.TH "Lexing" 3 2024-05-31 OCamldoc "OCaml library"
.SH NAME
Lexing \- The run-time library for lexers generated by ocamllex.
.SH Module
Module   Lexing
.SH Documentation
.sp
Module
.BI "Lexing"
 : 
.B sig end

.sp
The run\-time library for lexers generated by 
.ft B
ocamllex
.ft R
\&.

.sp

.sp
.sp

.PP
.SS Positions

.PP
.I type position 
= {
 pos_fname : 
.B string
;
 pos_lnum : 
.B int
;
 pos_bol : 
.B int
;
 pos_cnum : 
.B int
;
 }

.sp
A value of type 
.ft B
position
.ft R
describes a point in a source file\&.
.ft B
pos_fname
.ft R
is the file name; 
.ft B
pos_lnum
.ft R
is the line number;
.ft B
pos_bol
.ft R
is the offset of the beginning of the line (number
of characters between the beginning of the lexbuf and the beginning
of the line); 
.ft B
pos_cnum
.ft R
is the offset of the position (number of
characters between the beginning of the lexbuf and the position)\&.
The difference between 
.ft B
pos_cnum
.ft R
and 
.ft B
pos_bol
.ft R
is the character
offset within the line (i\&.e\&. the column number, assuming each
character is one column wide)\&.
.sp
See the documentation of type 
.ft B
lexbuf
.ft R
for information about
how the lexing engine will manage positions\&.

.sp

.I val dummy_pos 
: 
.B position
.sp
A value of type 
.ft B
position
.ft R
, guaranteed to be different from any
valid position\&.

.sp

.PP
.SS Lexer buffers

.PP
.I type lexbuf 
= {
 refill_buff : 
.B lexbuf -> unit
;

.B mutable 
lex_buffer : 
.B bytes
;

.B mutable 
lex_buffer_len : 
.B int
;

.B mutable 
lex_abs_pos : 
.B int
;

.B mutable 
lex_start_pos : 
.B int
;

.B mutable 
lex_curr_pos : 
.B int
;

.B mutable 
lex_last_pos : 
.B int
;

.B mutable 
lex_last_action : 
.B int
;

.B mutable 
lex_eof_reached : 
.B bool
;

.B mutable 
lex_mem : 
.B int array
;

.B mutable 
lex_start_p : 
.B position
;

.B mutable 
lex_curr_p : 
.B position
;
 }

.sp
The type of lexer buffers\&. A lexer buffer is the argument passed
to the scanning functions defined by the generated scanners\&.
The lexer buffer holds the current state of the scanner, plus
a function to refill the buffer from the input\&.
.sp
Lexers can optionally maintain the 
.ft B
lex_curr_p
.ft R
and 
.ft B
lex_start_p
.ft R
position fields\&.  This "position tracking" mode is the default, and
it corresponds to passing 
.ft B
~with_position:true
.ft R
to functions that
create lexer buffers\&. In this mode, the lexing engine and lexer
actions are co\-responsible for properly updating the position
fields, as described in the next paragraph\&.  When the mode is
explicitly disabled (with 
.ft B
~with_position:false
.ft R
), the lexing
engine will not touch the position fields and the lexer actions
should be careful not to do it either; the 
.ft B
lex_curr_p
.ft R
and
.ft B
lex_start_p
.ft R
field will then always hold the 
.ft B
dummy_pos
.ft R
invalid
position\&.  Not tracking positions avoids allocations and memory
writes and can significantly improve the performance of the lexer
in contexts where 
.ft B
lex_start_p
.ft R
and 
.ft B
lex_curr_p
.ft R
are not needed\&.
.sp
Position tracking mode works as follows\&.  At each token, the lexing
engine will copy 
.ft B
lex_curr_p
.ft R
to 
.ft B
lex_start_p
.ft R
, then change the
.ft B
pos_cnum
.ft R
field of 
.ft B
lex_curr_p
.ft R
by updating it with the number of
characters read since the start of the 
.ft B
lexbuf
.ft R
\&.  The other fields
are left unchanged by the lexing engine\&.  In order to keep them
accurate, they must be initialised before the first use of the
lexbuf, and updated by the relevant lexer actions (i\&.e\&. at each end
of line \-\- see also 
.ft B
new_line
.ft R
)\&.

.sp

.I val from_channel 
: 
.B ?with_positions:bool -> in_channel -> lexbuf
.sp
Create a lexer buffer on the given input channel\&.
.ft B
Lexing\&.from_channel inchan
.ft R
returns a lexer buffer which reads
from the input channel 
.ft B
inchan
.ft R
, at the current reading position\&.

