Hi, I have been toying around with the LaTeX printing over the last couple of days and in the process have put together a few (potentially idiosyncratic) changes to the printing class. I would be interested to know what everyone thinks of this (i.e., is there enough interest that it is worth making them mainstream).
Firstly, I added support for fractional power folding. If an exponent
is of the form p/q where both p and q are rational (and q is not 2, as
then a square-root is appropriate) then the fraction is folded. Here
is an example:
>>> ex = sympify('x^(3/4)')
>>> latex(ex)
'$x^{\\frac{3}{4}}$'
>>> latex(ex,foldFracPowers=True)
'$x^{3/4}$'
As you can see instead of a fraction a plain old division symbol is
used. This is often nice when handling fractions of the form p/q^a as
it saves all-important vertical space and makes the power much more
readable (as the font size will be larger).
Secondly, I have added support for function bracket folding. If a
function (lets say tan) has one and only one argument which does not
itself need brackets, then it is possible to fold them. Another
example if you will:
>>> ex = sympify('tan(2*x)')
>>> latex(ex)
'$\\operatorname{tan}\\left(2 x\\right)$'
>>> latex(ex,foldFuncBrackets=True)
'$\\operatorname{tan}2 x$'
This is done by the addition of a new function
_needs_function_brackets, which is a more generalised form of
needs_brackets, namely one that allows Mul and Pow to not require
them. I am still testing this, however, so there may be some cases
where it (erroneously) omits brackets. I am still working on a way to
handle tan(x)*cos(x) (currently '$\\operatorname{cos}x \
\operatorname{tan}2 x$') but it might come down to either a) folding
tan(x*cos(x)) to tan xcos x or b) folding tan(x)*cos(x) to tan x cos x.
Finally, I would like to apologise for my use of camelCase, I'll fix
that soon.
Regards, Freddie.
PGP.sig
Description: This is a digitally signed message part
