/* GNU Emacs routines to deal with syntax tables; also word and list parsing.
- Copyright (C) 1985, 1987, 1993, 1994, 1995 Free Software Foundation, Inc.
+ Copyright (C) 1985, 87, 93, 94, 95, 97, 1998 Free Software Foundation, Inc.
This file is part of GNU Emacs.
#include "commands.h"
#include "buffer.h"
#include "charset.h"
+
+/* Make syntax table lookup grant data in gl_state. */
+#define SYNTAX_ENTRY_VIA_PROPERTY
+
#include "syntax.h"
+#include "intervals.h"
+
+/* We use these constants in place for comment-style and
+ string-ender-char to distinguish comments/strings started by
+ comment_fence and string_fence codes. */
+
+#define ST_COMMENT_STYLE (256 + 1)
+#define ST_STRING_STYLE (256 + 2)
#include "category.h"
Lisp_Object Qsyntax_table_p, Qsyntax_table, Qscan_error;
-static void scan_sexps_forward ();
-static int char_quoted ();
-
int words_include_escapes;
+int parse_sexp_lookup_properties;
/* Used as a temporary in SYNTAX_ENTRY and other macros in syntax.h,
if not compiled with GCC. No need to mark it, since it is used
struct lisp_parse_state
{
- int depth; /* Depth at end of parsing */
- int instring; /* -1 if not within string, else desired terminator. */
- int incomment; /* Nonzero if within a comment at end of parsing */
- int comstyle; /* comment style a=0, or b=1 */
+ int depth; /* Depth at end of parsing. */
+ int instring; /* -1 if not within string, else desired terminator. */
+ int incomment; /* Nonzero if within a comment at end of parsing. */
+ int comstyle; /* comment style a=0, or b=1, or ST_COMMENT_STYLE. */
int quoted; /* Nonzero if just after an escape char at end of parsing */
int thislevelstart; /* Char number of most recent start-of-expression at current level */
int prevlevelstart; /* Char number of start of containing expression */
- int location; /* Char number at which parsing stopped. */
+ int location; /* Char number at which parsing stopped. */
int mindepth; /* Minimum depth seen while scanning. */
- int comstart; /* Position just after last comment starter. */
+ int comstr_start; /* Position just after last comment/string starter. */
};
\f
/* These variables are a cache for finding the start of a defun.
find_start_pos is the place for which the defun start was found.
find_start_value is the defun start position found for it.
+ find_start_value_byte is the corresponding byte position.
find_start_buffer is the buffer it was found in.
find_start_begv is the BEGV value when it was found.
find_start_modiff is the value of MODIFF when it was found. */
static int find_start_pos;
static int find_start_value;
+static int find_start_value_byte;
static struct buffer *find_start_buffer;
static int find_start_begv;
static int find_start_modiff;
+
+static int find_defun_start P_ ((int, int));
+static int back_comment P_ ((int, int, int, int, int *, int *));
+static int char_quoted P_ ((int, int));
+static Lisp_Object skip_chars P_ ((int, int, Lisp_Object, Lisp_Object));
+static Lisp_Object scan_lists P_ ((int, int, int, int));
+static void scan_sexps_forward P_ ((struct lisp_parse_state *,
+ int, int, int, int,
+ int, Lisp_Object, int));
+\f
+
+struct gl_state_s gl_state; /* Global state of syntax parser. */
+
+INTERVAL interval_of ();
+#define INTERVALS_AT_ONCE 10 /* 1 + max-number of intervals
+ to scan to property-change. */
+
+/* Update gl_state to an appropriate interval which contains CHARPOS. The
+ sign of COUNT give the relative position of CHARPOS wrt the previously
+ valid interval. If INIT, only [be]_property fields of gl_state are
+ valid at start, the rest is filled basing on OBJECT.
+
+ `gl_state.*_i' are the intervals, and CHARPOS is further in the search
+ direction than the intervals - or in an interval. We update the
+ current syntax-table basing on the property of this interval, and
+ update the interval to start further than CHARPOS - or be
+ NULL_INTERVAL. We also update lim_property to be the next value of
+ charpos to call this subroutine again - or be before/after the
+ start/end of OBJECT. */
+
+void
+update_syntax_table (charpos, count, init, object)
+ int charpos, count, init;
+ Lisp_Object object;
+{
+ Lisp_Object tmp_table;
+ int cnt = 0, doing_extra = 0, invalidate = 1;
+ INTERVAL i, oldi;
+
+ if (init)
+ {
+ gl_state.start = gl_state.b_property;
+ gl_state.stop = gl_state.e_property;
+ gl_state.forward_i = interval_of (charpos, object);
+ i = gl_state.backward_i = gl_state.forward_i;
+ gl_state.left_ok = gl_state.right_ok = 1;
+ invalidate = 0;
+ if (NULL_INTERVAL_P (i))
+ return;
+ /* interval_of () updates only ->position of the return value,
+ update the parents manually to speed up update_interval. */
+ while (!NULL_PARENT (i))
+ {
+ if (AM_RIGHT_CHILD (i))
+ i->parent->position = i->position
+ - LEFT_TOTAL_LENGTH (i) + TOTAL_LENGTH (i) /* right end */
+ - TOTAL_LENGTH (i->parent)
+ + LEFT_TOTAL_LENGTH (i->parent);
+ else
+ i->parent->position = i->position - LEFT_TOTAL_LENGTH (i)
+ + TOTAL_LENGTH (i);
+ i = i->parent;
+ }
+ i = gl_state.forward_i;
+ gl_state.b_property = i->position - 1 - gl_state.offset;
+ gl_state.e_property = INTERVAL_LAST_POS (i) - gl_state.offset;
+ goto update;
+ }
+ oldi = i = count > 0 ? gl_state.forward_i : gl_state.backward_i;
+
+ /* We are guarantied to be called with CHARPOS either in i,
+ or further off. */
+ if (NULL_INTERVAL_P (i))
+ error ("Error in syntax_table logic for to-the-end intervals");
+ else if (charpos < i->position) /* Move left. */
+ {
+ if (count > 0)
+ error ("Error in syntax_table logic for intervals <-.");
+ /* Update the interval. */
+ i = update_interval (i, charpos);
+ if (oldi->position != INTERVAL_LAST_POS (i))
+ {
+ invalidate = 0;
+ gl_state.right_ok = 1; /* Invalidate the other end. */
+ gl_state.forward_i = i;
+ gl_state.e_property = INTERVAL_LAST_POS (i) - gl_state.offset;
+ }
+ }
+ else if (charpos >= INTERVAL_LAST_POS (i)) /* Move right. */
+ {
+ if (count < 0)
+ error ("Error in syntax_table logic for intervals ->.");
+ /* Update the interval. */
+ i = update_interval (i, charpos);
+ if (i->position != INTERVAL_LAST_POS (oldi))
+ {
+ invalidate = 0;
+ gl_state.left_ok = 1; /* Invalidate the other end. */
+ gl_state.backward_i = i;
+ gl_state.b_property = i->position - 1 - gl_state.offset;
+ }
+ }
+ else if (count > 0 ? gl_state.right_ok : gl_state.left_ok)
+ {
+ /* We do not need to recalculate tmp_table. */
+ tmp_table = gl_state.old_prop;
+ }
+
+ update:
+ tmp_table = textget (i->plist, Qsyntax_table);
+
+ if (invalidate)
+ invalidate = !EQ (tmp_table, gl_state.old_prop); /* Need to invalidate? */
+
+ if (invalidate) /* Did not get to adjacent interval. */
+ { /* with the same table => */
+ /* invalidate the old range. */
+ if (count > 0)
+ {
+ gl_state.backward_i = i;
+ gl_state.left_ok = 1; /* Invalidate the other end. */
+ gl_state.b_property = i->position - 1 - gl_state.offset;
+ }
+ else
+ {
+ gl_state.forward_i = i;
+ gl_state.right_ok = 1; /* Invalidate the other end. */
+ gl_state.e_property = INTERVAL_LAST_POS (i) - gl_state.offset;
+ }
+ }
+
+ gl_state.current_syntax_table = tmp_table;
+ gl_state.old_prop = tmp_table;
+ if (EQ (Fsyntax_table_p (tmp_table), Qt))
+ {
+ gl_state.use_global = 0;
+ }
+ else if (CONSP (tmp_table))
+ {
+ gl_state.use_global = 1;
+ gl_state.global_code = tmp_table;
+ }
+ else
+ {
+ gl_state.use_global = 0;
+ gl_state.current_syntax_table = current_buffer->syntax_table;
+ }
+
+ while (!NULL_INTERVAL_P (i))
+ {
+ if (cnt && !EQ (tmp_table, textget (i->plist, Qsyntax_table)))
+ {
+ if (count > 0)
+ gl_state.right_ok = 0;
+ else
+ gl_state.left_ok = 0;
+ break;
+ }
+ else if (cnt == INTERVALS_AT_ONCE)
+ {
+ if (count > 0)
+ gl_state.right_ok = 1;
+ else
+ gl_state.left_ok = 1;
+ break;
+ }
+ cnt++;
+ i = count > 0 ? next_interval (i) : previous_interval (i);
+ }
+ if (NULL_INTERVAL_P (i))
+ { /* This property goes to the end. */
+ if (count > 0)
+ gl_state.e_property = gl_state.stop;
+ else
+ gl_state.b_property = gl_state.start;
+ }
+ else
+ {
+ if (count > 0)
+ {
+ gl_state.e_property = i->position - gl_state.offset;
+ gl_state.forward_i = i;
+ }
+ else
+ {
+ gl_state.b_property = i->position + LENGTH (i) - 1 - gl_state.offset;
+ gl_state.backward_i = i;
+ }
+ }
+}
+\f
+/* Returns TRUE if char at CHARPOS is quoted.
+ Global syntax-table data should be set up already to be good at CHARPOS
+ or after. On return global syntax data is good for lookup at CHARPOS. */
+
+static int
+char_quoted (charpos, bytepos)
+ register int charpos, bytepos;
+{
+ register enum syntaxcode code;
+ register int beg = BEGV;
+ register int quoted = 0;
+ int orig = charpos;
+
+ DEC_BOTH (charpos, bytepos);
+
+ while (bytepos >= beg)
+ {
+ UPDATE_SYNTAX_TABLE_BACKWARD (charpos);
+ code = SYNTAX (FETCH_CHAR (bytepos));
+ if (! (code == Scharquote || code == Sescape))
+ break;
+
+ DEC_BOTH (charpos, bytepos);
+ quoted = !quoted;
+ }
+
+ UPDATE_SYNTAX_TABLE (orig);
+ return quoted;
+}
+
+/* Return the bytepos one character after BYTEPOS.
