RE: unique-sequence-numbering

> Short of an elegant solution, here's a simple way to do it:
> 
> Step 1: go through your XML file and create another with the student 
> elements duplicated as necessary.
> 
> Step 2: go through the new file and create the output -- sequence 
> numbering is now trivial.

Doing a transformation in two phases often *is* the elegant solution.

Of course with the node-set() extension you can do this with one
stylesheet.

There's another solution here, which is to calculate the sequence number
as position() plus the number of preceding students that have
address="YYY":

position() + count(preceding-sibling::student[address='YYY'])

But this may have O(n^2) performance (i.e. it won't run at "wire-speed"
unless you have a very good optimizer). So the two phase solution may be
better.

Michael Kay
Software AG
home: Michael.H.Kay@xxxxxxxxxxxx
work: Michael.Kay@xxxxxxxxxxxxxx 
> 
> 
> On Wed, 7 Aug 2002 subbu@xxxxxxxxxxxx wrote:
> 
> ..Sorry for reposting..The subject was wrong!!
> ..i have an xml file which looks like below.
> ..<?xml version="1.0"?>
> ..<students>
> ..	<student id = "101">
> ..		<name>A</name>
> ..		<address>XXX</address>
> ..	</student>
> ..	<student id = "201">
> ..		<name>B</name>
> ..		<address>YYY</address>
> ..	</student>
> ..	<student id = "301">
> ..		<name>C</name>
> ..		<address>XXX</address>
> ..	</student>
> ..	<student id = "401">
> ..		<name>E</name>
> ..		<address>YYY<address>
> ..	</student>
> ..</students>
> ..
> ..And i have to deliver a CSV file that has the following 
> fields. .. ..Student-id , SequenceNumber , name, address. .. 
> ..Sequence number is an automatically generated  number ( 
> should start from 1 to 
> ..end of Students )..
> ..
> ..so a typical CSV FIle would be..
> ..
> ..101,1,A,abc
> ..102,2,B,bcd
> ..103,3,C,cde
> ..104,4,D,efg
> ....
> ..Now There is a business rule that says that if a student 
> has an address value 
> ..that is equal to 'YYY',  then show the complete information 
> *line* twice, with 
> ..only the sequence number changed..
> ..That is
> ..101,1,A,abc
> ..102,2,B,bcd
> ..102,3,B,bcd  <-- repeated again with sequence number 
> changed ..103,4,C,cde ..104,5,D,efg ..104,6,D,efg <-- 
> repeated again with sequence number changed .. .. ..I am 
> using the position() function to get the serial number..This 
> is where the 
> ..problem starts..
> ..If i put a condition that
> .. if address == 'YYY' ( just pseudo code. dont mind the 
> syntax ) ..{ ..value-of-select --> @id,position(),name,address
> ..		  @id,position()+1,name,address
> ..}
> ..else
> ..{
> ..value-of-select --> @id,position(),name,address
> ..}
> ..
> ..I end up getting something like this
> ..
> ..101,1,A,abc
> ..102,2,B,bcd
> ..102,3,B,bcd  <-- repeated again with sequence number 
> changed ..103,3,C,cde  <-- same sequence number repeated... 
> That is the position .. ..104,4,D,efg ..104,5,D,efg <-- 
> repeated again with sequence number changed .. ..and what i 
> want should look like .. ..101,1,A,abc ..102,2,B,bcd 
> ..102,3,B,bcd  <-- repeated again with sequence number 
> changed ..103,4,C,cde  <-- this is not showing up position 
> but prev-seq+1 ..104,5,D,efg ..104,6,D,efg <-- repeated again 
> with sequence number changed .. ..so the requirement tells, 
> if you find address equal to 'YYY', then add it show 
> ..it once more with a sequence number -> prev-seq-num +1  and 
> then every 
> ..subsequent sequence number should be consequitive. The 
> sequence number is a 
> ..unique number ( from 1 to End of students ).
> ..
> ..I can not think of a way to do this.This looks totally like 
> a dead end to me. ..I shall be highly thankful if some one 
> can please help me out of this 
> ..situation...
> ..Please let me know if i am missing out any information that 
> would make the 
> ..question clearer..
> ..
> ..Thanks a ton!!!!!!!!! 
> ..
> ..--------------------------------------------------------------
> ..Sent with "Me-Mail", Boltblue's FREE mobile messaging 
> service. ..http://www.boltblue.com .. .. .. XSL-List info and 
> archive:  http://www.mulberrytech.com/xsl/xsl-list
> ..
> 
> -- 
> 
> Jeff Kenton
> DataPower Technology, Inc.
>  *** Wire Speed XSLT ***
> 
http://www.datapower.com/products.shtml



 XSL-List info and archive:  http://www.mulberrytech.com/xsl/xsl-list


 XSL-List info and archive:  http://www.mulberrytech.com/xsl/xsl-list