The flex manual is placed under the same licensing conditions as the rest of flex:
Copyright (C) 1990, 1997 The Regents of the University of California. All rights reserved.
This code is derived from software contributed to Berkeley by Vern Paxson.
The United States Government has rights in this work pursuant to contract no. DE-AC03-76SF00098 between the United States Department of Energy and the University of California.
Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met:
Neither the name of the University nor the names of its contributors may be used to endorse or promote products derived from this software without specific prior written permission.
THIS SOFTWARE IS PROVIDED "AS IS" AND WITHOUT ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE.
If you have problems with flex or think you have found a bug,
please send mail detailing your problem to
help-flex@gnu.org. Patches are always welcome.
flex is a tool for generating scanners. A scanner is a
program which recognizes lexical patterns in text. The flex
program reads the given input files, or its standard input if no file
names are given, for a description of a scanner to generate. The
description is in the form of pairs of regular expressions and C code,
called rules. flex generates as output a C source file,
`lex.yy.c' by default, which defines a routine yylex().
This file can be compiled and linked with the flex runtime library to
produce an executable. When the executable is run, it analyzes its
input for occurrences of the regular expressions. Whenever it finds
one, it executes the corresponding C code.
First some simple examples to get the flavor of how one uses
flex.
The following flex input specifies a scanner which, when it
encounters the string `username' will replace it with the user's
login name:
@verbatim
%%
username printf( "%s", getlogin() );
By default, any text not matched by a flex scanner is copied to
the output, so the net effect of this scanner is to copy its input file
to its output with each occurrence of `username' expanded. In this
input, there is just one rule. `username' is the pattern and
the `printf' is the action. The `%%' symbol marks the
beginning of the rules.
Here's another simple example:
@verbatim
int num_lines = 0, num_chars = 0;
%%
\n ++num_lines; ++num_chars;
. ++num_chars;
%%
main()
{
yylex();
printf( "# of lines = %d, # of chars = %d\n",
num_lines, num_chars );
}
This scanner counts the number of characters and the number of lines in
its input. It produces no output other than the final report on the
character and line counts. The first line declares two globals,
num_lines and num_chars, which are accessible both inside
yylex() and in the main() routine declared after the
second `%%'. There are two rules, one which matches a newline
(`\n') and increments both the line count and the character count,
and one which matches any character other than a newline (indicated by
the `.' regular expression).
A somewhat more complicated example:
@verbatim
/* scanner for a toy Pascal-like language */
%{
/* need this for the call to atof() below */
#include math.h>
%}
DIGIT [0-9]
ID [a-z][a-z0-9]*
%%
{DIGIT}+ {
printf( "An integer: %s (%d)\n", yytext,
atoi( yytext ) );
}
{DIGIT}+"."{DIGIT}* {
printf( "A float: %s (%g)\n", yytext,
atof( yytext ) );
}
if|then|begin|end|procedure|function {
printf( "A keyword: %s\n", yytext );
}
{ID} printf( "An identifier: %s\n", yytext );
"+"|"-"|"*"|"/" printf( "An operator: %s\n", yytext );
"{"[\^{}}\n]*"}" /* eat up one-line comments */
[ \t\n]+ /* eat up whitespace */
. printf( "Unrecognized character: %s\n", yytext );
%%
main( argc, argv )
int argc;
char **argv;
{
++argv, --argc; /* skip over program name */
if ( argc > 0 )
yyin = fopen( argv[0], "r" );
else
yyin = stdin;
yylex();
}
This is the beginnings of a simple scanner for a language like Pascal. It identifies different types of tokens and reports on what it has seen.
The details of this example will be explained in the following sections.
The flex input file consists of three sections, separated by a
line containing only `%%'.
@verbatim
definitions
%%
rules
%%
user code
The definitions section contains declarations of simple name definitions to simplify the scanner specification, and declarations of start conditions, which are explained in a later section.
Name definitions have the form:
@verbatim
name definition
The `name' is a word beginning with a letter or an underscore (`_') followed by zero or more letters, digits, `_', or `-' (dash). The definition is taken to begin at the first non-whitespace character following the name and continuing to the end of the line. The definition can subsequently be referred to using `{name}', which will expand to `(definition)'. For example,
@verbatim
DIGIT [0-9]
ID [a-z][a-z0-9]*
Defines `DIGIT' to be a regular expression which matches a single digit, and `ID' to be a regular expression which matches a letter followed by zero-or-more letters-or-digits. A subsequent reference to
@verbatim
{DIGIT}+"."{DIGIT}*
is identical to
@verbatim
([0-9])+"."([0-9])*
and matches one-or-more digits followed by a `.' followed by zero-or-more digits.
An unindented comment (i.e., a line beginning with `/*') is copied verbatim to the output up to the next `*/'.
Any indented text or text enclosed in `%{' and `%}' is also copied verbatim to the output (with the %{ and %} symbols removed). The %{ and %} symbols must appear unindented on lines by themselves.
The rules section of the flex input contains a series of
rules of the form:
@verbatim
pattern action
where the pattern must be unindented and the action must begin on the same line. See section Patterns, for a further description of patterns and actions.
In the rules section, any indented or %{ %} enclosed text appearing before the first rule may be used to declare variables which are local to the scanning routine and (after the declarations) code which is to be executed whenever the scanning routine is entered. Other indented or %{ %} text in the rule section is still copied to the output, but its meaning is not well-defined and it may well cause compile-time errors (this feature is present for @acronym{POSIX} compliance. See section Incompatibilities with Lex and Posix, for other such features).
Any indented text or text enclosed in `%{' and `%}' is copied verbatim to the output (with the %{ and %} symbols removed). The %{ and %} symbols must appear unindented on lines by themselves.
The user code section is simply copied to `lex.yy.c' verbatim. It is used for companion routines which call or are called by the scanner. The presence of this section is optional; if it is missing, the second `%%' in the input file may be skipped, too.
Flex supports C-style comments, that is, anything between /* and */ is considered a comment. Whenever flex encounters a comment, it copies the entire comment verbatim to the generated source code. Comments may appear just about anywhere, but with the following exceptions:
If you want to follow a simple rule, then always begin a comment on a new line, with one or more whitespace characters before the initial `/*'). This rule will work anywhere in the input file.
All the comments in the following example are valid:
@verbatim
%{
/* code block */
%}
/* Definitions Section */
%x STATE_X
%%
/* Rules Section */
ruleA /* after regex */ { /* code block */ } /* after code block */
/* Rules Section (indented) */
<STATE_X>{
ruleC ECHO;
ruleD ECHO;
%{
/* code block */
%}
}
%%
/* User Code Section */
The patterns in the input (see section Format of the Rules Section) are written using an extended set of regular expressions. These are:
flex's notion of "newline" is exactly
whatever the C compiler used to compile flex
interprets `\n' as; in particular, on some DOS
systems you must either filter out `\r's in the
input yourself, or explicitly use `r/\r\n' for `r$'.
s (see section Start Conditions for discussion of start conditions).
s1, s2, or s3.
s1 or s2
Note that inside of a character class, all regular expression operators lose their special meaning except escape (`\') and the character class operators, `-', `]]', and, at the beginning of the class, `^'.
The regular expressions listed above are grouped according to precedence, from highest precedence at the top to lowest at the bottom. Those grouped together have equal precedence (see special note on the precedence of the repeat operator, `{}', under the documentation for the `--posix' POSIX compliance option). For example,
@verbatim
foo|bar*
is the same as
@verbatim
(foo)|(ba(r*))
since the `*' operator has higher precedence than concatenation, and concatenation higher than alternation (`|'). This pattern therefore matches either the string `foo' or the string `ba' followed by zero-or-more `r''s. To match `foo' or zero-or-more repetitions of the string `bar', use:
@verbatim
foo|(bar)*
And to match a sequence of zero or more repetitions of `foo' and `bar':
@verbatim
(foo|bar)*
In addition to characters and ranges of characters, character classes can also contain character class expressions. These are expressions enclosed inside `[': and `:]' delimiters (which themselves must appear between the `[' and `]' of the character class. Other elements may occur inside the character class, too). The valid expressions are:
@verbatim
[:alnum:] [:alpha:] [:blank:]
[:cntrl:] [:digit:] [:graph:]
[:lower:] [:print:] [:punct:]
[:space:] [:upper:] [:xdigit:]
These expressions all designate a set of characters equivalent to the
corresponding standard C isXXX function. For example,
`[:alnum:]' designates those characters for which isalnum()
returns true - i.e., any alphabetic or numeric character. Some systems
don't provide isblank(), so flex defines `[:blank:]' as a
blank or a tab.
For example, the following character classes are all equivalent:
@verbatim
[[:alnum:]]
[[:alpha:][:digit:]]
[[:alpha:][0-9]]
[a-zA-Z0-9]
If your scanner is case-insensitive (the `-i' flag), then `[:upper:]' and `[:lower:]' are equivalent to `[:alpha:]'.
Some notes on patterns are in order.
@verbatim
foo/bar$
<sc1>foo<sc2>bar
Note that the first of these can be written `foo/bar\n'.
@verbatim
foo|(bar$)
foo|^bar
If the desired meaning is a `foo' or a
`bar'-followed-by-a-newline, the following could be used (the
special | action is explained below, see section Actions):
@verbatim
foo |
bar$ /* action goes here */
A similar trick will work for matching a `foo' or a
`bar'-at-the-beginning-of-a-line.
When the generated scanner is run, it analyzes its input looking for
strings which match any of its patterns. If it finds more than one
match, it takes the one matching the most text (for trailing context
rules, this includes the length of the trailing part, even though it
will then be returned to the input). If it finds two or more matches of
the same length, the rule listed first in the flex input file is
chosen.
Once the match is determined, the text corresponding to the match
(called the token) is made available in the global character
pointer yytext, and its length in the global integer
yyleng. The action corresponding to the matched pattern is
then executed (see section Actions), and then the remaining input is scanned
for another match.
If no match is found, then the default rule is executed: the next
character in the input is considered matched and copied to the standard
output. Thus, the simplest valid flex input is:
@verbatim
%%
which generates a scanner that simply copies its input (one character at a time) to its output.
Note that yytext can be defined in two different ways: either as
a character pointer or as a character array. You can
control which definition flex uses by including one of the
special directives %pointer or %array in the first
(definitions) section of your flex input. The default is
%pointer, unless you use the `-l' lex compatibility option,
in which case yytext will be an array. The advantage of using
%pointer is substantially faster scanning and no buffer overflow
when matching very large tokens (unless you run out of dynamic memory).
The disadvantage is that you are restricted in how your actions can
modify yytext (see section Actions), and calls to the unput()
function destroys the present contents of yytext, which can be a
considerable porting headache when moving between different lex
versions.
The advantage of %array is that you can then modify yytext
to your heart's content, and calls to unput() do not destroy
yytext (see section Actions). Furthermore, existing lex
programs sometimes access yytext externally using declarations of
the form:
@verbatim
extern char yytext[];
This definition is erroneous when used with %pointer, but correct
for %array.
The %array declaration defines yytext to be an array of
YYLMAX characters, which defaults to a fairly large value. You
can change the size by simply #define'ing YYLMAX to a different
value in the first section of your flex input. As mentioned
above, with %pointer yytext grows dynamically to accommodate
large tokens. While this means your %pointer scanner can
accommodate very large tokens (such as matching entire blocks of
comments), bear in mind that each time the scanner must resize
yytext it also must rescan the entire token from the beginning,
so matching such tokens can prove slow. yytext presently does
not dynamically grow if a call to unput() results in too
much text being pushed back; instead, a run-time error results.
Also note that you cannot use %array with C++ scanner classes
(see section Generating C++ Scanners).
Each pattern in a rule has a corresponding action, which can be any arbitrary C statement. The pattern ends at the first non-escaped whitespace character; the remainder of the line is its action. If the action is empty, then when the pattern is matched the input token is simply discarded. For example, here is the specification for a program which deletes all occurrences of `zap me' from its input:
@verbatim
%%
"zap me"
(It will copy all other characters in the input to the output since they will be matched by the default rule.)
Here is a program which compresses multiple blanks and tabs down to a single blank, and throws away whitespace found at the end of a line:
@verbatim
%%
[ \t]+ putchar( ' ' );
[ \t]+$ /* ignore this token */
If the action contains a `}', then the action spans till the balancing `}'
is found, and the action may cross multiple lines.
flex
knows about C strings and comments and won't be fooled by braces found
within them, but also allows actions to begin with
`%{'
and will consider the action to be all the text up to the next
`%}'
(regardless of ordinary braces inside the action).
An action consisting solely of a vertical bar ('|') means "same as the action for the next rule". See below for an illustration.
Actions can include arbitrary C code, including return statements
to return a value to whatever routine called yylex(). Each time
yylex() is called it continues processing tokens from where it
last left off until it either reaches the end of the file or executes a
return.
Actions are free to modify yytext except for lengthening it
(adding characters to its end--these will overwrite later characters in
the input stream). This however does not apply when using %array
(see section How the Input Is Matched). In that case, yytext may be freely modified in
any way.
Actions are free to modify
yyleng
except they should not do so if the action also includes use of
yymore()
(see below).
There are a number of special directives which can be included within an action:
ECHO
BEGIN
REJECT
yytext and yyleng set
up appropriately. It may either be one which matched as much text as
the originally chosen rule but came later in the flex input file,
or one which matched less text. For example, the following will both
count the words in the input and call the routine special()
whenever `frob' is seen:
@verbatim
int word_count = 0;
%%
frob special(); REJECT;
[^ \t\n]+ ++word_count;
Without the REJECT, any occurences of `frob' in the input
would not be counted as words, since the scanner normally executes only
one action per token. Multiple uses of REJECT are allowed, each
one finding the next best choice to the currently active rule. For
example, when the following scanner scans the token `abcd', it will
write `abcdabcaba' to the output:
@verbatim
%%
a |
ab |
abc |
abcd ECHO; REJECT;
.|\n /* eat up any unmatched character */
(The first three rules share the fourth's action since they use
the special '|' action.)
REJECT
is a particularly expensive feature in terms of scanner performance;
if it is used in
any
of the scanner's actions it will slow down
all
of the scanner's matching. Furthermore,
REJECT
cannot be used with the
`-Cf'
or
`-CF'
options (see section Scanner Options).
Note also that unlike the other special actions, REJECT is a
branch. code immediately following it in the action will
not be executed.
yymore()
yytext rather than replacing it. For example, given the input
`mega-kludge' the following will write `mega-mega-kludge' to
the output:
@verbatim
%%
mega- ECHO; yymore();
kludge ECHO;
First `mega-' is matched and echoed to the output. Then `kludge'
is matched, but the previous `mega-' is still hanging around at the
beginning of
yytext
so the
ECHO
for the `kludge' rule will actually write `mega-kludge'.
Two notes regarding use of yymore(). First, yymore()
depends on the value of yyleng correctly reflecting the size of
the current token, so you must not modify yyleng if you are using
yymore(). Second, the presence of yymore() in the
scanner's action entails a minor performance penalty in the scanner's
matching speed.
yyless(n) returns all but the first n characters of the
current token back to the input stream, where they will be rescanned
when the scanner looks for the next match. yytext and
yyleng are adjusted appropriately (e.g., yyleng will now be
equal to n). For example, on the input `foobar' the
following will write out `foobarbar':
@verbatim
%%
foobar ECHO; yyless(3);
[a-z]+ ECHO;
An argument of 0 to
yyless()
will cause the entire current input string to be scanned again. Unless you've
changed how the scanner will subsequently process its input (using
BEGIN,
for example), this will result in an endless loop.
Note that
yyless()
is a macro and can only be used in the flex input file, not from
other source files.
unput(c)
puts the character
c
back onto the input stream. It will be the next character scanned.
The following action will take the current token and cause it
to be rescanned enclosed in parentheses.
@verbatim
{
int i;
/* Copy yytext because unput() trashes yytext */
char *yycopy = strdup( yytext );
unput( ')' );
for ( i = yyleng - 1; i >= 0; --i )
unput( yycopy[i] );
unput( '(' );
free( yycopy );
}
Note that since each unput() puts the given character back at the
beginning of the input stream, pushing back strings must be done
back-to-front.
An important potential problem when using
unput()
is that if you are using
%pointer
(the default), a call to
unput()
destroys
the contents of
yytext,
starting with its rightmost character and devouring one character to
the left with each call. If you need the value of yytext preserved
after a call to
unput()
(as in the above example),
you must either first copy it elsewhere, or build your scanner using
%array
instead (see section How the Input Is Matched).
Finally, note that you cannot put back `EOF' to attempt to mark the input stream with an end-of-file.
input() reads the next character from the input stream. For
example, the following is one way to eat up C comments:
@verbatim
%%
"/*" {
register int c;
for ( ; ; )
{
while ( (c = input()) != '*' &&
c != EOF )
; /* eat up text of comment */
if ( c == '*' )
{
while ( (c = input()) == '*' )
;
if ( c == '/' )
break; /* found the end */
}
if ( c == EOF )
{
error( "EOF in comment" );
break;
}
}
}
(Note that if the scanner is compiled using C++, then
input() is instead referred to as yyinput(), in order to
avoid a name clash with the C++ stream by the name of
input.)
YY_FLUSH_BUFFER()
flushes the scanner's internal buffer
so that the next time the scanner attempts to match a token, it will
first refill the buffer using
YY_INPUT()
(see section The Generated Scanner). This action is a special case
of the more general
yy_flush_buffer()
function, described below (see section Multiple Input Buffers)
yyterminate()
can be used in lieu of a return statement in an action. It terminates
the scanner and returns a 0 to the scanner's caller, indicating "all done".
By default,
yyterminate()
is also called when an end-of-file is encountered. It is a macro and
may be redefined.
The output of flex is the file `lex.yy.c', which contains
the scanning routine yylex(), a number of tables used by it for
matching tokens, and a number of auxiliary routines and macros. By
default, yylex() is declared as follows:
@verbatim
int yylex()
{
... various definitions and the actions in here ...
}
(If your environment supports function prototypes, then it will
be
int yylex( void ).) This definition may be changed by defining
the YY_DECL macro. For example, you could use:
@verbatim
#define YY_DECL float lexscan( a, b ) float a, b;
to give the scanning routine the name lexscan, returning a float,
and taking two floats as arguments. Note that if you give arguments to
the scanning routine using a K&R-style/non-prototyped function
declaration, you must terminate the definition with a semi-colon (;).
flex generates `C99' function definitions by default. However flex
does have the ability to generate obsolete, er, `traditional', function
definitions. This is to support bootstrapping gcc on old systems.
Unfortunately, traditional definitions prevent us from using any standard data
types smaller than int (such as short, char, or bool) as function arguments.
For this reason, future versions of flex may generate standard C99 code
only, leaving K&R-style functions to the historians. Currently, if you do
not want `C99' definitions, then you must define
YY_TRADITIONAL_FUNC_DEFS.
Whenever yylex() is called, it scans tokens from the global input
file `yyin' (which defaults to stdin). It continues until it
either reaches an end-of-file (at which point it returns the value 0) or
one of its actions executes a return statement.
If the scanner reaches an end-of-file, subsequent calls are undefined
unless either `yyin' is pointed at a new input file (in which case
scanning continues from that file), or yyrestart() is called.
yyrestart() takes one argument, a FILE * pointer (which
can be nil, if you've set up YY_INPUT to scan from a source other
than yyin), and initializes `yyin' for scanning from that
file. Essentially there is no difference between just assigning
`yyin' to a new input file or using yyrestart() to do so;
the latter is available for compatibility with previous versions of
flex, and because it can be used to switch input files in the
middle of scanning. It can also be used to throw away the current input
buffer, by calling it with an argument of `yyin'; but it would be
better to use YY_FLUSH_BUFFER (see section Actions). Note that
yyrestart() does not reset the start condition to
INITIAL (see section Start Conditions).
If
yylex()
stops scanning due to executing a
return
statement in one of the actions, the scanner may then be called again and it
will resume scanning where it left off.
By default (and for purposes of efficiency), the scanner uses
block-reads rather than simple getc() calls to read characters
from `yyin'. The nature of how it gets its input can be controlled
by defining the YY_INPUT macro. The calling sequence for
YY_INPUT() is YY_INPUT(buf,result,max_size). Its action
is to place up to max_size characters in the character array
buf and return in the integer variable result either the
number of characters read or the constant YY_NULL (0 on Unix
systems) to indicate `EOF'. The default YY_INPUT reads from
the global file-pointer `yyin'.
Here is a sample definition of YY_INPUT (in the definitions
section of the input file):
@verbatim
%{
#define YY_INPUT(buf,result,max_size) \
{ \
int c = getchar(); \
result = (c == EOF) ? YY_NULL : (buf[0] = c, 1); \
}
%}
This definition will change the input processing to occur one character at a time.
When the scanner receives an end-of-file indication from YY_INPUT, it
then checks the yywrap() function. If yywrap() returns
false (zero), then it is assumed that the function has gone ahead and
set up `yyin' to point to another input file, and scanning
continues. If it returns true (non-zero), then the scanner terminates,
returning 0 to its caller. Note that in either case, the start
condition remains unchanged; it does not revert to
INITIAL.
If you do not supply your own version of yywrap(), then you must
either use %option noyywrap (in which case the scanner behaves as
though yywrap() returned 1), or you must link with `-lfl' to
obtain the default version of the routine, which always returns 1.