+ We assume that BYTEPOS is not at the end of the buffer. */
+
+INLINE int
+inc_bytepos (bytepos)
+ int bytepos;
+{
+ if (NILP (current_buffer->enable_multibyte_characters))
+ return bytepos + 1;
+
+ INC_POS (bytepos);
+ return bytepos;
+}
+
+/* Return the bytepos one character before BYTEPOS.
+ We assume that BYTEPOS is not at the start of the buffer. */
+
+INLINE int
+dec_bytepos (bytepos)
+ int bytepos;
+{
+ if (NILP (current_buffer->enable_multibyte_characters))
+ return bytepos - 1;
+
+ DEC_POS (bytepos);
+ return bytepos;
+}
+\f
/* Find a defun-start that is the last one before POS (or nearly the last).
We record what we find, so that another call in the same area
- can return the same value right away. */
+ can return the same value right away.
+
+ There is no promise at which position the global syntax data is
+ valid on return from the subroutine, so the caller should explicitly
+ update the global data. */
static int
-find_defun_start (pos)
- int pos;
+find_defun_start (pos, pos_byte)
+ int pos, pos_byte;
{
int tem;
int shortage;
+ int opoint = PT, opoint_byte = PT_BYTE;
/* Use previous finding, if it's valid and applies to this inquiry. */
if (current_buffer == find_start_buffer
return find_start_value;
/* Back up to start of line. */
- tem = scan_buffer ('\n', pos, BEGV, -1, &shortage, 1);
-
- while (tem > BEGV)
+ scan_newline (pos, pos_byte, BEGV, BEGV_BYTE, -1, 1);
+
+ /* We optimize syntax-table lookup for rare updates. Thus we accept
+ only those `^\s(' which are good in global _and_ text-property
+ syntax-tables. */
+ gl_state.current_syntax_table = current_buffer->syntax_table;
+ gl_state.use_global = 0;
+ while (PT > BEGV)
{
/* Open-paren at start of line means we found our defun-start. */
- if (SYNTAX (FETCH_CHAR (tem)) == Sopen)
- break;
+ if (SYNTAX (FETCH_CHAR (PT_BYTE)) == Sopen)
+ {
+ SETUP_SYNTAX_TABLE (PT + 1, -1); /* Try again... */
+ if (SYNTAX (FETCH_CHAR (PT_BYTE)) == Sopen)
+ break;
+ /* Now fallback to the default value. */
+ gl_state.current_syntax_table = current_buffer->syntax_table;
+ gl_state.use_global = 0;
+ }
/* Move to beg of previous line. */
- tem = scan_buffer ('\n', tem, BEGV, -2, &shortage, 1);
+ scan_newline (PT, PT_BYTE, BEGV, BEGV_BYTE, -2, 1);
}
/* Record what we found, for the next try. */
- find_start_value = tem;
+ find_start_value = PT;
+ find_start_value_byte = PT_BYTE;
find_start_buffer = current_buffer;
find_start_modiff = MODIFF;
find_start_begv = BEGV;
find_start_pos = pos;
+ TEMP_SET_PT_BOTH (opoint, opoint_byte);
+
return find_start_value;
}
\f
+/* Checks whether charpos FROM is at the end of a comment.
+ FROM_BYTE is the bytepos corresponding to FROM.
+ Do not move back before STOP.
+
+ Return a positive value if we find a comment ending at FROM/FROM_BYTE;
+ return -1 otherwise.
+
+ If successful, store the charpos of the comment's beginning
+ into *CHARPOS_PTR, and the bytepos into *BYTEPOS_PTR.
+
+ Global syntax data remains valid for backward search starting at
+ the returned value (or at FROM, if the search was not successful). */
+
+static int
+back_comment (from, from_byte, stop, comstyle, charpos_ptr, bytepos_ptr)
+ int from, from_byte, stop;
+ int comstyle;
+ int *charpos_ptr, *bytepos_ptr;
+{
+ /* Look back, counting the parity of string-quotes,
+ and recording the comment-starters seen.
+ When we reach a safe place, assume that's not in a string;
+ then step the main scan to the earliest comment-starter seen
+ an even number of string quotes away from the safe place.
+
+ OFROM[I] is position of the earliest comment-starter seen
+ which is I+2X quotes from the comment-end.
+ PARITY is current parity of quotes from the comment end. */
+ int parity = 0;
+ int my_stringend = 0;
+ int string_lossage = 0;
+ int comment_end = from;
+ int comment_end_byte = from_byte;
+ int comstart_pos = 0;
+ int comstart_byte;
+ /* Value that PARITY had, when we reached the position
+ in COMSTART_POS. */
+ int comstart_parity = 0;
+ int scanstart = from - 1;
+ /* Place where the containing defun starts,
+ or 0 if we didn't come across it yet. */
+ int defun_start = 0;
+ int defun_start_byte = 0;
+ register enum syntaxcode code;
+ int c;
+
+ /* At beginning of range to scan, we're outside of strings;
+ that determines quote parity to the comment-end. */
+ while (from != stop)
+ {
+ int temp_byte;
+
+ /* Move back and examine a character. */
+ DEC_BOTH (from, from_byte);
+ UPDATE_SYNTAX_TABLE_BACKWARD (from);
+
+ c = FETCH_CHAR (from_byte);
+ code = SYNTAX (c);
+
+ /* If this char is the second of a 2-char comment end sequence,
+ back up and give the pair the appropriate syntax. */
+ if (from > stop && SYNTAX_COMEND_SECOND (c)
+ && (temp_byte = dec_bytepos (from_byte),
+ SYNTAX_COMEND_FIRST (FETCH_CHAR (temp_byte))))
+ {
+ code = Sendcomment;
+ DEC_BOTH (from, from_byte);
+ /* This is apparently the best we can do: */
+ UPDATE_SYNTAX_TABLE_BACKWARD (from);
+ c = FETCH_CHAR (from_byte);
+ }
+
+ /* If this char starts a 2-char comment start sequence,
+ treat it like a 1-char comment starter. */
+ if (from < scanstart && SYNTAX_COMSTART_FIRST (c)
+ && (temp_byte = inc_bytepos (from_byte),
+ (SYNTAX_COMSTART_SECOND (FETCH_CHAR (temp_byte))
+ && comstyle == SYNTAX_COMMENT_STYLE (FETCH_CHAR (temp_byte)))))
+ code = Scomment;
+
+ /* Ignore escaped characters, except comment-enders. */
+ if (code != Sendcomment && char_quoted (from, from_byte))
+ continue;
+
+ /* Track parity of quotes. */
+ if (code == Sstring)
+ {
+ parity ^= 1;
+ if (my_stringend == 0)
+ my_stringend = c;
+ /* If we have two kinds of string delimiters.
+ There's no way to grok this scanning backwards. */
+ else if (my_stringend != c)
+ string_lossage = 1;
+ }
+
+ if (code == Sstring_fence || code == Scomment_fence)
+ {
+ parity ^= 1;
+ if (my_stringend == 0)
+ my_stringend
+ = code == Sstring_fence ? ST_STRING_STYLE : ST_COMMENT_STYLE;
+ /* If we have two kinds of string delimiters.
+ There's no way to grok this scanning backwards. */
+ else if (my_stringend != (code == Sstring_fence
+ ? ST_STRING_STYLE : ST_COMMENT_STYLE))
+ string_lossage = 1;
+ }
+
+ /* Record comment-starters according to that
+ quote-parity to the comment-end. */
+ if (code == Scomment)
+ {
+ comstart_parity = parity;
+ comstart_pos = from;
+ comstart_byte = from_byte;
+ }
+
+ /* If we find another earlier comment-ender,
+ any comment-starts earlier than that don't count
+ (because they go with the earlier comment-ender). */
+ if (code == Sendcomment
+ && SYNTAX_COMMENT_STYLE (FETCH_CHAR (from_byte)) == comstyle)
+ break;
+
+ /* Assume a defun-start point is outside of strings. */
+ if (code == Sopen
+ && (from == stop
+ || (temp_byte = dec_bytepos (from_byte),
+ FETCH_CHAR (temp_byte) == '\n')))
+ {
+ defun_start = from;
+ defun_start_byte = from_byte;
+ break;
+ }
+ }
+
+ if (comstart_pos == 0)
+ {
+ from = comment_end;
+ from_byte = comment_end_byte;
+ UPDATE_SYNTAX_TABLE_FORWARD (comment_end - 1);
+ }
+ /* If the earliest comment starter
+ is followed by uniform paired string quotes or none,
+ we know it can't be inside a string
+ since if it were then the comment ender would be inside one.
+ So it does start a comment. Skip back to it. */
+ else if (comstart_parity == 0 && !string_lossage)
+ {
+ from = comstart_pos;
+ from_byte = comstart_byte;
+ /* Globals are correct now. */
+ }
+ else
+ {
+ /* We had two kinds of string delimiters mixed up
+ together. Decode this going forwards.
+ Scan fwd from the previous comment ender
+ to the one in question; this records where we
+ last passed a comment starter. */
+ struct lisp_parse_state state;
+ /* If we did not already find the defun start, find it now. */
+ if (defun_start == 0)
+ {
+ defun_start = find_defun_start (comment_end, comment_end_byte);
+ defun_start_byte = find_start_value_byte;
+ }
+ scan_sexps_forward (&state,
+ defun_start, defun_start_byte,
+ comment_end - 1, -10000, 0, Qnil, 0);
+ if (state.incomment)
+ {
+ /* scan_sexps_forward changed the direction of search in
+ global variables, so we need to update it completely. */
+
+ from = state.comstr_start;
+ }
+ else
+ {
+ from = comment_end;
+ }
+ from_byte = CHAR_TO_BYTE (from);
+ UPDATE_SYNTAX_TABLE_FORWARD (from - 1);
+ }
+
+ *charpos_ptr = from;
+ *bytepos_ptr = from_byte;
+
+ return from;
+}
+\f
DEFUN ("syntax-table-p", Fsyntax_table_p, Ssyntax_table_p, 1, 1, 0,
"Return t if OBJECT is a syntax table.\n\
Currently, any char-table counts as a syntax table.")
