/* GNU Emacs routines to deal with syntax tables; also word and list parsing.
- Copyright (C) 1985, 87, 93, 94, 95, 1997 Free Software Foundation, Inc.
+ Copyright (C) 1985, 87, 93, 94, 95, 97, 1998, 1999 Free Software Foundation, Inc.
This file is part of GNU Emacs.
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;
+/* Nonzero means `scan-sexps' treat all multibyte characters as symbol. */
+int multibyte_syntax_as_symbol;
+
/* 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
only very temporarily. */
Lisp_Object syntax_temp;
+/* Non-zero means an open parenthesis in column 0 is always considered
+ to be the start of a defun. Zero means an open parenthesis in
+ column 0 has no special meaning. */
+
+int open_paren_in_column_0_is_defun_start;
+
/* This is the internal form of the parse state used in parse-partial-sexp. */
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 incomment; /* -1 if in unnestable comment else comment nesting */
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 location; /* Char number at which parsing stopped. */
int mindepth; /* Minimum depth seen while scanning. */
int comstr_start; /* Position just after last comment/string starter. */
+ Lisp_Object levelstarts; /* Char numbers of starts-of-expression
+ of levels (starting from outermost). */
};
\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 *, 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. */
#define INTERVALS_AT_ONCE 10 /* 1 + max-number of intervals
to scan to property-change. */
-/*
- Update gl_state to an appropriate interval which contains POS. The
- sign of COUNT give the relative position of POS wrt the previously
+/* 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.
+ valid at start, the rest is filled basing on OBJECT.
- `gl_state.*_i' are the intervals, and pos is further in the search
+ `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 POS - or be
+ update the interval to start further than CHARPOS - or be
NULL_INTERVAL. We also update lim_property to be the next value of
- pos to call this subroutine again - or be before/after the
+ charpos to call this subroutine again - or be before/after the
start/end of OBJECT. */
void
-update_syntax_table (pos, count, init, object)
- int pos, count, init;
+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;
+ int cnt = 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 (pos, object);
+ 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;
- gl_state.b_property = i->position - 1;
- gl_state.e_property = INTERVAL_LAST_POS (i);
+ /* interval_of updates only ->position of the return value, so
+ update the parents manually to speed up update_interval. */
+ while (!NULL_PARENT (i))
+ {
+ if (AM_RIGHT_CHILD (i))
+ INTERVAL_PARENT (i)->position = i->position
+ - LEFT_TOTAL_LENGTH (i) + TOTAL_LENGTH (i) /* right end */
+ - TOTAL_LENGTH (INTERVAL_PARENT (i))
+ + LEFT_TOTAL_LENGTH (INTERVAL_PARENT (i));
+ else
+ INTERVAL_PARENT (i)->position = i->position - LEFT_TOTAL_LENGTH (i)
+ + TOTAL_LENGTH (i);
+ i = INTERVAL_PARENT (i);
+ }
+ 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 pos either in i, of further off. */
+ /* 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 (pos < i->position) /* Move left. */
+ else if (charpos < i->position) /* Move left. */
{
if (count > 0)
- error ("Error in syntax_table logic for intervals <-.");
+ error ("Error in syntax_table logic for intervals <-");
/* Update the interval. */
- i = update_interval (i, pos);
+ 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.e_property = INTERVAL_LAST_POS (i) - gl_state.offset;
}
}
- else if (pos >= INTERVAL_LAST_POS (i)) /* Move right. */
+ else if (charpos >= INTERVAL_LAST_POS (i)) /* Move right. */
{
if (count < 0)
- error ("Error in syntax_table logic for intervals ->.");
+ error ("Error in syntax_table logic for intervals ->");
/* Update the interval. */
- i = update_interval (i, pos);
+ 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.b_property = i->position - 1 - gl_state.offset;
}
}
else if (count > 0 ? gl_state.right_ok : gl_state.left_ok)
{
gl_state.backward_i = i;
gl_state.left_ok = 1; /* Invalidate the other end. */
- gl_state.b_property = i->position - 1;
+ 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.e_property = INTERVAL_LAST_POS (i) - gl_state.offset;
}
}
gl_state.current_syntax_table = tmp_table;
gl_state.old_prop = tmp_table;
- if (Fsyntax_table_p (tmp_table) == Qt)
+ if (EQ (Fsyntax_table_p (tmp_table), Qt))
{
gl_state.use_global = 0;
}
{
if (count > 0)
{
- gl_state.e_property = i->position;
+ 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.b_property = i->position + LENGTH (i) - 1 - gl_state.offset;
gl_state.backward_i = i;
}
}
}
\f
-/* Returns TRUE if char at POS is quoted.
