preparser does not know about python notation for complex numbers

[jasongrout]

sage: 1j
------------------------------------------------------------
   File "<ipython console>", line 1
     Integer(1)j
               ^
SyntaxError: invalid syntax

but in python:

sage: preparser(False)
sage: 1j
1j
sage: type(1j)
<type 'complex'>

Note that this does work now:

sage: 1rj
1j
sage: 1rj == complex('j')
True

Component: basic arithmetic

Issue created by migration from https://trac.sagemath.org/ticket/4501

[jasongrout]

comment:1

According to http://docs.python.org/reference/lexical_analysis.html#id7, python supports imaginary numbers being declared as:

imagnumber ::= (floatnumber | intpart) ("j" | "J")

[jasongrout]

comment:2

[10:57] <mhansen> jason-: I don't have time to make a patch now, but the following two lines around line 805 in sage/misc/preparser.py are the correct fix if we want 1j to return a <type 'complex'>:
[10:57] <mhansen>                 elif i < len(line) and line[i] == 'j':
[10:57] <mhansen>                     pass
[10:57] <jason-> I can make a quick patch
[10:57] <jason-> It needs to support "J"
[10:57] <jason-> and also work for floating numbers too
[10:58] <jason-> and then there's the issue of if we want to construct Sage complex numbers instead of python complex numbers

[jasongrout]

comment:3

and the next line:

[10:58] <mhansen> Just change == "j" to in 'jJ'

[robertwb]

comment:4

I think 1j should be a Sage complex number.

[williamstein]

comment:5

I think 1j should be a Sage complex number.

And I think it should return a Python complex number. :-)

[robertwb]

comment:6

Your reasoning (that the userbase for this feature is almost entirely numeric users) has convinced me.

[robertwb]

comment:8

See #5079