Lisp_Object object;
{
if (CHAR_TABLE_P (object)
- && XCHAR_TABLE (object)->purpose == Qsyntax_table)
+ && EQ (XCHAR_TABLE (object)->purpose, Qsyntax_table))
return Qt;
return Qnil;
}
Lisp_Object obj;
{
if (!(CHAR_TABLE_P (obj)
- && XCHAR_TABLE (obj)->purpose == Qsyntax_table))
+ && EQ (XCHAR_TABLE (obj)->purpose, Qsyntax_table)))
wrong_type_argument (Qsyntax_table_p, obj);
}
\f
/* Convert a letter which signifies a syntax code
into the code it signifies.
- This is used by modify-syntax-entry, and other things. */
+ This is used by modify-syntax-entry, and other things. */
unsigned char syntax_spec_code[0400] =
{ 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377,
0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377,
0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377,
0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377,
- (char) Swhitespace, 0377, (char) Sstring, 0377,
+ (char) Swhitespace, (char) Scomment_fence, (char) Sstring, 0377,
(char) Smath, 0377, 0377, (char) Squote,
(char) Sopen, (char) Sclose, 0377, 0377,
0377, (char) Swhitespace, (char) Spunct, (char) Scharquote,
0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377, /* `, a, ... */
0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377,
0377, 0377, 0377, 0377, 0377, 0377, 0377, (char) Sword,
- 0377, 0377, 0377, 0377, 0377, 0377, 0377, 0377
+ 0377, 0377, 0377, 0377, (char) Sstring_fence, 0377, 0377, 0377
};
-/* Indexed by syntax code, give the letter that describes it. */
+/* Indexed by syntax code, give the letter that describes it. */
-char syntax_code_spec[14] =
+char syntax_code_spec[16] =
{
- ' ', '.', 'w', '_', '(', ')', '\'', '\"', '$', '\\', '/', '<', '>', '@'
+ ' ', '.', 'w', '_', '(', ')', '\'', '\"', '$', '\\', '/', '<', '>', '@',
+ '!', '|'
};
/* Indexed by syntax code, give the object (cons of syntax code and
Lisp_Object character;
{
int char_int;
+ gl_state.current_syntax_table = current_buffer->syntax_table;
+
+ gl_state.use_global = 0;
CHECK_NUMBER (character, 0);
char_int = XINT (character);
return make_number (syntax_code_spec[(int) SYNTAX (char_int)]);
Lisp_Object character;
{
int char_int, code;
+ gl_state.current_syntax_table = current_buffer->syntax_table;
+ gl_state.use_global = 0;
CHECK_NUMBER (character, 0);
char_int = XINT (character);
code = SYNTAX (char_int);
if (code == Sopen || code == Sclose)
- return make_number (SYNTAX_MATCH (char_int));
+ return SYNTAX_MATCH (char_int);
return Qnil;
}
3 means CHAR is the start of a two-char comment end sequence.\n\
4 means CHAR is the second character of such a sequence.\n\
\n\
-There can be up to two orthogonal comment sequences. This is to support\n\
+There can be up to two orthogonal comment sequences. This is to support\n\
language modes such as C++. By default, all comment sequences are of style\n\
a, but you can set the comment sequence style to b (on the second character\n\
of a comment-start, or the first character of a comment-end sequence) using\n\
p = XSTRING (newentry)->data;
code = (enum syntaxcode) syntax_spec_code[*p++];
if (((int) code & 0377) == 0377)
- error ("invalid syntax description letter: %c", c);
+ error ("invalid syntax description letter: %c", p[-1]);
if (code == Sinherit)
{
- SET_RAW_SYNTAX_ENTRY (syntax_table, c, Qnil);
+ SET_RAW_SYNTAX_ENTRY (syntax_table, XINT (c), Qnil);
return Qnil;
}
/* Since we can't use a shared object, let's make a new one. */
newentry = Fcons (make_number (val), match);
- SET_RAW_SYNTAX_ENTRY (syntax_table, c, newentry);
+ SET_RAW_SYNTAX_ENTRY (syntax_table, XINT (c), newentry);
return Qnil;
}
return;
}
- code = (enum syntaxcode) (first & 0377);
+ code = (enum syntaxcode) (XINT (first) & 0377);
start1 = (XINT (first) >> 16) & 1;
start2 = (XINT (first) >> 17) & 1;
end1 = (XINT (first) >> 18) & 1;
{
struct buffer *old = current_buffer;
set_buffer_internal (XBUFFER (Vstandard_output));
- describe_vector (vector, Qnil, describe_syntax, 0, Qnil, Qnil);
+ describe_vector (vector, Qnil, describe_syntax, 0, Qnil, Qnil, (int *) 0, 0);
while (! NILP (XCHAR_TABLE (vector)->parent))
{
vector = XCHAR_TABLE (vector)->parent;
insert_string ("\nThe parent syntax table is:");
- describe_vector (vector, Qnil, describe_syntax, 0, Qnil, Qnil);
+ describe_vector (vector, Qnil, describe_syntax, 0, Qnil, Qnil,
+ (int *) 0, 0);
}
call0 (intern ("help-mode"));
return Qnil;
}
\f
+int parse_sexp_ignore_comments;
+
/* Return the position across COUNT words from FROM.
If that many words cannot be found before the end of the buffer, return 0.
COUNT negative means scan backward and stop at word beginning. */
+int
scan_words (from, count)
register int from, count;
{
register int beg = BEGV;
register int end = ZV;
+ register int from_byte = CHAR_TO_BYTE (from);
register enum syntaxcode code;
int ch0, ch1;
- int temp_pos;
immediate_quit = 1;
QUIT;
+ SETUP_SYNTAX_TABLE (from, count);
+
while (count > 0)
{
while (1)
immediate_quit = 0;
return 0;
}
- ch0 = FETCH_CHAR (from);
+ UPDATE_SYNTAX_TABLE_FORWARD (from);
+ ch0 = FETCH_CHAR (from_byte);
code = SYNTAX (ch0);
- INC_POS (from);
+ INC_BOTH (from, from_byte);
if (words_include_escapes
&& (code == Sescape || code == Scharquote))
break;
while (1)
{
if (from == end) break;
- ch1 = FETCH_CHAR (from);
+ UPDATE_SYNTAX_TABLE_FORWARD (from);
+ ch1 = FETCH_CHAR (from_byte);
code = SYNTAX (ch1);
if (!(words_include_escapes
&& (code == Sescape || code == Scharquote)))
if (code != Sword || WORD_BOUNDARY_P (ch0, ch1))
break;
- INC_POS (from);
+ INC_BOTH (from, from_byte);
ch0 = ch1;
}
count--;
immediate_quit = 0;
return 0;
}
- DEC_POS (from);
- ch1 = FETCH_CHAR (from);
+ DEC_BOTH (from, from_byte);
+ UPDATE_SYNTAX_TABLE_BACKWARD (from);
+ ch1 = FETCH_CHAR (from_byte);
code = SYNTAX (ch1);
if (words_include_escapes
&& (code == Sescape || code == Scharquote))
position of it. */
while (1)
{
- if (from == beg) break;
- temp_pos = from;
- DEC_POS (temp_pos);
- ch0 = FETCH_CHAR (temp_pos);
+ int temp_byte;
+
+ if (from == beg)
+ break;
+ temp_byte = dec_bytepos (from_byte);
+ UPDATE_SYNTAX_TABLE_BACKWARD (from);
+ ch0 = FETCH_CHAR (temp_byte);
code = SYNTAX (ch0);
if (!(words_include_escapes
&& (code == Sescape || code == Scharquote)))
if (code != Sword || WORD_BOUNDARY_P (ch0, ch1))
break;
- from = temp_pos;
+ DEC_BOTH (from, from_byte);
ch1 = ch0;
}
count++;
return Qt;
}
\f
+Lisp_Object skip_chars ();
+
+DEFUN ("skip-chars-forward", Fskip_chars_forward, Sskip_chars_forward, 1, 2, 0,
+ "Move point forward, stopping before a char not in STRING, or at pos LIM.\n\
+STRING is like the inside of a `[...]' in a regular expression\n\
+except that `]' is never special and `\\' quotes `^', `-' or `\\'.\n\
+Thus, with arg \"a-zA-Z\", this skips letters stopping before first nonletter.\n\
+With arg \"^a-zA-Z\", skips nonletters stopping before first letter.\n\
+Returns the distance traveled, either zero or positive.")
+ (string, lim)
+ Lisp_Object string, lim;
+{
+ return skip_chars (1, 0, string, lim);
+}
+
+DEFUN ("skip-chars-backward", Fskip_chars_backward, Sskip_chars_backward, 1, 2, 0,
+ "Move point backward, stopping after a char not in STRING, or at pos LIM.\n\
+See `skip-chars-forward' for details.\n\
+Returns the distance traveled, either zero or negative.")
+ (string, lim)
+ Lisp_Object string, lim;
+{
+ return skip_chars (0, 0, string, lim);
+}
+
+DEFUN ("skip-syntax-forward", Fskip_syntax_forward, Sskip_syntax_forward, 1, 2, 0,
+ "Move point forward across chars in specified syntax classes.\n\
+SYNTAX is a string of syntax code characters.\n\
+Stop before a char whose syntax is not in SYNTAX, or at position LIM.\n\
+If SYNTAX starts with ^, skip characters whose syntax is NOT in SYNTAX.\n\
+This function returns the distance traveled, either zero or positive.")
+ (syntax, lim)
+ Lisp_Object syntax, lim;
+{
+ return skip_chars (1, 1, syntax, lim);
+}
+
+DEFUN ("skip-syntax-backward", Fskip_syntax_backward, Sskip_syntax_backward, 1, 2, 0,
+ "Move point backward across chars in specified syntax classes.\n\
+SYNTAX is a string of syntax code characters.\n\
+Stop on reaching a char whose syntax is not in SYNTAX, or at position LIM.\n\
+If SYNTAX starts with ^, skip characters whose syntax is NOT in SYNTAX.\n\
+This function returns the distance traveled, either zero or negative.")