For scanning from in-memory buffers (e.g., scanning strings), see @xref{Scanning Strings} See section Multiple Input Buffers.
The scanner writes its
ECHO
output to the
`yyout'
global (default, `stdout'), which may be redefined by the user simply
by assigning it to some other
FILE
pointer.
flex
provides a mechanism for conditionally activating rules. Any rule
whose pattern is prefixed with `<sc>' will only be active when
the scanner is in the start condition named sc. For example,
@verbatim
<STRING>[^"]* { /* eat up the string body ... */
...
}
will be active only when the scanner is in the STRING start
condition, and
@verbatim
<INITIAL,STRING,QUOTE>\. { /* handle an escape ... */
...
}
will be active only when the current start condition is either
INITIAL, STRING, or QUOTE.
Start conditions are declared in the definitions (first) section of the
input using unindented lines beginning with either `%s' or
`%x' followed by a list of names. The former declares
inclusive start conditions, the latter exclusive start
conditions. A start condition is activated using the BEGIN
action. Until the next BEGIN action is executed, rules with the
given start condition will be active and rules with other start
conditions will be inactive. If the start condition is inclusive, then
rules with no start conditions at all will also be active. If it is
exclusive, then only rules qualified with the start condition
will be active. A set of rules contingent on the same exclusive start
condition describe a scanner which is independent of any of the other
rules in the flex input. Because of this, exclusive start
conditions make it easy to specify "mini-scanners" which scan portions
of the input that are syntactically different from the rest (e.g.,
comments).
If the distinction between inclusive and exclusive start conditions is still a little vague, here's a simple example illustrating the connection between the two. The set of rules:
@verbatim
%s example
%%
<example>foo do_something();
bar something_else();
is equivalent to
@verbatim
%x example
%%
<example>foo do_something();
<INITIAL,example>bar something_else();
Without the <INITIAL,example> qualifier, the bar pattern in
the second example wouldn't be active (i.e., couldn't match) when in
start condition example. If we just used example> to
qualify bar, though, then it would only be active in
example and not in INITIAL, while in the first example
it's active in both, because in the first example the example
start condition is an inclusive (%s) start condition.
Also note that the special start-condition specifier
<*>
matches every start condition. Thus, the above example could also
have been written:
@verbatim
%x example
%%
<example>foo do_something();
<*>bar something_else();
The default rule (to ECHO any unmatched character) remains active
in start conditions. It is equivalent to:
@verbatim
<*>.|\n ECHO;
BEGIN(0) returns to the original state where only the rules with
no start conditions are active. This state can also be referred to as
the start-condition INITIAL, so BEGIN(INITIAL) is
equivalent to BEGIN(0). (The parentheses around the start
condition name are not required but are considered good style.)
BEGIN actions can also be given as indented code at the beginning
of the rules section. For example, the following will cause the scanner
to enter the SPECIAL start condition whenever yylex() is
called and the global variable enter_special is true:
@verbatim
int enter_special;
%x SPECIAL
%%
if ( enter_special )
BEGIN(SPECIAL);
<SPECIAL>blahblahblah
...more rules follow...
To illustrate the uses of start conditions, here is a scanner which provides two different interpretations of a string like `123.456'. By default it will treat it as three tokens, the integer `123', a dot (`.'), and the integer `456'. But if the string is preceded earlier in the line by the string `expect-floats' it will treat it as a single token, the floating-point number `123.456':
@verbatim
%{
#include <math.h>
%}
%s expect
%%
expect-floats BEGIN(expect);
<expect>[0-9]+`.'[0-9]+ {
printf( "found a float, = %f\n",
atof( yytext ) );
}
<expect>\n {
/* that's the end of the line, so
* we need another "expect-number"
* before we'll recognize any more
* numbers
*/
BEGIN(INITIAL);
}
[0-9]+ {
printf( "found an integer, = %d\n",
atoi( yytext ) );
}
"." printf( "found a dot\n" );
Here is a scanner which recognizes (and discards) C comments while maintaining a count of the current input line.
@verbatim
%x comment
%%
int line_num = 1;
"/*" BEGIN(comment);
<comment>[^*\n]* /* eat anything that's not a '*' */
<comment>"*"+[^*/\n]* /* eat up '*'s not followed by '/'s */
<comment>\n ++line_num;
<comment>"*"+"/" BEGIN(INITIAL);
This scanner goes to a bit of trouble to match as much text as possible with each rule. In general, when attempting to write a high-speed scanner try to match as much possible in each rule, as it's a big win.
Note that start-conditions names are really integer values and can be stored as such. Thus, the above could be extended in the following fashion:
@verbatim
%x comment foo
%%
int line_num = 1;
int comment_caller;
"/*" {
comment_caller = INITIAL;
BEGIN(comment);
}
...
<foo>"/*" {
comment_caller = foo;
BEGIN(comment);
}
<comment>[^*\n]* /* eat anything that's not a '*' */
<comment>"*"+[^*/\n]* /* eat up '*'s not followed by '/'s */
<comment>\n ++line_num;
<comment>"*"+"/" BEGIN(comment_caller);
Furthermore, you can access the current start condition using the
integer-valued YY_START macro. For example, the above
assignments to comment_caller could instead be written
@verbatim
comment_caller = YY_START;
Flex provides YYSTATE as an alias for YY_START (since that
is what's used by AT&T lex).
For historical reasons, start conditions do not have their own name-space within the generated scanner. The start condition names are unmodified in the generated scanner and generated header. @xref{option-header}. @xref{option-prefix}.
Finally, here's an example of how to match C-style quoted strings using exclusive start conditions, including expanded escape sequences (but not including checking for a string that's too long):
@verbatim
%x str
%%
char string_buf[MAX_STR_CONST];
char *string_buf_ptr;
\" string_buf_ptr = string_buf; BEGIN(str);
<str>\" { /* saw closing quote - all done */
BEGIN(INITIAL);
*string_buf_ptr = '\0';
/* return string constant token type and
* value to parser
*/
}
<str>\n {
/* error - unterminated string constant */
/* generate error message */
}
<str>\\[0-7]{1,3} {
/* octal escape sequence */
int result;
(void) sscanf( yytext + 1, "%o", &result );
if ( result > 0xff )
/* error, constant is out-of-bounds */
*string_buf_ptr++ = result;
}
<str>\\[0-9]+ {
/* generate error - bad escape sequence; something
* like '\48' or '\0777777'
*/
}
<str>\\n *string_buf_ptr++ = '\n';
<str>\\t *string_buf_ptr++ = '\t';
<str>\\r *string_buf_ptr++ = '\r';
<str>\\b *string_buf_ptr++ = '\b';
<str>\\f *string_buf_ptr++ = '\f';
<str>\\(.|\n) *string_buf_ptr++ = yytext[1];
<str>[^\\\n\"]+ {
char *yptr = yytext;
while ( *yptr )
*string_buf_ptr++ = *yptr++;
}
Often, such as in some of the examples above, you wind up writing a whole bunch of rules all preceded by the same start condition(s). Flex makes this a little easier and cleaner by introducing a notion of start condition scope. A start condition scope is begun with:
@verbatim
<SCs>{
where SCs is a list of one or more start conditions. Inside the
start condition scope, every rule automatically has the prefix
SCs> applied to it, until a `}' which matches the initial
`{'. So, for example,
@verbatim
<ESC>{
"\\n" return '\n';
"\\r" return '\r';
"\\f" return '\f';
"\\0" return '\0';
}
is equivalent to:
@verbatim
<ESC>"\\n" return '\n';
<ESC>"\\r" return '\r';
<ESC>"\\f" return '\f';
<ESC>"\\0" return '\0';
Start condition scopes may be nested.
The following routines are available for manipulating stacks of start conditions:
new_state )
new_state
as though you had used
BEGIN new_state
(recall that start condition names are also integers).
BEGIN.
The start condition stack grows dynamically and so has no built-in size limitation. If memory is exhausted, program execution aborts.
To use start condition stacks, your scanner must include a %option
stack directive (see section Scanner Options).
Some scanners (such as those which support "include" files) require
reading from several input streams. As flex scanners do a large
amount of buffering, one cannot control where the next input will be
read from by simply writing a YY_INPUT() which is sensitive to
the scanning context. YY_INPUT() is only called when the scanner
reaches the end of its buffer, which may be a long time after scanning a
statement such as an include statement which requires switching
the input source.
To negotiate these sorts of problems, flex provides a mechanism
for creating and switching between multiple input buffers. An input
buffer is created by using:
which takes a FILE pointer and a size and creates a buffer
associated with the given file and large enough to hold size
characters (when in doubt, use YY_BUF_SIZE for the size). It
returns a YY_BUFFER_STATE handle, which may then be passed to
other routines (see below).
The YY_BUFFER_STATE type is a
pointer to an opaque struct yy_buffer_state structure, so you may
safely initialize YY_BUFFER_STATE variables to ((YY_BUFFER_STATE)
0) if you wish, and also refer to the opaque structure in order to
correctly declare input buffers in source files other than that of your
scanner. Note that the FILE pointer in the call to
yy_create_buffer is only used as the value of `yyin' seen by
YY_INPUT. If you redefine YY_INPUT() so it no longer uses
`yyin', then you can safely pass a nil FILE pointer to
yy_create_buffer. You select a particular buffer to scan from
using:
The above switches the scanner's input buffer so subsequent tokens will
come from new_buffer. Note that yy_switch_to_buffer() may
be used by yywrap() to set things up for continued scanning,
instead of opening a new file and pointing `yyin' at it. Note also
that switching input sources via either yy_switch_to_buffer() or
yywrap() does not change the start condition.
is used to reclaim the storage associated with a buffer. (buffer
can be nil, in which case the routine does nothing.) You can also clear
the current contents of a buffer using:
This function discards the buffer's contents,
so the next time the scanner attempts to match a token from the
buffer, it will first fill the buffer anew using
YY_INPUT().
is an alias for yy_create_buffer(),
provided for compatibility with the C++ use of new and
delete for creating and destroying dynamic objects.
Finally, the macro YY_CURRENT_BUFFER macro returns a
YY_BUFFER_STATE handle to the current buffer.
Here is an example of using these features for writing a scanner
which expands include files (the
<<EOF>>
feature is discussed below):
@verbatim
/* the "incl" state is used for picking up the name
* of an include file
*/
%x incl
%{
#define MAX_INCLUDE_DEPTH 10
YY_BUFFER_STATE include_stack[MAX_INCLUDE_DEPTH];
int include_stack_ptr = 0;
%}
%%
include BEGIN(incl);
[a-z]+ ECHO;
[^a-z\n]*\n? ECHO;
<incl>[ \t]* /* eat the whitespace */
<incl>[^ \t\n]+ { /* got the include file name */
if ( include_stack_ptr >= MAX_INCLUDE_DEPTH )
{
fprintf( stderr, "Includes nested too deeply" );
exit( 1 );
}
include_stack[include_stack_ptr++] =
YY_CURRENT_BUFFER;
yyin = fopen( yytext, "r" );
if ( ! yyin )
error( ... );
yy_switch_to_buffer(
yy_create_buffer( yyin, YY_BUF_SIZE ) );
BEGIN(INITIAL);
}
<<EOF>> {
if ( --include_stack_ptr 0 )
{
yyterminate();
}
else
{
yy_delete_buffer( YY_CURRENT_BUFFER );
yy_switch_to_buffer(
include_stack[include_stack_ptr] );
}
}
@anchor{Scanning Strings}
The following routines are available for setting up input buffers for
scanning in-memory strings instead of files. All of them create a new
input buffer for scanning the string, and return a corresponding
YY_BUFFER_STATE handle (which you should delete with
yy_delete_buffer() when done with it). They also switch to the
new buffer using yy_switch_to_buffer(), so the next call to
yylex() will start scanning the string.
len bytes (including possibly NULs) starting at location
bytes.
Note that both of these functions create and scan a copy of the
string or bytes. (This may be desirable, since yylex() modifies
the contents of the buffer it is scanning.) You can avoid the copy by
using:
base, consisting of
size bytes, the last two bytes of which must be
YY_END_OF_BUFFER_CHAR (ASCII NUL). These last two bytes are not
scanned; thus, scanning consists of base[0] through
base[size-2], inclusive.
If you fail to set up base in this manner (i.e., forget the final
two YY_END_OF_BUFFER_CHAR bytes), then yy_scan_buffer()
returns a nil pointer instead of creating a new input buffer.
The special rule <<EOF>> indicates
actions which are to be taken when an end-of-file is
encountered and yywrap() returns non-zero (i.e., indicates
no further files to process). The action must finish
by doing one of the following things:
flex, after doing the assignment you had to call the special
action YY_NEW_FILE. This is no longer necessary.)
return statement;
yyterminate() action.
yy_switch_to_buffer() as
shown in the example above.
<<EOF>> rules may not be used with other patterns; they may only be qualified with a list of start conditions. If an unqualified <<EOF>> rule is given, it applies to all start conditions which do not already have <<EOF>> actions. To specify an <<EOF>> rule for only the initial start condition, use:
@verbatim
<INITIAL><<EOF>>
These rules are useful for catching things like unclosed comments. An example:
@verbatim
%x quote
%%
...other rules for dealing with quotes...
<quote><<EOF>> {
error( "unterminated quote" );
yyterminate();
}
<<EOF>> {
if ( *++filelist )
yyin = fopen( *filelist, "r" );
else
yyterminate();
}
The macro YY_USER_ACTION can be defined to provide an action
which is always executed prior to the matched rule's action. For
example, it could be #define'd to call a routine to convert yytext to
lower-case. When YY_USER_ACTION is invoked, the variable
yy_act gives the number of the matched rule (rules are numbered
starting with 1). Suppose you want to profile how often each of your
rules is matched. The following would do the trick:
@verbatim
#define YY_USER_ACTION ++ctr[yy_act]
where ctr is an array to hold the counts for the different rules.
Note that the macro YY_NUM_RULES gives the total number of rules
(including the default rule), even if you use `-s)', so a correct
declaration for ctr is:
@verbatim
int ctr[YY_NUM_RULES];
The macro YY_USER_INIT may be defined to provide an action which
is always executed before the first scan (and before the scanner's
internal initializations are done). For example, it could be used to
call a routine to read in a data table or open a logging file.
The macro yy_set_interactive(is_interactive) can be used to
control whether the current buffer is considered interactive. An
interactive buffer is processed more slowly, but must be used when the
scanner's input source is indeed interactive to avoid problems due to
waiting to fill buffers (see the discussion of the `-I' flag in
section Scanner Options). A non-zero value in the macro invocation marks
the buffer as interactive, a zero value as non-interactive. Note that
use of this macro overrides %option always-interactive or
%option never-interactive (see section Scanner Options).
yy_set_interactive() must be invoked prior to beginning to scan
the buffer that is (or is not) to be considered interactive.
The macro yy_set_bol(at_bol) can be used to control whether the
current buffer's scanning context for the next token match is done as
though at the beginning of a line. A non-zero macro argument makes
rules anchored with `^' active, while a zero argument makes
`^' rules inactive.
The macro YY_AT_BOL() returns true if the next token scanned from
the current buffer will have `^' rules active, false otherwise.
In the generated scanner, the actions are all gathered in one large
switch statement and separated using YY_BREAK, which may be
redefined. By default, it is simply a break, to separate each
rule's action from the following rule's. Redefining YY_BREAK
allows, for example, C++ users to #define YY_BREAK to do nothing (while
being very careful that every rule ends with a break" or a
return!) to avoid suffering from unreachable statement warnings
where because a rule's action ends with return, the
YY_BREAK is inaccessible.
This chapter summarizes the various values available to the user in the rule actions.
char *yytext
%array appears in the first section of
the scanner description, then yytext is instead declared
char yytext[YYLMAX], where YYLMAX is a macro definition
that you can redefine in the first section if you don't like the default
value (generally 8KB). Using %array results in somewhat slower
scanners, but the value of yytext becomes immune to calls to
unput(), which potentially destroy its value when yytext is
a character pointer. The opposite of %array is %pointer,
which is the default.
You cannot use %array when generating C++ scanner classes (the
`-+' flag).
int yyleng
FILE *yyin
flex reads from. It may be
redefined but doing so only makes sense before scanning begins or after
an EOF has been encountered. Changing it in the midst of scanning will
have unexpected results since flex buffers its input; use
yyrestart() instead. Once scanning terminates because an
end-of-file has been seen, you can assign `yyin' at the new input
file and then call the scanner again to continue scanning.
void yyrestart( FILE *new_file )
yyrestart() with `yyin'
as an argument thus throws away the current input buffer and continues
scanning the same input file.
FILE *yyout
ECHO actions are done. It can be reassigned
by the user.
YY_CURRENT_BUFFER
YY_BUFFER_STATE handle to the current buffer.
YY_START
BEGIN to return to that
start condition.
One of the main uses of flex is as a companion to the yacc
parser-generator. yacc parsers expect to call a routine named
yylex() to find the next input token. The routine is supposed to
return the type of the next token as well as putting any associated
value in the global yylval. To use flex with yacc,
one specifies the `-d' option to yacc to instruct it to
generate the file `y.tab.h' containing definitions of all the
%tokens appearing in the yacc input. This file is then
included in the flex scanner. For example, if one of the tokens
is TOK_NUMBER, part of the scanner might look like:
@verbatim
%{
#include "y.tab.h"
%}
%%
[0-9]+ yylval = atoi( yytext ); return TOK_NUMBER;
The various flex options are categorized by function in the following
menu. If you want to lookup a particular option by name, See section Index of Scanner Options.
Even though there are many scanner options, a typical scanner might only specify the following options:
@verbatim %option 8-bit reentrant bison-bridge %option warn nodefault %option yylineno %option outfile="scanner.c" header-file="scanner.h"
The first line specifies the general type of scanner we want. The second line specifies that we are being careful. The third line asks flex to track line numbers. The last line tells flex what to name the files. (The options can be specified in any order. We just dividied them.)
flex also provides a mechanism for controlling options within the
scanner specification itself, rather than from the flex command-line.
This is done by including %option directives in the first section
of the scanner specification. You can specify multiple options with a
single %option directive, and multiple directives in the first
section of your flex input file.
Most options are given simply as names, optionally preceded by the word `no' (with no intervening whitespace) to negate their meaning. The names are the same as their long-option equivalents (but without the leading `--' ).
flex scans your rule actions to determine whether you use the
REJECT or yymore() features. The REJECT and
yymore options are available to override its decision as to
whether you use the options, either by setting them (e.g., %option
reject) to indicate the feature is indeed used, or unsetting them to
indicate it actually is not used (e.g., %option noyymore).
A number of options are available for lint purists who want to suppress
the appearance of unneeded routines in the generated scanner. Each of
the following, if unset (e.g., %option nounput), results in the
corresponding routine not appearing in the generated scanner:
@verbatim
input, unput
yy_push_state, yy_pop_state, yy_top_state
yy_scan_buffer, yy_scan_bytes, yy_scan_string
yyget_extra, yyset_extra, yyget_leng, yyget_text,
yyget_lineno, yyset_lineno, yyget_in, yyset_in,
yyget_out, yyset_out, yyget_lval, yyset_lval,
yyget_lloc, yyset_lloc, yyget_debug, yyset_debug
(though yy_push_state() and friends won't appear anyway unless
you use %option stack).
%option header-file="FILE"'
yyIN_HEADER is defined, where `yy'
is substituted with the appropriate prefix.
The `--header-file' option is not compatible with the `--c++' option,
since the C++ scanner provides its own header in `yyFlexLexer.h'.
@anchor{option-outfile}
%option outfile="FILE"'
#line
directives (see the `-l' option above) refer to the file
`FILE'.
@anchor{option-stdout}
%option stdout'
flex to write the scanner it generates to standard
output instead of `lex.yy.c'.
flex
constructs its scanners. You'll never need this option unless you are doing
flex
maintenance or development.
%option case-insensitive'
flex to generate a case-insensitive scanner. The
case of letters given in the flex input patterns will be ignored,
and tokens in the input will be matched regardless of case. The matched
text given in yytext will have the preserved case (i.e., it will
not be folded).
@anchor{option-lex-compat}
%option lex-compat'
lex
implementation. Note that this does not mean full compatibility.
Use of this option costs a considerable amount of performance, and it
cannot be used with the `--c++', `--full', `--fast', `-Cf', or
`-CF' options. For details on the compatibilities it provides, see
section Incompatibilities with Lex and Posix. This option also results in the name
YY_FLEX_LEX_COMPAT being #define'd in the generated scanner.
@anchor{option-batch}
%option batch'
flex to generate a batch scanner, the opposite of
interactive scanners generated by `--interactive' (see below). In
general, you use `-B' when you are certain that your scanner
will never be used interactively, and you want to squeeze a
little more performance out of it. If your goal is instead to
squeeze out a lot more performance, you should be using the
`-Cf' or `-CF' options, which turn on `--batch' automatically
anyway.
