About This Book

Syntax Conventions for the SAS Language

Overview of Syntax Conventions for the SAS Language

SAS uses standard conventions in the documentation of syntax for SAS language elements. These conventions enable you to easily identify the components of SAS syntax. The conventions can be divided into these parts:

syntax components
style conventions
special characters
references to SAS libraries and external files

Syntax Components

The components of the syntax for most language elements include a keyword and arguments. For some language elements, only a keyword is necessary. For other language elements, the keyword is followed by an equal sign (=). The syntax for arguments has multiple forms in order to demonstrate the syntax of multiple arguments, with and without punctuation.

keyword

specifies the name of the SAS language element that you use when you write your program. Keyword is a literal that is usually the first word in the syntax. In a CALL routine, the first two words are keywords.

In these examples of SAS syntax, the keywords are bold:

CHAR (string, position)

CALL RANBIN (seed, n, p, x);

ALTER (alter-password)

BEST w.

REMOVE <data-set-name>

In this example, the first two words of the CALL routine are the keywords:

CALL RANBIN(seed, n, p, x)

The syntax of some SAS statements consists of a single keyword without arguments:

DO;
... SAS code ...

END;

Some system options require that one of two keyword values be specified:

DUPLEX | NODUPLEX

Some procedure statements have multiple keywords throughout the statement syntax:

CREATE <UNIQUE> INDEX index-name ON table-name (column-1 <, column-2, ...>)

argument

specifies a numeric or character constant, variable, or expression. Arguments follow the keyword or an equal sign after the keyword. The arguments are used by SAS to process the language element. Arguments can be required or optional. In the syntax, optional arguments are enclosed in angle brackets ( < > ).

In this example, string and position follow the keyword CHAR. These arguments are required arguments for the CHAR function:

CHAR (string, position)

Each argument has a value. In this example of SAS code, the argument string has a value of 'summer', and the argument position has a value of 4:

x=char('summer', 4);

In this example, string and substring are required arguments, whereas modifiers and startpos are optional.

FIND(string, substring <, modifiers> <, startpos>

argument(s)

specifies that one argument is required and that multiple arguments are allowed. Separate arguments with a space. Punctuation, such as a comma ( , ) is not required between arguments.

The MISSING statement is an example of this form of multiple arguments:

MISSING character(s);

<LITERAL_ARGUMENT> argument-1 <<LITERAL_ARGUMENT> argument-2 ... >

specifies that one argument is required and that a literal argument can be associated with the argument. You can specify multiple literals and argument pairs. No punctuation is required between the literal and argument pairs. The ellipsis (...) indicates that additional literals and arguments are allowed.

The BY statement is an example of this argument:

BY <DESCENDING> variable-1 <<DESCENDING> variable-2 ...>;

argument-1 <option(s)> <argument-2 <option(s)> ...>

specifies that one argument is required and that one or more options can be associated with the argument. You can specify multiple arguments and associated options. No punctuation is required between the argument and the option. The ellipsis (...) indicates that additional arguments with an associated option are allowed.

The FORMAT procedure PICTURE statement is an example of this form of multiple arguments:

PICTURE name <(format-option(s))>
<value-range-set-1 <(picture-1-option(s))>
<value-range-set-2 <(picture-2-option(s))> ...>>;

argument-1=value-1 <argument-2=value-2 ...>

specifies that the argument must be assigned a value and that you can specify multiple arguments. The ellipsis (...) indicates that additional arguments are allowed. No punctuation is required between arguments.

The LABEL statement is an example of this form of multiple arguments:

LABEL variable-1=label-1 <variable-2=label-2 ...>;

argument-1 <, argument-2, ...>

specifies that one argument is required and that you can specify multiple arguments that are separated by a comma or other punctuation. The ellipsis (...) indicates a continuation of the arguments, separated by a comma. Both forms are used in the SAS documentation.

Here are examples of this form of multiple arguments:

AUTHPROVIDERDOMAIN (provider-1:domain-1 <, provider-2:domain-2, ...>

INTO :macro-variable-specification-1 <, :macro-variable-specification-2, ...>

Note: In most cases, example code in SAS documentation is written in lowercase with a monospace font. You can use uppercase, lowercase, or mixed case in the code that you write.

Style Conventions

The style conventions that are used in documenting SAS syntax include uppercase bold, uppercase, and italic:

UPPERCASE BOLD

identifies SAS keywords such as the names of functions or statements. In this example, the keyword ERROR is written in uppercase bold:

ERROR <message>;

UPPERCASE

identifies arguments that are literals.

In this example of the CMPMODEL= system option, the literals include BOTH, CATALOG, and XML:

CMPMODEL=BOTH | CATALOG | XML |

italic

identifies arguments or values that you supply. Items in italic represent user-supplied values that are either one of the following:

nonliteral arguments. In this example of the LINK statement, the argument label is a user-supplied value and therefore appears in italic:

LINK label;
nonliteral values that are assigned to an argument.

In this example of the FORMAT statement, the argument DEFAULT is assigned the variable default-format:

FORMAT variable(s) <format > <DEFAULT = default-format>;

Special Characters

The syntax of SAS language elements can contain the following special characters:

an equal sign identifies a value for a literal in some language elements such as system options.

In this example of the MAPS system option, the equal sign sets the value of MAPS:

MAPS=location-of-maps

< >

angle brackets identify optional arguments. A required argument is not enclosed in angle brackets.

In this example of the CAT function, at least one item is required:

CAT (item-1 <, item-2, ...>)

a vertical bar indicates that you can choose one value from a group of values. Values that are separated by the vertical bar are mutually exclusive.

In this example of the CMPMODEL= system option, you can choose only one of the arguments:

CMPMODEL=BOTH | CATALOG | XML

...

an ellipsis indicates that the argument can be repeated. If an argument and the ellipsis are enclosed in angle brackets, then the argument is optional. The repeated argument must contain punctuation if it appears before or after the argument.

In this example of the CAT function, multiple item arguments are allowed, and they must be separated by a comma:

CAT (item-1 <, item-2, ...>)

'value' or "value"

indicates that an argument that is enclosed in single or double quotation marks must have a value that is also enclosed in single or double quotation marks.

In this example of the FOOTNOTE statement, the argument text is enclosed in quotation marks:

FOOTNOTE <n> <ods-format-options 'text' | "text">;

;

a semicolon indicates the end of a statement or CALL routine.

In this example, each statement ends with a semicolon:

data namegame;
   length color name $8;
   color = 'black';
   name = 'jack';
   game = trim(color) || name;
run;

References to SAS Libraries and External Files

Many SAS statements and other language elements refer to SAS libraries and external files. You can choose whether to make the reference through a logical name (a libref or fileref) or use the physical filename enclosed in quotation marks. If you use a logical name, you typically have a choice of using a SAS statement (LIBNAME or FILENAME) or the operating environment's control language to make the reference. Several methods of referring to SAS libraries and external files are available, and some of these methods depend on your operating environment.

In the examples that use external files, SAS documentation uses the italicized phrase file-specification. In the examples that use SAS libraries, SAS documentation uses the italicized phrase SAS-library enclosed in quotation marks:

infile file-specification obs = 100;
libname libref 'SAS-library';