NAME

lex − Generates a C Language program that matches patterns for simple lexical analysis of an input stream

SYNOPSIS

lex [−cnrtv] [−V] [−Qy|−Qn] [file ...]

The lex command reads file or standard input, generates a C Language program, and writes it to a file named lex.yy.c, a compilable C Language program. 

FLAGS

If the environment variable CMD_ENV is set to svr4, all flags listed in the synopsis are legal. Otherwise n, t, v are the only legal flags, and they may be upper or lower case. 

-cWrites C code to the file lex.yy.c. This is the default. 

-nSuppresses the statistics summary.  When you set your own table sizes for the finite state machine, lex automatically produces this summary if you do not select this flag. 

-rWrites RATFOR code to the file lex.yy.r. Note: there is no RATFOR compiler for DEC OSF/1. 

-tWrites to standard output instead of to a file. 

-vProvides a summary of the generated finite state machine statistics. 

-VOutputs lex version number to standard error. Requires the environment variable CMD_ENV to be set to svr4. 

−Q[y|n]Determines whether the lex version number is written to the output file.  −Qn does not do so, and is the default.  Requires the environment variable CMD_ENV to be set to svr4. 

DESCRIPTION

The lex command uses the rules and actions contained in file to generate a program, lex.yy.c, which can be compiled with the cc command.  That program can then receive input, break the input into the logical pieces defined by the rules in file, and run program fragments contained in the actions in file. 

The generated program is a C Language function called yylex().  The lex command stores yylex() in a file named lex.yy.c.  You can use yylex() alone to recognize simple, 1-word input, or you can use it with other C Language programs to perform more difficult input analysis functions.  For example, you can use lex to generate a program that tokenizes an input stream before sending it to a parser program generated by the yacc command. 

The yylex() function analyzes the input stream using a program structure called a finite state machine.  This structure allows the program to exist in only one state (or condition) at a time.  There is a finite number of states allowed.  The rules in file determine how the program moves from one state to another based on the input the program receives. 

The lex command reads its skeleton finite state machine from the file /usr/ccs/lib/ncform. Use the environment variable LEXER to specify another location for lex to read from. 

If you do not specify a file, lex reads standard input.  It treats multiple files as a single file. 

Input File Format

The input file can contain three sections:  definitions, rules, and user subroutines.  Each section must be separated from the others by a line containing only the delimiter, %%.  The format is as follows:

definitions
%%
rules
%%
user_subroutines

The purpose and format of each are described in the following sections. 

Definitions

If you want to use variables in rules, you must define them in this section.  The variables make up the left column, and their definitions make up the right column.  For example, to define D as a numerical digit, enter:

D[0-9]

You can use a defined variable in the rules section by enclosing the variable name in braces, {D}. 

In the definitions section, you can also set table sizes for the resulting finite state machine.  The default sizes are large enough for small programs.  You may want to set larger sizes for more complex programs. 

%p  numberNumber of positions is number (default 5000)

%n  numberNumber of states is number (default 2500)

%e  numberNumber of parse tree nodes is number (default 2000)

%a  numberNumber of transitions is number (default 5000)

%k  numberNumber of packed character classes is number (default 1000)

%o  numberNumber of output slots is number (default 5000)

If extended characters appear in regular expression strings, you may need to reset the output array size with the %o parameter (possibly to array sizes in the range 10,000 to 20,000).  This reset reflects the much larger number of characters relative to the number of ASCII characters. 

Rules

Once you have defined your terms, you can write the rules section.  In this section, the left column contains the pattern to be recognized in an input file to yylex().  The right column contains the C program fragment executed when that pattern is recognized.  This section is required, and it must be preceded by the %% delimiter, whether or not you have a definitions section.  The lex command does not recognize rules without this delimiter. 

Patterns can include extended characters with one exception: these characters may not appear in range specifications within character class expressions surrounded by brackets. 

The columns are separated by a tab.  For example, to search files for the word LEAD and replace it with GOLD, perform the following steps:

Create a file called transmute.l containing the lines:

%%
(LEAD)  printf("GOLD");

Then issue the following commands to the shell:

lex transmute.l
cc -o transmute lex.yy.c -ll

You can test the resulting program with the command:

transmute <transmute.l

This command echoes the contents of transmute.l, with the occurrences of LEAD changed to GOLD. 

Each pattern may have a corresponding action, that is, a fragment of C source code to execute when the pattern is matched.  Each statement must end with a ; (semicolon).  If you use more than one statement in an action, you must enclose all of them in {} (braces). A second delimiter, %%, must follow the rules section if you have a user subroutine section. 

When yylex() matches a string in the input stream, it copies the matched text to an external character array, yytext, before it executes any actions in the rules section. 

You can use the following operators to form patterns that you want to match:

x, yMatches the characters written. 

[ ]Matches any one character in the enclosed range ([.-.]) or the enclosed list ([...]).  [abcx-z] matches a,b,c,x,y, or z. 

"" ""Matches the enclosed character or string even if it is an operator.  "$" prevents lex from interpreting the $ character as an operator. 

\Acts the same as double quotes.  \$ prevents lex from interpreting the $ character as an operator. 

∗Matches zero or more occurrences of the character immediately preceding it.  x∗ matches zero or more repeated literal characters x. 

+Matches one or more occurrences of the character immediately preceding it. 

?Matches either zero or one occurrence of the character immediately preceding it. 

^Matches the character only at the beginning of a line.  ^x matches an x at the beginning of a line. 

[^]Matches any character but the ones following the ^.  For example, [^xyz] matches any character but x, y, or z. 