+ (syntax, lim)
+ Lisp_Object syntax, lim;
+{
+ return skip_chars (0, 1, syntax, lim);
+}
+
+static Lisp_Object
+skip_chars (forwardp, syntaxp, string, lim)
+ int forwardp, syntaxp;
+ Lisp_Object string, lim;
+{
+ register unsigned char *p, *pend;
+ register unsigned int c;
+ register int ch;
+ unsigned char fastmap[0400];
+ /* If SYNTAXP is 0, STRING may contain multi-byte form of characters
+ of which codes don't fit in FASTMAP. In that case, we set the
+ first byte of multibyte form (i.e. base leading-code) in FASTMAP
+ and set the actual ranges of characters in CHAR_RANGES. In the
+ form "X-Y" of STRING, both X and Y must belong to the same
+ character set because a range striding across character sets is
+ meaningless. */
+ int *char_ranges;
+ int n_char_ranges = 0;
+ int negate = 0;
+ register int i, i_byte;
+ int multibyte = !NILP (current_buffer->enable_multibyte_characters);
+ int string_multibyte = STRING_MULTIBYTE (string);
+
+ CHECK_STRING (string, 0);
+ char_ranges = (int *) alloca (XSTRING (string)->size * (sizeof (int)) * 2);
+
+ if (NILP (lim))
+ XSETINT (lim, forwardp ? ZV : BEGV);
+ else
+ CHECK_NUMBER_COERCE_MARKER (lim, 0);
+
+ /* In any case, don't allow scan outside bounds of buffer. */
+ if (XINT (lim) > ZV)
+ XSETFASTINT (lim, ZV);
+ if (XINT (lim) < BEGV)
+ XSETFASTINT (lim, BEGV);
+
+ bzero (fastmap, sizeof fastmap);
+
+ i = 0, i_byte = 0;
+
+ if (i < XSTRING (string)->size
+ && XSTRING (string)->data[0] == '^')
+ {
+ negate = 1; i++, i_byte++;
+ }
+
+ /* Find the characters specified and set their elements of fastmap.
+ If syntaxp, each character counts as itself.
+ Otherwise, handle backslashes and ranges specially. */
+
+ while (i < XSTRING (string)->size)
+ {
+ int c_leading_code;
+
+ if (string_multibyte)
+ {
+ c_leading_code = XSTRING (string)->data[i_byte];
+ FETCH_STRING_CHAR_ADVANCE (c, string, i, i_byte);
+ }
+ else
+ c = c_leading_code = XSTRING (string)->data[i++];
+
+ /* Convert multibyteness between what the string has
+ and what the buffer has. */
+ if (multibyte)
+ c = unibyte_char_to_multibyte (c);
+ else
+ c &= 0377;
+
+ if (syntaxp)
+ fastmap[syntax_spec_code[c & 0377]] = 1;
+ else
+ {
+ if (c == '\\')
+ {
+ if (i == XSTRING (string)->size)
+ break;
+
+ if (string_multibyte)
+ FETCH_STRING_CHAR_ADVANCE (c, string, i, i_byte);
+ else
+ c = XSTRING (string)->data[i++];
+ }
+ if (i < XSTRING (string)->size && XSTRING (string)->data[i] == '-')
+ {
+ unsigned int c2;
+
+ /* Skip over the dash. */
+ i++, i_byte++;
+
+ if (i == XSTRING (string)->size)
+ break;
+
+ /* Get the end of the range. */
+ if (string_multibyte)
+ FETCH_STRING_CHAR_ADVANCE (c2, string, i, i_byte);
+ else
+ c2 = XSTRING (string)->data[i++];
+
+ if (SINGLE_BYTE_CHAR_P (c))
+ while (c <= c2)
+ {
+ fastmap[c] = 1;
+ c++;
+ }
+ else
+ {
+ fastmap[c_leading_code] = 1;
+ if (c <= c2)
+ {
+ char_ranges[n_char_ranges++] = c;
+ char_ranges[n_char_ranges++] = c2;
+ }
+ }
+ }
+ else
+ {
+ fastmap[c_leading_code] = 1;
+ if (!SINGLE_BYTE_CHAR_P (c))
+ {
+ char_ranges[n_char_ranges++] = c;
+ char_ranges[n_char_ranges++] = c;
+ }
+ }
+ }
+ }
+
+ /* If ^ was the first character, complement the fastmap. In
+ addition, as all multibyte characters have possibility of
+ matching, set all entries for base leading codes, which is
+ harmless even if SYNTAXP is 1. */
+
+ if (negate)
+ for (i = 0; i < sizeof fastmap; i++)
+ {
+ if (!multibyte || !BASE_LEADING_CODE_P (i))
+ fastmap[i] ^= 1;
+ else
+ fastmap[i] = 1;
+ }
+
+ {
+ int start_point = PT;
+ int pos = PT;
+ int pos_byte = PT_BYTE;
+
+ immediate_quit = 1;
+ if (syntaxp)
+ {
+ SETUP_SYNTAX_TABLE (pos, forwardp ? 1 : -1);
+ if (forwardp)
+ {
+ if (multibyte)
+ {
+ if (pos < XINT (lim))
+ while (fastmap[(int) SYNTAX (FETCH_CHAR (pos_byte))])
+ {
+ /* Since we already checked for multibyteness,
+ avoid using INC_BOTH which checks again. */
+ INC_POS (pos_byte);
+ pos++;
+ if (pos >= XINT (lim))
+ break;
+ UPDATE_SYNTAX_TABLE_FORWARD (pos);
+ }
+ }
+ else
+ {
+ while (pos < XINT (lim)
+ && fastmap[(int) SYNTAX (FETCH_BYTE (pos))])
+ {
+ pos++;
+ UPDATE_SYNTAX_TABLE_FORWARD (pos);
+ }
+ }
+ }
+ else
+ {
+ if (multibyte)
+ {
+ while (pos > XINT (lim))
+ {
+ int savepos = pos_byte;
+ /* Since we already checked for multibyteness,
+ avoid using DEC_BOTH which checks again. */
+ pos--;
+ DEC_POS (pos_byte);
+ UPDATE_SYNTAX_TABLE_BACKWARD (pos);
+ if (!fastmap[(int) SYNTAX (FETCH_CHAR (pos_byte))])
+ {
+ pos++;
+ pos_byte = savepos;
+ break;
+ }
+ }
+ }
+ else
+ {
+ if (pos > XINT (lim))
+ while (fastmap[(int) SYNTAX (FETCH_BYTE (pos - 1))])
+ {
+ pos--;
+ if (pos <= XINT (lim))
+ break;
+ UPDATE_SYNTAX_TABLE_BACKWARD (pos - 1);
+ }
+ }
+ }
+ }
+ else
+ {
+ if (forwardp)
+ {
+ if (multibyte)
+ while (pos < XINT (lim) && fastmap[(c = FETCH_BYTE (pos_byte))])
+ {
+ if (!BASE_LEADING_CODE_P (c))
+ INC_BOTH (pos, pos_byte);
+ else if (n_char_ranges)
+ {
+ /* We much check CHAR_RANGES for a multibyte
+ character. */
+ ch = FETCH_MULTIBYTE_CHAR (pos_byte);
+ for (i = 0; i < n_char_ranges; i += 2)
+ if ((ch >= char_ranges[i] && ch <= char_ranges[i + 1]))
+ break;
+ if (!(negate ^ (i < n_char_ranges)))
+ break;
+
+ INC_BOTH (pos, pos_byte);
+ }
+ else
+ {
+ if (!negate) break;
+ INC_BOTH (pos, pos_byte);
+ }
+ }
+ else
+ while (pos < XINT (lim) && fastmap[FETCH_BYTE (pos)])
+ pos++;
+ }
+ else
+ {
+ if (multibyte)
+ while (pos > XINT (lim))
+ {
+ int savepos = pos_byte;
+ DEC_BOTH (pos, pos_byte);
+ if (fastmap[(c = FETCH_BYTE (pos_byte))])
+ {
+ if (!BASE_LEADING_CODE_P (c))
+ ;
+ else if (n_char_ranges)
+ {
+ /* We much check CHAR_RANGES for a multibyte
+ character. */
+ ch = FETCH_MULTIBYTE_CHAR (pos_byte);
+ for (i = 0; i < n_char_ranges; i += 2)
+ if (ch >= char_ranges[i] && ch <= char_ranges[i + 1])
+ break;
+ if (!(negate ^ (i < n_char_ranges)))
+ {
+ pos++;
+ pos_byte = savepos;
+ break;
+ }
+ }
+ else
+ if (!negate)
+ {
+ pos++;
+ pos_byte = savepos;
+ break;
+ }
+ }
+ else
+ {
+ pos++;
+ pos_byte = savepos;
+ break;
+ }
+ }
+ else
+ while (pos > XINT (lim) && fastmap[FETCH_BYTE (pos - 1)])
+ pos--;
+ }
+ }
+
+#if 0 /* Not needed now that a position in mid-character
+ cannot be specified in Lisp. */
+ if (multibyte
+ /* INC_POS or DEC_POS might have moved POS over LIM. */
+ && (forwardp ? (pos > XINT (lim)) : (pos < XINT (lim))))
+ pos = XINT (lim);
+#endif
+
+ if (! multibyte)
+ pos_byte = pos;
+
+ SET_PT_BOTH (pos, pos_byte);
+ immediate_quit = 0;
+
+ return make_number (PT - start_point);
+ }
+}
+\f
DEFUN ("forward-comment", Fforward_comment, Sforward_comment, 1, 1, 0,
"Move forward across up to N comments. If N is negative, move backward.\n\
Stop scanning if we find something other than a comment or whitespace.\n\
Lisp_Object count;
{
register int from;
+ int from_byte;
register int stop;
register int c, c1;
register enum syntaxcode code;
int found;
int count1;
int temp_pos;
+ int out_charpos, out_bytepos;
CHECK_NUMBER (count, 0);
count1 = XINT (count);
+ stop = count1 > 0 ? ZV : BEGV;
immediate_quit = 1;
QUIT;
from = PT;
+ from_byte = PT_BYTE;
+ SETUP_SYNTAX_TABLE (from, count1);
while (count1 > 0)
{
- stop = ZV;
do
{
if (from == stop)
{
- SET_PT (from);
+ SET_PT_BOTH (from, from_byte);
immediate_quit = 0;
return Qnil;
}
- c = FETCH_CHAR (from);