@anchor{option-interactive}
%option interactive'
flex to generate an interactive scanner. An
interactive scanner is one that only looks ahead to decide what token
has been matched if it absolutely must. It turns out that always
looking one extra character ahead, even if the scanner has already seen
enough text to disambiguate the current token, is a bit faster than only
looking ahead when necessary. But scanners that always look ahead give
dreadful interactive performance; for example, when a user types a
newline, it is not recognized as a newline token until they enter
another token, which often means typing in another whole line.
flex scanners default to interactive unless you use the
`-Cf' or `-CF' table-compression options
(see section Performance Considerations). That's because if you're looking for
high-performance you should be using one of these options, so if you
didn't, flex assumes you'd rather trade off a bit of run-time
performance for intuitive interactive behavior. Note also that you
cannot use `--interactive' in conjunction with `-Cf' or
`-CF'. Thus, this option is not really needed; it is on by default
for all those cases in which it is allowed.
You can force a scanner to
not
be interactive by using
`--batch'
@anchor{option-7bit}
%option 7bit'
flex to generate a 7-bit scanner, i.e., one which can
only recognize 7-bit characters in its input. The advantage of using
`--7bit' is that the scanner's tables can be up to half the size of
those generated using the `--8bit'. The disadvantage is that such
scanners often hang or crash if their input contains an 8-bit character.
Note, however, that unless you generate your scanner using the
`-Cf' or `-CF' table compression options, use of `--7bit'
will save only a small amount of table space, and make your scanner
considerably less portable. Flex's default behavior is to
generate an 8-bit scanner unless you use the `-Cf' or `-CF',
in which case flex defaults to generating 7-bit scanners unless
your site was always configured to generate 8-bit scanners (as will
often be the case with non-USA sites). You can tell whether flex
generated a 7-bit or an 8-bit scanner by inspecting the flag summary in
the `--verbose' output as described above.
Note that if you use `-Cfe' or `-CFe' flex still
defaults to generating an 8-bit scanner, since usually with these
compression options full 8-bit tables are not much more expensive than
7-bit tables.
@anchor{option-8bit}
%option 8bit'
flex to generate an 8-bit scanner, i.e., one which can
recognize 8-bit characters. This flag is only needed for scanners
generated using `-Cf' or `-CF', as otherwise flex defaults to
generating an 8-bit scanner anyway.
See the discussion of
`--7bit'
above for flex's default behavior and the tradeoffs between 7-bit
and 8-bit scanners.
@anchor{option-default}
%option default'
%option always-interactive'
isatty() in an attempt to determine whether the scanner's input
source is interactive and thus should be read a character at a time.
When this option is used, however, then no such call is made.
--never-interactive'
always-interactive.
@anchor{option-posix}
%option posix'
lex. Since flex was originally designed to implement the
POSIX definition of lex this generally involves very few changes
in behavior. At the current writing the known differences between
flex and the POSIX standard are:
lex, the repeat operator, `{}', has lower
precedence than concatenation (thus `ab{3}' yields `ababab').
Most POSIX utilities use an Extended Regular Expression (ERE) precedence
that has the precedence of the repeat operator higher than concatenation
(which causes `ab{3}' to yield `abbb'). By default, flex
places the precedence of the repeat operator higher than concatenation
which matches the ERE processing of other POSIX utilities. When either
`--posix' or `-l' are specified, flex will use the
traditional AT&T and POSIX-compliant precedence for the repeat operator
where concatenation has higher precedence than the repeat operator.
%option stack'
%option stdinit'
yyin and
yyout to `stdin' and `stdout', instead of the default of
`nil'. Some existing lex programs depend on this behavior,
even though it is not compliant with ANSI C, which does not require
`stdin' and `stdout' to be compile-time constant. In a
reentrant scanner, however, this is not a problem since initialization
is performed in yylex_init at runtime.
@anchor{option-yylineno}
%option yylineno'
flex to generate a scanner
that maintains the number of the current line read from its input in the
global variable yylineno. This option is implied by %option
lex-compat. In a reentrant C scanner, the macro yylineno is
accessible regardless of the value of %option yylineno, however, its
value is not modified by flex unless %option yylineno is enabled.
@anchor{option-yywrap}
%option yywrap'
--noyywrap), makes the scanner not call
yywrap() upon an end-of-file, but simply assume that there are no
more files to scan (until the user points `yyin' at a new file and
calls yylex() again).
%option bison-bridge'
GNU bison
parser. The scanner has minor API changes for
bison
compatibility. In particular, the declaration of
yylex
is modified, and support for
yylval
and
yylloc
is incorporated. See section C Scanners with Bison Parsers.
@anchor{option-noline}
%option noline'
flex
not to generate
#line
directives. Without this option,
flex
peppers the generated scanner
with #line directives so error messages in the actions will be correctly
located with respect to either the original
flex
input file (if the errors are due to code in the input file), or
`lex.yy.c'
(if the errors are
flex's
fault -- you should report these sorts of errors to the email address
given in section Reporting Bugs).
@anchor{option-reentrant}
%option reentrant'
flex scanners, non-reentrant flex
code must be modified before it is suitable for use with this option.
This option is not compatible with the `--c++' option.
The option `--reentrant' does not affect the performance of
the scanner.
@anchor{option-c++}
%option c++'
%option array'
%option pointer'
yytext should be a char *, not an array.
This default is char *.
@anchor{option-prefix}
%option prefix="PREFIX"'
flex for all
globally-visible variable and function names to instead be
`PREFIX'. For example, `--prefix=foo' changes the name of
yytext to footext. It also changes the name of the default
output file from `lex.yy.c' to `lex.foo.c'. Here is a partial
list of the names affected:
@verbatim
yy_create_buffer
yy_delete_buffer
yy_flex_debug
yy_init_buffer
yy_flush_buffer
yy_load_buffer_state
yy_switch_to_buffer
yyin
yyleng
yylex
yylineno
yyout
yyrestart
yytext
yywrap
yyalloc
yyrealloc
yyfree
(If you are using a C++ scanner, then only yywrap and
yyFlexLexer are affected.) Within your scanner itself, you can
still refer to the global variables and functions using either version
of their name; but externally, they have the modified name.
This option lets you easily link together multiple
flex
programs into the same executable. Note, though, that using this
option also renames
yywrap(),
so you now
must
either
provide your own (appropriately-named) version of the routine for your
scanner, or use
%option noyywrap,
as linking with
`-lfl'
no longer provides one for you by default.
@anchor{option-main}
%option main'
main() program for the
scanner, which simply calls yylex(). This option implies
noyywrap (see below).
@anchor{option-nounistd}
%option nounistd'
isatty(), read().)
If you wish to use these functions, you will have to inform your compiler where
to find them.
@xref{option-always-interactive}. @xref{option-read}.
@anchor{option-yyclass}
%option yyclass="NAME"'
flex that you have derived foo as a subclass of
yyFlexLexer, so flex will place your actions in the member
function foo::yylex() instead of yyFlexLexer::yylex(). It
also generates a yyFlexLexer::yylex() member function that emits
a run-time error (by invoking yyFlexLexer::LexerError()) if
called. See section Generating C++ Scanners.
%option align'
%option ecs'
flex to construct equivalence classes, i.e., sets
of characters which have identical lexical properties (for example, if
the only appearance of digits in the flex input is in the
character class "[0-9]" then the digits '0', '1', ..., '9' will all be
put in the same equivalence class). Equivalence classes usually give
dramatic reductions in the final table/object file sizes (typically a
factor of 2-5) and are pretty cheap performance-wise (one array look-up
per character scanned).
flex should not compress the tables by taking advantages of
similar transition functions for different states.
%option meta-ecs'
flex
to construct
meta-equivalence classes,
which are sets of equivalence classes (or characters, if equivalence
classes are not being used) that are commonly used together. Meta-equivalence
classes are often a big win when using compressed tables, but they
have a moderate performance impact (one or two if tests and one
array look-up per character scanned).
@anchor{option-read}
%option read'
stdio) for input. Instead of calling fread() or
getc(), the scanner will use the read() system call,
resulting in a performance gain which varies from system to system, but
in general is probably negligible unless you are also using `-Cf'
or `-CF'. Using `-Cr' can cause strange behavior if, for
example, you read from `yyin' using stdio prior to calling
the scanner (because the scanner will miss whatever text your previous
reads left in the stdio input buffer). `-Cr' has no effect
if you define YY_INPUT() (see section The Generated Scanner).
flex
should generate equivalence classes and meta-equivalence classes. This
setting provides the highest degree of table compression. You can trade
off faster-executing scanners at the cost of larger tables with the
following generally being true:
@verbatim
slowest & smallest
-Cem
-Cm
-Ce
-C
-C{f,F}e
-C{f,F}
-C{f,F}a
fastest & largest
Note that scanners with the smallest tables are usually generated and
compiled the quickest, so during development you will usually want to
use the default, maximal compression.
`-Cfe' is often a good compromise between speed and size for
production scanners.
@anchor{option-full}
%option full'
stdio is bypassed.
The result is large but fast. This option is equivalent to
`--Cfr'
@anchor{option-fast}
%option fast'
stdio bypassed). This representation is about as fast
as the full table representation `--full', and for some sets of
patterns will be considerably smaller (and for others, larger). In
general, if the pattern set contains both keywords and a
catch-all, identifier rule, such as in the set:
@verbatim
"case" return TOK_CASE;
"switch" return TOK_SWITCH;
...
"default" return TOK_DEFAULT;
[a-z]+ return TOK_ID;
then you're better off using the full table representation. If only
the identifier rule is present and you then use a hash table or some such
to detect the keywords, you're better off using
`--fast'.
This option is equivalent to `-CFr' (see below). It cannot be used
with `--c++'.
%option backup'
-CF
is used, the generated scanner will run faster (see the `--perf-report' flag).
Only users who wish to squeeze every last cycle out of their scanners
need worry about this option. (see section Performance Considerations).
@anchor{option-debug}
%option debug'
yy_flex_debug is non-zero
(which is the default), the scanner will write to `stderr' a line
of the form:
@verbatim
-accepting rule at line 53 ("the matched text")
The line number refers to the location of the rule in the file defining
the scanner (i.e., the file that was fed to flex). Messages are also
generated when the scanner backs up, accepts the default rule, reaches
the end of its input buffer (or encounters a NUL; at this point, the two
look the same as far as the scanner's concerned), or reaches an
end-of-file.
@anchor{option-perf-report}
%option perf-report'
flex input file which will
cause a serious loss of performance in the resulting scanner. If you
give the flag twice, you will also get comments regarding features that
lead to minor performance losses.
Note that the use of REJECT, and
variable trailing context (see section Limitations) entails a substantial
performance penalty; use of yymore(), the `^' operator, and
the `--interactive' flag entail minor performance penalties.
@anchor{option-nodefault}
%option nodefault'
%option trace'
flex run in trace mode. It will generate a lot of
messages to `stderr' concerning the form of the input and the
resultant non-deterministic and deterministic finite automata. This
option is mostly for use in maintaining flex.
@anchor{option-nowarn}
%option nowarn'
%option verbose'
flex should write to `stderr' a summary of
statistics regarding the scanner it generates. Most of the statistics
are meaningless to the casual flex user, but the first line
identifies the version of flex (same as reported by `--version'),
and the next line the flags used when generating the scanner, including
those that are on by default.
@anchor{option-warn}
%option warn'
flex's options to `stdout'
and then exits.
The main design goal of flex is that it generate high-performance
scanners. It has been optimized for dealing well with large sets of
rules. Aside from the effects on scanner speed of the table compression
`-C' options outlined above, there are a number of options/actions
which degrade performance. These are, from most expensive to least:
@verbatim
REJECT
arbitrary trailing context
pattern sets that require backing up
%option yylineno
%array
%option interactive
%option always-interactive
`^' beginning-of-line operator
yymore()
with the first two all being quite expensive and the last two being
quite cheap. Note also that unput() is implemented as a routine
call that potentially does quite a bit of work, while yyless() is
a quite-cheap macro. So if you are just putting back some excess text
you scanned, use ss().
REJECT should be avoided at all costs when performance is
important. It is a particularly expensive option.
There is one case when %option yylineno can be expensive. That is when
your patterns match long tokens that could possibly contain a newline
character. There is no performance penalty for rules that can not possibly
match newlines, since flex does not need to check them for newlines. In
general, you should avoid rules such as [^f]+, which match very long
tokens, including newlines, and may possibly match your entire file! A better
approach is to separate [^f]+ into two rules:
@verbatim
%option yylineno
%%
[^f\n]+
\n+
The above scanner does not incur a performance penalty.
Getting rid of backing up is messy and often may be an enormous amount of work for a complicated scanner. In principal, one begins by using the `-b' flag to generate a `lex.backup' file. For example, on the input:
@verbatim
%%
foo return TOK_KEYWORD;
foobar return TOK_KEYWORD;
the file looks like:
@verbatim
State #6 is non-accepting -
associated rule line numbers:
2 3
out-transitions: [ o ]
jam-transitions: EOF [ \001-n p-\177 ]
State #8 is non-accepting -
associated rule line numbers:
3
out-transitions: [ a ]
jam-transitions: EOF [ \001-` b-\177 ]
State #9 is non-accepting -
associated rule line numbers:
3
out-transitions: [ r ]
jam-transitions: EOF [ \001-q s-\177 ]
Compressed tables always back up.
The first few lines tell us that there's a scanner state in which it can make a transition on an 'o' but not on any other character, and that in that state the currently scanned text does not match any rule. The state occurs when trying to match the rules found at lines 2 and 3 in the input file. If the scanner is in that state and then reads something other than an 'o', it will have to back up to find a rule which is matched. With a bit of headscratching one can see that this must be the state it's in when it has seen `fo'. When this has happened, if anything other than another `o' is seen, the scanner will have to back up to simply match the `f' (by the default rule).
The comment regarding State #8 indicates there's a problem when `foob' has been scanned. Indeed, on any character other than an `a', the scanner will have to back up to accept "foo". Similarly, the comment for State #9 concerns when `fooba' has been scanned and an `r' does not follow.
The final comment reminds us that there's no point going to all the trouble of removing backing up from the rules unless we're using `-Cf' or `-CF', since there's no performance gain doing so with compressed scanners.
The way to remove the backing up is to add "error" rules:
@verbatim
%%
foo return TOK_KEYWORD;
foobar return TOK_KEYWORD;
fooba |
foob |
fo {
/* false alarm, not really a keyword */
return TOK_ID;
}
Eliminating backing up among a list of keywords can also be done using a "catch-all" rule:
@verbatim
%%
foo return TOK_KEYWORD;
foobar return TOK_KEYWORD;
[a-z]+ return TOK_ID;
This is usually the best solution when appropriate.
Backing up messages tend to cascade. With a complicated set of rules
it's not uncommon to get hundreds of messages. If one can decipher
them, though, it often only takes a dozen or so rules to eliminate the
backing up (though it's easy to make a mistake and have an error rule
accidentally match a valid token. A possible future flex feature
will be to automatically add rules to eliminate backing up).
It's important to keep in mind that you gain the benefits of eliminating backing up only if you eliminate every instance of backing up. Leaving just one means you gain nothing.
Variable trailing context (where both the leading and trailing
parts do not have a fixed length) entails almost the same performance
loss as REJECT (i.e., substantial). So when possible a rule
like:
@verbatim
%%
mouse|rat/(cat|dog) run();
is better written:
@verbatim
%%
mouse/cat|dog run();
rat/cat|dog run();
or as
@verbatim
%%
mouse|rat/cat run();
mouse|rat/dog run();
Note that here the special '|' action does not provide any savings, and can even make things worse (see section Limitations).
Another area where the user can increase a scanner's performance (and
one that's easier to implement) arises from the fact that the longer the
tokens matched, the faster the scanner will run. This is because with
long tokens the processing of most input characters takes place in the
(short) inner scanning loop, and does not often have to go through the
additional work of setting up the scanning environment (e.g.,
yytext) for the action. Recall the scanner for C comments:
@verbatim
%x comment
%%
int line_num = 1;
"/*" BEGIN(comment);
<comment>[^*\n]*
<comment>"*"+[^*/\n]*
<comment>\n ++line_num;
<comment>"*"+"/" BEGIN(INITIAL);
This could be sped up by writing it as:
@verbatim
%x comment
%%
int line_num = 1;
"/*" BEGIN(comment);
<comment>[^*\n]*
<comment>[^*\n]*\n ++line_num;
<comment>"*"+[^*/\n]*
<comment>"*"+[^*/\n]*\n ++line_num;
<comment>"*"+"/" BEGIN(INITIAL);
Now instead of each newline requiring the processing of another action, recognizing the newlines is distributed over the other rules to keep the matched text as long as possible. Note that adding rules does not slow down the scanner! The speed of the scanner is independent of the number of rules or (modulo the considerations given at the beginning of this section) how complicated the rules are with regard to operators such as `*' and `|'.
A final example in speeding up a scanner: suppose you want to scan through a file containing identifiers and keywords, one per line and with no other extraneous characters, and recognize all the keywords. A natural first approach is:
@verbatim
%%
asm |
auto |
break |
... etc ...
volatile |
while /* it's a keyword */
.|\n /* it's not a keyword */
To eliminate the back-tracking, introduce a catch-all rule:
@verbatim
%%
asm |
auto |
break |
... etc ...
volatile |
while /* it's a keyword */
[a-z]+ |
.|\n /* it's not a keyword */
Now, if it's guaranteed that there's exactly one word per line, then we can reduce the total number of matches by a half by merging in the recognition of newlines with that of the other tokens:
@verbatim
%%
asm\n |
auto\n |
break\n |
... etc ...
volatile\n |
while\n /* it's a keyword */
[a-z]+\n |
.|\n /* it's not a keyword */
One has to be careful here, as we have now reintroduced backing up
into the scanner. In particular, while
we
know that there will never be any characters in the input stream
other than letters or newlines,
flex
can't figure this out, and it will plan for possibly needing to back up
when it has scanned a token like `auto' and then the next character
is something other than a newline or a letter. Previously it would
then just match the `auto' rule and be done, but now it has no `auto'
rule, only a `auto\n' rule. To eliminate the possibility of backing up,
we could either duplicate all rules but without final newlines, or,
since we never expect to encounter such an input and therefore don't
how it's classified, we can introduce one more catch-all rule, this
one which doesn't include a newline:
@verbatim
%%
asm\n |
auto\n |
break\n |
... etc ...
volatile\n |
while\n /* it's a keyword */
[a-z]+\n |
[a-z]+ |
.|\n /* it's not a keyword */
Compiled with `-Cf', this is about as fast as one can get a
flex scanner to go for this particular problem.
A final note: flex is slow when matching NULs,
particularly when a token contains multiple NULs. It's best to
write rules which match short amounts of text if it's anticipated
that the text will often include NULs.
Another final note regarding performance: as mentioned in
section How the Input Is Matched, dynamically resizing yytext to accommodate huge
tokens is a slow process because it presently requires that the (huge)
token be rescanned from the beginning. Thus if performance is vital,
you should attempt to match "large" quantities of text but not
"huge" quantities, where the cutoff between the two is at about 8K
characters per token.
IMPORTANT: the present form of the scanning class is experimental and may change considerably between major releases.
flex provides two different ways to generate scanners for use
with C++. The first way is to simply compile a scanner generated by
flex using a C++ compiler instead of a C compiler. You should
not encounter any compilation errors (see section Reporting Bugs). You can
then use C++ code in your rule actions instead of C code. Note that the
default input source for your scanner remains `yyin', and default
echoing is still done to `yyout'. Both of these remain FILE
* variables and not C++ streams.
You can also use flex to generate a C++ scanner class, using the
`-+' option (or, equivalently, %option c++), which is
automatically specified if the name of the flex executable ends
in a '+', such as flex++. When using this option, flex
defaults to generating the scanner to the file `lex.yy.cc' instead
of `lex.yy.c'. The generated scanner includes the header file
`FlexLexer.h', which defines the interface to two C++ classes.
The first class,
FlexLexer,
provides an abstract base class defining the general scanner class
interface. It provides the following member functions:
const char* YYText()
yytext.
int YYLeng()
yyleng.
int lineno() const
%option yylineno), or
1 if %option yylineno was not used.
void set_debug( int flag )
yy_flex_debug (see section Scanner Options). Note that you must build
the scannerusing %option debug to include debugging information
in it.
int debug() const
Also provided are member functions equivalent to
yy_switch_to_buffer(), yy_create_buffer() (though the
first argument is an istream* object pointer and not a
FILE*), yy_flush_buffer(), yy_delete_buffer(), and
yyrestart() (again, the first argument is a istream*
object pointer).
The second class defined in `FlexLexer.h' is yyFlexLexer,
which is derived from FlexLexer. It defines the following
additional member functions:
yyFlexLexer( istream* arg_yyin = 0, ostream* arg_yyout = 0 )
yyFlexLexer object using the given streams for input
and output. If not specified, the streams default to cin and
cout, respectively.
virtual int yylex()
yylex() does for ordinary flex
scanners: it scans the input stream, consuming tokens, until a rule's
action returns a value. If you derive a subclass S from
yyFlexLexer and want to access the member functions and variables
of S inside yylex(), then you need to use %option
yyclass="S" to inform flex that you will be using that subclass
instead of yyFlexLexer. In this case, rather than generating
yyFlexLexer::yylex(), flex generates S::yylex()
(and also generates a dummy yyFlexLexer::yylex() that calls
yyFlexLexer::LexerError() if called).
virtual void switch_streams(istream* new_in = 0, ostream* new_out = 0)
yyin to new_in (if non-nil) and yyout to
new_out (if non-nil), deleting the previous input buffer if
yyin is reassigned.
int yylex( istream* new_in, ostream* new_out = 0 )
switch_streams( new_in,
new_out ) and then returns the value of yylex().
