IBM Unveils XML Search Engine, Parser for C, Security Tools.

By Rachel Chalmers The XML 99 Europe show in Granada, Spain, has provided the platform for three XML technology announcements from IBM Corp. First is xCentral, a smart search engine in the tradition of IBM's jCentral Java search engine. IBM XML expert Simon Phipps explains that xCentral rests on top of "a monster web crawler. We point a harvesting application at the webcrawler's results and interpret and understand them in a way that is useful." Users of xCentral get hits classified by content and type. "It knows what sort of thing it found," says Phipps, "whether that's a document type definition, metadata or an XML document. If it is a document, xCentral will check that the XML is well-formed and valid. If you're looking for XML information for development or specification work, I expect you'll want to put this straight onto your Favorites list." Next up is Big Blue's XML Parser for C and C++ (XP4C). "We've heard rumors from third or fourth parties that there are still people out there who are not using Java," quips Phipps. XP4C is being made available under the same license agreement as IBM's XML Parser for Java. Phipps reports that the parser has already attracted positive comments from developers and competitors at the XML 99 show. The third and final announcement concerns IBM's XML Security Tools. "This goes to show how active a part IBM is playing in the standards debates," says Phipps, "for example, how do you digitally sign an XML document? Ordinarily, you would take any file and use a hashing algorithm on it." Next, the sender would encrypt the result as a signature. Both signature and hashed document would be sent to the recipient, who would compare the two files. If they're still the same, the document is signed. "The trouble is that XML doesn't care about whitespace," Phipps explains. "That means two documents can be equivalent but not identical." Accordingly, IBM has proposed a method of digitally signing XML documents using the Document Object Model (DOM). A function called DOMHASH can be used to produce a canonical form of the document. The user then signs that version. Other elements of IBM's new XML security toolkit include the ability to encrypt individual elements of a document. A research firm, for example, could freely distribute abstracts but charge for keys to decrypt the body text. A prototype of the toolkit is now available for free download from IBM's alphaworks site for developers. Big Blue hopes to incorporate feedback from alphaworks, add more encryption and access control features, and release a full version of the XML Security Suite later this year.