTML editors
• Text, script and style sheet extraction utilities
• Text filters and spell checkers>
• Web site search engines
• Web site robots
• Proxy servers

The ATagParser v2.04+ package costs $39.95 for a single-user license with clean, well commented source code, extensive help files and support for the lifetime of the component. Site licenses cost $99.95 and include all features of the single-user license.