In addition, yyFlexLexer defines the following protected virtual
functions which you can redefine in derived classes to tailor the
scanner:
virtual int LexerInput( char* buf, int max_size )
max_size characters into buf and returns the
number of characters read. To indicate end-of-input, return 0
characters. Note that interactive scanners (see the `-B'
and `-I' flags in section Scanner Options) define the macro
YY_INTERACTIVE. If you redefine LexerInput() and need to
take different actions depending on whether or not the scanner might be
scanning an interactive input source, you can test for the presence of
this name via #ifdef statements.
virtual void LexerOutput( const char* buf, int size )
size characters from the buffer buf, which, while
NUL-terminated, may also contain internal NULs if the
scanner's rules can match text with NULs in them.
virtual void LexerError( const char* msg )
cerr and exits.
Note that a yyFlexLexer object contains its entire
scanning state. Thus you can use such objects to create reentrant
scanners, but see also section Reentrant C Scanners. You can instantiate multiple
instances of the same yyFlexLexer class, and you can also combine
multiple C++ scanner classes together in the same program using the
`-P' option discussed above.
Finally, note that the %array feature is not available to C++
scanner classes; you must use %pointer (the default).
Here is an example of a simple C++ scanner:
@verbatim
// An example of using the flex C++ scanner class.
%{
int mylineno = 0;
%}
string \"[^\n"]+\"
ws [ \t]+
alpha [A-Za-z]
dig [0-9]
name ({alpha}|{dig}|\$)({alpha}|{dig}|[_.\-/$])*
num1 [-+]?{dig}+\.?([eE][-+]?{dig}+)?
num2 [-+]?{dig}*\.{dig}+([eE][-+]?{dig}+)?
number {num1}|{num2}
%%
{ws} /* skip blanks and tabs */
"/*" {
int c;
while((c = yyinput()) != 0)
{
if(c == '\n')
++mylineno;
else if(c == `*')
{
if((c = yyinput()) == '/')
break;
else
unput(c);
}
}
}
{number} cout "number " YYText() '\n';
\n mylineno++;
{name} cout "name " YYText() '\n';
{string} cout "string " YYText() '\n';
%%
int main( int /* argc */, char** /* argv */ )
{
flexLexer* lexer = new yyFlexLexer;
while(lexer->yylex() != 0)
;
return 0;
}
If you want to create multiple (different) lexer classes, you use the
`-P' flag (or the prefix= option) to rename each
yyFlexLexer to some other `xxFlexLexer'. You then can
include `FlexLexer.h>' in your other sources once per lexer class,
first renaming yyFlexLexer as follows:
@verbatim
#undef yyFlexLexer
#define yyFlexLexer xxFlexLexer
#include <FflexLexer.h>
#undef yyFlexLexer
#define yyFlexLexer zzFlexLexer
#include FlexLexer.h>
if, for example, you used %option prefix="xx" for one of your
scanners and %option prefix="zz" for the other.
flex has the ability to generate a reentrant C scanner. This is
accomplished by specifying %option reentrant (`-R') The generated
scanner is both portable, and safe to use in one or more separate threads of
control. The most common use for reentrant scanners is from within
multi-threaded applications. Any thread may create and execute a reentrant
flex scanner without the need for synchronization with other threads.
However, there are other uses for a reentrant scanner. For example, you
could scan two or more files simultaneously to implement a diff at
the token level (i.e., instead of at the character level):
@verbatim
/* Example of maintaining more than one active scanner. */
do {
int tok1, tok2;
tok1 = yylex( scanner_1 );
tok2 = yylex( scanner_2 );
if( tok1 != tok2 )
printf("Files are different.");
} while ( tok1 && tok2 );
Another use for a reentrant scanner is recursion. (Note that a recursive scanner can also be created using a non-reentrant scanner and buffer states. See section Multiple Input Buffers.)
The following crude scanner supports the `eval' command by invoking another instance of itself.
@verbatim
/* Example of recursive invocation. */
%option reentrant
%%
"eval(".+")" {
yyscan_t scanner;
YY_BUFFER_STATE buf;
yylex_init( &scanner );
yytext[yyleng-1] = ' ';
buf = yy_scan_string( yytext + 5, scanner );
yylex( scanner );
yy_delete_buffer(buf,scanner);
yylex_destroy( scanner );
}
...
%%
The API for reentrant scanners is different than for non-reentrant scanners. Here is a quick overview of the API:
%option reentrant must be specified.
yyscanner
yylex_init and yylex_destroy must be called before and
after yylex, respectively.
flex variables.
yyextra.
First, an example of a reentrant scanner:
@verbatim
/* This scanner prints "//" comments. */
%option reentrant stack
%x COMMENT
%%
"//" yy_push_state( COMMENT, yyscanner);
.|\n
<COMMENT>\n yy_pop_state( yyscanner );
<COMMENT>[^\n]+ fprintf( yyout, "%s\n", yytext);
%%
int main ( int argc, char * argv[] )
{
yyscan_t scanner;
yylex_init ( &scanner );
yylex ( scanner );
yylex_destroy ( scanner );
return 0;
}
Here are the things you need to do or know to use the reentrant C API of
flex.
%option reentrant (--reentrant) must be specified.
Notice that %option reentrant is specified in the above example
(see section Reentrant Example. Had this option not been specified,
flex would have happily generated a non-reentrant scanner without
complaining. You may explicitly specify %option noreentrant, if
you do not want a reentrant scanner, although it is not
necessary. The default is to generate a non-reentrant scanner.
All functions take one additional argument: yyscanner.
Notice that the calls to yy_push_state and yy_pop_state
both have an argument, yyscanner , that is not present in a
non-reentrant scanner. Here are the declarations of
yy_push_state and yy_pop_state in the generated scanner:
@verbatim
static void yy_push_state ( int new_state , yyscan_t yyscanner ) ;
static void yy_pop_state ( yyscan_t yyscanner ) ;
Notice that the argument yyscanner appears in the declaration of
both functions. In fact, all flex functions in a reentrant
scanner have this additional argument. It is always the last argument
in the argument list, it is always of type yyscan_t (which is
typedef'd to void *) and it is
always named yyscanner. As you may have guessed,
yyscanner is a pointer to an opaque data structure encapsulating
the current state of the scanner. For a list of function declarations,
see section Functions and Macros Available in Reentrant C Scanners. Note that preprocessor macros, such as
BEGIN, ECHO, and REJECT, do not take this
additional argument.
All global variables in traditional flex have been replaced by macro equivalents.
Note that in the above example, yyout and yytext are
not plain variables. These are macros that will expand to their equivalent lvalue.
All of the familiar flex globals have been replaced by their macro
equivalents. In particular, yytext, yyleng, yylineno,
yyin, yyout, yyextra, yylval, and yylloc
are macros. You may safely use these macros in actions as if they were plain
variables. We only tell you this so you don't expect to link to these variables
externally. Currently, each macro expands to a member of an internal struct, e.g.,
@verbatim #define yytext (((struct yyguts_t*)yyscanner)->yytext_r)
One important thing to remember about
yytext
and friends is that
yytext
is not a global variable in a reentrant
scanner, you can not access it directly from outside an action or from
other functions. You must use an accessor method, e.g.,
yyget_text,
to accomplish this. (See below).
yylex_init and yylex_destroy must be called before and
after yylex, respectively.
@verbatim
int yylex_init ( yyscan_t * ptr_yy_globals ) ;
int yylex ( yyscan_t yyscanner ) ;
int yylex_destroy ( yyscan_t yyscanner ) ;
The function yylex_init must be called before calling any other
function. The argument to yylex_init is the address of an
uninitialized pointer to be filled in by flex. The contents of
ptr_yy_globals need not be initialized, since flex will
overwrite it anyway. The value stored in ptr_yy_globals should
thereafter be passed to yylex() and yylex_destroy(). Flex
does not save the argument passed to yylex_init, so it is safe to
pass the address of a local pointer to yylex_init. The function
yylex should be familiar to you by now. The reentrant version
takes one argument, which is the value returned (via an argument) by
yylex_init. Otherwise, it behaves the same as the non-reentrant
version of yylex.
yylex_init returns 0 (zero) on success, or non-zero on failure,
in which case, errno is set to one of the following values:
The function yylex_destroy should be
called to free resources used by the scanner. After yylex_destroy
is called, the contents of yyscanner should not be used. Of
course, there is no need to destroy a scanner if you plan to reuse it.
A flex scanner (both reentrant and non-reentrant) may be
restarted by calling yyrestart.
Below is an example of a program that creates a scanner, uses it, then destroys it when done:
@verbatim
int main ()
{
yyscan_t scanner;
int tok;
yylex_init(&scanner);
while ((tok=yylex()) > 0)
printf("tok=%d yytext=%s\n", tok, yyget_text(scanner));
yylex_destroy(scanner);
return 0;
}
Accessor methods (get/set functions) provide access to common
flex variables.
Many scanners that you build will be part of a larger project. Portions
of your project will need access to flex values, such as
yytext. In a non-reentrant scanner, these values are global, so
there is no problem accessing them. However, in a reentrant scanner, there are no
global flex values. You can not access them directly. Instead,
you must access flex values using accessor methods (get/set
functions). Each accessor method is named yyget_NAME or
yyset_NAME, where NAME is the name of the flex
variable you want. For example:
@verbatim
/* Set the last character of yytext to NULL. */
void chop ( yyscan_t scanner )
{
int len = yyget_leng( scanner );
yyget_text( scanner )[len - 1] = '\0';
}
The above code may be called from within an action like this:
@verbatim
%%
.+\n { chop( yyscanner );}
You may find that %option header-file is particularly useful for generating
prototypes of all the accessor functions. @xref{option-header}.
User-specific data can be stored in yyextra.
In a reentrant scanner, it is unwise to use global variables to
communicate with or maintain state between different pieces of your program.
However, you may need access to external data or invoke external functions
from within the scanner actions.
Likewise, you may need to pass information to your scanner
(e.g., open file descriptors, or database connections).
In a non-reentrant scanner, the only way to do this would be through the
use of global variables.
Flex allows you to store arbitrary, "extra" data in a scanner.
This data is accessible through the accessor methods
yyget_extra
and
yyset_extra
from outside the scanner, and through the shortcut macro
yyextra
from within the scanner itself. They are defined as follows:
@verbatim
#define YY_EXTRA_TYPE void*
YY_EXTRA_TYPE yyget_extra ( yyscan_t scanner );
void yyset_extra ( YY_EXTRA_TYPE arbitrary_data , yyscan_t scanner);
By default, YY_EXTRA_TYPE is defined as type void *. You
will have to cast yyextra and the return value from
yyget_extra to the appropriate value each time you access the
extra data. To avoid casting, you may override the default type by
defining YY_EXTRA_TYPE in section 1 of your scanner:
@verbatim
/* An example of overriding YY_EXTRA_TYPE. */
%{
#include <sys/stat.h>
#include <unistd.h>
#define YY_EXTRA_TYPE struct stat*
%}
%option reentrant
%%
__filesize__ printf( "%ld", yyextra->st_size );
__lastmod__ printf( "%ld", yyextra->st_mtime );
%%
void scan_file( char* filename )
{
yyscan_t scanner;
struct stat buf;
yylex_init ( &scanner );
yyset_in( fopen(filename,"r"), scanner );
stat( filename, &buf);
yyset_extra( &buf, scanner );
yylex ( scanner );
yylex_destroy( scanner );
}
@verbatim
typedef void* yyscan_t;
It is initialized by yylex_init() to point to
an internal structure. You should never access this value
directly. In particular, you should never attempt to free it
(use yylex_destroy() instead.)
The following Functions are available in a reentrant scanner:
@verbatim
char *yyget_text ( yyscan_t scanner );
int yyget_leng ( yyscan_t scanner );
FILE *yyget_in ( yyscan_t scanner );
FILE *yyget_out ( yyscan_t scanner );
int yyget_lineno ( yyscan_t scanner );
YY_EXTRA_TYPE yyget_extra ( yyscan_t scanner );
int yyget_debug ( yyscan_t scanner );
void yyset_debug ( int flag, yyscan_t scanner );
void yyset_in ( FILE * in_str , yyscan_t scanner );
void yyset_out ( FILE * out_str , yyscan_t scanner );
void yyset_lineno ( int line_number , yyscan_t scanner );
void yyset_extra ( YY_EXTRA_TYPE user_defined , yyscan_t scanner );
There are no "set" functions for yytext and yyleng. This is intentional.
The following Macro shortcuts are available in actions in a reentrant scanner:
@verbatim
yytext
yyleng
yyin
yyout
yylineno
yyextra
yy_flex_debug
In a reentrant C scanner, support for yylineno is always present
(i.e., you may access yylineno), but the value is never modified by
flex unless %option yylineno is enabled. This is to allow
the user to maintain the line count independently of flex.
The following functions and macros are made available when %option
bison-bridge (`--bison-bridge') is specified:
@verbatim
YYSTYPE * yyget_lval ( yyscan_t scanner );
YYLTYPE *yyget_lloc ( yyscan_t scanner );
void yyset_lval ( YYSTYPE * yylvalp , yyscan_t scanner );
void yyset_lloc ( YYLTYPE * yyllocp , yyscan_t scanner );
yylval
yylloc
Support for yylloc is dependent upon the presence of the preprocessor
symbol YYLTYPE. Support for yylval relies on the type
YYSTYPE to be defined. Typically, these definitions are generated
by bison, in a .h file, and are included in section 1 of the
flex input.
flex is a rewrite of the AT&T Unix lex tool (the two
implementations do not share any code, though), with some extensions and
incompatibilities, both of which are of concern to those who wish to
write scanners acceptable to both implementations. flex is fully
compliant with the POSIX lex specification, except that when
using %pointer (the default), a call to unput() destroys
the contents of yytext, which is counter to the POSIX
specification. In this section we discuss all of the known areas of
incompatibility between flex, AT&T lex, and the POSIX
specification. flex's `-l' option turns on maximum
compatibility with the original AT&T lex implementation, at the
cost of a major loss in the generated scanner's performance. We note
below which incompatibilities can be overcome using the `-l'
option. flex is fully compatible with lex with the
following exceptions:
lex scanner internal variable yylineno is
not supported unless `-l' or %option yylineno is used.
yylineno should be maintained on a per-buffer basis, rather than
a per-scanner (single global variable) basis.
yylineno is not part of the POSIX specification.
input() routine is not redefinable, though it may be called
to read characters following whatever has been matched by a rule. If
input() encounters an end-of-file the normal yywrap()
processing is done. A "real" end-of-file is returned by
input() as EOF.
YY_INPUT() macro.
flex restriction that input() cannot be redefined is
in accordance with the POSIX specification, which simply does not
specify any way of controlling the scanner's input other than by making
an initial assignment to `yyin'.
unput() routine is not redefinable. This restriction is in
accordance with POSIX.
flex scanners are not as reentrant as lex scanners. In
particular, if you have an interactive scanner and an interrupt handler
which long-jumps out of the scanner, and the scanner is subsequently
called again, you may get the following message:
@verbatim
fatal flex scanner internal error--end of buffer missed
To reenter the scanner, first use:
@verbatim
yyrestart( yyin );
Note that this call will throw away any buffered input; usually this
isn't a problem with an interactive scanner. See section Reentrant C Scanners, for
flex's reentrant API.
flex C++ scanner classes
are
reentrant, so if using C++ is an option for you, you should use
them instead. See section Generating C++ Scanners, and section Reentrant C Scanners for details.
output() is not supported. Output from the ECHO macro is
done to the file-pointer yyout (default `stdout)'.
output() is not part of the POSIX specification.
lex does not support exclusive start conditions (%x), though they
are in the POSIX specification.
flex encloses them in parentheses.
With lex, the following:
@verbatim
NAME [A-Z][A-Z0-9]*
%%
foo{NAME}? printf( "Found it\n" );
%%
will not match the string `foo' because when the macro is expanded
the rule is equivalent to `foo[A-Z][A-Z0-9]*?' and the precedence
is such that the `?' is associated with `[A-Z0-9]*'. With
flex, the rule will be expanded to `foo([A-Z][A-Z0-9]*)?'
and so the string `foo' will match.
<<EOF>> operators
cannot be used in a flex definition.
lex behavior of no parentheses
around the definition.
lex allow a rule's action to begin on a
separate line, if the rule's pattern has trailing whitespace:
@verbatim
%%
foo|bar<space here>
{ foobar_action();}
flex does not support this feature.
lex %r (generate a Ratfor scanner) option is not
supported. It is not part of the POSIX specification.
unput(), yytext is undefined until the
next token is matched, unless the scanner was built using %array.
This is not the case with lex or the POSIX specification. The
`-l' option does away with this incompatibility.
lex
interpret `abc{1,3}' as match one, two,
or three occurrences of `abc'", whereas flex interprets it
as "match `ab' followed by one, two, or three occurrences of
`c'". The `-l' and `--posix' options do away with this
incompatibility.
lex
interprets `^foo|bar' as "match either 'foo' at the beginning of a
line, or 'bar' anywhere", whereas flex interprets it as "match
either `foo' or `bar' if they come at the beginning of a
line". The latter is in agreement with the POSIX specification.
%a supported by
lex are not required by flex scanners.. flex
ignores them.
FLEX_SCANNER is #define'd so scanners may be
written for use with either flex or lex. Scanners also
include YY_FLEX_MAJOR_VERSION, YY_FLEX_MINOR_VERSION
and YY_FLEX_SUBMINOR_VERSION
indicating which version of flex generated the scanner. For
example, for the 2.5.22 release, these defines would be 2, 5 and 22
respectively. If the version of flex being used is a beta
version, then the symbol FLEX_BETA is defined.
The following flex features are not included in lex or the
POSIX specification:
flex command-line options
The feature "multiple actions on a line"
refers to the fact that with flex you can put multiple actions on
the same line, separated with semi-colons, while with lex, the
following:
@verbatim
foo handle_foo(); ++num_foos_seen;
is (rather surprisingly) truncated to
@verbatim
foo handle_foo();
flex does not truncate the action. Actions that are not enclosed
in braces are simply terminated at the end of the line.
@anchor{memory-management} This chapter describes how flex handles dynamic memory, and how you can override the default behavior.
Flex allocates dynamic memory during initialization, and once in a while from
within a call to yylex(). Initialization takes place during the first call to
yylex(). Thereafter, flex may reallocate more memory if it needs to enlarge a
buffer. As of version 2.5.9 Flex will clean up all memory when you call yylex_destroy
@xref{faq-memory-leak}.
Flex allocates dynamic memory for four purposes, listed below (1)
%option stack is not
specified. You will rarely need to tune this buffer. The ideal size for this
stack is the maximum depth expected. The memory for this stack is
automatically destroyed when you call yylex_destroy(). @xref{option-stack}.
Flex calls the functions yyalloc, yyrealloc, and yyfree
when it needs to allocate or free memory. By default, these functions are
wrappers around the standard C functions, malloc, realloc, and
free, respectively. You can override the default implementations by telling
flex that you will provide your own implementations.
To override the default implementations, you must do two things:
@verbatim // For a non-reentrant scanner void * yyalloc (size_t bytes); void * yyrealloc (void * ptr, size_t bytes); void yyfree (void * ptr); // For a reentrant scanner void * yyalloc (size_t bytes, void * yyscanner); void * yyrealloc (void * ptr, size_t bytes, void * yyscanner); void yyfree (void * ptr, void * yyscanner);
In the following example, we will override all three memory routines. We assume
that there is a custom allocator with garbage collection. In order to make this
example interesting, we will use a reentrant scanner, passing a pointer to the
custom allocator through yyextra.
@verbatim
%{
#include "some_allocator.h"
%}
/* Suppress the default implementations. */
%option noyyalloc noyyrealloc noyyfree
%option reentrant
/* Initialize the allocator. */
#define YY_EXTRA_TYPE struct allocator*
#define YY_USER_INIT yyextra = allocator_create();
%%
.|\n ;
%%
/* Provide our own implementations. */
void * yyalloc (size_t bytes, void* yyscanner) {
return allocator_alloc (yyextra, bytes);
}
void * yyrealloc (void * ptr, size_t bytes, void* yyscanner) {
return allocator_realloc (yyextra, bytes);
}
void yyfree (void * ptr, void * yyscanner) {
/* Do nothing -- we leave it to the garbage collector. */
}
When flex finds a match, yytext points to the first character of the
match in the input buffer. The string itself is part of the input buffer, and
is NOT allocated separately. The value of yytext will be overwritten the next
time yylex() is called. In short, the value of yytext is only valid from within
the matched rule's action.
Often, you want the value of yytext to persist for later processing, i.e., by a parser with non-zero lookahead. In order to preserve yytext, you will have to copy it with strdup() or a similar function. But this introduces some headache because your parser is now responsible for freeing the copy of yytext. If you use a yacc or bison parser, (commonly used with flex), you will discover that the error recovery mechanisms can cause memory to be leaked.
