How to engineer a scientific paper system
Are there products out for this, should we produce one?
In short, I would like to have Zope assist me in producing a
So how can I produce pointers from within the text to objects holding
more information about these pointers, i.e. a referene within a
paragraph is linked to the bibliography.
My setup so far is a folder structure, with DTML Documents for
content that have Properties holding the references in token
(semi-colon) separeted lines.
Methods are used to parse the folder (having chapter properties) and
DTML Documents in order to get a table of content structure and
html-styled reference list respectively.
The method I wrote are much of the type of dtml_tree using dtml-in to
identify the object, produce something with that object, and then
calling the method again from within the method. Sometimes, though, I
get an infinite recursion. This requires lots of debugging, as
the loads of methods intermingled with standard puplishing methods are
hard to overlook. The parsing techniques would need exception
handling and whenever I split a method up, the value is passed on the
the request namespace (that might produce further havoc in a newbie
scope of understanding).
A class object system instead of scripts should
work better and should also be easier to enrich with additional
qualities later in the development process (as it still is the
paper first that should get done instead of the software system).
So maybe it is time for a full revision, letting classes produce the
work. Are there other ways to achieve this? Maybe should I a produce
something such as a lightened ZWiki product for that purpose. Or,
should all content go into SQL tables?
What steps should I take next? Any further ideas of
Zope maillist - [EMAIL PROTECTED]
** No cross posts or HTML encoding! **
(Related lists -