SAMPLE QUESTIONS FOR THE XML LECTURES

Is it possible to validate an XML document that is not well-structured?

Is it possible to validate an XML document that does not specify a DTD?

Is it possible for a well-structured XML document not to specify a DTD?

Can an element in XML have two sub-elements with the same name (tag)?

Can an element in XML have two attributes with the same name?

You are given a relational database with a single table:
EMP(ID[32 bit integer],
    Name[String with length uniformly distributed 0-30 characters],
    Dept[String with length exactly 8 characters])
If this table has 1000 tuples, estimate its size.
Estimate the size of an (uncompressed) XML representation of this table.

Consider an XML database structured as a complete 16-ary tree with 4 levels.
(If you draw this tree correctly, you should have 1 + 16 + 256 + 4096 nodes
in all).  Let each node require 256 bytes of storage.  Assume 4K as page
size.  Consider a query that asks for all (immediate) children of a specified
node (which is already in memory through some previous operation).  
If nodes are stored in pre-order (depth-first), how many pages must be
fetched in response to this query?
Can you think of an order in which to store the nodes so that exactly one
page is fetched in response to this query?

You should be able to work out what the result would be for an XQuery
expression similar to those considered in class applied to a small data set.

Consider a directory of law firms expressed as a single XML document, as
follows:
<law-dir>
  <firm>
    <partner> eagle </partner>
    <partner> eagle </partner>
    <partner> beagle </partner>
    <address>
       <street> 123 Main </street>
       <city> Ann Arbor </city>
    </address>
  </firm>
  <firm>
    <partner> vulture </partner>
    <partner> eagle </partner>
    <phone> 1-800-SUE-THEM </phone>
    <address>
       <street> 999 Main </street>
       <city> Ann Arbor </city>
    </address>
  </firm>
</law-dir>


a. Suppose the element "vulture" has been assigned an interval of (26, 29).
   What can  you say about the start and end positions of the interval
   corresponding to each firm, and for each "Ann Arbor" city element?
b. Evaluate the query, where the above data is in "ex.xml"
   FOR $c in doc("ex.xml")//city
   LET $f = doc("ex.xml")//firm[city=$c]
   RETURN {Count($f/partner)}
c. Compare with the output of:
   FOR $f in doc("ex.xml")//firm
   RETURN {Count($f/partner)}