To prevent memory leaks from strdup'd yytext, you will have to track the memory somehow. Our experience has shown that a garbage collection mechanism or a pooled memory mechanism will save you a lot of grief when writing parsers.
@anchor{serialization}
A flex scanner has the ability to save the DFA tables to a file, and
load them at runtime when needed. The motivation for this feature is to reduce
the runtime memory footprint. Traditionally, these tables have been compiled into
the scanner as C arrays, and are sometimes quite large. Since the tables are
compiled into the scanner, the memory used by the tables can never be freed.
This is a waste of memory, especially if an application uses several scanners,
but none of them at the same time.
The serialization feature allows the tables to be loaded at runtime, before scanning begins. The tables may be discarded when scanning is finished.
You may create a scanner with serialized tables by specifying:
@verbatim
%option tables-file=FILE
or
--tables-file=FILE
These options instruct flex to save the DFA tables to the file FILE. The tables will not be embedded in the generated scanner. The scanner will not function on its own. The scanner will be dependent upon the serialized tables. You must load the tables from this file at runtime before you can scan anything.
If you do not specify a filename to --tables-file, the tables will be
saved to `lex.yy.tables', where `yy' is the appropriate prefix.
If your project uses several different scanners, you can concatenate the serialized tables into one file, and flex will find the correct set of tables, using the scanner prefix as part of the lookup key. An example follows:
@verbatim $ flex --tables-file --prefix=cpp cpp.l $ flex --tables-file --prefix=c c.l $ cat lex.cpp.tables lex.c.tables > all.tables
The above example created two scanners, `cpp', and `c'. Since we did not specify a filename, the tables were serialized to `lex.c.tables' and `lex.cpp.tables', respectively. Then, we concatenated the two files together into `all.tables', which we will distribute with our project. At runtime, we will open the file and tell flex to load the tables from it. Flex will find the correct tables automatically. (See next section).
If you've built your scanner with %option tables-file, then you must
load the scanner tables at runtime. This can be accomplished with the following
function:
yyalloc. You must call this
function before the first call to yylex. The argument scanner
only appears in the reentrant scanner.
This function returns `0' (zero) on success, or non-zero on error.
The loaded tables are not automatically destroyed (unloaded) when you
call yylex_destroy. The reason is that you may create several scanners
of the same type (in a reentrant scanner), each of which needs access to these
tables. To avoid a nasty memory leak, you must call the following function:
The functions yytables_fload and yytables_destroy are not
thread-safe. You must ensure that these functions are called exactly once (for
each scanner type) in a threaded program, before any thread calls yylex.
After the tables are loaded, they are never written to, and no thread
protection is required thereafter -- until you destroy them.
This section defines the file format of serialized flex tables.
The tables format allows for one or more sets of tables to be specified, where each set corresponds to a given scanner. Scanners are indexed by name, as described below. The file format is as follows:
@verbatim
TABLE SET 1
+-------------------------------+
Header | uint32 th_magic; |
| uint32 th_hsize; |
| uint32 th_ssize; |
| uint16 th_flags; |
| char th_version[]; |
| char th_name[]; |
| uint8 th_pad64[]; |
+-------------------------------+
Table 1 | uint16 td_id; |
| uint16 td_flags; |
| uint32 td_lolen; |
| uint32 td_hilen; |
| void td_data[]; |
| uint8 td_pad64[]; |
+-------------------------------+
Table 2 | |
. . .
. . .
. . .
. . .
Table n | |
+-------------------------------+
TABLE SET 2
.
.
.
TABLE SET N
The above diagram shows that a complete set of tables consists of a header followed by multiple individual tables. Furthermore, multiple complete sets may be present in the same file, each set with its own header and tables. The sets are contiguous in the file. The only way to know if another set follows is to check the next four bytes for the magic number (or check for EOF). The header and tables sections are padded to 64-bit boundaries. Below we describe each field in detail. This format does not specify how the scanner will expand the given data, i.e., data may be serialized as int8, but expanded to an int32 array at runtime. This is to reduce the size of the serialized data where possible. Remember, all integer values are in network byte order.
Fields of a table header:
th_magic
th_hsize
th_ssize
th_flags
th_version[]
th_name[]
th_pad64[]
Fields of a table:
td_id
YYTD_ID_ACCEPT (0x01)
yy_accept
YYTD_ID_BASE (0x02)
yy_base
YYTD_ID_CHK (0x03)
yy_chk
YYTD_ID_DEF (0x04)
yy_def
YYTD_ID_EC (0x05)
yy_ec
YYTD_ID_META (0x06)
yy_meta
YYTD_ID_NUL_TRANS (0x07)
yy_NUL_trans
YYTD_ID_NXT (0x08)
yy_nxt. This array may be two dimensional. See the td_hilen
field below.
YYTD_ID_RULE_CAN_MATCH_EOL (0x09)
yy_rule_can_match_eol
YYTD_ID_START_STATE_LIST (0x0A)
yy_start_state_list. This array is handled specially because it is an
array of pointers to structs. See the td_flags field below.
YYTD_ID_TRANSITION (0x0B)
yy_transition. This array is handled specially because it is an array of
structs. See the td_lolen field below.
YYTD_ID_ACCLIST (0x0C)
yy_acclist
td_flags
td_data.
The data arrays are one-dimensional by default, but may be
two dimensional as specified in the td_hilen field.
YYTD_DATA8 (0x01)
YYTD_DATA16 (0x02)
YYTD_DATA32 (0x04)
YYTD_PTRANS (0x08)
yy_transition
array. Each index should be expanded to a pointer to the corresponding entry
in the yy_transition array. We count on the fact that the
yy_transition array has already been seen.
YYTD_STRUCT (0x10)
YYTD_DATA* bits.
td_lolen
td_flags field.
td_hilen
td_hilen is non-zero, then the data is a two-dimensional array.
Otherwise, the data is a one-dimensional array. td_hilen contains the
number of elements in the higher dimensional array, and td_lolen contains
the number of elements in the lowest dimension.
Conceptually, td_data is either sometype td_data[td_lolen], or
sometype td_data[td_hilen][td_lolen], where sometype is specified
by the td_flags field. It is possible for both td_lolen and
td_hilen to be zero, in which case td_data is a zero length
array, and no data is loaded, i.e., this table is simply skipped. Flex does not
currently generate tables of zero length.
td_data[]
int8, int16, int32, struct yy_trans_info, or
struct yy_trans_info*, depending upon the values in the
td_flags, td_lolen, and td_hilen fields.
td_pad64[]
The following is a list of flex diagnostic messages:
@verbatim
[a-z]+ got_identifier();
foo got_foo();
Using REJECT in a scanner suppresses this warning.
reject_used_but_not_detected undefined or
yymore_used_but_not_detected undefined. These errors can occur
at compile time. They indicate that the scanner uses REJECT or
yymore() but that flex failed to notice the fact, meaning
that flex scanned the first two sections looking for occurrences
of these actions and failed to find any, but somehow you snuck some in
(via a #include file, for example). Use %option reject or
%option yymore to indicate to flex that you really do use
these features.
%array
and one of its rules matched a string longer than the YYLMAX
constant (8K bytes by default). You can increase the value by
#define'ing YYLMAX in the definitions section of your flex
input.
unput() to push
back so much text that the scanner's buffer could not hold both the
pushed-back text and the current token in yytext. Ideally the
scanner should dynamically resize the buffer in this case, but at
present it does not.
REJECT.
@verbatim
yyrestart( yyin );
or, as noted above, switch to using the C++ scanner class.
Some trailing context patterns cannot be properly matched and generate warning messages (`dangerous trailing context'). These are patterns where the ending of the first part of the rule matches the beginning of the second part, such as `zx*/xy*', where the 'x*' matches the 'x' at the beginning of the trailing context. (Note that the POSIX draft states that the text matched by such patterns is undefined.) For some trailing context rules, parts which are actually fixed-length are not recognized as such, leading to the abovementioned performance loss. In particular, parts using `|' or `{n}' (such as `foo{3}') are always considered variable-length. Combining trailing context with the special `|' action can result in fixed trailing context being turned into the more expensive variable trailing context. For example, in the following:
@verbatim
%%
abc |
xyz/def
Use of unput() invalidates yytext and yyleng, unless the
%array directive or the `-l' option has been used.
Pattern-matching of NULs is substantially slower than matching
other characters. Dynamic resizing of the input buffer is slow, as it
entails rescanning all the text matched so far by the current (generally
huge) token. Due to both buffering of input and read-ahead, you cannot
intermix calls to `<stdio.h>' routines, such as, getchar(),
with flex rules and expect it to work. Call input()
instead. The total table entries listed by the `-v' flag excludes
the number of table entries needed to determine what rule has been
matched. The number of entries is equal to the number of DFA states if
the scanner does not use REJECT, and somewhat greater than the
number of states if it does. REJECT cannot be used with the
`-f' or `-F' options.
The flex internal algorithms need documentation.
You may wish to read more about the following programs:
The following books may contain material of interest:
John Levine, Tony Mason, and Doug Brown, Lex & Yacc, O'Reilly and Associates. Be sure to get the 2nd edition.
M. E. Lesk and E. Schmidt, LEX -- Lexical Analyzer Generator
Alfred Aho, Ravi Sethi and Jeffrey Ullman, Compilers: Principles,
Techniques and Tools, Addison-Wesley (1986). Describes the
pattern-matching techniques used by flex (deterministic finite
automata).
From time to time, the flex maintainer receives certain
questions. Rather than repeat answers to well-understood problems, we
publish them here.
Vern Paxson took over the Software Tools lex project from Jef Poskanzer in 1982. At that point it was written in Ratfor. Around 1987 or so, Paxson translated it into C, and a legend was born :-).
A key point when scanning quoted strings is that you cannot (easily) write a single rule that will precisely match the string if you allow things like embedded escape sequences and newlines. If you try to match strings with a single rule then you'll wind up having to rescan the string anyway to find any escape sequences.
Instead you can use exclusive start conditions and a set of rules, one for
matching non-escaped text, one for matching a single escape, one for
matching an embedded newline, and one for recognizing the end of the
string. Each of these rules is then faced with the question of where to
put its intermediary results. The best solution is for the rules to
append their local value of yytext to the end of a "string literal"
buffer. A rule like the escape-matcher will append to the buffer the
meaning of the escape sequence rather than the literal text in yytext.
In this way, yytext does not need to be modified at all.
Flex scanners call fileno() in order to get the file descriptor
corresponding to yyin. The file descriptor may be passed to
isatty() or read(), depending upon which %options you specified.
If your system does not have fileno() support, to get rid of the
read() call, do not specify %option read. To get rid of the isatty()
call, you must specify one of %option always-interactive or
%option never-interactive.
e.g.,
@verbatim
%%
block "{"({block}|{statement})*"}"
No. You cannot have recursive definitions. The pattern-matching power of regular expressions in general (and therefore flex scanners, too) is limited. In particular, regular expressions cannot "balance" parentheses to an arbitrary degree. For example, it's impossible to write a regular expression that matches all strings containing the same number of '{'s as '}'s. For more powerful pattern matching, you need a parser, such as GNU bison.
Use fseek() (or lseek()) to position yyin, then call yyrestart().
flex picks the
rule that matches the most text (i.e., the longest possible input string).
This is because flex uses an entirely different matching technique
("deterministic finite automata") that actually does all of the matching
simultaneously, in parallel. (Seems impossible, but it's actually a fairly
simple technique once you understand the principles.)
A side-effect of this parallel matching is that when the input matches more
than one rule, flex scanners pick the rule that matched the most text. This
is explained further in the manual, in the section See section How the Input Is Matched.
If you want flex to choose a shorter match, then you can work around this
behavior by expanding your short
rule to match more text, then put back the extra:
@verbatim data_.* yyless( 5 ); BEGIN BLOCKIDSTATE;
Another fix would be to make the second rule active only during the
<BLOCKIDSTATE> start condition, and make that start condition exclusive
by declaring it with %x instead of %s.
A final fix is to change the input language so that the ambiguity for `data_' is removed, by adding characters to it that don't match the identifier rule, or by removing characters (such as `_') from the identifier rule so it no longer matches `data_'. (Of course, you might also not have the option of changing the input language.)
Most likely, you have (in error) placed the opening `{' of the action block on a different line than the rule, e.g.,
@verbatim
^(foo|bar)
{ <<<--- WRONG!
}
flex requires that the opening `{' of an action associated with a rule
begin on the same line as does the rule. You need instead to write your rules
as follows:
@verbatim
^(foo|bar) { // CORRECT!
}
If ...
flex's `-b' flag),
YY_INPUT to do so,
then every time it matches a token, it will have exhausted its input
buffer (because the scanner is free of backtracking). This means you
can safely use select() at the point and only call yylex() for another
token if select() indicates there's data available.
That is, move the select() out from the input function to a point where
it determines whether yylex() gets called for the next token.
With this approach, you will still have problems if your input can arrive
piecemeal; select() could inform you that the beginning of a token is
available, you call yylex() to get it, but it winds up blocking waiting
for the later characters in the token.
Here's another way: Move your input multiplexing inside of YY_INPUT. That
is, whenever YY_INPUT is called, it select()'s to see where input is
available. If input is available for the scanner, it reads and returns the
next byte. If input is available from another source, it calls whatever
function is responsible for reading from that source. (If no input is
available, it blocks until some input is available.) I've used this technique in an
interpreter I wrote that both reads keyboard input using a flex scanner and
IPC traffic from sockets, and it works fine.
This is not going to work without some additional effort. The reason is
that flex block-buffers the input it reads from yyin. This means that the
"outermost" yylex(), when called, will automatically slurp up the first 8K
of input available on yyin, and subsequent calls to other yylex()'s won't
see that input. You might be tempted to work around this problem by
redefining YY_INPUT to only return a small amount of text, but it turns out
that that approach is quite difficult. Instead, the best solution is to
combine all of your scanners into one large scanner, using a different
exclusive start condition for each.
There is no way to write a rule which is "match this text, but only if
it comes at the end of the file". You can fake it, though, if you happen
to have a character lying around that you don't allow in your input.
Then you redefine YY_INPUT to call your own routine which, if it sees
an `EOF', returns the magic character first (and remembers to return a
real EOF next time it's called). Then you could write:
@verbatim
<COMMENT>(.|\n)*{EOF_CHAR} /* saw comment at EOF */
You can do this as follows. Suppose you have a start condition `A', and after exhausting all of the possible matches in `<A>', you want to try matches in `<INITIAL>'. Then you could use the following:
@verbatim
%x A
%%
<A>rule_that_is_long ...; REJECT;
<A>rule ...; REJECT; /* shorter rule */
<A>etc.
...
<A>.|\n {
/* Shortest and last rule in <A>, so
* cascaded REJECT's will eventually
* wind up matching this rule. We want
* to now switch to the initial state
* and try matching from there instead.
*/
yyless(0); /* put back matched text */
BEGIN(INITIAL);
}
One of the assumptions flex makes is that interactive applications are inherently slow (they're waiting on a human after all). It has to do with how the scanner detects that it must be finished scanning a token. For interactive scanners, after scanning each character the current state is looked up in a table (essentially) to see whether there's a chance of another input character possibly extending the length of the match. If not, the scanner halts. For non-interactive scanners, the end-of-token test is much simpler, basically a compare with 0, so no memory bus cycles. Since the test occurs in the innermost scanning loop, one would like to make it go as fast as possible.
Still, it seems reasonable to allow the user to choose to trade off a bit of performance in this area to gain the corresponding flexibility. There might be another reason, though, why fast scanners don't support the interactive option.
Much faster (factor of 2-3).
Is your grammar recursive? That's almost always a sign that you're better off using a parser/scanner rather than just trying to use a scanner alone.
There are two reasons. The first is that there might
be programs that rely on the start state not changing across file changes.
The second is that beginning with flex version 2.4, use of yyrestart() is no longer required,
so fixing the problem there doesn't solve the more general problem.
You might be tempted to try something like this:
@verbatim "/*".*"*/" // WRONG!
or, worse, this:
@verbatim "/*"(.|\n)"*/" // WRONG!
The above rules will eat too much input, and blow up on things like:
@verbatim /* a comment */ do_my_thing( "oops */" );
Here is one way which allows you to track line information:
@verbatim
<INITIAL>{
"/*" BEGIN(IN_COMMENT);
}
<IN_COMMENT>{
"*/" BEGIN(INITIAL);
[^*\n]+ // eat comment in chunks
"*" // eat the lone star
\n yylineno++;
}
Here are some tips for using `.':
(foo|bar)+ and NOT this (foo|bar+).
The first pattern matches the words `foo' or `bar' any number of
times, e.g., it matches the text `barfoofoobarfoo'. The
second pattern matches a single instance of foo or a single instance of
bar followed by one or more `r's, e.g., it matches the text barrrr .
(.|\n)
Beware that the regex (.|\n)+ will match your entire input!
The flex source distribution includes a texinfo manual. You are
free to convert that texinfo into whatever format you desire. The
texinfo package includes tools for conversion to a number of formats.
There's no way around the potential exponential running time - it can take you exponential time just to enumerate all of the DFA states. In practice, though, the running time is closer to linear, or sometimes quadratic.
There are two big speed wins that flex uses:
Flex is compiled with an upper limit of 8192 rules per scanner.
If you need more than 8192 rules in your scanner, you'll have to recompile flex
with the following changes in `flexdef.h':
@verbatim < #define YY_TRAILING_MASK 0x2000 < #define YY_TRAILING_HEAD_MASK 0x4000 -- > #define YY_TRAILING_MASK 0x20000000 > #define YY_TRAILING_HEAD_MASK 0x40000000
This should work okay as long as your C compiler uses 32 bit integers. But you might want to think about whether using such a huge number of rules is the best way to solve your problem.
The following may also be relevant:
With luck, you should be able to increase the definitions in flexdef.h for:
@verbatim #define JAMSTATE -32766 /* marks a reference to the state that always jams */ #define MAXIMUM_MNS 31999 #define BAD_SUBSCRIPT -32767
recompile everything, and it'll all work. Flex only has these 16-bit-like values built into it because a long time ago it was developed on a machine with 16-bit ints. I've given this advice to others in the past but haven't heard back from them whether it worked okay or not...
Just call yyrestart(newfile). Be sure to reset the start state if you want a
"fresh start, since yyrestart does NOT reset the start state back to INITIAL.
You can specify an initial action by defining the macro YY_USER_INIT (though
note that yyout may not be available at the time this macro is executed). Or you
can add to the beginning of your rules section:
@verbatim
%%
/* Must be indented! */
static int did_init = 0;
if ( ! did_init ){
do_my_init();
did_init = 1;
}
You can specify an action for the <<EOF>> rule.
The help-flex email list is served by GNU. See http://www.gnu.org/ for
details on how to subscribe or search the archives.
Yes, just about anywhere you want to. See the manual for the specific syntax.
You must supply a yywrap() function of your own, or link to `libfl.a'
(which provides one), or use
@verbatim %option noyywrap
in your source to say you don't want a yywrap() function.
You can't, it's compiled into a static table when flex builds the scanner.
The best way to approach this problem is at a higher level, e.g., in the parser.
However, you can do this using multiple input buffers.
@verbatim
%%
macro/[a-z]+ {
/* Saw the macro "macro" followed by extra stuff. */
main_buffer = YY_CURRENT_BUFFER;
expansion_buffer = yy_scan_string(expand(yytext));
yy_switch_to_buffer(expansion_buffer);
}
<<EOF>> {
if ( expansion_buffer )
{
// We were doing an expansion, return to where
// we were.
yy_switch_to_buffer(main_buffer);
yy_delete_buffer(expansion_buffer);
expansion_buffer = 0;
}
else
yyterminate();
}
You probably will want a stack of expansion buffers to allow nested macros. From the above though hopefully the idea is clear.
One way to do it is to filter the first pass to a temporary file, then process the temporary file on the second pass. You will probably see a performance hit, do to all the disk I/O.
When you need to look ahead far forward like this, it almost always means that the right solution is to build a parse tree of the entire input, then walk it after the parse in order to generate the output. In a sense, this is a two-pass approach, once through the text and once through the parse tree, but the performance hit for the latter is usually an order of magnitude smaller, since everything is already classified, in binary format, and residing in memory.
One way to assign precedence, is to place the more specific rules first. If
two rules would match the same input (same sequence of characters) then the
first rule listed in the flex input wins. e.g.,
@verbatim %% foo[a-zA-Z_]+ return FOO_ID; bar[a-zA-Z_]+ return BAR_ID; [a-zA-Z_]+ return GENERIC_ID;
Note that the rule [a-zA-Z_]+ must come *after* the others. It will match the
same amount of text as the more specific rules, and in that case the
flex scanner will pick the first rule listed in your scanner as the
one to match.
Those are internal variables pointing into the AT&T scanner's input buffer. I
imagine they're being manipulated in user versions of the input() and unput()
functions. If so, what you need to do is analyze those functions to figure out
what they're doing, and then replace input() with an appropriate definition of
YY_INPUT. You shouldn't need to (and must not) replace
flex's unput() function.
Yes, `\0' and `\x00' should both do the trick. Perhaps you have an ancient
version of flex. The latest release is version 2.5.23.
You need to add a rule that matches the otherwise-unmatched text. e.g.,
@verbatim
%option yylineno
%%
[[a bunch of rules here]]
. printf("bad input character '%s' at line %d\n", yytext, yylineno);
See %option default for more information.