+ UPDATE_SYNTAX_TABLE_FORWARD (from);
+ c = FETCH_CHAR (from_byte);
code = SYNTAX (c);
- INC_POS (from);
+ INC_BOTH (from, from_byte);
comstyle = 0;
if (from < stop && SYNTAX_COMSTART_FIRST (c)
- && (c1 = FETCH_CHAR (from),
+ && (c1 = FETCH_CHAR (from_byte),
SYNTAX_COMSTART_SECOND (c1)))
{
/* We have encountered a comment start sequence and we
the comment section. */
code = Scomment;
comstyle = SYNTAX_COMMENT_STYLE (c1);
- INC_POS (from);
+ INC_BOTH (from, from_byte);
}
}
while (code == Swhitespace || code == Sendcomment);
- if (code != Scomment)
+
+ if (code != Scomment && code != Scomment_fence)
{
immediate_quit = 0;
- DEC_POS (from);
- SET_PT (from);
+ DEC_BOTH (from, from_byte);
+ SET_PT_BOTH (from, from_byte);
return Qnil;
}
/* We're at the start of a comment. */
if (from == stop)
{
immediate_quit = 0;
- SET_PT (from);
+ SET_PT_BOTH (from, from_byte);
return Qnil;
}
- c = FETCH_CHAR (from);
- INC_POS (from);
+ UPDATE_SYNTAX_TABLE_FORWARD (from);
+ c = FETCH_CHAR (from_byte);
+ INC_BOTH (from, from_byte);
if (SYNTAX (c) == Sendcomment
&& SYNTAX_COMMENT_STYLE (c) == comstyle)
/* we have encountered a comment end of the same style
as the comment sequence which began this comment
section */
break;
+ if (SYNTAX (c) == Scomment_fence
+ && comstyle == ST_COMMENT_STYLE)
+ /* we have encountered a comment end of the same style
+ as the comment sequence which began this comment
+ section. */
+ break;
if (from < stop && SYNTAX_COMEND_FIRST (c)
- && (c1 = FETCH_CHAR (from),
+ && (c1 = FETCH_CHAR (from_byte),
SYNTAX_COMEND_SECOND (c1))
&& SYNTAX_COMMENT_STYLE (c) == comstyle)
/* we have encountered a comment end of the same style
as the comment sequence which began this comment
section */
- { INC_POS (from); break; }
+ {
+ INC_BOTH (from, from_byte);
+ break;
+ }
}
/* We have skipped one comment. */
count1--;
while (count1 < 0)
{
- stop = BEGV;
- while (from > stop)
+ while (1)
{
int quoted;
+ if (from <= stop)
+ {
+ SET_PT_BOTH (BEGV, BEGV_BYTE);
+ immediate_quit = 0;
+ return Qnil;
+ }
- DEC_POS (from);
- quoted = char_quoted (from);
+ DEC_BOTH (from, from_byte);
+ quoted = char_quoted (from, from_byte);
if (quoted)
- DEC_POS (from);
- c = FETCH_CHAR (from);
+ {
+ DEC_BOTH (from, from_byte);
+ goto leave;
+ }
+ UPDATE_SYNTAX_TABLE_BACKWARD (from);
+ c = FETCH_CHAR (from_byte);
code = SYNTAX (c);
comstyle = 0;
if (code == Sendcomment)
comstyle = SYNTAX_COMMENT_STYLE (c);
- temp_pos = from;
- DEC_POS (temp_pos);
+ temp_pos = dec_bytepos (from_byte);
if (from > stop && SYNTAX_COMEND_SECOND (c)
&& (c1 = FETCH_CHAR (temp_pos),
SYNTAX_COMEND_FIRST (c1))
- && !char_quoted (temp_pos))
+ && !char_quoted (from - 1, temp_pos))
{
/* We must record the comment style encountered so that
later, we can match only the proper comment begin
sequence of the same style. */
code = Sendcomment;
comstyle = SYNTAX_COMMENT_STYLE (c1);
- from = temp_pos;
+ DEC_BOTH (from, from_byte);
+ }
+ if (from > stop && SYNTAX_COMSTART_SECOND (c)
+ && (c1 = FETCH_CHAR (temp_pos),
+ SYNTAX_COMSTART_FIRST (c1))
+ && !char_quoted (from - 1, temp_pos))
+ {
+ /* We must record the comment style encountered so that
+ later, we can match only the proper comment begin
+ sequence of the same style. */
+ code = Scomment;
+ DEC_BOTH (from, from_byte);
}
- if (code == Sendcomment && !quoted)
+ if (code == Scomment_fence)
{
-#if 0
- if (code != SYNTAX (c))
- /* For a two-char comment ender, we can assume
- it does end a comment. So scan back in a simple way. */
+ /* Skip until first preceding unquoted comment_fence. */
+ int found = 0, ini = from, ini_byte = from_byte;
+
+ while (1)
{
- if (from != stop) DEC_POS (from);
- while (1)
+ DEC_BOTH (from, from_byte);
+ if (from == stop)
+ break;
+ UPDATE_SYNTAX_TABLE_BACKWARD (from);
+ c = FETCH_CHAR (from_byte);
+ if (SYNTAX (c) == Scomment_fence
+ && !char_quoted (from, from_byte))
{
- if ((c = FETCH_CHAR (from),
- SYNTAX (c) == Scomment)
- && SYNTAX_COMMENT_STYLE (c) == comstyle)
- break;
- if (from == stop)
- {
- immediate_quit = 0;
- SET_PT (from);
- return Qnil;
- }
- DEC_POS (from);
- if (SYNTAX_COMSTART_SECOND (c)
- && (c1 = FETCH_CHAR (from),
- SYNTAX_COMSTART_FIRST (c1))
- && SYNTAX_COMMENT_STYLE (c) == comstyle
- && !char_quoted (from))
- break;
+ found = 1;
+ break;
}
- break;
}
-#endif /* 0 */
-
- /* Look back, counting the parity of string-quotes,
- and recording the comment-starters seen.
- When we reach a safe place, assume that's not in a string;
- then step the main scan to the earliest comment-starter seen
- an even number of string quotes away from the safe place.
-
- OFROM[I] is position of the earliest comment-starter seen
- which is I+2X quotes from the comment-end.
- PARITY is current parity of quotes from the comment end. */
- {
- int parity = 0;
- char my_stringend = 0;
- int string_lossage = 0;
- int comment_end = from;
- int comstart_pos = 0;
- int comstart_parity = 0;
- int scanstart = from;
-
- DEC_POS (scanstart);
- /* At beginning of range to scan, we're outside of strings;
- that determines quote parity to the comment-end. */
- while (from != stop)
- {
- /* Move back and examine a character. */
- DEC_POS (from);
-
- c = FETCH_CHAR (from);
- code = SYNTAX (c);
-
- /* If this char is the second of a 2-char comment sequence,
- back up and give the pair the appropriate syntax. */
- temp_pos = from;
- DEC_POS (temp_pos);
- if (from > stop && SYNTAX_COMEND_SECOND (c)
- && (c1 = FETCH_CHAR (temp_pos),
- SYNTAX_COMEND_FIRST (c1)))
- {
- code = Sendcomment;
- from = temp_pos;
- c = c1;
- }
-
- temp_pos = from;
- INC_POS (temp_pos);
- /* If this char starts a 2-char comment start sequence,
- treat it like a 1-char comment starter. */
- if (from < scanstart && SYNTAX_COMSTART_FIRST (c)
- && (c1 = FETCH_CHAR (temp_pos),
- SYNTAX_COMSTART_SECOND (c1))
- && comstyle == SYNTAX_COMMENT_STYLE (c1))
- code = Scomment;
-
- /* Ignore escaped characters. */
- if (char_quoted (from))
- continue;
-
- /* Track parity of quotes. */
- if (code == Sstring)
- {
- parity ^= 1;
- if (my_stringend == 0)
- my_stringend = c;
- /* If we have two kinds of string delimiters.
- There's no way to grok this scanning backwards. */
- else if (my_stringend != c)
- string_lossage = 1;
- }
-
- /* Record comment-starters according to that
- quote-parity to the comment-end. */
- if (code == Scomment)
- {
- comstart_parity = parity;
- comstart_pos = from;
- }
-
- /* If we find another earlier comment-ender,
- any comment-starts earlier than that don't count
- (because they go with the earlier comment-ender). */
- if (code == Sendcomment
- && SYNTAX_COMMENT_STYLE (FETCH_CHAR (from)) == comstyle)
- break;
-
- /* Assume a defun-start point is outside of strings. */
- if (code == Sopen
- && (from == stop || FETCH_BYTE (from - 1) == '\n'))
- break;
- }
-
- if (comstart_pos == 0)
- from = comment_end;
- /* If the earliest comment starter
- is followed by uniform paired string quotes or none,
- we know it can't be inside a string
- since if it were then the comment ender would be inside one.
- So it does start a comment. Skip back to it. */
- else if (comstart_parity == 0 && !string_lossage)
- from = comstart_pos;
- else
- {
- /* We had two kinds of string delimiters mixed up
- together. Decode this going forwards.
