Obtain a Sage object from the input string by evaluating it using Sage. This means calling eval after preparsing and with globals equal to everything included in the scope of from sage.all import *.).
INPUT:
EXAMPLES: This example illustrates that preparsing is applied.
sage: eval('2^3')
1
sage: sage_eval('2^3')
8
However, preparsing can be turned off.
sage: sage_eval('2^3', preparse=False)
1
Note that you can explicitly define variables and pass them as the second option:
sage: x = PolynomialRing(RationalField(),"x").gen()
sage: sage_eval('x^2+1', locals={'x':x})
x^2 + 1
This example illustrates that evaluation occurs in the context of from sage.all import *. Even though bernoulli has been redefined in the local scope, when calling sage_eval the default value meaning of bernoulli is used. Likewise for QQ below.
sage: bernoulli = lambda x : x^2
sage: bernoulli(6)
36
sage: eval('bernoulli(6)')
36
sage: sage_eval('bernoulli(6)')
1/42
sage: QQ = lambda x : x^2
sage: QQ(2)
4
sage: sage_eval('QQ(2)')
2
sage: parent(sage_eval('QQ(2)'))
Rational Field
This example illustrates setting a variable for use in evaluation.
sage: x = 5
sage: eval('4/3 + x', {'x':25})
26
sage: sage_eval('4/3 + x', locals={'x':25})
79/3
You can also specify a sequence of commands to be run before the expression is evaluated:
sage: sage_eval('p', cmds='K.<x> = QQ[]\np = x^2 + 1')
x^2 + 1
If you give commands to execute and a dictionary of variables, then the dictionary will be modified by assignments in the commands:
sage: vars = {}
sage: sage_eval('None', cmds='y = 3', locals=vars)
sage: vars['y'], parent(vars['y'])
(3, Integer Ring)
You can also specify the object to evaluate as a tuple. A 2-tuple is assumed to be a pair of a command sequence and an expression; a 3-tuple is assumed to be a triple of a command sequence, an expression, and a dictionary holding local variables. (In this case, the given dictionary will not be modified by assignments in the commands.)
sage: sage_eval(('f(x) = x^2', 'f(3)'))
9
sage: vars = {'rt2': sqrt(2.0)}
sage: sage_eval(('rt2 += 1', 'rt2', vars))
2.41421356237309
sage: vars['rt2']
1.41421356237310
This example illustrates how sage_eval can be useful when evaluating the output of other computer algebra systems.
sage: R.<x> = PolynomialRing(RationalField())
sage: gap.eval('R:=PolynomialRing(Rationals,["x"]);')
'Rationals[x]'
sage: ff = gap.eval('x:=IndeterminatesOfPolynomialRing(R);; f:=x^2+1;'); ff
'x^2+1'
sage: sage_eval(ff, locals={'x':x})
x^2 + 1
sage: eval(ff)
...
RuntimeError: Use ** for exponentiation, not '^', which means xor
in Python, and has the wrong precedence.
Here you can see eval simply will not work but sage_eval will.
TESTS:
We get a nice minimal error message for syntax errors, that still points to the location of the error (in the input string):
sage: sage_eval('RR(22/7]')
...
File "<string>", line 1
RR(Integer(22)/Integer(7)]
^
SyntaxError: unexpected EOF while parsing
sage: sage_eval('None', cmds='$x = $y[3] # Does Perl syntax work?')
...
File "<string>", line 1
$x = $y[Integer(3)] # Does Perl syntax work?
^
SyntaxError: invalid syntax
Return a native Sage object associated to x, if possible and implemented.
If the object has an _sage_ method it is called and the value is returned. Otherwise str is called on the object, and all preparsing is applied and the resulting expression is evaluated in the context of from sage.all import *. To evaluate the expression with certain variables set, use the vars argument, which should be a dictionary.
EXAMPLES:
sage: type(sageobj(gp('34/56')))
<type 'sage.rings.rational.Rational'>
sage: n = 5/2
sage: sageobj(n) is n
True
sage: k = sageobj('Z(8^3/1)', {'Z':ZZ}); k
512
sage: type(k)
<type 'sage.rings.integer.Integer'>
This illustrates interfaces:
sage: f = gp('2/3')
sage: type(f)
<class 'sage.interfaces.gp.GpElement'>
sage: f._sage_()
2/3
sage: type(f._sage_())
<type 'sage.rings.rational.Rational'>
sage: a = gap(939393/2433)
sage: a._sage_()
313131/811
sage: type(a._sage_())
<type 'sage.rings.rational.Rational'>