A DFA can do a non-greedy match by stopping the first time it enters an accepting state, instead of consuming input until it determines that no further matching is possible (a "jam" state). This is actually easier to implement than longest leftmost match (which flex does).
But it's also much less useful than longest leftmost match. In general, when you find yourself wishing for non-greedy matching, that's usually a sign that you're trying to make the scanner do some parsing. That's generally the wrong approach, since it lacks the power to do a decent job. Better is to either introduce a separate parser, or to split the scanner into multiple scanners using (exclusive) start conditions.
You might have a separate start state once you've seen the `BEGIN'. In that state, you might then have a regex that will match `END' (to kick you out of the state), and perhaps `(.|\n)' to get a single character within the chunk ...
This approach also has much better error-reporting properties.
@anchor{faq-memory-leak}
UPDATED 2002-07-10: As of flex version 2.5.9, this leak means that you did not
call yylex_destroy(). If you are using an earlier version of flex, then read
on.
The leak is about 16426 bytes. That is, (8192 * 2 + 2) for the read-buffer, and
about 40 for struct yy_buffer_state (depending upon alignment). The leak is in
the non-reentrant C scanner only (NOT in the reentrant scanner, NOT in the C++
scanner). Since flex doesn't know when you are done, the buffer is never freed.
However, the leak won't multiply since the buffer is reused no matter how many
times you call yylex().
If you want to reclaim the memory when you are completely done scanning, then you might try this:
@verbatim /* For non-reentrant C scanner only. */ yy_delete_buffer(yy_current_buffer); yy_init = 1;
Note: yy_init is an "internal variable", and hasn't been tested in this
situation. It is possible that some other globals may need resetting as well.
@verbatim > We thought that it would be possible to have this number through the > evaluation of the following expression: > > seek_position = (no_buffers)*YY_READ_BUF_SIZE + yy_c_buf_p - yy_current_buffer->yy_ch_buf
While this is the right idea, it has two problems. The first is that
it's possible that flex will request less than YY_READ_BUF_SIZE during
an invocation of YY_INPUT (or that your input source will return less
even though YY_READ_BUF_SIZE bytes were requested). The second problem
is that when refilling its internal buffer, flex keeps some characters
from the previous buffer (because usually it's in the middle of a match,
and needs those characters to construct yytext for the match once it's
done). Because of this, yy_c_buf_p - yy_current_buffer->yy_ch_buf won't
be exactly the number of characters already read from the current buffer.
An alternative solution is to count the number of characters you've matched
since starting to scan. This can be done by using YY_USER_ACTION. For
example,
@verbatim #define YY_USER_ACTION num_chars += yyleng;
(You need to be careful to update your bookkeeping if you use yymore(),
yyless(), unput(), or input().)
When the flex C++ scanning class rewrite finally happens, then this sort of thing should become much easier.
You can do this by passing the various functions (such as LexerInput()
and LexerOutput()) nil iostream*'s, and then
dealing with your own I/O classes surreptitiously (i.e., stashing them in
special member variables). This works because the only assumption about
the lexer regarding what's done with the iostream's is that they're
ultimately passed to LexerInput() and LexerOutput, which then do whatever
is necessary with them.
How do I skip as many chars as possible -- without interfering with the other patterns?
In the example below, we want to skip over characters until we see the phrase "endskip". The following will NOT work correctly (do you see why not?)
@verbatim /* INCORRECT SCANNER */ %x SKIP %% <INITIAL>startskip BEGIN(SKIP); ... <SKIP>"endskip" BEGIN(INITIAL); <SKIP>.* ;
The problem is that the pattern .* will eat up the word "endskip." The simplest (but slow) fix is:
@verbatim <SKIP>"endskip" BEGIN(INITIAL); <SKIP>. ;
The fix involves making the second rule match more, without making it match "endskip" plus something else. So for example:
@verbatim <SKIP>"endskip" BEGIN(INITIAL); <SKIP>[^e]+ ; <SKIP>. ;/* so you eat up e's, too */
@verbatim QUESTION: When was flex born? Vern Paxson took over the Software Tools lex project from Jef Poskanzer in 1982. At that point it was written in Ratfor. Around 1987 or so, Paxson translated it into C, and a legend was born :-).
@verbatim
To: Adoram Rogel <adoram@orna.hybridge.com>
Subject: Re: Flex 2.5.2 performance questions
In-reply-to: Your message of Wed, 18 Sep 96 11:12:17 EDT.
Date: Wed, 18 Sep 96 10:51:02 PDT
From: Vern Paxson <vern>
[Note, the most recent flex release is 2.5.4, which you can get from
ftp.ee.lbl.gov. It has bug fixes over 2.5.2 and 2.5.3.]
> 1. Using the pattern
> ([Ff](oot)?)?[Nn](ote)?(\.)?
> instead of
> (((F|f)oot(N|n)ote)|((N|n)ote)|((N|n)\.)|((F|f)(N|n)(\.)))
> (in a very complicated flex program) caused the program to slow from
> 300K+/min to 100K/min (no other changes were done).
These two are not equivalent. For example, the first can match "footnote."
but the second can only match "footnote". This is almost certainly the
cause in the discrepancy - the slower scanner run is matching more tokens,
and/or having to do more backing up.
> 2. Which of these two are better: [Ff]oot or (F|f)oot ?
From a performance point of view, they're equivalent (modulo presumably
minor effects such as memory cache hit rates; and the presence of trailing
context, see below). From a space point of view, the first is slightly
preferable.
> 3. I have a pattern that look like this:
> pats {p1}|{p2}|{p3}|...|{p50} (50 patterns ORd)
>
> running yet another complicated program that includes the following rule:
> <snext>{and}/{no4}{bb}{pats}
>
> gets me to "too complicated - over 32,000 states"...
I can't tell from this example whether the trailing context is variable-length
or fixed-length (it could be the latter if {and} is fixed-length). If it's
variable length, which flex -p will tell you, then this reflects a basic
performance problem, and if you can eliminate it by restructuring your
scanner, you will see significant improvement.
> so I divided {pats} to {pats1}, {pats2},..., {pats5} each consists of about
> 10 patterns and changed the rule to be 5 rules.
> This did compile, but what is the rule of thumb here ?
The rule is to avoid trailing context other than fixed-length, in which for
a/b, either the 'a' pattern or the 'b' pattern have a fixed length. Use
of the '|' operator automatically makes the pattern variable length, so in
this case '[Ff]oot' is preferred to '(F|f)oot'.
> 4. I changed a rule that looked like this:
> <snext8>{and}{bb}/{ROMAN}[^A-Za-z] { BEGIN...
>
> to the next 2 rules:
> <snext8>{and}{bb}/{ROMAN}[A-Za-z] { ECHO;}
> <snext8>{and}{bb}/{ROMAN} { BEGIN...
>
> Again, I understand the using [^...] will cause a great performance loss
Actually, it doesn't cause any sort of performance loss. It's a surprising
fact about regular expressions that they always match in linear time
regardless of how complex they are.
> but are there any specific rules about it ?
See the "Performance Considerations" section of the man page, and also
the example in MISC/fastwc/.
Vern
@verbatim
To: Adoram Rogel <adoram@hybridge.com>
Subject: Re: Flex 2.5.2 performance questions
In-reply-to: Your message of Thu, 19 Sep 96 10:16:04 EDT.
Date: Thu, 19 Sep 96 09:58:00 PDT
From: Vern Paxson <vern>
> a lot about the backing up problem.
> I believe that there lies my biggest problem, and I'll try to improve
> it.
Since you have variable trailing context, this is a bigger performance
problem. Fixing it is usually easier than fixing backing up, which in a
complicated scanner (yours seems to fit the bill) can be extremely
difficult to do correctly.
You also don't mention what flags you are using for your scanner.
-f makes a large speed difference, and -Cfe buys you nearly as much
speed but the resulting scanner is considerably smaller.
> I have an | operator in {and} and in {pats} so both of them are variable
> length.
-p should have reported this.
> Is changing one of them to fixed-length is enough ?
Yes.
> Is it possible to change the 32,000 states limit ?
Yes. I've appended instructions on how. Before you make this change,
though, you should think about whether there are ways to fundamentally
simplify your scanner - those are certainly preferable!
Vern
To increase the 32K limit (on a machine with 32 bit integers), you increase
the magnitude of the following in flexdef.h:
#define JAMSTATE -32766 /* marks a reference to the state that always jams */
#define MAXIMUM_MNS 31999
#define BAD_SUBSCRIPT -32767
#define MAX_SHORT 32700
Adding a 0 or two after each should do the trick.
@verbatim
To: Heeman_Lee@hp.com
Subject: Re: flex - multi-byte support?
In-reply-to: Your message of Thu, 03 Oct 1996 17:24:04 PDT.
Date: Fri, 04 Oct 1996 11:42:18 PDT
From: Vern Paxson <vern>
> I assume as long as my *.l file defines the
> range of expected character code values (in octal format), flex will
> scan the file and read multi-byte characters correctly. But I have no
> confidence in this assumption.
Your lack of confidence is justified - this won't work.
Flex has in it a widespread assumption that the input is processed
one byte at a time. Fixing this is on the to-do list, but is involved,
so it won't happen any time soon. In the interim, the best I can suggest
(unless you want to try fixing it yourself) is to write your rules in
terms of pairs of bytes, using definitions in the first section:
X \xfe\xc2
...
%%
foo{X}bar found_foo_fe_c2_bar();
etc. Definitely a pain - sorry about that.
By the way, the email address you used for me is ancient, indicating you
have a very old version of flex. You can get the most recent, 2.5.4, from
ftp.ee.lbl.gov.
Vern
@verbatim To: moleary@primus.com Subject: Re: Flex / Unicode compatibility question In-reply-to: Your message of Tue, 22 Oct 1996 10:15:42 PDT. Date: Tue, 22 Oct 1996 11:06:13 PDT From: Vern Paxson <vern> Unfortunately flex at the moment has a widespread assumption within it that characters are processed 8 bits at a time. I don't see any easy fix for this (other than writing your rules in terms of double characters - a pain). I also don't know of a wider lex, though you might try surfing the Plan 9 stuff because I know it's a Unicode system, and also the PCCT toolkit (try searching say Alta Vista for "Purdue Compiler Construction Toolkit"). Fixing flex to handle wider characters is on the long-term to-do list. But since flex is a strictly spare-time project these days, this probably won't happen for quite a while, unless someone else does it first. Vern
@verbatim To: Johan Linde <jl@theophys.kth.se> Subject: Re: translation of flex In-reply-to: Your message of Sun, 10 Nov 1996 09:16:36 PST. Date: Mon, 11 Nov 1996 10:33:50 PST From: Vern Paxson <vern> > I'm working for the Swedish team translating GNU program, and I'm currently > working with flex. I have a few questions about some of the messages which > I hope you can answer. All of the things you're wondering about, by the way, concerning flex internals - probably the only person who understands what they mean in English is me! So I wouldn't worry too much about getting them right. That said ... > #: main.c:545 > msgid " %d protos created\n" > > Does proto mean prototype? Yes - prototypes of state compression tables. > #: main.c:539 > msgid " %d/%d (peak %d) template nxt-chk entries created\n" > > Here I'm mainly puzzled by 'nxt-chk'. I guess it means 'next-check'. (?) > However, 'template next-check entries' doesn't make much sense to me. To be > able to find a good translation I need to know a little bit more about it. There is a scheme in the Aho/Sethi/Ullman compiler book for compressing scanner tables. It involves creating two pairs of tables. The first has "base" and "default" entries, the second has "next" and "check" entries. The "base" entry is indexed by the current state and yields an index into the next/check table. The "default" entry gives what to do if the state transition isn't found in next/check. The "next" entry gives the next state to enter, but only if the "check" entry verifies that this entry is correct for the current state. Flex creates templates of series of next/check entries and then encodes differences from these templates as a way to compress the tables. > #: main.c:533 > msgid " %d/%d base-def entries created\n" > > The same problem here for 'base-def'. See above. Vern
@verbatim To: Xinying Li <xli@npac.syr.edu> Subject: Re: FLEX ? In-reply-to: Your message of Wed, 13 Nov 1996 17:28:38 PST. Date: Wed, 13 Nov 1996 19:51:54 PST From: Vern Paxson <vern> > "unput()" them to input flow, question occurs. If I do this after I scan > a carriage, the variable "yy_current_buffer->yy_at_bol" is changed. That > means the carriage flag has gone. You can control this by calling yy_set_bol(). It's described in the manual. > And if in pre-reading it goes to the end of file, is anything done > to control the end of curren buffer and end of file? No, there's no way to put back an end-of-file. > By the way I am using flex 2.5.2 and using the "-l". The latest release is 2.5.4, by the way. It fixes some bugs in 2.5.2 and 2.5.3. You can get it from ftp.ee.lbl.gov. Vern
@verbatim To: Alain.ISSARD@st.com Subject: Re: Start condition with FLEX In-reply-to: Your message of Mon, 18 Nov 1996 09:45:02 PST. Date: Mon, 18 Nov 1996 10:41:34 PST From: Vern Paxson <vern> > I am not able to use the start condition scope and to use the | (OR) with > rules having start conditions. The problem is that if you use '|' as a regular expression operator, for example "a|b" meaning "match either 'a' or 'b'", then it must *not* have any blanks around it. If you instead want the special '|' *action* (which from your scanner appears to be the case), which is a way of giving two different rules the same action: foo | bar matched_foo_or_bar(); then '|' *must* be separated from the first rule by whitespace and *must* be followed by a new line. You *cannot* write it as: foo | bar matched_foo_or_bar(); even though you might think you could because yacc supports this syntax. The reason for this unfortunately incompatibility is historical, but it's unlikely to be changed. Your problems with start condition scope are simply due to syntax errors from your use of '|' later confusing flex. Let me know if you still have problems. Vern
@verbatim To: Gregory Margo <gmargo@newton.vip.best.com> Subject: Re: flex-2.5.3 bug report In-reply-to: Your message of Sat, 23 Nov 1996 16:50:09 PST. Date: Sat, 23 Nov 1996 17:07:32 PST From: Vern Paxson <vern> > Enclosed is a lex file that "real" lex will process, but I cannot get > flex to process it. Could you try it and maybe point me in the right direction? Your problem is that some of the definitions in the scanner use the '/' trailing context operator, and have it enclosed in ()'s. Flex does not allow this operator to be enclosed in ()'s because doing so allows undefined regular expressions such as "(a/b)+". So the solution is to remove the parentheses. Note that you must also be building the scanner with the -l option for AT&T lex compatibility. Without this option, flex automatically encloses the definitions in parentheses. Vern
@verbatim
To: Thomas Hadig <hadig@toots.physik.rwth-aachen.de>
Subject: Re: Flex Bug ?
In-reply-to: Your message of Tue, 26 Nov 1996 14:35:01 PST.
Date: Tue, 26 Nov 1996 11:15:05 PST
From: Vern Paxson <vern>
> In my lexer code, i have the line :
> ^\*.* { }
>
> Thus all lines starting with an astrix (*) are comment lines.
> This does not work !
I can't get this problem to reproduce - it works fine for me. Note
though that if what you have is slightly different:
COMMENT ^\*.*
%%
{COMMENT} { }
then it won't work, because flex pushes back macro definitions enclosed
in ()'s, so the rule becomes
(^\*.*) { }
and now that the '^' operator is not at the immediate beginning of the
line, it's interpreted as just a regular character. You can avoid this
behavior by using the "-l" lex-compatibility flag, or "%option lex-compat".
Vern
@verbatim To: Adoram Rogel <adoram@hybridge.com> Subject: Re: Flex 2.5.4 BOF ??? In-reply-to: Your message of Tue, 26 Nov 1996 16:10:41 PST. Date: Wed, 27 Nov 1996 10:56:25 PST From: Vern Paxson <vern> > Organization(s)?/[a-z] > > This matched "Organizations" (looking in debug mode, the trailing s > was matched with trailing context instead of the optional (s) in the > end of the word. That should only happen with lex. Flex can properly match this pattern. (That might be what you're saying, I'm just not sure.) > Is there a way to avoid this dangerous trailing context problem ? Unfortunately, there's no easy way. On the other hand, I don't see why it should be a problem. Lex's matching is clearly wrong, and I'd hope that usually the intent remains the same as expressed with the pattern, so flex's matching will be correct. Vern
@verbatim To: Cameron MacKinnon <mackin@interlog.com> Subject: Re: Flex documentation bug In-reply-to: Your message of Mon, 02 Dec 1996 00:07:08 PST. Date: Sun, 01 Dec 1996 22:29:39 PST From: Vern Paxson <vern> > I'm not sure how or where to submit bug reports (documentation or > otherwise) for the GNU project stuff ... Well, strictly speaking flex isn't part of the GNU project. They just distribute it because no one's written a decent GPL'd lex replacement. So you should send bugs directly to me. Those sent to the GNU folks sometimes find there way to me, but some may drop between the cracks. > In GNU Info, under the section 'Start Conditions', and also in the man > page (mine's dated April '95) is a nice little snippet showing how to > parse C quoted strings into a buffer, defined to be MAX_STR_CONST in > size. Unfortunately, no overflow checking is ever done ... This is already mentioned in the manual: Finally, here's an example of how to match C-style quoted strings using exclusive start conditions, including expanded escape sequences (but not including checking for a string that's too long): The reason for not doing the overflow checking is that it will needlessly clutter up an example whose main purpose is just to demonstrate how to use flex. The latest release is 2.5.4, by the way, available from ftp.ee.lbl.gov. Vern
@verbatim To: tsv@cs.UManitoba.CA Subject: Re: Flex (reg).. In-reply-to: Your message of Thu, 06 Mar 1997 23:50:16 PST. Date: Thu, 06 Mar 1997 15:54:19 PST From: Vern Paxson <vern> > [:alpha:] ([:alnum:] | \\_)* If your rule really has embedded blanks as shown above, then it won't work, as the first blank delimits the rule from the action. (It wouldn't even compile ...) You need instead: [:alpha:]([:alnum:]|\\_)* and that should work fine - there's no restriction on what can go inside of ()'s except for the trailing context operator, '/'. Vern
@verbatim To: "Mike Stolnicki" <mstolnic@ford.com> Subject: Re: FLEX help In-reply-to: Your message of Fri, 30 May 1997 13:33:27 PDT. Date: Fri, 30 May 1997 10:46:35 PDT From: Vern Paxson <vern> > We'd like to add "if-then-else", "while", and "for" statements to our > language ... > We've investigated many possible solutions. The one solution that seems > the most reasonable involves knowing the position of a TOKEN in yyin. I strongly advise you to instead build a parse tree (abstract syntax tree) and loop over that instead. You'll find this has major benefits in keeping your interpreter simple and extensible. That said, the functionality you mention for get_position and set_position have been on the to-do list for a while. As flex is a purely spare-time project for me, no guarantees when this will be added (in particular, it for sure won't be for many months to come). Vern
@verbatim To: Colin Paul Adams <colin@colina.demon.co.uk> Subject: Re: Flex C++ classes and Bison In-reply-to: Your message of 09 Aug 1997 17:11:41 PDT. Date: Fri, 15 Aug 1997 10:48:19 PDT From: Vern Paxson <vern> > #define YY_DECL int yylex (YYSTYPE *lvalp, struct parser_control > *parm) > > I have been trying to get this to work as a C++ scanner, but it does > not appear to be possible (warning that it matches no declarations in > yyFlexLexer, or something like that). > > Is this supposed to be possible, or is it being worked on (I DID > notice the comment that scanner classes are still experimental, so I'm > not too hopeful)? What you need to do is derive a subclass from yyFlexLexer that provides the above yylex() method, squirrels away lvalp and parm into member variables, and then invokes yyFlexLexer::yylex() to do the regular scanning. Vern
@verbatim
To: Mikael.Latvala@lmf.ericsson.se
Subject: Re: Possible mistake in Flex v2.5 document
In-reply-to: Your message of Fri, 05 Sep 1997 16:07:24 PDT.
Date: Fri, 05 Sep 1997 10:01:54 PDT
From: Vern Paxson <vern>
> In that example you show how to count comment lines when using
> C style /* ... */ comments. My question is, shouldn't you take into
> account a scenario where end of a comment marker occurs inside
> character or string literals?
The scanner certainly needs to also scan character and string literals.
However it does that (there's an example in the man page for strings), the
lexer will recognize the beginning of the literal before it runs across the
embedded "/*". Consequently, it will finish scanning the literal before it
even considers the possibility of matching "/*".
Example:
'([^']*|{ESCAPE_SEQUENCE})'
will match all the text between the ''s (inclusive). So the lexer
considers this as a token beginning at the first ', and doesn't even
attempt to match other tokens inside it.
I thinnk this subtlety is not worth putting in the manual, as I suspect
it would confuse more people than it would enlighten.
Vern
@verbatim
To: "Marty Leisner" <leisner@sdsp.mc.xerox.com>
Subject: Re: flex limitations
In-reply-to: Your message of Sat, 06 Sep 1997 11:27:21 PDT.
