shithub: riscv

ref: db809f2d4786af8fbdf221d59f638c6d0d0d439c
dir: /sys/man/1/mpc/

View raw version
.TH MPC 1 
.SH NAME
mpc \- extended precision arithmetic code generator
.SH SYNOPSIS
.B mpc
[
.I file ...
]
.SH DESCRIPTION
.I Mpc
generates C functions from a simple language that operates on
extended precision integers using the
.IR mp (2)
library.
.SH LANGUAGE
The language consists of a series of function definitions of the form:
.IP
.I name
(
.I "parameter list"
) {
.I statements
}
.PP
All variables and parameters are extended precision integers and are
passed by reference. Statements are separated by semicolon and the
following statements are defined:
.IP
.I name
.B =
.I expression
.IP
.B if
(
.I condition
) {
.I statements
}
.B "else if"
(
.I condition
) {
.I statements
}
.B else
{
.I statements
}
.IP
.B while
(
.I condition
) {
.I statements
}
.IP
.B break
.IP
.I
name
(
.I "parameter list"
)
.IP
.B
mod
(
.I modulus
) {
.I statements
}
.PP
There is no distinction between input and output parameters, but
conventionally, the outputs are put at the end of the
.I "parameter list"
and the language allows one to write
.IP
.I F
(
.IR X ,
.IR Y ,
.I Z
)
as
.IR Y ,
.I Z
.B =
.I F
(
.I X
)
.PP
Expressions are composed out of the following arithmetic operations:
.RS
.TF _____________
.TP
.B +
addition.
.TP
.B -
subtraction.
.TP
.B *
multiplication.
.TP
.B /
division, or multiplicative inverse when enclosed in
.B mod
block.
.TP
.B %
division remainder.
.TP
.B ^
exponentiation.
.TP
.BI >> constant
right shift by a constant.
.TP
.BI << constant
left shift by a constant.
.TP
.IB condition ? a : b
pick
.I a
when
.I condition is true, otherwise
.I b
when false.
.RE
.PD
.PP
Conditions can use the following operations:
.RS
.TF _____________
.TP
.B ==
equality.
.TP
.B !=
inequality.
.TP
.B >
bigger than.
.TP
.B <
smaller than.
.TP
.BI ! condition
negation.
.RE
.SH SOURCE
.B /sys/src/cmd/mpc.y
.SH "SEE ALSO"
.IR mp (2)