- Scan fwd from the previous comment ender
- to the one in question; this records where we
- last passed a comment starter. */
- struct lisp_parse_state state;
- scan_sexps_forward (&state, find_defun_start (comment_end),
- comment_end - 1, -10000, 0, Qnil, 0);
- if (state.incomment)
- from = state.comstart;
- else
- /* We can't grok this as a comment; scan it normally. */
- from = comment_end;
- }
- }
+ if (found == 0)
+ {
+ from = ini; /* Set point to ini + 1. */
+ from_byte = ini_byte;
+ goto leave;
+ }
+ }
+ else if (code == Sendcomment)
+ {
+ found = back_comment (from, from_byte, stop, comstyle,
+ &out_charpos, &out_bytepos);
+ if (found != -1)
+ from = out_charpos, from_byte = out_bytepos;
/* We have skipped one comment. */
break;
}
- else if ((code != Swhitespace && code != Scomment) || quoted)
+ else if (code != Swhitespace && code != Scomment)
{
+ leave:
immediate_quit = 0;
- INC_POS (from);
- SET_PT (from);
+ INC_BOTH (from, from_byte);
+ SET_PT_BOTH (from, from_byte);
return Qnil;
}
}
count1++;
}
- SET_PT (from);
+ SET_PT_BOTH (from, from_byte);
immediate_quit = 0;
return Qt;
}
\f
-int parse_sexp_ignore_comments;
-
-Lisp_Object
+static Lisp_Object
scan_lists (from, count, depth, sexpflag)
register int from;
int count, depth, sexpflag;
{
Lisp_Object val;
- register int stop;
+ register int stop = count > 0 ? ZV : BEGV;
register int c, c1;
int stringterm;
int quoted;
int mathexit = 0;
register enum syntaxcode code, temp_code;
- int min_depth = depth; /* Err out if depth gets less than this. */
+ int min_depth = depth; /* Err out if depth gets less than this. */
int comstyle = 0; /* style of comment encountered */
int temp_pos;
int last_good = from;
+ int found;
+ int from_byte = CHAR_TO_BYTE (from);
+ int out_bytepos, out_charpos;
if (depth > 0) min_depth = 0;
immediate_quit = 1;
QUIT;
+ SETUP_SYNTAX_TABLE (from, count);
while (count > 0)
{
- stop = ZV;
while (from < stop)
{
- c = FETCH_CHAR (from);
+ UPDATE_SYNTAX_TABLE_FORWARD (from);
+ c = FETCH_CHAR (from_byte);
code = SYNTAX (c);
if (depth == min_depth)
last_good = from;
- INC_POS (from);
+ INC_BOTH (from, from_byte);
+ UPDATE_SYNTAX_TABLE_FORWARD (from);
if (from < stop && SYNTAX_COMSTART_FIRST (c)
- && SYNTAX_COMSTART_SECOND (FETCH_CHAR (from))
+ && SYNTAX_COMSTART_SECOND (FETCH_CHAR (from_byte))
&& parse_sexp_ignore_comments)
{
/* we have encountered a comment start sequence and we
- are ignoring all text inside comments. we must record
+ are ignoring all text inside comments. We must record
the comment style this sequence begins so that later,
only a comment end of the same style actually ends
the comment section */
code = Scomment;
- comstyle = SYNTAX_COMMENT_STYLE (FETCH_CHAR (from));
- INC_POS (from);
+ comstyle = SYNTAX_COMMENT_STYLE (FETCH_CHAR (from_byte));
+ INC_BOTH (from, from_byte);
}
+ UPDATE_SYNTAX_TABLE_FORWARD (from);
if (SYNTAX_PREFIX (c))
continue;
case Sescape:
case Scharquote:
if (from == stop) goto lose;
- INC_POS (from);
+ INC_BOTH (from, from_byte);
/* treat following character as a word constituent */
case Sword:
case Ssymbol:
if (depth || !sexpflag) break;
- /* This word counts as a sexp; return at end of it. */
+ /* This word counts as a sexp; return at end of it. */
while (from < stop)
{
- switch (SWITCH_ENUM_CAST (SYNTAX (FETCH_CHAR (from))))
+ UPDATE_SYNTAX_TABLE_FORWARD (from);
+ switch (SWITCH_ENUM_CAST (SYNTAX (FETCH_CHAR (from_byte))))
{
case Scharquote:
case Sescape:
- INC_POS (from);
+ INC_BOTH (from, from_byte);
if (from == stop) goto lose;
break;
case Sword:
default:
goto done;
}
- INC_POS (from);
+ INC_BOTH (from, from_byte);
}
goto done;
case Scomment:
+ case Scomment_fence:
if (!parse_sexp_ignore_comments) break;
while (1)
{
goto done;
goto lose;
}
- c = FETCH_CHAR (from);
- if (SYNTAX (c) == Sendcomment
- && SYNTAX_COMMENT_STYLE (c) == comstyle)
+ UPDATE_SYNTAX_TABLE_FORWARD (from);
+ c = FETCH_CHAR (from_byte);
+ if (code == Scomment
+ ? (SYNTAX (c) == Sendcomment
+ && SYNTAX_COMMENT_STYLE (c) == comstyle)
+ : (SYNTAX (c) == Scomment_fence))
/* we have encountered a comment end of the same style
as the comment sequence which began this comment
section */
break;
- INC_POS (from);
+ INC_BOTH (from, from_byte);
if (from < stop && SYNTAX_COMEND_FIRST (c)
- && SYNTAX_COMEND_SECOND (FETCH_CHAR (from))
- && SYNTAX_COMMENT_STYLE (c) == comstyle)
+ && SYNTAX_COMEND_SECOND (FETCH_CHAR (from_byte))
+ && SYNTAX_COMMENT_STYLE (c) == comstyle
+ && code == Scomment)
/* we have encountered a comment end of the same style
as the comment sequence which began this comment
section */
- { INC_POS (from); break; }
+ {
+ INC_BOTH (from, from_byte);
+ break;
+ }
}
break;
case Smath:
if (!sexpflag)
break;
- if (from != stop && c == FETCH_CHAR (from))
- INC_POS (from);
+ if (from != stop && c == FETCH_CHAR (from_byte))
+ {
+ INC_BOTH (from, from_byte);
+ }
if (mathexit)
{
mathexit = 0;
break;
case Sstring:
- temp_pos = from;
- DEC_POS (temp_pos);
+ case Sstring_fence:
+ temp_pos = dec_bytepos (from_byte);
stringterm = FETCH_CHAR (temp_pos);
while (1)
{
if (from >= stop) goto lose;
- if (FETCH_CHAR (from) == stringterm) break;
- switch (SWITCH_ENUM_CAST (SYNTAX (FETCH_CHAR (from))))
+ UPDATE_SYNTAX_TABLE_FORWARD (from);
+ if (code == Sstring
+ ? (FETCH_CHAR (from_byte) == stringterm)
+ : SYNTAX (FETCH_CHAR (from_byte)) == Sstring_fence)
+ break;
+ switch (SWITCH_ENUM_CAST (SYNTAX (FETCH_CHAR (from_byte))))
{
case Scharquote:
case Sescape:
- INC_POS (from);
+ INC_BOTH (from, from_byte);
}
- INC_POS (from);
+ INC_BOTH (from, from_byte);
}
- INC_POS (from);
+ INC_BOTH (from, from_byte);
if (!depth && sexpflag) goto done;
break;
}
while (count < 0)
{
- stop = BEGV;
while (from > stop)
{
- DEC_POS (from);
- if (quoted = char_quoted (from))
- DEC_POS (from);
- c = FETCH_CHAR (from);
+ DEC_BOTH (from, from_byte);
+ UPDATE_SYNTAX_TABLE_BACKWARD (from);
+ c = FETCH_CHAR (from_byte);
code = SYNTAX (c);
if (depth == min_depth)
last_good = from;
comstyle = 0;
if (code == Sendcomment)
comstyle = SYNTAX_COMMENT_STYLE (c);
- temp_pos = from;
- DEC_POS (temp_pos);
+ temp_pos = from_byte;
+ if (! NILP (current_buffer->enable_multibyte_characters))
+ DEC_POS (temp_pos);
+ else
+ temp_pos--;
if (from > stop && SYNTAX_COMEND_SECOND (c)
&& (c1 = FETCH_CHAR (temp_pos), SYNTAX_COMEND_FIRST (c1))
- && !char_quoted (temp_pos)
&& parse_sexp_ignore_comments)
{
/* we must record the comment style encountered so that
sequence of the same style */
code = Sendcomment;
comstyle = SYNTAX_COMMENT_STYLE (c1);
- from = temp_pos;
+ DEC_BOTH (from, from_byte);
}
- if (SYNTAX_PREFIX (c))
+ /* Quoting turns anything except a comment-ender
+ into a word character. */
+ if (code != Sendcomment && char_quoted (from, from_byte))
+ code = Sword;
+ else if (SYNTAX_PREFIX (c))
continue;
- switch (SWITCH_ENUM_CAST (quoted ? Sword : code))
+ switch (SWITCH_ENUM_CAST (code))
{
case Sword:
case Ssymbol:
+ case Sescape:
+ case Scharquote:
if (depth || !sexpflag) break;
- /* This word counts as a sexp; count object finished after passing it. */
+ /* This word counts as a sexp; count object finished
+ after passing it. */
while (from > stop)
{
- temp_pos = from;
- DEC_POS (temp_pos);
- quoted = char_quoted (temp_pos);
+ temp_pos = from_byte;
+ if (! NILP (current_buffer->enable_multibyte_characters))
+ DEC_POS (temp_pos);
+ else
+ temp_pos--;
+ UPDATE_SYNTAX_TABLE_BACKWARD (from - 1);
+ c1 = FETCH_CHAR (temp_pos);
+ temp_code = SYNTAX (c1);
+ /* Don't allow comment-end to be quoted. */
+ if (temp_code == Sendcomment)
+ goto done2;
+ quoted = char_quoted (from - 1, temp_pos);
if (quoted)
{
- from = temp_pos;
- DEC_POS (temp_pos);
+ DEC_BOTH (from, from_byte);
+ temp_pos = dec_bytepos (temp_pos);
+ UPDATE_SYNTAX_TABLE_BACKWARD (from - 1);
}
c1 = FETCH_CHAR (temp_pos);
temp_code = SYNTAX (c1);
|| temp_code == Ssymbol
|| temp_code == Squote))
goto done2;
- from = temp_pos;
+ DEC_BOTH (from, from_byte);
}
goto done2;
case Smath:
if (!sexpflag)
break;
- temp_pos = from;
- DEC_POS (temp_pos);
+ temp_pos = dec_bytepos (from_byte);
+ UPDATE_SYNTAX_TABLE_BACKWARD (from - 1);
if (from != stop && c == FETCH_CHAR (temp_pos))
- from = temp_pos;
+ DEC_BOTH (from, from_byte);
if (mathexit)
{
mathexit = 0;
case Sendcomment:
if (!parse_sexp_ignore_comments)
break;
-#if 0
- if (code != SYNTAX (c))
- /* For a two-char comment ender, we can assume
- it does end a comment. So scan back in a simple way. */
+ found = back_comment (from, from_byte, stop, comstyle,
+ &out_charpos, &out_bytepos);
+ if (found != -1)
+ from = out_charpos, from_byte = out_bytepos;
+ break;
+
+ case Scomment_fence:
+ case Sstring_fence:
+ while (1)
{
- if (from != stop) DEC_POS (from);
- while (1)
- {
- if (SYNTAX (c = FETCH_CHAR (from)) == Scomment
- && SYNTAX_COMMENT_STYLE (c) == comstyle)
- break;
- if (from == stop)
- {
- if (depth == 0)
- goto done2;
- goto lose;
- }
- DEC_POS (from);
- if (SYNTAX_COMSTART_SECOND (c)
- && SYNTAX_COMSTART_FIRST (FETCH_CHAR (from))
- && SYNTAX_COMMENT_STYLE (c) == comstyle
- && !char_quoted (from))
- break;
- }
- break;
+ DEC_BOTH (from, from_byte);
+ if (from == stop) goto lose;
+ UPDATE_SYNTAX_TABLE_BACKWARD (from);
+ if (!char_quoted (from, from_byte)
+ && SYNTAX (FETCH_CHAR (from_byte)) == code)
+ break;
}
-#endif /* 0 */
-
- /* Look back, counting the parity of string-quotes,
- and recording the comment-starters seen.
- When we reach a safe place, assume that's not in a string;
- then step the main scan to the earliest comment-starter seen
- an even number of string quotes away from the safe place.
-
- OFROM[I] is position of the earliest comment-starter seen
- which is I+2X quotes from the comment-end.