Date: Mon, 08 Sep 1997 11:38:08 PDT
From: Vern Paxson <vern>
> %%
> [a-zA-Z]+ /* skip a line */
> { printf("got %s\n", yytext); }
> %%
What version of flex are you using? If I feed this to 2.5.4, it complains:
"bug.l", line 5: EOF encountered inside an action
"bug.l", line 5: unrecognized rule
"bug.l", line 5: fatal parse error
Not the world's greatest error message, but it manages to flag the problem.
(With the introduction of start condition scopes, flex can't accommodate
an action on a separate line, since it's ambiguous with an indented rule.)
You can get 2.5.4 from ftp.ee.lbl.gov.
Vern
@verbatim To: uocarroll@deagostini.co.uk (Ultan O'Carroll) Subject: Re: Flex repositries In-reply-to: Your message of Fri, 12 Sep 1997 15:02:28 PDT. Date: Fri, 12 Sep 1997 10:31:50 PDT From: Vern Paxson <vern> > before I start beavering away I wonder if you know of any > place/libraries for flex > desciption files that might already do this or give me a head start ? Unfortunately, no, I don't. You might try asking on comp.compilers. Vern
@verbatim To: Adoram Rogel <adoram@hybridge.com> Subject: Re: Conditional compiling in the definitions section In-reply-to: Your message of Thu, 25 Sep 1997 11:22:42 PDT. Date: Thu, 25 Sep 1997 10:56:31 PDT From: Vern Paxson <vern> > I'm trying to combine two large lex files that now differ only in > about 10 lines in the definitions section. > I would like to have something like this: > #ifdef FFF > it \<IT\> > #else > it \<I\> > #endif > > Now, I can't add states for these, as I have already too many states > and the program is very complicated, and I won't be able to handle > 10 or 20 more states. > > Any trick to do this ? You might try using m4, or the C preprocessor plus a sed script to clean up the result (strip out the #line's). Vern
@verbatim To: Steve Antoch <SteveAn@visio.com> Subject: Re: lex and yacc grammars In-reply-to: Your message of Mon, 17 Nov 1997 15:31:25 PST. Date: Mon, 17 Nov 1997 15:27:01 PST From: Vern Paxson <vern> > Would you happen to know where I can find grammars for lex and yacc? The flex sources have a grammar for (f)lex. Dunno about yacc, Vern
@verbatim To: Bryan Housel <bryan@drawcomp.com> Subject: Re: Question about Flex v2.5 In-reply-to: Your message of Tue, 11 Nov 1997 21:30:23 PST. Date: Mon, 17 Nov 1997 17:12:21 PST From: Vern Paxson <vern> > It prints one of those "end of buffer.." messages for each character in the > token... This will happen if your LexerInput() function returns only one character at a time, which can happen either if you're scanner is "interactive", or if the streams library on your platform always returns 1 for yyin->gcount(). Solution: override LexerInput() with a version that returns whole buffers. Vern
@verbatim To: Georg.Rehm@CL-KI.Uni-Osnabrueck.DE Subject: Re: Flex maximums In-reply-to: Your message of Mon, 17 Nov 1997 17:16:06 PST. Date: Mon, 17 Nov 1997 17:16:15 PST From: Vern Paxson <vern> > I took a quick look into the flex-sources and altered some #defines in > flexdefs.h: > > #define INITIAL_MNS 64000 > #define MNS_INCREMENT 1024000 > #define MAXIMUM_MNS 64000 The things to fix are to add a couple of zeroes to: #define JAMSTATE -32766 /* marks a reference to the state that always jams */ #define MAXIMUM_MNS 31999 #define BAD_SUBSCRIPT -32767 #define MAX_SHORT 32700 and, if you get complaints about too many rules, make the following change too: #define YY_TRAILING_MASK 0x200000 #define YY_TRAILING_HEAD_MASK 0x400000 - Vern
@verbatim To: jimmey@lexis-nexis.com (Jimmey Todd) Subject: Re: FLEX question regarding istream vs ifstream In-reply-to: Your message of Mon, 08 Dec 1997 15:54:15 PST. Date: Mon, 15 Dec 1997 13:21:35 PST From: Vern Paxson <vern> > stdin_handle = YY_CURRENT_BUFFER; > ifstream fin( "aFile" ); > yy_switch_to_buffer( yy_create_buffer( fin, YY_BUF_SIZE ) ); > > What I'm wanting to do, is pass the contents of a file thru one set > of rules and then pass stdin thru another set... It works great if, I > don't use the C++ classes. But since everything else that I'm doing is > in C++, I thought I'd be consistent. > > The problem is that 'yy_create_buffer' is expecting an istream* as it's > first argument (as stated in the man page). However, fin is a ifstream > object. Any ideas on what I might be doing wrong? Any help would be > appreciated. Thanks!! You need to pass &fin, to turn it into an ifstream* instead of an ifstream. Then its type will be compatible with the expected istream*, because ifstream is derived from istream. Vern
@verbatim To: Enda Fadian <fadiane@piercom.ie> Subject: Re: Question related to Flex man page? In-reply-to: Your message of Tue, 16 Dec 1997 15:17:34 PST. Date: Tue, 16 Dec 1997 14:17:09 PST From: Vern Paxson <vern> > Can you explain to me what is ment by a long-jump in relation to flex? Using the longjmp() function while inside yylex() or a routine called by it. > what is the flex activation frame. Just yylex()'s stack frame. > As far as I can see yyrestart will bring me back to the sart of the input > file and using flex++ isnot really an option! No, yyrestart() doesn't imply a rewind, even though its name might sound like it does. It tells the scanner to flush its internal buffers and start reading from the given file at its present location. Vern
@verbatim To: hassan@larc.info.uqam.ca (Hassan Alaoui) Subject: Re: Need urgent Help In-reply-to: Your message of Sat, 20 Dec 1997 19:38:19 PST. Date: Sun, 21 Dec 1997 21:30:46 PST From: Vern Paxson <vern> > /usr/lib/yaccpar: In function `int yyparse()': > /usr/lib/yaccpar:184: warning: implicit declaration of function `int yylex(...)' > > ld: Undefined symbol > _yylex > _yyparse > _yyin This is a known problem with Solaris C++ (and/or Solaris yacc). I believe the fix is to explicitly insert some 'extern "C"' statements for the corresponding routines/symbols. Vern
@verbatim To: mc0307@mclink.it Cc: gnu@prep.ai.mit.edu Subject: Re: [mc0307@mclink.it: Help request] In-reply-to: Your message of Fri, 12 Dec 1997 17:57:29 PST. Date: Sun, 21 Dec 1997 22:33:37 PST From: Vern Paxson <vern> > This is my definition for float and integer types: > . . . > NZD [1-9] > ... > I've tested my program on other lex version (on UNIX Sun Solaris an HP > UNIX) and it work well, so I think that my definitions are correct. > There are any differences between Lex and Flex? There are indeed differences, as discussed in the man page. The one you are probably running into is that when flex expands a name definition, it puts parentheses around the expansion, while lex does not. There's an example in the man page of how this can lead to different matching. Flex's behavior complies with the POSIX standard (or at least with the last POSIX draft I saw). Vern
@verbatim To: hassan@larc.info.uqam.ca (Hassan Alaoui) Subject: Re: Thanks In-reply-to: Your message of Mon, 22 Dec 1997 16:06:35 PST. Date: Mon, 22 Dec 1997 14:35:05 PST From: Vern Paxson <vern> > Thank you very much for your help. I compile and link well with C++ while > declaring 'yylex ...' extern, But a little problem remains. I get a > segmentation default when executing ( I linked with lfl library) while it > works well when using LEX instead of flex. Do you have some ideas about the > reason for this ? The one possible reason for this that comes to mind is if you've defined yytext as "extern char yytext[]" (which is what lex uses) instead of "extern char *yytext" (which is what flex uses). If it's not that, then I'm afraid I don't know what the problem might be. Vern
@verbatim To: "Bart Niswonger" <NISWONGR@almaden.ibm.com> Subject: Re: flex 2.5: c++ scanners & start conditions In-reply-to: Your message of Tue, 06 Jan 1998 10:34:21 PST. Date: Tue, 06 Jan 1998 19:19:30 PST From: Vern Paxson <vern> > The problem is that when I do this (using %option c++) start > conditions seem to not apply. The BEGIN macro modifies the yy_start variable. For C scanners, this is a static with scope visible through the whole file. For C++ scanners, it's a member variable, so it only has visible scope within a member function. Your lexbegin() routine is not a member function when you build a C++ scanner, so it's not modifying the correct yy_start. The diagnostic that indicates this is that you found you needed to add a declaration of yy_start in order to get your scanner to compile when using C++; instead, the correct fix is to make lexbegin() a member function (by deriving from yyFlexLexer). Vern
@verbatim To: "Boris Zinin" <boris@ippe.rssi.ru> Subject: Re: current position in flex buffer In-reply-to: Your message of Mon, 12 Jan 1998 18:58:23 PST. Date: Mon, 12 Jan 1998 12:03:15 PST From: Vern Paxson <vern> > The problem is how to determine the current position in flex active > buffer when a rule is matched.... You will need to keep track of this explicitly, such as by redefining YY_USER_ACTION to count the number of characters matched. The latest flex release, by the way, is 2.5.4, available from ftp.ee.lbl.gov. Vern
@verbatim To: Bik.Dhaliwal@bis.org Subject: Re: Flex question In-reply-to: Your message of Mon, 26 Jan 1998 13:05:35 PST. Date: Tue, 27 Jan 1998 22:41:52 PST From: Vern Paxson <vern> > That requirement involves knowing > the character position at which a particular token was matched > in the lexer. The way you have to do this is by explicitly keeping track of where you are in the file, by counting the number of characters scanned for each token (available in yyleng). It may prove convenient to do this by redefining YY_USER_ACTION, as described in the manual. Vern
@verbatim To: Vladimir Alexiev <vladimir@cs.ualberta.ca> Subject: Re: flex: how to control start condition from parser? In-reply-to: Your message of Mon, 26 Jan 1998 05:50:16 PST. Date: Tue, 27 Jan 1998 22:45:37 PST From: Vern Paxson <vern> > It seems useful for the parser to be able to tell the lexer about such > context dependencies, because then they don't have to be limited to > local or sequential context. One way to do this is to have the parser call a stub routine that's included in the scanner's .l file, and consequently that has access ot BEGIN. The only ugliness is that the parser can't pass in the state it wants, because those aren't visible - but if you don't have many such states, then using a different set of names doesn't seem like to much of a burden. While generating a .h file like you suggests is certainly cleaner, flex development has come to a virtual stand-still :-(, so a workaround like the above is much more pragmatic than waiting for a new feature. Vern
@verbatim
To: Barbara Denny <denny@3com.com>
Subject: Re: freebsd flex bug?
In-reply-to: Your message of Fri, 30 Jan 1998 12:00:43 PST.
Date: Fri, 30 Jan 1998 12:42:32 PST
From: Vern Paxson <vern>
> lex.yy.c:1996: parse error before `='
This is the key, identifying this error. (It may help to pinpoint
it by using flex -L, so it doesn't generate #line directives in its
output.) I will bet you heavy money that you have a start condition
name that is also a variable name, or something like that; flex spits
out #define's for each start condition name, mapping them to a number,
so you can wind up with:
%x foo
%%
...
%%
void bar()
{
int foo = 3;
}
and the penultimate will turn into "int 1 = 3" after C preprocessing,
since flex will put "#define foo 1" in the generated scanner.
Vern
@verbatim To: Maurice Petrie <mpetrie@infoscigroup.com> Subject: Re: Lost flex .l file In-reply-to: Your message of Mon, 02 Feb 1998 14:10:01 PST. Date: Mon, 02 Feb 1998 11:15:12 PST From: Vern Paxson <vern> > I am curious as to > whether there is a simple way to backtrack from the generated source to > reproduce the lost list of tokens we are searching on. In theory, it's straight-forward to go from the DFA representation back to a regular-expression representation - the two are isomorphic. In practice, a huge headache, because you have to unpack all the tables back into a single DFA representation, and then write a program to munch on that and translate it into an RE. Sorry for the less-than-happy news ... Vern
@verbatim To: jimmey@lexis-nexis.com (Jimmey Todd) Subject: Re: Flex performance question In-reply-to: Your message of Thu, 19 Feb 1998 11:01:17 PST. Date: Thu, 19 Feb 1998 08:48:51 PST From: Vern Paxson <vern> > What I have found, is that the smaller the data chunk, the faster the > program executes. This is the opposite of what I expected. Should this be > happening this way? This is exactly what will happen if your input file has embedded NULs. From the man page: A final note: flex is slow when matching NUL's, particularly when a token contains multiple NUL's. It's best to write rules which match short amounts of text if it's anticipated that the text will often include NUL's. So that's the first thing to look for. Vern
@verbatim To: jimmey@lexis-nexis.com (Jimmey Todd) Subject: Re: Flex performance question In-reply-to: Your message of Thu, 19 Feb 1998 11:01:17 PST. Date: Thu, 19 Feb 1998 15:42:25 PST From: Vern Paxson <vern> So there are several problems. First, to go fast, you want to match as much text as possible, which your scanners don't in the case that what they're scanning is *not* a <RN> tag. So you want a rule like: [^<]+ Second, C++ scanners are particularly slow if they're interactive, which they are by default. Using -B speeds it up by a factor of 3-4 on my workstation. Third, C++ scanners that use the istream interface are slow, because of how poorly implemented istream's are. I built two versions of the following scanner: %% .*\n .* %% and the C version inhales a 2.5MB file on my workstation in 0.8 seconds. The C++ istream version, using -B, takes 3.8 seconds. Vern
@verbatim To: "Frescatore, David (CRD, TAD)" <frescatore@exc01crdge.crd.ge.com> Subject: Re: FLEX 2.5 & THE YEAR 2000 In-reply-to: Your message of Wed, 03 Jun 1998 11:26:22 PDT. Date: Wed, 03 Jun 1998 10:22:26 PDT From: Vern Paxson <vern> > I am researching the Y2K problem with General Electric R&D > and need to know if there are any known issues concerning > the above mentioned software and Y2K regardless of version. There shouldn't be, all it ever does with the date is ask the system for it and then print it out. Vern
@verbatim To: "Hans Dermot Doran" <htd@ibhdoran.com> Subject: Re: flex problem In-reply-to: Your message of Wed, 15 Jul 1998 21:30:13 PDT. Date: Tue, 21 Jul 1998 14:23:34 PDT From: Vern Paxson <vern> > To overcome this, I gets() the stdin into a string and lex the string. The > string is lexed OK except that the end of string isn't lexed properly > (yy_scan_string()), that is the lexer dosn't recognise the end of string. Flex doesn't contain mechanisms for recognizing buffer endpoints. But if you use fgets instead (which you should anyway, to protect against buffer overflows), then the final \n will be preserved in the string, and you can scan that in order to find the end of the string. Vern
@verbatim
To: soumen@almaden.ibm.com
Subject: Re: Flex++ 2.5.3 instance member vs. static member
In-reply-to: Your message of Mon, 27 Jul 1998 02:10:04 PDT.
Date: Tue, 28 Jul 1998 01:10:34 PDT
From: Vern Paxson <vern>
> %{
> int mylineno = 0;
> %}
> ws [ \t]+
> alpha [A-Za-z]
> dig [0-9]
> %%
>
> Now you'd expect mylineno to be a member of each instance of class
> yyFlexLexer, but is this the case? A look at the lex.yy.cc file seems to
> indicate otherwise; unless I am missing something the declaration of
> mylineno seems to be outside any class scope.
>
> How will this work if I want to run a multi-threaded application with each
> thread creating a FlexLexer instance?
Derive your own subclass and make mylineno a member variable of it.
Vern
@verbatim To: Adoram Rogel <adoram@hybridge.com> Subject: Re: More than 32K states change hangs In-reply-to: Your message of Tue, 04 Aug 1998 16:55:39 PDT. Date: Tue, 04 Aug 1998 22:28:45 PDT From: Vern Paxson <vern> > Vern Paxson, > > I followed your advice, posted on Usenet bu you, and emailed to me > personally by you, on how to overcome the 32K states limit. I'm running > on Linux machines. > I took the full source of version 2.5.4 and did the following changes in > flexdef.h: > #define JAMSTATE -327660 > #define MAXIMUM_MNS 319990 > #define BAD_SUBSCRIPT -327670 > #define MAX_SHORT 327000 > > and compiled. > All looked fine, including check and bigcheck, so I installed. Hmmm, you shouldn't increase MAX_SHORT, though looking through my email archives I see that I did indeed recommend doing so. Try setting it back to 32700; that should suffice that you no longer need -Ca. If it still hangs, then the interesting question is - where? > Compiling the same hanged program with a out-of-the-box (RedHat 4.2 > distribution of Linux) > flex 2.5.4 binary works. Since Linux comes with source code, you should diff it against what you have to see what problems they missed. > Should I always compile with the -Ca option now ? even short and simple > filters ? No, definitely not. It's meant to be for those situations where you absolutely must squeeze every last cycle out of your scanner. Vern
@verbatim To: "Schmackpfeffer, Craig" <Craig.Schmackpfeffer@usa.xerox.com> Subject: Re: flex output for static code portion In-reply-to: Your message of Tue, 11 Aug 1998 11:55:30 PDT. Date: Mon, 17 Aug 1998 23:57:42 PDT From: Vern Paxson <vern> > I would like to use flex under the hood to generate a binary file > containing the data structures that control the parse. This has been on the wish-list for a long time. In principle it's straight-forward - you redirect mkdata() et al's I/O to another file, and modify the skeleton to have a start-up function that slurps these into dynamic arrays. The concerns are (1) the scanner generation code is hairy and full of corner cases, so it's easy to get surprised when going down this path :-( ; and (2) being careful about buffering so that when the tables change you make sure the scanner starts in the correct state and reading at the right point in the input file. > I was wondering if you know of anyone who has used flex in this way. I don't - but it seems like a reasonable project to undertake (unlike numerous other flex tweaks :-). Vern
@verbatim
Received: from 131.173.17.11 (131.173.17.11 [131.173.17.11])
by ee.lbl.gov (8.9.1/8.9.1) with ESMTP id AAA03838
for <vern@ee.lbl.gov>; Thu, 20 Aug 1998 00:47:57 -0700 (PDT)
Received: from hal.cl-ki.uni-osnabrueck.de (hal.cl-ki.Uni-Osnabrueck.DE [131.173.141.2])
by deimos.rz.uni-osnabrueck.de (8.8.7/8.8.8) with ESMTP id JAA34694
for <vern@ee.lbl.gov>; Thu, 20 Aug 1998 09:47:55 +0200
Received: (from georg@localhost) by hal.cl-ki.uni-osnabrueck.de (8.6.12/8.6.12) id JAA34834 for vern@ee.lbl.gov; Thu, 20 Aug 1998 09:47:54 +0200
From: Georg Rehm <georg@hal.cl-ki.uni-osnabrueck.de>
Message-Id: <199808200747.JAA34834@hal.cl-ki.uni-osnabrueck.de>
Subject: "flex scanner push-back overflow"
To: vern@ee.lbl.gov
Date: Thu, 20 Aug 1998 09:47:54 +0200 (MEST)
Reply-To: Georg.Rehm@CL-KI.Uni-Osnabrueck.DE
X-NoJunk: Do NOT send commercial mail, spam or ads to this address!
X-URL: http://www.cl-ki.uni-osnabrueck.de/~georg/
X-Mailer: ELM [version 2.4ME+ PL28 (25)]
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Hi Vern,
Yesterday, I encountered a strange problem: I use the macro processor m4
to include some lengthy lists into a .l file. Following is a flex macro
definition that causes some serious pain in my neck:
AUTHOR ("A. Boucard / L. Boucard"|"A. Dastarac / M. Levent"|"A.Boucaud / L.Boucaud"|"Abderrahim Lamchichi"|"Achmat Dangor"|"Adeline Toullier"|"Adewale Maja-Pearce"|"Ahmed Ziri"|"Akram Ellyas"|"Alain Bihr"|"Alain Gresh"|"Alain Guillemoles"|"Alain Joxe"|"Alain Morice"|"Alain Renon"|"Alain Zecchini"|"Albert Memmi"|"Alberto Manguel"|"Alex De Waal"|"Alfonso Artico"| [...])
The complete list contains about 10kB. When I try to "flex" this file
(on a Solaris 2.6 machine, using a modified flex 2.5.4 (I only increased
some of the predefined values in flexdefs.h) I get the error:
myflex/flex -8 sentag.tmp.l
flex scanner push-back overflow
When I remove the slashes in the macro definition everything works fine.
As I understand it, the double quotes escape the slash-character so it
really means "/" and not "trailing context". Furthermore, I tried to
escape the slashes with backslashes, but with no use, the same error message
appeared when flexing the code.
Do you have an idea what's going on here?
Greetings from Germany,
Georg
--
Georg Rehm georg@cl-ki.uni-osnabrueck.de
Institute for Semantic Information Processing, University of Osnabrueck, FRG
@verbatim To: Georg.Rehm@CL-KI.Uni-Osnabrueck.DE Subject: Re: "flex scanner push-back overflow" In-reply-to: Your message of Thu, 20 Aug 1998 09:47:54 PDT. Date: Thu, 20 Aug 1998 07:05:35 PDT From: Vern Paxson <vern> > myflex/flex -8 sentag.tmp.l > flex scanner push-back overflow Flex itself uses a flex scanner. That scanner is running out of buffer space when it tries to unput() the humongous macro you've defined. When you remove the '/'s, you make it small enough so that it fits in the buffer; removing spaces would do the same thing. The fix is to either rethink how come you're using such a big macro and perhaps there's another/better way to do it; or to rebuild flex's own scan.c with a larger value for #define YY_BUF_SIZE 16384 - Vern
@verbatim
To: Jan Kort <jan@research.techforce.nl>
Subject: Re: Flex
In-reply-to: Your message of Fri, 04 Sep 1998 12:18:43 +0200.
