'\" et
.TH EXPR "1P" 2017 "IEEE/The Open Group" "POSIX Programmer's Manual"
.\"
.SH PROLOG
This manual page is part of the POSIX Programmer's Manual.
The Linux implementation of this interface may differ (consult
the corresponding Linux manual page for details of Linux behavior),
or the interface may not be implemented on Linux.
.\"
.SH NAME
expr
\(em evaluate arguments as an expression
.SH SYNOPSIS
.LP
.nf
expr \fIoperand\fR...
.fi
.SH DESCRIPTION
The
.IR expr
utility shall evaluate an expression and write the result to standard
output.
.SH OPTIONS
None.
.SH OPERANDS
The single expression evaluated by
.IR expr
shall be formed from the
.IR operand
operands, as described in the EXTENDED DESCRIPTION section. The
application shall ensure that each of the expression operator symbols:
.sp
.RS 4
.nf

(  )  |  &  =  >  >=  <  <=  !=  +  -  *  /  %  :
.fi
.P
.RE
.P
and the symbols
.IR integer
and
.IR string
in the table are provided as separate arguments to
.IR expr .
.SH STDIN
Not used.
.SH "INPUT FILES"
None.
.SH "ENVIRONMENT VARIABLES"
The following environment variables shall affect the execution of
.IR expr :
.IP "\fILANG\fP" 10
Provide a default value for the internationalization variables that are
unset or null. (See the Base Definitions volume of POSIX.1\(hy2017,
.IR "Section 8.2" ", " "Internationalization Variables"
for the precedence of internationalization variables used to determine
the values of locale categories.)
.IP "\fILC_ALL\fP" 10
If set to a non-empty string value, override the values of all the
other internationalization variables.
.IP "\fILC_COLLATE\fP" 10
.br
Determine the locale for the behavior of ranges, equivalence classes,
and multi-character collating elements within regular expressions and
by the string comparison operators.
.IP "\fILC_CTYPE\fP" 10
Determine the locale for the interpretation of sequences of bytes of
text data as characters (for example, single-byte as opposed to
multi-byte characters in arguments) and the behavior of character
classes within regular expressions.
.IP "\fILC_MESSAGES\fP" 10
.br
Determine the locale that should be used to affect the format and
contents of diagnostic messages written to standard error.
.IP "\fINLSPATH\fP" 10
Determine the location of message catalogs for the processing of
.IR LC_MESSAGES .
.SH "ASYNCHRONOUS EVENTS"
Default.
.SH STDOUT
The
.IR expr
utility shall evaluate the expression and write the result, followed by
a
<newline>,
to standard output.
.SH STDERR
The standard error shall be used only for diagnostic messages.
.SH "OUTPUT FILES"
None.
.SH "EXTENDED DESCRIPTION"
The formation of the expression to be evaluated is shown in the
following table. The symbols
.IR expr ,
.IR expr1 ,
and
.IR expr2
represent expressions formed from
.IR integer
and
.IR string
symbols and the expression operator symbols (all separate arguments) by
recursive application of the constructs described in the table. The
expressions are listed in order of decreasing precedence, with
equal-precedence operators grouped between horizontal lines. All of
the operators shall be left-associative.
.TS
center tab(@) box;
cB | cB
l | lw(4i).
Expression@Description
_
\fIinteger\fP@T{
An argument consisting only of an (optional) unary minus followed
by digits.
T}
\fIstring\fP@T{
A string argument; see below.
T}
_
(\ \fIexpr\fR\ )@T{
Grouping symbols. Any expression can be placed within parentheses.
Parentheses can be nested to a depth of
{EXPR_NEST_MAX}.
T}
_
\fIexpr1\fP\ :\ \fIexpr2\fP@T{
Matching expression; see below.
T}
_
\fIexpr1\fP\ *\ \fIexpr2\fP@T{
Multiplication of decimal integer-valued arguments.
T}
\fIexpr1\fP\ /\ \fIexpr2\fP@T{
Integer division of decimal integer-valued arguments, producing
an integer result.
T}
\fIexpr1\fP\ %\ \fIexpr2\fP@T{
Remainder of integer division of decimal integer-valued arguments.
T}
_
\fIexpr1\fP\ +\ \fIexpr2\fP@T{
Addition of decimal integer-valued arguments.
T}
\fIexpr1\fP\ \-\ \fIexpr2\fP@T{
Subtraction of decimal integer-valued arguments.
T}
_
@T{
Returns the result of a decimal integer comparison if both arguments