.sp

.I val from_string 
: 
.B ?with_positions:bool -> string -> lexbuf
.sp
Create a lexer buffer which reads from
the given string\&. Reading starts from the first character in
the string\&. An end\-of\-input condition is generated when the
end of the string is reached\&.

.sp

.I val from_function 
: 
.B ?with_positions:bool -> (bytes -> int -> int) -> lexbuf
.sp
Create a lexer buffer with the given function as its reading method\&.
When the scanner needs more characters, it will call the given
function, giving it a byte sequence 
.ft B
s
.ft R
and a byte
count 
.ft B
n
.ft R
\&. The function should put 
.ft B
n
.ft R
bytes or fewer in 
.ft B
s
.ft R
,
starting at index 0, and return the number of bytes
provided\&. A return value of 0 means end of input\&.

.sp

.I val set_position 
: 
.B lexbuf -> position -> unit
.sp
Set the initial tracked input position for 
.ft B
lexbuf
.ft R
to a custom value\&.
Ignores 
.ft B
pos_fname
.ft R
\&. See 
.ft B
Lexing\&.set_filename
.ft R
for changing this field\&.

.sp
.B "Since"
4.11

.sp

.I val set_filename 
: 
.B lexbuf -> string -> unit
.sp
Set filename in the initial tracked position to 
.ft B
file
.ft R
in
.ft B
lexbuf
.ft R
\&.

.sp
.B "Since"
4.11

.sp

.I val with_positions 
: 
.B lexbuf -> bool
.sp
Tell whether the lexer buffer keeps track of position fields
.ft B
lex_curr_p
.ft R
/ 
.ft B
lex_start_p
.ft R
, as determined by the corresponding
optional argument for functions that create lexer buffers
(whose default value is 
.ft B
true
.ft R
)\&.
.sp
When 
.ft B
with_positions
.ft R
is 
.ft B
false
.ft R
, lexer actions should not
modify position fields\&.  Doing it nevertheless could
re\-enable the 
.ft B
with_position
.ft R
mode and degrade performances\&.

.sp

.PP
.SS Functions for lexer semantic actions

.PP

.PP
The following functions can be called from the semantic actions
of lexer definitions (the ML code enclosed in braces that
computes the value returned by lexing functions)\&. They give
access to the character string matched by the regular expression
associated with the semantic action\&. These functions must be
applied to the argument 
.ft B
lexbuf
.ft R
, which, in the code generated by
.ft B
ocamllex
.ft R
, is bound to the lexer buffer passed to the parsing
function\&.
.PP

.I val lexeme 
: 
.B lexbuf -> string
.sp

.ft B
Lexing\&.lexeme lexbuf
.ft R
returns the string matched by
the regular expression\&.

.sp

.I val lexeme_char 
: 
.B lexbuf -> int -> char
.sp

.ft B
Lexing\&.lexeme_char lexbuf i
.ft R
returns character number 
.ft B
i
.ft R
in
the matched string\&.

.sp

.I val lexeme_start 
: 
.B lexbuf -> int
.sp

.ft B
Lexing\&.lexeme_start lexbuf
.ft R
returns the offset in the
input stream of the first character of the matched string\&.
The first character of the stream has offset 0\&.

.sp

.I val lexeme_end 
: 
.B lexbuf -> int
.sp

.ft B
Lexing\&.lexeme_end lexbuf
.ft R
returns the offset in the input stream
of the character following the last character of the matched
string\&. The first character of the stream has offset 0\&.

.sp

.I val lexeme_start_p 
: 
.B lexbuf -> position
.sp
Like 
.ft B
lexeme_start
.ft R
, but return a complete 
.ft B
position
.ft R
instead
of an offset\&.  When position tracking is disabled, the function
returns 
.ft B
dummy_pos
.ft R
\&.

.sp

.I val lexeme_end_p 
: 
.B lexbuf -> position
.sp
Like 
.ft B
lexeme_end
.ft R
, but return a complete 
.ft B
position
.ft R
instead
of an offset\&.  When position tracking is disabled, the function
returns 
.ft B
dummy_pos
.ft R
\&.

.sp

.I val new_line 
: 
.B lexbuf -> unit
.sp
Update the 
.ft B
lex_curr_p
.ft R
field of the lexbuf to reflect the start
of a new line\&.  You can call this function in the semantic action
of the rule that matches the end\-of\-line character\&.  The function
does nothing when position tracking is disabled\&.

.sp
.B "Since"
3.11

.sp

.PP
.SS Miscellaneous functions

.PP

.I val flush_input 
: 
.B lexbuf -> unit
.sp
Discard the contents of the buffer and reset the current
position to 0\&.  The next use of the lexbuf will trigger a
refill\&.

.sp