cdd-dev team mailing list archive

[Timothy Cook <timothywayne.cook@xxxxxxxxx>]

Hi Eduardo,

2012/4/8 Eduardo César <xcesar@xxxxxxxxx>:
> Hi Tim,
>
> The process that takes place in the conversion is basically a big loop that
> translates from the xmind file to the XSD files.
>
> It goes from one xmind sheet to the next, and each sheet corresponds to a
> XSD file that will be generated. The name of the XSD file is the name of the
> main topic of each sheet.
>

Okay, actually since I had played with the mekk.xmind tool
http://pypi.python.org/pypi/mekk.xmind  before, I knew this.  However,
I thought it would be good to start the public thread this way.

First thing to note is that you should only be concerned with the CCD
sheet and the MetaData sheet.

> It converts a specific structure that is present in the CDD file: At the
> first level we have the names that translates to the each 'topic'. This name
> here is actually each one of the the subtopics of the main topic of the
> current processed sheet of the xmind file. Attached to the 'topic' in the
> second level, we have the 'subtopics' and in the third level, attached to
> the second, we have the 'attributes'. At some point, I will make it more
> general, capable of converting a more diverse xmind structure.

Hopefully, only needing to do the two sheets will make it easier.  I
put a simple-test.xmind file in the ccd_repository folder as a first
stage test.  At this point, the converter doesn't follow the subtopics
at all.  The CCD.xsd only contains:

<?xml version='1.0' encoding='ASCII'?>
<xsd:schema xmlns:xsd="http://www.w3.org/2001/XMLSchema"/>

>
> It generates a very simple (but valid) XSD, one that is not yet the most
> adequate to express the complexity of the xmind file. I need to do a smarter
> algorithm that will take in account the variety of the XSD data strucures
> and, for that, I'm taking a more in-depth look at the XML Schema
> specification.

A thought "might" be to just output a text file first that lists all
the topics/subtopics and attribute levels.  Then be concerned about
XSD output.  At least in my opinion, getting the parsing correct of
the various levels is the most difficult.  Your thoughts?


>
> Regarding the comments, I was already planning to write then in english the
> next time. This code that I sent is just a first draft, more for me to
> experiment. As you could see, there is almost no comments and the code is
> not exacly descriptive. Those too will be improved in the upcoming versions.
> :)

Thanks.  :-)

Cheers,
Tim