are integers; otherwise, returns the result of a string comparison
using the locale-specific collation sequence. The result of each
comparison is 1 if the specified relationship is true, or 0 if the
relationship is false.
T}
\fIexpr1\fP\ =\ \fIexpr2\fR@Equal.
\fIexpr1\fP\ >\ \fIexpr2\fR@Greater than.
\fIexpr1\fP\ >=\ \fIexpr2\fR@Greater than or equal.
\fIexpr1\fP\ <\ \fIexpr2\fR@Less than.
\fIexpr1\fP\ <=\ \fIexpr2\fR@Less than or equal.
\fIexpr1\fP\ !=\ \fIexpr2\fR@Not equal.
_
\fIexpr1\fP\ &\ \fIexpr2\fP@T{
Returns the evaluation of
.IR expr1
if neither expression evaluates to null or zero; otherwise, returns zero.
T}
_
\fIexpr1\fP\ |\ \fIexpr2\fP@T{
Returns the evaluation of
.IR expr1
if it is neither null nor zero; otherwise, returns the evaluation of
.IR expr2
if it is not null; otherwise, zero.
T}
.TE
.SS "Matching Expression"
.P
The
.BR ':' 
matching operator shall compare the string resulting from the
evaluation of
.IR expr1
with the regular expression pattern resulting from the evaluation of
.IR expr2 .
Regular expression syntax shall be that defined in the Base Definitions volume of POSIX.1\(hy2017,
.IR "Section 9.3" ", " "Basic Regular Expressions",
except that all patterns are anchored to the beginning of the string (that
is, only sequences starting at the first character of a string are matched
by the regular expression) and, therefore, it is unspecified whether
.BR '\(ha' 
is a special character in that context. Usually, the matching operator
shall return a string representing the number of characters matched (\c
.BR '0' 
on failure). Alternatively, if the pattern contains at least one
regular expression subexpression
.BR \(dq[\e(...\e)]\(dq ,
the string matched by the back-reference expression
.BR \(dq\e1\(dq 
shall be returned. If the back-reference expression
.BR \(dq\e1\(dq 
does not match, then the null string shall be returned.
.SS "Identification as Integer or String"
.P
An argument or the value of a subexpression that consists only of an
optional unary minus followed by digits is a candidate for treatment
as an integer if it is used as the left argument to the
.IR |
operator or as either argument to any of the following operators:
.IR "& = > >= < <= != + - * / %" .
Otherwise, the argument or subexpression value shall be treated as a string.
.P
The use of string arguments
.BR length ,
.BR substr ,
.BR index ,
or
.BR match
produces unspecified results.
.SH "EXIT STATUS"
The following exit values shall be returned:
.IP "\00" 6
The
.IR expression
evaluates to neither null nor zero.
.IP "\01" 6
The
.IR expression
evaluates to null or zero.
.IP "\02" 6
Invalid
.IR expression .
.IP >2 6
An error occurred.
.SH "CONSEQUENCES OF ERRORS"
Default.
.LP
.IR "The following sections are informative."
.SH "APPLICATION USAGE"
The
.IR expr
utility has a rather difficult syntax:
.IP " *" 4
Many of the operators are also shell control operators or reserved
words, so they have to be escaped on the command line.
.IP " *" 4
Each part of the expression is composed of separate arguments, so
liberal usage of
<blank>
characters is required. For example:
.TS
center tab(@) box;
cB | cB
lf5 | lf5.
Invalid@Valid
_
\fIexpr\fR 1+2@\fIexpr\fR 1 + 2
\fIexpr\fR "1 + 2"@\fIexpr\fR 1 + 2
\fIexpr\fR 1 + (2 * 3)@\fIexpr\fR 1 + \e( 2 \e* 3 \e)\fR
.TE
.P
In many cases, the arithmetic and string features provided as part of
the shell command language are easier to use than their equivalents in
.IR expr .
Newly written scripts should avoid
.IR expr
in favor of the new features within the shell; see
.IR "Section 2.5" ", " "Parameters and Variables"
and
.IR "Section 2.6.4" ", " "Arithmetic Expansion".
.P
After argument processing by the shell,
.IR expr
is not required to be able to tell the difference between an operator
and an operand except by the value. If
.BR \(dq$a\(dq 
is
.BR '=' ,
the command:
.sp
.RS 4
.nf

expr "$a" = \(aq=\(aq
.fi
.P
.RE
.P
looks like:
.sp
.RS 4
.nf

expr = = =
.fi
.P
.RE
.P
as the arguments are passed to
.IR expr
(and they all may be taken as the
.BR '=' 
operator). The following works reliably:
.sp
.RS 4
.nf