.Matches any character except the newline character. 

$Matches the end of a line. 

|Matches either of two characters.  x | y matches either x or y. 

/Matches one character only when followed by a second character.  It reads only the first character into yytext.  x/y matches x when it is followed by y, and reads x into yytext. 

( )Matches the pattern in the ( ) (parentheses).  This is used for grouping.  It reads the whole pattern into yytext.  A group in parentheses can be used in place of any single character in any other pattern.  (xyz123) matches the pattern xyz123 and reads the whole string into yytext. 

{}Matches the character as defined in the Definitions section.  If D is defined as numeric digits, {D} matches all numeric digits. 

{m,n}Matches m to n occurrences of the character.  x{2,4} matches 2, 3, or 4 occurrences of x. 

If a line begins with only a space, lex copies it to the lex.yy.c output file.  If the line is in the definitions section of file, lex copies it to the declarations section of lex.yy.c.  If the line is in the rules section, lex copies it to the program code section of lex.yy.c. 

User Subroutines

The lex library has three subroutines defined as macros that you can use in the rules. 

input( )Reads a character from yyin. 

unput( )Replaces a character after it is read. 

output( )Writes an output character to yyout. 

You can override these three macros by writing your own code for these routines in the user subroutines section.  But if you write your own, you must undefine these macros in the definitions section as follows:

%{
#undef input
#undef unput
#undef output
}%

When you are using lex as a simple transformer/recognizer for stdin to stdout piping, you can avoid writing the framework by using libl.a (the lex library).  It has a main routine that calls yylex() for you. 

External names generated by lex all begin with the prefix yy, as in yyin, yyout, yylex, and yytext. 

Putting Spaces in an Expression

Normally, spaces or tabs end a rule and therefore, the expression that defines a rule.  However, you can enclose the spaces or tab characters in "" (double quotes) to include them in the expression.  Use quotes around all spaces in expressions that are not already within sets of [ ] (brackets). 

Other Special Characters

The lex program recognizes many of the normal C language special characters.  These character sequences are as follows:

SequenceMeaning
\nNewline
\tTab
\bBackspace
\\Backslash

Do not use the actual newline character in an expression. 

When using these special characters in an expression, you do not need to enclose them in quotes.  Every character, except these special characters and the previously described operator symbols, is always a text character. 

Matching Rules

When more than one expression can match the current input, lex chooses the longest match first.  Among rules that match the same number of characters, the rule that occurs first is chosen.  For example:

integer keyword action...;
[a-z]+ identifier action...;

If the preceding rules are given in that order, and integers is the input word, lex matches the input as an identifier because [a-z]+ matches eight characters, while integer matches only seven.  However, if the input is integer, both rules match seven characters.  The keyword rule is selected because it occurs first.  A shorter input, such as int, does not match the expression rule integer and so lex selects the rule identifier. 

Matching a String with Wildcard Characters

Because lex chooses the longest match first, do not use rules containing expressions like .∗.  For example:

’.∗’

The preceding rule might seem like a good way to recognize a string in single quotes.  However, the lexical analyzer reads far ahead, looking for a distant single quote to complete the long match.  If a lexical analyzer with such a rule gets the following input, it matches the whole string:

’first’ quoted string here, ’second’ here

To find the smaller strings, first and second, use the following rule:

’[^’\n]∗’

This rule stops after matching ’first’. 

Errors of this type are not far-reaching because the . (dot) operator does not match a newline character.  Therefore, expressions like .∗ stop on the current line.  Do not try to defeat this with expressions like [.\n] +.  The lexical analyzer tries to read the entire input file and an internal buffer overflow occurs. 

Finding Strings within Strings

The lex program partitions the input stream, and does not search for all possible matches of each expression.  Each character is accounted for once and only once.  For example, to count occurrences of both she and he in an input text, try the following rules:

shes++;
heh++;
\n|
.;

The last two rules ignore everything besides he and she.  However, because she includes he, lex does not recognize the instances of he that are included in she. 

To override this choice, use the REJECT action.  This directive tells lex to go to the next rule.  lex then adjusts the position of the input pointer to where it was before the first rule was executed, and executes the second choice rule.  For example, to count the included instances of he, use the following rules:

she{s++; REJECT;}
he{h++; REJECT;}
\n|
.;

After counting the occurrences of she, lex rejects the input stream and then counts the occurrences of he.  Because in this case, she includes he, but not vice versa, you can omit the REJECT action on he.  In other cases, it may be difficult to determine which input characters are in both classes. 

In general, REJECT is useful whenever the purpose of lex is not to partition the input stream but to detect all examples of some items in the input, and the instances of these items may overlap or include each other. 

EXAMPLES

     1.The following command draws lex instructions from the file lexcommands, and places the output in lex.yy.c:

lex lexcommands

     2.The contents of the file lexcommands are an example of a lex program that would be put into a lex command file.  This program converts uppercase to lowercase, removes spaces at the end of a line, and replaces multiple spaces with single spaces:

%%
[A-Z] putchar(tolower(yytext[0]));
[ ]+$ ;
[ ]+ putchar(’ ’);

NOTES

Because lex uses fixed names for intermediate and output files, you can have only one lex-generated program in a given directory. 

FILES

/usr/ccs/lib/libl.aRun-time library. 

/usr/ccs/lib/ncformDefault C language skeleton finite state machine for lex. 

/usr/ccs/lib/nrformDefault RATFOR language skeleton finite state machine for lex. 

RELATED INFORMATION

Commands:  yacc(1). 

Programming Support Tools