On Mon, Sep 7, 2009 at 8:23 AM, John H Palmieri<jhpalmier...@gmail.com> wrote:
>
> On Sep 7, 12:54 am, Ondrej Certik <ond...@certik.cz> wrote:
>> On Fri, Sep 4, 2009 at 2:28 PM, John H Palmieri<jhpalmier...@gmail.com>
>> wrote:
>>
>> I tried your patch (I made a sphinx extension out of it), but it
>> didn't work for me --- the backsubstitution to docstringlines failed
>> in the extension (maybe sphinx is inconsistent here), but this can be
>> fixed easily by just doing
>>
>> docstringlines[:] = [s]
>
> That's strange. Maybe the issue is that mine is interacting with the
> Sphinx autodoc extension, which passes "docstringlines" as an
> argument: this is a list of strings, the lines of the doc string that
> autodoc has extracted. You're using "source-read" which passes a
> list containing one element, the entire docstring. Maybe you can
> change the first line of your code to
>
> s = source[0]
>
> and then the last line would be
>
> source[0] = s or source = [s]
>
> I couldn't figure out how to use "source-read" for the Sage reference
> manual, since "source" in our case is some autogenerated string that
> doesn't contain the actual docstring -- you have to interact with
> autodoc to get that.
>
>> worse problem is that I want to change it to math:`sd`, not just `sdf`
>> (I have not figured it out how to do that using your indices
>> approach). In any case, I rewrote it so that it works for me,
>> attached.
>
> Oh, I like your approach. I see that you're not worrying about `$ or
> $` in the code, but there are strings like that in the Sage code --
> e.g., ``$SAGE_ROOT...`` -- so I had to test for this and not replace
> such dollar signs.
Yeah, I will use \$ for that, if I ever need it. But yes, I agree it
should be more clever --- I think we can even combine my and your
approach.
>
>> So that's a very convenient solution, as I can ship it with my sphinx
>> notes and it will just work. Later on, after I gain more experience,
>> I'll ask on the sphinx list, if they want to ship this with sphinx
>> itself.
>
> Sounds good. I'm curious: for your own use, why don't you use
>
> default_role = 'math'
>
> so you can use `x=y` instead of :math:`x=y` ? Maybe a better
I just wasn't sure if this can't break other people's code, let's say
if they take my .rst file and include it in their project (where they
use a different definition for `...`).
> question, which just occurred to me: is it worth adding some sort of
> parameter to your extension so that it replaces $x=y$ by
> either :math:`x=y` or `x=y`, depending on that parameter?
Yeah, that should be easy to implement. Why would you ever need it
though? I thought you always need :math:`...` at the end of the day.
Btw, I just found a bug in my (and your) code, I have to change:
- s = re.sub(r"({[^{}$]*\$[^{}$]*\$[^{}$]*})", repl, s)
+ s = re.sub(r"({[^{}$]*\$[^{}$]*\$[^{}]*})", repl, s)
that's because there could be things like \mbox{adflj $x$ and $y$ and $z$}.
After this, here I converted some of my notes (I had them in xml and
used my own converter to generate latex or html, that was from before
sphinx was born). So I just wrote a simple converter to rst and
everything just works. It's awesome:
http://certik.github.com/theoretical-physics/book/src/qft.html
I will now convert all my physics and math notes into sphinx. Right
now I have them scattered over dozens latex documents, sometimes I
even lost the source, only have pdf. So this will now change, I put
everything into one git repository, and host it on github.
Ondrej
--~--~---------~--~----~------------~-------~--~----~
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to
sage-devel-unsubscr...@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---
