numlit

exp

Introduced in the 1977 ANSI M[UMPS] language standard.

This metalanguage symbol represents an exponent. An exponent looks like a capital (!) E, then optionally a plus (+) or minus (-) sign, and then one or more digits.

Valid exponents are:

E0
E12
E+12
E-12
E0012

but not:

E
E12.34
E---5
e+13

intlit

Introduced in the 1977 ANSI M[UMPS] language standard.

This metalanguage symbol represents an integer numeric literal, i.e. a string of digits.

An intlit is not necessarily a canonic representation of a number.

Valid intlits are:

12345
0
00123
00000

mant

Introduced in the 1977 ANSI M[UMPS] language standard.

This metalanguage symbol represents a mantissa. A mantissa looks like either one or more digits, optionally followed by a period followed by one or more digits, or a period followed by one or more digits.

This means that valid mantissas are:

123.456
.456
123

but not

.
1234.

numlit

Introduced in the 1977 ANSI M[UMPS] language standard.

This metalanguage symbol represents a numeric literal string. A numeric literal string consists of a mantissa (see metalanguage symbol mant), and optionally an exponent (see metalanguage symbol exp.
A numlit is not necessarily a canonic representation of a number.

Valid numeric literals are:

0
123
00000
00123
.00000
.10100
.12345
1.23E+20
1.23E-5
1.23E-005
12E3

but not:

0.
1.
123e-4
12.3E4.5

The ANSI standard guarantees 9 digits of precision, and a numeric range from -1*10**25 through -1*10**-25, then 0, then +1*10**-25 through +1*10**25.

Modification in the 1984 ANSI M[UMPS] language standard.

The ANSI standard guarantees 12 digits of precision, and a numeric range from -1*10**25 through -1*10**-25, then 0, then +1*10**-25 through +1*10**25.

Modification in the 1995 ANSI M[UMPS] language standard.

The ANSI standard guarantees 15 digits of precision, a computational accuracy of 1 in 10**12 for the operators +, -, *, #, / and \, a computational accuracy of 1 in 10**9 for the operator **, and a numeric range from -1*10**25 through -1*10**-25, then 0, then +1*10**-25 through +1*10**25.