- Global syntax-table data should be set up already to be good at pos
- or after. On return global syntax data is good for lookup at POS. */
+/* 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 (pos)
- register int pos;
+char_quoted (charpos, bytepos)
+ register int charpos, bytepos;
{
register enum syntaxcode code;
register int beg = BEGV;
register int quoted = 0;
- int temp_pos = pos;
+ int orig = charpos;
+
+ DEC_BOTH (charpos, bytepos);
- DEC_POS (temp_pos);
- while (temp_pos >= beg
- && ( UPDATE_SYNTAX_TABLE_BACKWARD (temp_pos), 1)
- && ((code = SYNTAX (FETCH_CHAR (temp_pos))) == Scharquote
- || code == Sescape))
+ while (bytepos >= beg)
{
- temp_pos--, quoted = !quoted;
+ 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 (pos);
+
+ 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
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);
+ 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 (tem > BEGV)
+ if (open_paren_in_column_0_is_defun_start)
{
- /* Open-paren at start of line means we found our defun-start. */
- if (SYNTAX (FETCH_CHAR (tem)) == Sopen)
+ while (PT > BEGV)
{
- SETUP_SYNTAX_TABLE (tem + 1, -1); /* Try again... */
- if (SYNTAX (FETCH_CHAR (tem)) == Sopen)
- break;
- /* Now fallback to the default value. */
- gl_state.current_syntax_table = current_buffer->syntax_table;
- gl_state.use_global = 0;
+ /* Open-paren at start of line means we may have found our
+ defun-start. */
+ 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. */
+ scan_newline (PT, PT_BYTE, BEGV, BEGV_BYTE, -2, 1);
}
- /* Move to beg of previous line. */
- tem = scan_buffer ('\n', tem, BEGV, -2, &shortage, 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 FROM is the end of comment. Does not try to
- fallback more than to STOP.
- Returns -1 if cannot find comment ending at from, otherwise start
- of comment. Global syntax data remains valid for
- backward search starting at the returned value (or at FROM, if
- the search was not successful). */
+/* Return the SYNTAX_COMEND_FIRST of the character before POS, POS_BYTE. */
static int
-back_comment (from, stop, comstyle)
- int from, stop, comstyle;
+prev_char_comend_first (pos, pos_byte)
+ int pos, pos_byte;
+{
+ int c, val;
+
+ DEC_BOTH (pos, pos_byte);
+ UPDATE_SYNTAX_TABLE_BACKWARD (pos);
+ c = FETCH_CHAR (pos_byte);
+ val = SYNTAX_COMEND_FIRST (c);
+ UPDATE_SYNTAX_TABLE_FORWARD (pos + 1);
+ return val;
+}
+
+/* Return the SYNTAX_COMSTART_FIRST of the character before POS, POS_BYTE. */
+
+static int
+prev_char_comstart_first (pos, pos_byte)
+ int pos, pos_byte;
+{
+ int c, val;
+
+ DEC_BOTH (pos, pos_byte);
+ UPDATE_SYNTAX_TABLE_BACKWARD (pos);
+ c = FETCH_CHAR (pos_byte);
+ val = SYNTAX_COMSTART_FIRST (c);
+ UPDATE_SYNTAX_TABLE_FORWARD (pos + 1);
+ return val;
+}
+
+/* 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, comnested, comstyle, charpos_ptr, bytepos_ptr)
+ int from, from_byte, stop;
+ int comnested, comstyle;
+ int *charpos_ptr, *bytepos_ptr;
{
/* Look back, counting the parity of string-quotes,
and recording the comment-starters seen.
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_style = -1; /* Presumed outside of any string. */
int string_lossage = 0;
+ /* Not a real lossage: indicates that we have passed a matching comment
+ starter plus an non-matching comment-ender, meaning that any matching
+ comment-starter we might see later could be a false positive (hidden
+ inside another comment).
+ Test case: { a (* b } c (* d *) */
+ int comment_lossage = 0;
int comment_end = from;
+ int comment_end_byte = from_byte;
int comstart_pos = 0;
- int comstart_parity = 0;
+ int comstart_byte;
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 nesting = 1; /* current comment nesting */
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_POS (from);
+ DEC_BOTH (from, from_byte);
UPDATE_SYNTAX_TABLE_BACKWARD (from);
- c = FETCH_CHAR (from);
+ c = FETCH_CHAR (from_byte);
code = SYNTAX (c);
- /* If this char is the second of a 2-char comment sequence,
+ /* 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)
- && SYNTAX_COMEND_FIRST (FETCH_CHAR (from - 1)))
+ && prev_char_comend_first (from, from_byte))
{
code = Sendcomment;
- DEC_POS (from);
- /* This is apparently the best we can do: */
+ DEC_BOTH (from, from_byte);
UPDATE_SYNTAX_TABLE_BACKWARD (from);
- c = FETCH_CHAR (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_SECOND (c)
- && SYNTAX_COMSTART_FIRST (FETCH_CHAR (from - 1))
- && comstyle == SYNTAX_COMMENT_STYLE (c))
+ if (from < scanstart && SYNTAX_COMSTART_FIRST (c))
{
- code = Scomment;
- DEC_POS (from);
- /* This is apparently the best we can do: */
+ temp_byte = inc_bytepos (from_byte);
+ UPDATE_SYNTAX_TABLE_FORWARD (from + 1);
+ if (SYNTAX_COMSTART_SECOND (FETCH_CHAR (temp_byte))
+ && comstyle == SYNTAX_COMMENT_STYLE (FETCH_CHAR (temp_byte)))
+ code = Scomment;
UPDATE_SYNTAX_TABLE_BACKWARD (from);
- c = FETCH_CHAR (from);
}
+ else if (code == Scomment && comstyle != SYNTAX_COMMENT_STYLE (c))
+ /* Ignore comment starters of a different style. */
+ continue;
- /* Ignore escaped characters. */
- if (char_quoted (from))
+ /* Ignore escaped characters, except comment-enders. */
+ if (code != Sendcomment && char_quoted (from, from_byte))
continue;
- /* Track parity of quotes. */
- if (code == Sstring)
+ switch (code)
{
- 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)
+ case Sstring_fence:
+ case Scomment_fence:
+ c = (code == Sstring_fence ? ST_STRING_STYLE : ST_COMMENT_STYLE);
+ case Sstring:
+ /* Track parity of quotes. */
+ if (string_style == -1)
+ /* Entering a string. */
+ string_style = c;
+ else if (string_style == c)
+ /* Leaving the string. */
+ string_style = -1;
+ else
+ /* If we have two kinds of string delimiters.
