Hi,

Le samedi 22 février 2020 à 09:33 -0800, Nils Bruin a écrit :
> Sorry, code to produce the sobj is already included in the package.
> From jones.py:

>     def _init(self, path):
>         """
>         Create the database from scratch from the PARI files on John
> Jones's
>         web page, downloaded (e.g., via wget) to a local directory,
> which
>         is specified as path above.
> 
>         INPUT:
> 
> 
>         -  ``path`` - (default works on William Stein install.)
>            path must be the path to Jones's Number_Fields directory
>            http://hobbes.la.asu.edu/Number_Fields These files should
> have
>            been downloaded using wget.
> 
> 
>         EXAMPLES: This is how to create the database from scratch,
> assuming
>         that the number fields are in the default directory above:
> From a
>         cold start of Sage::
> 
>                 sage: J = JonesDatabase()
>                 sage: J._init()   # not tested
>                 ...
> 
>         This takes about 5 seconds.
>         """
>         from sage.misc.misc import sage_makedirs
>         x = PolynomialRing(RationalField(), 'x').gen()
>         self.root = {}
>         self.root[tuple([])] = [x - 1]
>         if not os.path.exists(path):
>             raise IOError("Path %s does not exist." % path)
>         for X in os.listdir(path):
>             if X[-4:] == "solo":
>                 Z = path + "/" + X
>                 print(X)
>                 for Y in os.listdir(Z):
>                     if Y[-3:] == ".gp":
>                         self._load(Z, Y)
>         sage_makedirs(JONESDATA)
>         save(self.root, JONESDATA + "/jones.sobj")
> 
> Modulo collecting the text source that aren't included (but readily
> available on a web site), the code to compile the database is already
> there! You might want to check with the author if he/she is OK with
> you downloading and distributing the text files separately, but for
> robustness I think you'd be doing the community a favour by wrapping
> the data in a "source distribution" too.

If I understand well, there's a series of build steps :
  pari code -> data files -> sobj
and the last step is jones.py.

But then what I want is the pari code, as that's the source.

(Unless computing the data files is incredibly long, in which case I'll
ship both the pari code and the data files, with a nice notice to
explain that I ship pre-build for a good reason.)

It's not clear which data files have been used to generate the current
sagemath package : https://hobbes.la.asu.edu/Number_Fields/ points to
several pages (including a 404), and on each of these pages, there's a
link to .gp files which must be the ones getting used, if I understand
well.

And how is versioning managed?

JP

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/f36e52ce492d702904f9817f6c227da4e0ff546b.camel%40laposte.net.

Reply via email to