- PARITY is current parity of quotes from the comment end. */
- {
- int parity = 0;
- char my_stringend = 0;
- int string_lossage = 0;
- int comment_end = from;
- int comstart_pos = 0;
- int comstart_parity = 0;
- int scanstart = from;
-
- DEC_POS (scanstart);
-
- /* At beginning of range to scan, we're outside of strings;
- that determines quote parity to the comment-end. */
- while (from != stop)
- {
- /* Move back and examine a character. */
- DEC_POS (from);
-
- c = FETCH_CHAR (from);
- code = SYNTAX (c);
-
- /* If this char is the second of a 2-char comment sequence,
- back up and give the pair the appropriate syntax. */
- temp_pos = from;
- DEC_POS (temp_pos);
- if (from > stop && SYNTAX_COMEND_SECOND (c)
- && (c1 = FETCH_CHAR (temp_pos),
- SYNTAX_COMEND_FIRST (c1)))
- {
- code = Sendcomment;
- from = temp_pos;
- c = c1;
- }
-
- /* If this char starts a 2-char comment start sequence,
- treat it like a 1-char comment starter. */
- temp_pos = from;
- INC_POS (temp_pos);
- if (from < scanstart && SYNTAX_COMSTART_FIRST (c)
- && (c1 = FETCH_CHAR (temp_pos),
- SYNTAX_COMSTART_SECOND (c1))
- && comstyle == SYNTAX_COMMENT_STYLE (c1))
- code = Scomment;
-
- /* Ignore escaped characters. */
- if (char_quoted (from))
- continue;
-
- /* Track parity of quotes. */
- if (code == Sstring)
- {
- parity ^= 1;
- if (my_stringend == 0)
- my_stringend = c;
- /* If we have two kinds of string delimiters.
- There's no way to grok this scanning backwards. */
- else if (my_stringend != c)
- string_lossage = 1;
- }
-
- /* Record comment-starters according to that
- quote-parity to the comment-end. */
- if (code == Scomment)
- {
- comstart_parity = parity;
- comstart_pos = from;
- }
-
- /* If we find another earlier comment-ender,
- any comment-starts earlier than that don't count
- (because they go with the earlier comment-ender). */
- if (code == Sendcomment
- && SYNTAX_COMMENT_STYLE (FETCH_CHAR (from)) == comstyle)
- break;
-
- /* Assume a defun-start point is outside of strings. */
- if (code == Sopen
- && (from == stop || FETCH_BYTE (from - 1) == '\n'))
- break;
- }
-
- if (comstart_pos == 0)
- from = comment_end;
- /* If the earliest comment starter
- is followed by uniform paired string quotes or none,
- we know it can't be inside a string
- since if it were then the comment ender would be inside one.
- So it does start a comment. Skip back to it. */
- else if (comstart_parity == 0 && !string_lossage)
- from = comstart_pos;
- else
- {
- /* We had two kinds of string delimiters mixed up
- together. Decode this going forwards.
- Scan fwd from the previous comment ender
- to the one in question; this records where we
- last passed a comment starter. */
- struct lisp_parse_state state;
- scan_sexps_forward (&state, find_defun_start (comment_end),
- comment_end - 1, -10000, 0, Qnil, 0);
- if (state.incomment)
- from = state.comstart;
- else
- /* We can't grok this as a comment; scan it normally. */
- from = comment_end;
- }
- }
+ if (code == Sstring_fence && !depth && sexpflag) goto done2;
break;
-
+
case Sstring:
- stringterm = FETCH_CHAR (from);
+ stringterm = FETCH_CHAR (from_byte);
while (1)
{
if (from == stop) goto lose;
- temp_pos = from;
- DEC_POS (temp_pos);
- if (!char_quoted (temp_pos)
+ temp_pos = from_byte;
+ if (! NILP (current_buffer->enable_multibyte_characters))
+ DEC_POS (temp_pos);
+ else
+ temp_pos--;
+ UPDATE_SYNTAX_TABLE_BACKWARD (from - 1);
+ if (!char_quoted (from - 1, temp_pos)
&& stringterm == FETCH_CHAR (temp_pos))
break;
- from = temp_pos;
+ DEC_BOTH (from, from_byte);
}
- DEC_POS (from);
+ DEC_BOTH (from, from_byte);
if (!depth && sexpflag) goto done2;
break;
}
/* NOTREACHED */
}
-static int
-char_quoted (pos)
- register int pos;
-{
- register enum syntaxcode code;
- register int beg = BEGV;
- register int quoted = 0;
- int temp_pos = pos;
-
- DEC_POS (temp_pos);
- while (pos > beg
- && ((code = SYNTAX (FETCH_CHAR (temp_pos))) == Scharquote
- || code == Sescape))
- {
- pos = temp_pos;
- quoted = !quoted;
- DEC_POS (temp_pos);
- }
- return quoted;
-}
-
DEFUN ("scan-lists", Fscan_lists, Sscan_lists, 3, 3, 0,
"Scan from character number FROM by COUNT lists.\n\
Returns the character number of the position thus found.\n\
()
{
int beg = BEGV;
+ int opoint = PT;
+ int opoint_byte = PT_BYTE;
int pos = PT;
+ int pos_byte = PT_BYTE;
int c;
- int temp_pos = pos;
- DEC_POS (temp_pos);
+ if (pos > beg)
+ {
+ SETUP_SYNTAX_TABLE (pos, -1);
+ }
+
+ DEC_BOTH (pos, pos_byte);
- while (pos > beg && !char_quoted (temp_pos)
- && ((c = FETCH_CHAR (temp_pos), SYNTAX (c) == Squote)
+ while (!char_quoted (pos, pos_byte)
+ /* Previous statement updates syntax table. */
+ && ((c = FETCH_CHAR (pos_byte), SYNTAX (c) == Squote)
|| SYNTAX_PREFIX (c)))
{
- pos = temp_pos;
- DEC_POS (temp_pos);
+ opoint = pos;
+ opoint_byte = pos_byte;
+
+ if (pos + 1 > beg)
+ DEC_BOTH (pos, pos_byte);
}
- SET_PT (pos);
+ SET_PT_BOTH (opoint, opoint_byte);
return Qnil;
}
\f
-/* Parse forward from FROM to END,
+/* Parse forward from FROM / FROM_BYTE to END,
assuming that FROM has state OLDSTATE (nil means FROM is start of function),
and return a description of the state of the parse at END.
If STOPBEFORE is nonzero, stop at the start of an atom.
- If COMMENTSTOP is nonzero, stop at the start of a comment. */
+ If COMMENTSTOP is 1, stop at the start of a comment.
+ If COMMENTSTOP is -1, stop at the start or end of a comment,
+ after the beginning of a string, or after the end of a string. */
static void
-scan_sexps_forward (stateptr, from, end, targetdepth,
+scan_sexps_forward (stateptr, from, from_byte, end, targetdepth,
stopbefore, oldstate, commentstop)
struct lisp_parse_state *stateptr;
register int from;
int start_quoted = 0; /* Nonzero means starting after a char quote */
Lisp_Object tem;
int prev_from; /* Keep one character before FROM. */
+ int prev_from_byte;
+ int prev_from_syntax;
+ int boundary_stop = commentstop == -1;
+ int nofence;
prev_from = from;
- DEC_POS (prev_from);
+ prev_from_byte = from_byte;
+ if (from != BEGV)
+ DEC_BOTH (prev_from, prev_from_byte);
/* Use this macro instead of `from++'. */
-#define INC_FROM do { prev_from = from; INC_POS (from); } while (0)
+#define INC_FROM \
+do { prev_from = from; \
+ prev_from_byte = from_byte; \
+ prev_from_syntax \
+ = SYNTAX_WITH_FLAGS (FETCH_CHAR (prev_from_byte)); \
+ INC_BOTH (from, from_byte); \
+ UPDATE_SYNTAX_TABLE_FORWARD (from); \
+ } while (0)
immediate_quit = 1;
QUIT;
depth = 0;
state.instring = -1;
state.incomment = 0;
- state.comstyle = 0; /* comment style a by default */
+ state.comstyle = 0; /* comment style a by default. */
+ state.comstr_start = -1; /* no comment/string seen. */
}
else
{
oldstate = Fcdr (oldstate);
oldstate = Fcdr (oldstate);
tem = Fcar (oldstate);
- state.instring = !NILP (tem) ? XINT (tem) : -1;
+ /* Check whether we are inside string_fence-style string: */
+ state.instring = ( !NILP (tem)
+ ? ( INTEGERP (tem) ? XINT (tem) : ST_STRING_STYLE)
+ : -1);
oldstate = Fcdr (oldstate);
tem = Fcar (oldstate);
start_quoted = !NILP (tem);
/* if the eight element of the list is nil, we are in comment
- style a. if it is non-nil, we are in comment style b */
+ style a. If it is non-nil, we are in comment style b */
oldstate = Fcdr (oldstate);
oldstate = Fcdr (oldstate);
tem = Fcar (oldstate);
- state.comstyle = !NILP (tem);
+ state.comstyle = NILP (tem) ? 0 : ( EQ (tem, Qsyntax_table)
+ ? ST_COMMENT_STYLE : 1 );
+
+ oldstate = Fcdr (oldstate);
+ tem = Fcar (oldstate);
+ state.comstr_start = NILP (tem) ? -1 : XINT (tem) ;
}
state.quoted = 0;
mindepth = depth;
curlevel->prev = -1;
curlevel->last = -1;
- /* Enter the loop at a place appropriate for initial state. */
+ /* Enter the loop at a place appropriate for initial state. */
if (state.incomment) goto startincomment;
if (state.instring >= 0)
{
+ nofence = state.instring != ST_STRING_STYLE;
if (start_quoted) goto startquotedinstring;
goto startinstring;
}
if (start_quoted) goto startquoted;
+
+ SETUP_SYNTAX_TABLE (prev_from, 1);
+ prev_from_syntax = SYNTAX_WITH_FLAGS (FETCH_CHAR (prev_from_byte));
+ UPDATE_SYNTAX_TABLE_FORWARD (from);
+
while (from < end)
{
- code = SYNTAX (FETCH_CHAR (from));
INC_FROM;
+ code = prev_from_syntax & 0xff;
+
if (code == Scomment)
- state.comstart = prev_from;
-
- else if (from < end && SYNTAX_COMSTART_FIRST (FETCH_CHAR (prev_from))
- && SYNTAX_COMSTART_SECOND (FETCH_CHAR (from)))
+ state.comstr_start = prev_from;
+ else if (code == Scomment_fence)
{
/* Record the comment style we have entered so that only
the comment-end sequence of the same style actually
terminates the comment section. */
+ state.comstyle = ( code == Scomment_fence