expr "X$a" = X=
.fi
.P
.RE
.P
Also note that this volume of POSIX.1\(hy2017 permits implementations to extend utilities. The
.IR expr
utility permits the integer arguments to be preceded with a unary
minus. This means that an integer argument could look like an option.
Therefore, the conforming application must employ the
.BR \(dq--\(dq 
construct of Guideline 10 of the Base Definitions volume of POSIX.1\(hy2017,
.IR "Section 12.2" ", " "Utility Syntax Guidelines"
to protect its operands if there is any chance the first operand might
be a negative integer (or any string with a leading minus).
.P
For testing string equality the
.IR test
utility is preferred over
.IR expr ,
as it is usually implemented as a shell built-in. However, the
functionality is not quite the same because the
.IR expr
.IR =
and
.IR !=
operators check whether strings collate equally, whereas
.IR test
checks whether they are identical. Therefore, they can produce
different results in locales where the collation sequence does not
have a total ordering of all characters (see the Base Definitions volume of POSIX.1\(hy2017,
.IR "Section 7.3.2" ", " "LC_COLLATE").
.br
.SH EXAMPLES
The following command:
.sp
.RS 4
.nf

a=$(expr "$a" + 1)
.fi
.P
.RE
.P
adds 1 to the variable
.IR a .
.P
The following command, for
.BR \(dq$a\(dq 
equal to either
.BR /usr/abc/file
or just
.BR file :
.sp
.RS 4
.nf

expr $a : \(aq.*/\e(.*\e)\(aq \e| $a
.fi
.P
.RE
.P
returns the last segment of a pathname (that is,
.BR file ).
Applications should avoid the character
.BR '/' 
used alone as an argument;
.IR expr
may interpret it as the division operator.
.P
The following command:
.sp
.RS 4
.nf

expr "//$a" : \(aq.*/\e(.*\e)\(aq
.fi
.P
.RE
.P
is a better representation of the previous example. The addition of
the
.BR \(dq//\(dq 
characters eliminates any ambiguity about the division operator and
simplifies the whole expression. Also note that pathnames may contain
characters contained in the
.IR IFS
variable and should be quoted to avoid having
.BR \(dq$a\(dq 
expand into multiple arguments.
.P
The following command:
.sp
.RS 4
.nf

expr "X$VAR" : \(aq.*\(aq - 1
.fi
.P
.RE
.P
returns the number of characters in
.IR VAR .
.SH RATIONALE
In an early proposal, EREs were used in the matching expression syntax.
This was changed to BREs to avoid breaking historical applications.
.P
The use of a leading
<circumflex>
in the BRE is unspecified because many historical implementations have
treated it as a special character, despite their system documentation. For
example:
.sp
.RS 4
.nf

expr foo : \(hafoo     expr \(hafoo : \(hafoo
.fi
.P
.RE
.P
return 3 and 0, respectively, on those systems; their documentation
would imply the reverse. Thus, the anchoring condition is left
unspecified to avoid breaking historical scripts relying on this
undocumented feature.
.SH "FUTURE DIRECTIONS"
None.
.SH "SEE ALSO"
.IR "Section 2.5" ", " "Parameters and Variables",
.IR "Section 2.6.4" ", " "Arithmetic Expansion"
.P
The Base Definitions volume of POSIX.1\(hy2017,
.IR "Section 7.3.2" ", " "LC_COLLATE",
.IR "Chapter 8" ", " "Environment Variables",
.IR "Section 9.3" ", " "Basic Regular Expressions",
.IR "Section 12.2" ", " "Utility Syntax Guidelines"
.\"
.SH COPYRIGHT
Portions of this text are reprinted and reproduced in electronic form
from IEEE Std 1003.1-2017, Standard for Information Technology
-- Portable Operating System Interface (POSIX), The Open Group Base
Specifications Issue 7, 2018 Edition,
Copyright (C) 2018 by the Institute of
Electrical and Electronics Engineers, Inc and The Open Group.
In the event of any discrepancy between this version and the original IEEE and
The Open Group Standard, the original IEEE and The Open Group Standard
is the referee document. The original Standard can be obtained online at
http://www.opengroup.org/unix/online.html .
.PP
Any typographical or formatting errors that appear
in this page are most likely
to have been introduced during the conversion of the source files to
man page format. To report such errors, see
https://www.kernel.org/doc/man-pages/reporting_bugs.html .