UNIX for Advanced & Expert Users

Hi
I have an xml file which has multiple xml records..
I don't know how to read those records and pipe them to another shell command

the file is like

<abc>z<def>y<ghi>x........</ghi></def></abc> (1st record)
<jkl>z<mno>y<pqr>x........</pqr></mno></jkl> (2nd record)

Each record end with a "newline" character i belive but i am not able to see the newline character..
I used cat -vte command to view it

I am planning to grep the record by means of a common string
say "Z" found in all the records..

take the count of the number of records and then send each record by reading one at a time..and then pipe it to say a "sed" command...


Is there a mechanism...can somebody help out..


Thanks in advance!!!

If the XML is not conveniently line-oriented, you best bet is probably to use a real XML parser to extract the stuff you want. You might end up not wanting to use the shell or sed at all once you figure it out, although the learning curve might seem a bit steep at first. Search for threads demonstrating the use of e.g. xsltproc

Ya thats fine ...but i want some script statements that would conveniently help me parse the file containing the xml records record by record..

Another way around if i think of the information that i have provided above is that..say
I have certain information within the tag

<PQRS><ABCD>hi how r u</ABCD></PQRS>
and i need to exract this info and pipe it to another sed command for futher formatting

can xsltproc as suggested by era help in any way? Can any one suggest doing that


I need to extract day from only 1 tags say <ABCD></ABCD> and this tags is enclosed within others..

Can some 1 suggest a way to extract onlt this 1 tag's data ...

I have already succesfully tried sed command and ...but the command is failng when the tag is occuring in a large xml record...I mean a particular record is
occuring which is unusually large in its size

Unless your XML document is very small, you should use an XML processor to extract the required elements and attributes. Classic UNIX tools such as sed and awk are not really suited to manipulating XML documents nor were they designed or intended for this purpose.

xsltproc is one such processor which uses an XSL stylesheet to transform the document into the required output. There are a large number of other XML document transformation tools available. Just do a Web search.