An Internet-Scale Service for Publishing and Locating XML Documents

Try it!

  • Main Page
  • Publish XML Documents
    • Complete the form below to locate XML documents.
    XPath query 1,2,3
    XML DTD 4
    Replica type5 Nearest Primary

    1. A user is returned the global docids of matching documents. The global id may contain the publisher's IP address, URL, etc., from where the document can be requested.
    2. The contents of the XML documents are not stored by psiX.
    3. Currently, a limited subset of XPath is supported. (See Grammar.) If you pose a query such as /*/* that has no elements, the system will pick just one of the tags that match the '*' of the first location step (given that a DTD is provided) to select an index. Your results may be incomplete.
    4. If DTD is absent, then the signature is computed using the distinct tags in the query. Thus, the structural information is ignored.
    5. Replica type indicates whether a PlanetLab node that process the query should fetch the (estimated) nearest replicas of index nodes or the primary copies of the index nodes. By fetching nearest replicas, the query may be processed quicker, but the results may be incomplete. The improvement in processing time depends on the accuracy of the network coordinates used by the DHT.