Hello!

I am working on a XML Software Use Case Specification markup.
My idea is to create XSLT schemas to transform this XML to HTML and PDF.

I done a HTML version of a XML use case to see how i would like it be transformed, then i started to develop a XSLT. There is a situation, on Use Case Extensions that i am facing with complex indexes. An Extension refer to a step in main flow. Example:

Main Flow
1. The actor inputs its username and password.
2. The system processes user login.

Extensions (non-well written)
1a. Username doesn't exist
1a1. The system notifies Actor ...
1a2. The actor modifies username ...
1b. Password is invalid
1b1. The actor informs password again
1b1a. Password invalid again (A SUBEXTENSION)
1b1a1. The system blocks user account
2a. ...

My XML representation for above scenario is:

<MainSuccessScenario>
<Step>The actor inputs its username and password.</Step>
<Step>The system processes user login.</Step>
</MainSuccessScenario>
<Extensions>
<Extension mainstep="1">
<Description>Username doesn't exist</Description>
<Step>The system notifies Actor ...</Step>
<Step>The actor modifies username ...</Step>
</Extension>
<Extension mainstep="1">
<Description>Password is invalid</Description>
<Step>
The actor informs password again
<Extension>
<Description>Password invalid again</Description>
<Step>The system blocks user account</Step>
</Extension>
</Step>
</Extension>
...
</Extensions>

The default is:
- Extension description index is mainstep attribute or parent index,
plus a sequential letter.
- Extension step index is description index, plus a sequential number.

Can XSLT generate these indexes for me? If so, how?
I am also opened for suggestions on xml markup to make things simpler.

Thank you