Date: Sat, 05 Sep 1998 00:59:49 PDT
From: Vern Paxson <vern>
> %%
>
> "TEST1\n" { fprintf(stderr, "TEST1\n"); yyless(5); }
> ^\n { fprintf(stderr, "empty line\n"); }
> . { }
> \n { fprintf(stderr, "new line\n"); }
>
> %%
> -- input ---------------------------------------
> TEST1
> -- output --------------------------------------
> TEST1
> empty line
> ------------------------------------------------
IMHO, it's not clear whether or not this is in fact a bug. It depends
on whether you view yyless() as backing up in the input stream, or as
pushing new characters onto the beginning of the input stream. Flex
interprets it as the latter (for implementation convenience, I'll admit),
and so considers the newline as in fact matching at the beginning of a
line, as after all the last token scanned an entire line and so the
scanner is now at the beginning of a new line.
I agree that this is counter-intuitive for yyless(), given its
functional description (it's less so for unput(), depending on whether
you're unput()'ing new text or scanned text). But I don't plan to
change it any time soon, as it's a pain to do so. Consequently,
you do indeed need to use yy_set_bol() and YY_AT_BOL() to tweak
your scanner into the behavior you desire.
Sorry for the less-than-completely-satisfactory answer.
Vern
@verbatim To: Patrick Krusenotto <krusenot@mac-info-link.de> Subject: Re: Problems with restarting flex-2.5.2-generated scanner In-reply-to: Your message of Thu, 24 Sep 1998 10:14:07 PDT. Date: Thu, 24 Sep 1998 23:28:43 PDT From: Vern Paxson <vern> > I am using flex-2.5.2 and bison 1.25 for Solaris and I am desperately > trying to make my scanner restart with a new file after my parser stops > with a parse error. When my compiler restarts, the parser always > receives the token after the token (in the old file!) that caused the > parser error. I suspect the problem is that your parser has read ahead in order to attempt to resolve an ambiguity, and when it's restarted it picks up with that token rather than reading a fresh one. If you're using yacc, then the special "error" production can sometimes be used to consume tokens in an attempt to get the parser into a consistent state. Vern
@verbatim To: Henric Jungheim <junghelh@pe-nelson.com> Subject: Re: flex 2.5.4a In-reply-to: Your message of Tue, 27 Oct 1998 16:41:42 PST. Date: Tue, 27 Oct 1998 16:50:14 PST From: Vern Paxson <vern> > This brings up a feature request: How about a command line > option to specify the filename when reading from stdin? That way one > doesn't need to create a temporary file in order to get the "#line" > directives to make sense. Use -o combined with -t (per the man page description of -o). > P.S., Is there any simple way to use non-blocking IO to parse multiple > streams? Simple, no. One approach might be to return a magic character on EWOULDBLOCK and have a rule .*<magic-character> // put back .*, eat magic character This is off the top of my head, not sure it'll work. Vern
@verbatim To: "Repko, Billy D" <billy.d.repko@intel.com> Subject: Re: Compiling scanners In-reply-to: Your message of Wed, 13 Jan 1999 10:52:47 PST. Date: Thu, 14 Jan 1999 00:25:30 PST From: Vern Paxson <vern> > It appears that maybe it cannot find the lfl library. The Makefile in the distribution builds it, so you should have it. It's exceedingly trivial, just a main() that calls yylex() and a yyrap() that always returns 1. > %% > \n ++num_lines; ++num_chars; > . ++num_chars; You can't indent your rules like this - that's where the errors are coming from. Flex copies indented text to the output file, it's how you do things like int num_lines_seen = 0; to declare local variables. Vern
@verbatim To: Erick Branderhorst <Erick.Branderhorst@asml.nl> Subject: Re: flex input buffer In-reply-to: Your message of Tue, 09 Feb 1999 13:53:46 PST. Date: Tue, 09 Feb 1999 21:03:37 PST From: Vern Paxson <vern> > In the flex.skl file the size of the default input buffers is set. Can you > explain why this size is set and why it is such a high number. It's large to optimize performance when scanning large files. You can safely make it a lot lower if needed. Vern
@verbatim To: "Guido Minnen" <guidomi@cogs.susx.ac.uk> Subject: Re: Flex error message In-reply-to: Your message of Wed, 24 Feb 1999 15:31:46 PST. Date: Thu, 25 Feb 1999 00:11:31 PST From: Vern Paxson <vern> > I'm extending a larger scanner written in Flex and I keep running into > problems. More specifically, I get the error message: > "flex: input rules are too complicated (>= 32000 NFA states)" Increase the definitions in flexdef.h for: #define JAMSTATE -32766 /* marks a reference to the state that always j ams */ #define MAXIMUM_MNS 31999 #define BAD_SUBSCRIPT -32767 recompile everything, and it should all work. Vern
@verbatim To: "Dmitriy Goldobin" <gold@ems.chel.su> Subject: Re: FLEX trouble In-reply-to: Your message of Mon, 31 May 1999 18:44:49 PDT. Date: Tue, 01 Jun 1999 00:15:07 PDT From: Vern Paxson <vern> > I have a trouble with FLEX. Why rule "/*".*"*/" work properly,=20 > but rule "/*"(.|\n)*"*/" don't work ? The second of these will have to scan the entire input stream (because "(.|\n)*" matches an arbitrary amount of any text) in order to see if it ends with "*/", terminating the comment. That potentially will overflow the input buffer. > More complex rule "/*"([^*]|(\*/[^/]))*"*/ give an error > 'unrecognized rule'. You can't use the '/' operator inside parentheses. It's not clear what "(a/b)*" actually means. > I now use workaround with state <comment>, but single-rule is > better, i think. Single-rule is nice but will always have the problem of either setting restrictions on comments (like not allowing multi-line comments) and/or running the risk of consuming the entire input stream, as noted above. Vern
@verbatim Received: from mc-qout4.whowhere.com (mc-qout4.whowhere.com [209.185.123.18]) by ee.lbl.gov (8.9.3/8.9.3) with SMTP id IAA05100 for <vern@ee.lbl.gov>; Tue, 15 Jun 1999 08:56:06 -0700 (PDT) Received: from Unknown/Local ([?.?.?.?]) by my-deja.com; Tue Jun 15 08:55:43 1999 To: vern@ee.lbl.gov Date: Tue, 15 Jun 1999 08:55:43 -0700 From: "Aki Niimura" <neko@my-deja.com> Message-ID: <KNONDOHDOBGAEAAA@my-deja.com> Mime-Version: 1.0 Cc: X-Sent-Mail: on Reply-To: X-Mailer: MailCity Service Subject: A question on flex C++ scanner X-Sender-Ip: 12.72.207.61 Organization: My Deja Email (http://www.my-deja.com:80) Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Dear Dr. Paxon, I have been using flex for years. It works very well on many projects. Most case, I used it to generate a scanner on C language. However, one project I needed to generate a scanner on C++ lanuage. Thanks to your enhancement, flex did the job. Currently, I'm working on enhancing my previous project. I need to deal with multiple input streams (recursive inclusion) in this scanner (C++). I did similar thing for another scanner (C) as you explained in your documentation. The generated scanner (C++) has necessary methods: - switch_to_buffer(struct yy_buffer_state *b) - yy_create_buffer(istream *is, int sz) - yy_delete_buffer(struct yy_buffer_state *b) However, I couldn't figure out how to access current buffer (yy_current_buffer). yy_current_buffer is a protected member of yyFlexLexer. I can't access it directly. Then, I thought yy_create_buffer() with is = 0 might return current stream buffer. But it seems not as far as I checked the source. (flex 2.5.4) I went through the Web in addition to Flex documentation. However, it hasn't been successful, so far. It is not my intention to bother you, but, can you comment about how to obtain the current stream buffer? Your response would be highly appreciated. Best regards, Aki Niimura --== Sent via Deja.com http://www.deja.com/ ==-- Share what you know. Learn what you don't.
@verbatim To: neko@my-deja.com Subject: Re: A question on flex C++ scanner In-reply-to: Your message of Tue, 15 Jun 1999 08:55:43 PDT. Date: Tue, 15 Jun 1999 09:04:24 PDT From: Vern Paxson <vern> > However, I couldn't figure out how to access current > buffer (yy_current_buffer). Derive your own subclass from yyFlexLexer. Vern
@verbatim To: "Stones, Darren" <Darren.Stones@nectech.co.uk> Subject: Re: You're the man to see? In-reply-to: Your message of Wed, 23 Jun 1999 11:10:29 PDT. Date: Wed, 23 Jun 1999 09:01:40 PDT From: Vern Paxson <vern> > I hope you can help me. I am using Flex and Bison to produce an interpreted > language. However all goes well until I try to implement an IF statement or > a WHILE. I cannot get this to work as the parser parses all the conditions > eg. the TRUE and FALSE conditons to check for a rule match. So I cannot > make a decision!! You need to use the parser to build a parse tree (= abstract syntax trwee), and when that's all done you recursively evaluate the tree, binding variables to values at that time. Vern
@verbatim To: Petr Danecek <petr@ics.cas.cz> Subject: Re: flex - question In-reply-to: Your message of Mon, 28 Jun 1999 19:21:41 PDT. Date: Fri, 02 Jul 1999 16:52:13 PDT From: Vern Paxson <vern> > file, it takes an enormous amount of time. It is funny, because the > source code has only 12 rules!!! I think it looks like an exponencial > growth. Right, that's the problem - some patterns (those with a lot of ambiguity, where yours has because at any given time the scanner can be in the middle of all sorts of combinations of the different rules) blow up exponentially. For your rules, there is an easy fix. Change the ".*" that comes fater the directory name to "[^ ]*". With that in place, the rules are no longer nearly so ambiguous, because then once one of the directories has been matched, no other can be matched (since they all require a leading blank). If that's not an acceptable solution, then you can enter a start state to pick up the .*\n after each directory is matched. Also note that for speed, you'll want to add a ".*" rule at the end, otherwise rules that don't match any of the patterns will be matched very slowly, a character at a time. Vern
@verbatim To: Tielman Koekemoer <tielman@spi.co.za> Subject: Re: Please help. In-reply-to: Your message of Thu, 08 Jul 1999 13:20:37 PDT. Date: Thu, 08 Jul 1999 08:20:39 PDT From: Vern Paxson <vern> > I was hoping you could help me with my problem. > > I tried compiling (gnu)flex on a Solaris 2.4 machine > but when I ran make (after configure) I got an error. > > -------------------------------------------------------------- > gcc -c -I. -I. -g -O parse.c > ./flex -t -p ./scan.l >scan.c > sh: ./flex: not found > *** Error code 1 > make: Fatal error: Command failed for target `scan.c' > ------------------------------------------------------------- > > What's strange to me is that I'm only > trying to install flex now. I then edited the Makefile to > and changed where it says "FLEX = flex" to "FLEX = lex" > ( lex: the native Solaris one ) but then it complains about > the "-p" option. Is there any way I can compile flex without > using flex or lex? > > Thanks so much for your time. You managed to step on the bootstrap sequence, which first copies initscan.c to scan.c in order to build flex. Try fetching a fresh distribution from ftp.ee.lbl.gov. (Or you can first try removing ".bootstrap" and doing a make again.) Vern
@verbatim To: Tielman Koekemoer <tielman@spi.co.za> Subject: Re: Please help. In-reply-to: Your message of Fri, 09 Jul 1999 09:16:14 PDT. Date: Fri, 09 Jul 1999 00:27:20 PDT From: Vern Paxson <vern> > First I removed .bootstrap (and ran make) - no luck. I downloaded the > software but I still have the same problem. Is there anything else I > could try. Try: cp initscan.c scan.c touch scan.c make scan.o If this last tries to first build scan.c from scan.l using ./flex, then your "make" is broken, in which case compile scan.c to scan.o by hand. Vern
@verbatim To: Sumanth Kamenani <skamenan@crl.nmsu.edu> Subject: Re: Error In-reply-to: Your message of Mon, 19 Jul 1999 23:08:41 PDT. Date: Tue, 20 Jul 1999 00:18:26 PDT From: Vern Paxson <vern> > I am getting a compilation error. The error is given as "unknown symbol- yylex". The parser relies on calling yylex(), but you're instead using the C++ scanning class, so you need to supply a yylex() "glue" function that calls an instance scanner of the scanner (e.g., "scanner->yylex()"). Vern
@verbatim
To: daniel@synchrods.synchrods.COM (Daniel Senderowicz)
Subject: Re: lex
In-reply-to: Your message of Mon, 22 Nov 1999 11:19:04 PST.
Date: Tue, 23 Nov 1999 15:54:30 PST
From: Vern Paxson <vern>
Well, your problem is the
switch (yybgin-yysvec-1) { /* witchcraft */
at the beginning of lex rules. "witchcraft" == "non-portable". It's
assuming knowledge of the AT&T lex's internal variables.
For flex, you can probably do the equivalent using a switch on YYSTATE.
Vern
@verbatim To: archow@hss.hns.com Subject: Re: Regarding distribution of flex and yacc based grammars In-reply-to: Your message of Sun, 19 Dec 1999 17:50:24 +0530. Date: Wed, 22 Dec 1999 01:56:24 PST From: Vern Paxson <vern> > When we provide the customer with an object code distribution, is it > necessary for us to provide source > for the generated C files from flex and bison since they are generated by > flex and bison ? For flex, no. I don't know what the current state of this is for bison. > Also, is there any requrirement for us to neccessarily provide source for > the grammar files which are fed into flex and bison ? Again, for flex, no. See the file "COPYING" in the flex distribution for the legalese. Vern
@verbatim
To: Martin Gallwey <gallweym@hyperion.moe.ul.ie>
Subject: Re: Flex, and self referencing rules
In-reply-to: Your message of Sun, 20 Feb 2000 01:01:21 PST.
Date: Sat, 19 Feb 2000 18:33:16 PST
From: Vern Paxson <vern>
> However, I do not use unput anywhere. I do use self-referencing
> rules like this:
>
> UnaryExpr ({UnionExpr})|("-"{UnaryExpr})
You can't do this - flex is *not* a parser like yacc (which does indeed
allow recursion), it is a scanner that's confined to regular expressions.
Vern
@verbatim
To: slg3@lehigh.edu (SAMUEL L. GULDEN)
Subject: Re: Flex problem
In-reply-to: Your message of Thu, 02 Mar 2000 12:29:04 PST.
Date: Thu, 02 Mar 2000 23:00:46 PST
From: Vern Paxson <vern>
If this is exactly your program:
> digit [0-9]
> digits {digit}+
> whitespace [ \t\n]+
>
> %%
> "[" { printf("open_brac\n");}
> "]" { printf("close_brac\n");}
> "+" { printf("addop\n");}
> "*" { printf("multop\n");}
> {digits} { printf("NUMBER = %s\n", yytext);}
> whitespace ;
then the problem is that the last rule needs to be "{whitespace}" !
Vern
In this appendix, we provide tips for writing Makefiles to build your scanners.
In a traditional build environment, we say that the `.c' files are the
sources, and the `.o' files are the intermediate files. When using
flex, however, the `.l' files are the sources, and the generated
`.c' files (along with the `.o' files) are the intermediate files.
This requires you to carefully plan your Makefile.
Modern @command{make} programs understand that `foo.l' is intended to generate `lex.yy.c' or `foo.c', and will behave accordingly(3) and GNU @command{automake} are two such programs that provide implicit rules for flex-generated scanners.}. The following Makefile does not explicitly instruct @command{make} how to build `foo.c' from `foo.l'. Instead, it relies on the implicit rules of the @command{make} program to build the intermediate file, `scan.c':
@verbatim
# Basic Makefile -- relies on implicit rules
# Creates "myprogram" from "scan.l" and "myprogram.c"
#
LEX=flex
myprogram: scan.o myprogram.o
scan.o: scan.l
For simple cases, the above may be sufficient. For other cases, you may have to explicitly instruct @command{make} how to build your scanner. The following is an example of a Makefile containing explicit rules:
@verbatim
# Basic Makefile -- provides explicit rules
# Creates "myprogram" from "scan.l" and "myprogram.c"
#
LEX=flex
myprogram: scan.o myprogram.o
$(CC) -o $ $(LDFLAGS) $^
myprogram.o: myprogram.c
$(CC) $(CPPFLAGS) $(CFLAGS) -o $ -c $^
scan.o: scan.c
$(CC) $(CPPFLAGS) $(CFLAGS) -o $ -c $^
scan.c: scan.l
$(LEX) $(LFLAGS) -o $ $^
clean:
$(RM) *.o scan.c
Notice in the above example that `scan.c' is in the clean target.
This is because we consider the file `scan.c' to be an intermediate file.
Finally, we provide a realistic example of a flex scanner used with a
bison parser(4).
There is a tricky problem we have to deal with. Since a flex scanner
will typically include a header file (e.g., `y.tab.h') generated by the
parser, we need to be sure that the header file is generated BEFORE the scanner
is compiled. We handle this case in the following example:
@verbatim
# Makefile example -- scanner and parser.
# Creates "myprogram" from "scan.l", "parse.y", and "myprogram.c"
#
LEX = flex
YACC = bison -y
YFLAGS = -d
objects = scan.o parse.o myprogram.o
myprogram: $(objects)
scan.o: scan.l parse.c
parse.o: parse.y
myprogram.o: myprogram.c
In the above example, notice the line,
@verbatim
scan.o: scan.l parse.c
, which lists the file `parse.c' (the generated parser) as a dependency of `scan.o'. We want to ensure that the parser is created before the scanner is compiled, and the above line seems to do the trick. Feel free to experiment with your specific implementation of @command{make}.
For more details on writing Makefiles, see section `Top' in The GNU Make Manual.
This section describes the flex features useful when integrating
flex with GNU bison(5).
Skip this section if you are not using
bison with your scanner. Here we discuss only the flex
half of the flex and bison pair. We do not discuss
bison in any detail. For more information about generating
bison parsers, see section `Top' in the GNU Bison Manual.
A compatible bison scanner is generated by declaring `%option
bison-bridge' or by supplying `--bison-bridge' when invoking flex
from the command line. This instructs flex that the macros
yylval and yylloc may be used. The data types for
yylval and yylloc, (YYSTYPE and YYLTYPE,
are typically defined in a header file, included in section 1 of the
flex input file. If %option bison-bridge is
specified, flex provides support for the functions
yyget_lval, yyset_lval, yyget_lloc, and
yyset_lloc, defined below, and the corresponding macros
yylval and yylloc, for use within actions.
Where yyscan_t is defined in the reentrant scanner (6). Accordingly, the declaration of yylex becomes one of the following:
@verbatim
int yylex ( YYSTYPE * lvalp, yyscan_t scanner );
int yylex ( YYSTYPE * lvalp, YYLTYPE * llocp, yyscan_t scanner );
Note that the macros yylval and yylloc evaluate to
pointers. Support for yylloc is optional in bison, so it
is optional in flex as well. This support is automatically
handled by flex. Specifically, support for yyloc is only
present in a flex scanner if the preprocessor symbol
YYLTYPE is defined. The following is an example of a flex
scanner that is compatible with bison.
@verbatim
/* Scanner for "C" assignment statements... sort of. */
%{
#include "y.tab.h" /* Generated by bison. */
%}
%option reentrant-bison
%
[[:digit:]]+ { yylval->num = atoi(yytext); return NUMBER;}
[[:alnum:]]+ { yylval->str = strdup(yytext); return STRING;}
"="|";" { return yytext[0];}
. {}
%
As you can see, there really is no magic here. We just use
yylval as we would any other variable. The data type of
yylval is generated by bison, and included in the file
`y.tab.h'. Here is the corresponding bison parser:
@verbatim
/* Parser to convert "C" assignments to lisp. */
%{
/* Pass the argument to yyparse through to yylex. */
#define YYPARSE_PARAM scanner
#define YYLEX_PARAM scanner
%}
%pure_parser
%union {
int num;
char* str;
}
%token <str> STRING
%token <num> NUMBER
%%
assignment:
STRING '=' NUMBER ';' {
printf( "(setf %s %d)", $1, $3 );
}
;
Jump to: $ - % - < - [ - ^ - a - b - c - d - e - f - g - h - i - k - l - m - n - o - p - r - s - t - u - v - w - y
This is an index of functions and preprocessor macros that look like functions. For macros that expand to variables or constants, see section Index of Variables.
This is an index of variables, constants, and preprocessor macros that expand to variables or constants.
This is an index of "hooks" that the user may define. These hooks typically correspond to specific locations in the generated scanner, and may be used to insert arbitrary code.
Jump to: y
Jump to: - - 7 - 8 - a - b - c - d - e - f - h - i - l - m - n - o - p - r - s - t - v - w - y
This document was generated on 9 November 2002 using texi2html 1.56k.