[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Axiom-developer] How was axiom's bookvol10.2.pamphlet file generate
Re: [Axiom-developer] How was axiom's bookvol10.2.pamphlet file generated?
Sun, 04 Sep 2011 18:09:35 -0400
It occurs to me that I didn't fully answer your question about
the complete graph information.
The complete graph is VERY complex and I could not find a way
to visualize the whole thing. I'm getting better at HTML5,
In any case, the complete dependence graph information is
available in src/algebra/Makefile.pamphlet. The layerpic
chunks have all of the edges listed although mostly as
comments. Only the "supporting edge" is not commented.
To construct a "supporting edge" you need to find out every
dependence and then find the one in the highest layer. This
forces a new domain to be one layer higher and is the edge
that supports the new domain. So everything in layer 14
depends on something in layer 13 and nothing in or above
Adding new algebra involves finding the supporting edge.
I have written some code for this but it is not very robust
so it is not part of the system.
If you really want to construct the "FULL" graph you can use
the dot information in src/algebra/Makefile.pamphlet. All you
need to do is uncomment the dot commands in the layerpic chunks.
On Sun, 2011-08-28 at 16:19 +0400, Grigory Sarnitskiy wrote:
> 08.08.2011, 00:25, address@hidden:
> > The pamphlet file was hand generated. Pamphlet files are the source
> > code of the system.All of the pamphlet files are source code and are
> > hand generated in latex.
> > The system is moving to using pure latex rather than using noweb for the
> > pamphlet format. A new latex "chunk" environment was written to support
> > this. All of the new latex macros are in the axiom.sty file. Volume 10.2
> > (the category structure) is still in noweb format for the moment but this
> > will change.
> > Axiom's source structure is being moved from the "tree-of-tiny-files"
> > structure into a "literate software" structure. The process involves
> > rewriting the source code into straight common lisp. Each function is
> > rewritten and placed into one of the pamphlets in the books directory.
> > This phase takes a while as there are about 1M "things of code" that
> > need to be rewritten. Work is currently being done in volume 5 (the
> > interpreter) and volume 9 (the compiler).
> > The next step is to organize and document the system using these books.
> > The ultimate idea is that you will be able to read the system as a
> > series of novels, moving from ideas to implementation.
> > Volume 10.2 (categories) is the category structure of the system.
> > It is currently in alphabetical order but that will change in whatever
> > way needed to support the human documentation.
> > Tim Daly
> And how were graphs generated from this pamphlet? I mean dotfull and dotabb.
> The latter is supplied with axiom sources (in axiom/src/axiom-website), but I
> want to have dot source for the full name graph.
> Axiom-developer mailing list
|[Prev in Thread]
||[Next in Thread]|
- Re: [Axiom-developer] How was axiom's bookvol10.2.pamphlet file generated?,