+ ? ST_COMMENT_STYLE
+ : SYNTAX_COMMENT_STYLE (FETCH_CHAR (from_byte)));
+ state.comstr_start = prev_from;
+ if (code != Scomment_fence)
+ INC_FROM;
code = Scomment;
- state.comstyle = SYNTAX_COMMENT_STYLE (FETCH_CHAR (from));
- state.comstart = prev_from;
- INC_FROM;
}
+ else if (from < end)
+ if (SYNTAX_FLAGS_COMSTART_FIRST (prev_from_syntax))
+ if (SYNTAX_COMSTART_SECOND (FETCH_CHAR (from_byte)))
+ /* Duplicate code to avoid a complex if-expression
+ which causes trouble for the SGI compiler. */
+ {
+ /* Record the comment style we have entered so that only
+ the comment-end sequence of the same style actually
+ terminates the comment section. */
+ state.comstyle = ( code == Scomment_fence
+ ? ST_COMMENT_STYLE
+ : SYNTAX_COMMENT_STYLE (FETCH_CHAR (from_byte)));
+ state.comstr_start = prev_from;
+ if (code != Scomment_fence)
+ INC_FROM;
+ code = Scomment;
+ }
- if (SYNTAX_PREFIX (FETCH_CHAR (prev_from)))
+ if (SYNTAX_FLAGS_PREFIX (prev_from_syntax))
continue;
switch (SWITCH_ENUM_CAST (code))
{
symstarted:
while (from < end)
{
- switch (SWITCH_ENUM_CAST (SYNTAX (FETCH_CHAR (from))))
+ switch (SWITCH_ENUM_CAST (SYNTAX (FETCH_CHAR (from_byte))))
{
case Scharquote:
case Sescape:
break;
startincomment:
- if (commentstop)
+ if (commentstop == 1)
goto done;
if (from != BEGV)
{
/* Enter the loop in the middle so that we find
a 2-char comment ender if we start in the middle of it. */
- prev = FETCH_CHAR (prev_from);
goto startincomment_1;
}
/* At beginning of buffer, enter the loop the ordinary way. */
+ state.incomment = 1;
+ goto commentloop;
case Scomment:
state.incomment = 1;
- if (commentstop)
- goto done;
+ if (commentstop || boundary_stop) goto done;
+ commentloop:
while (1)
{
if (from == end) goto done;
- prev = FETCH_CHAR (from);
+ prev = FETCH_CHAR (from_byte);
if (SYNTAX (prev) == Sendcomment
&& SYNTAX_COMMENT_STYLE (prev) == state.comstyle)
/* Only terminate the comment section if the endcomment
of the same style as the start sequence has been
encountered. */
break;
+ if (state.comstyle == ST_COMMENT_STYLE
+ && SYNTAX (prev) == Scomment_fence)
+ break;
INC_FROM;
startincomment_1:
- if (from < end && SYNTAX_COMEND_FIRST (prev)
- && SYNTAX_COMEND_SECOND (FETCH_CHAR (from))
- && SYNTAX_COMMENT_STYLE (prev) == state.comstyle)
+ if (from < end && SYNTAX_FLAGS_COMEND_FIRST (prev_from_syntax)
+ && SYNTAX_COMEND_SECOND (FETCH_CHAR (from_byte))
+ && (SYNTAX_FLAGS_COMMENT_STYLE (prev_from_syntax)
+ == state.comstyle))
/* Only terminate the comment section if the end-comment
sequence of the same style as the start sequence has
been encountered. */
- { INC_FROM; break; }
+ break;
}
+ INC_FROM;
state.incomment = 0;
state.comstyle = 0; /* reset the comment style */
+ if (boundary_stop) goto done;
break;
case Sopen:
break;
case Sstring:
+ case Sstring_fence:
+ state.comstr_start = from - 1;
if (stopbefore) goto stop; /* this arg means stop at sexp start */
curlevel->last = prev_from;
- state.instring = FETCH_CHAR (prev_from);
+ state.instring = (code == Sstring
+ ? (FETCH_CHAR (prev_from_byte))
+ : ST_STRING_STYLE);
+ if (boundary_stop) goto done;
startinstring:
- while (1)
- {
- int c;
+ {
+ nofence = state.instring != ST_STRING_STYLE;
+
+ while (1)
+ {
+ int c;
- if (from >= end) goto done;
- c = FETCH_CHAR (from);
- if (c == state.instring) break;
- switch (SWITCH_ENUM_CAST (SYNTAX (c)))
- {
- case Scharquote:
- case Sescape:
- INC_FROM;
- startquotedinstring:
- if (from >= end) goto endquoted;
- }
- INC_FROM;
- }
+ if (from >= end) goto done;
+ c = FETCH_CHAR (from_byte);
+ if (nofence && c == state.instring) break;
+ switch (SWITCH_ENUM_CAST (SYNTAX (c)))
+ {
+ case Sstring_fence:
+ if (!nofence) goto string_end;
+ break;
+ case Scharquote:
+ case Sescape:
+ INC_FROM;
+ startquotedinstring:
+ if (from >= end) goto endquoted;
+ }
+ INC_FROM;
+ }
+ }
+ string_end:
state.instring = -1;
curlevel->prev = curlevel->last;
INC_FROM;
+ if (boundary_stop) goto done;
break;
case Smath:
point is set to where parsing stops.\n\
If fifth arg STATE is omitted or nil,\n\
parsing assumes that FROM is the beginning of a function.\n\
-Value is a list of eight elements describing final state of parsing:\n\
+Value is a list of nine elements describing final state of parsing:\n\
0. depth in parens.\n\
1. character address of start of innermost containing list; nil if none.\n\
2. character address of start of last complete sexp terminated.\n\
3. non-nil if inside a string.\n\
- (it is the character that will terminate the string.)\n\
+ (it is the character that will terminate the string,\n\
+ or t if the string should be terminated by a generic string delimiter.)\n\
4. t if inside a comment.\n\
5. t if following a quote character.\n\
6. the minimum paren-depth encountered during this scan.\n\
- 7. t if in a comment of style `b'.\n\
+ 7. t if in a comment of style b; `syntax-table' if the comment\n\
+ should be terminated by a generic comment delimiter.\n\
+ 8. character address of start of comment or string; nil if not in one.\n\
If third arg TARGETDEPTH is non-nil, parsing stops if the depth\n\
in parentheses becomes equal to TARGETDEPTH.\n\
Fourth arg STOPBEFORE non-nil means stop when come to\n\
any character that starts a sexp.\n\
-Fifth arg STATE is an eight-list like what this function returns.\n\
-It is used to initialize the state of the parse. Its second and third
-elements are ignored.
-Sixth args COMMENTSTOP non-nil means stop at the start of a comment.")
+Fifth arg STATE is a nine-element list like what this function returns.\n\
+ It is used to initialize the state of the parse. Elements number 1, 2, 6\n\
+ and 8 are ignored; you can leave off element 8 (the last) entirely.\n\
+Sixth arg COMMENTSTOP non-nil means stop at the start of a comment.\n\
+ If it is `syntax-table', stop after the start of a comment or a string,\n\
+ or after end of a comment or a string.")
(from, to, targetdepth, stopbefore, state, commentstop)
*/
target = -100000; /* We won't reach this depth */
validate_region (&from, &to);
- scan_sexps_forward (&state, XINT (from), XINT (to),
+ scan_sexps_forward (&state, XINT (from), CHAR_TO_BYTE (XINT (from)),
+ XINT (to),
target, !NILP (stopbefore), oldstate,
- !NILP (commentstop));
+ (NILP (commentstop)
+ ? 0 : (EQ (commentstop, Qsyntax_table) ? -1 : 1)));
SET_PT (state.location);
return Fcons (make_number (state.depth),
Fcons (state.prevlevelstart < 0 ? Qnil : make_number (state.prevlevelstart),
Fcons (state.thislevelstart < 0 ? Qnil : make_number (state.thislevelstart),
- Fcons (state.instring >= 0 ? make_number (state.instring) : Qnil,
+ Fcons (state.instring >= 0
+ ? (state.instring == ST_STRING_STYLE
+ ? Qt : make_number (state.instring)) : Qnil,
Fcons (state.incomment ? Qt : Qnil,
Fcons (state.quoted ? Qt : Qnil,
- Fcons (make_number (state.mindepth),
- Fcons (state.comstyle ? Qt : Qnil,
- Qnil))))))));
+ Fcons (make_number (state.mindepth),
+ Fcons ((state.comstyle
+ ? (state.comstyle == ST_COMMENT_STYLE
+ ? Qsyntax_table : Qt) :
+ Qnil),
+ Fcons ((state.incomment || state.instring
+ ? make_number (state.comstr_start)
+ : Qnil),
+ Qnil)))))))));
}
\f
+void
init_syntax_once ()
{
register int i, c;
Qchar_table_extra_slots = intern ("char-table-extra-slots");
/* Create objects which can be shared among syntax tables. */
- Vsyntax_code_object = Fmake_vector (13, Qnil);
+ Vsyntax_code_object = Fmake_vector (make_number (13), Qnil);
for (i = 0; i < XVECTOR (Vsyntax_code_object)->size; i++)
XVECTOR (Vsyntax_code_object)->contents[i]
= Fcons (make_number (i), Qnil);
}
}
+void
syms_of_syntax ()
{
Qsyntax_table_p = intern ("syntax-table-p");
DEFVAR_BOOL ("parse-sexp-ignore-comments", &parse_sexp_ignore_comments,
"Non-nil means `forward-sexp', etc., should treat comments as whitespace.");
+ DEFVAR_BOOL ("parse-sexp-lookup-properties", &parse_sexp_lookup_properties,
+ "Non-nil means `forward-sexp', etc., grant `syntax-table' property.\n\
+The value of this property should be either a syntax table, or a cons\n\
+of the form (SYNTAXCODE . MATCHCHAR), SYNTAXCODE being the numeric\n\
+syntax code, MATCHCHAR being nil or the character to match (which is\n\
+relevant only for open/close type.");
+
words_include_escapes = 0;
DEFVAR_BOOL ("words-include-escapes", &words_include_escapes,
"Non-nil means `forward-word', etc., should treat escape chars part of words.");
defsubr (&Sforward_word);
+ defsubr (&Sskip_chars_forward);
+ defsubr (&Sskip_chars_backward);
+ defsubr (&Sskip_syntax_forward);
+ defsubr (&Sskip_syntax_backward);
+
defsubr (&Sforward_comment);
defsubr (&Sscan_lists);
defsubr (&Sscan_sexps);
HCoop / bpt / emacs.git / blobdiff