+ There's no way to grok this scanning backwards. */
string_lossage = 1;
- }
+ break;
+
+ case Scomment:
+ /* We've already checked that it is the relevant comstyle. */
+ if (string_style != -1 || comment_lossage || string_lossage)
+ /* There are odd string quotes involved, so let's be careful.
+ Test case in Pascal: " { " a { " } */
+ goto lossage;
- 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;
- }
+ if (!comnested)
+ {
+ /* Record best comment-starter so far. */
+ comstart_pos = from;
+ comstart_byte = from_byte;
+ }
+ else if (--nesting <= 0)
+ /* nested comments have to be balanced, so we don't need to
+ keep looking for earlier ones. We use here the same (slightly
+ incorrect) reasoning as below: since it is followed by uniform
+ paired string quotes, this comment-start has to be outside of
+ strings, else the comment-end itself would be inside a string. */
+ goto done;
+ break;
- /* Record comment-starters according to that
- quote-parity to the comment-end. */
- if (code == Scomment)
- {
- comstart_parity = parity;
- comstart_pos = from;
- }
+ case Sendcomment:
+ if (SYNTAX_COMMENT_STYLE (FETCH_CHAR (from_byte)) == comstyle)
+ /* This is the same style of comment ender as ours. */
+ {
+ if (comnested)
+ nesting++;
+ else
+ /* Anything before that can't count because it would match
+ this comment-ender rather than ours. */
+ from = stop; /* Break out of the loop. */
+ }
+ else if (comstart_pos != 0 || c != '\n')
+ /* We're mixing comment styles here, so we'd better be careful.
+ The (comstart_pos != 0 || c != '\n') check is not quite correct
+ (we should just always set comment_lossage), but removing it
+ would imply that any multiline comment in C would go through
+ lossage, which seems overkill.
+ The failure should only happen in the rare cases such as
+ { (* } *) */
+ comment_lossage = 1;
+ break;
- /* 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;
+ case Sopen:
+ /* Assume a defun-start point is outside of strings. */
+ if (open_paren_in_column_0_is_defun_start
+ && (from == stop
+ || (temp_byte = dec_bytepos (from_byte),
+ FETCH_CHAR (temp_byte) == '\n')))
+ {
+ defun_start = from;
+ defun_start_byte = from_byte;
+ from = stop; /* Break out of the loop. */
+ }
+ break;
- /* Assume a defun-start point is outside of strings. */
- if (code == Sopen
- && (from == stop || FETCH_CHAR (from - 1) == '\n'))
- break;
+ default:
+ 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)
+ /* If comstart_pos is set and we get here (ie. didn't jump to `lossage'
+ or `done'), then we've found the beginning of the non-nested comment. */
+ else if (1) /* !comnested */
{
from = comstart_pos;
+ from_byte = comstart_byte;
/* Globals are correct now. */
}
else
{
+ struct lisp_parse_state state;
+ lossage:
/* We had two kinds of string delimiters mixed up
together. Decode this going forwards.
- Scan fwd from the previous comment ender
+ Scan fwd from a known safe place (beginning-of-defun)
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)
+ /* If we did not already find the defun start, find it now. */
+ if (defun_start == 0)
{
- /* scan_sexps_forward changed the direction of search in
- global variables, so we need to update it completely. */
-
- from = state.comstr_start;
+ defun_start = find_defun_start (comment_end, comment_end_byte);
+ defun_start_byte = find_start_value_byte;
}
- else
+ do
{
- from = comment_end;
- }
+ scan_sexps_forward (&state,
+ defun_start, defun_start_byte,
+ comment_end, -10000, 0, Qnil, 0);
+ defun_start = comment_end;
+ if (state.incomment == (comnested ? 1 : -1)
+ && state.comstyle == comstyle)
+ from = state.comstr_start;
+ else
+ {
+ from = comment_end;
+ if (state.incomment)
+ /* If comment_end is inside some other comment, maybe ours
+ is nested, so we need to try again from within the
+ surrounding comment. Example: { a (* " *) */
+ {
+ /* FIXME: We should advance by one or two chars. */
+ defun_start = state.comstr_start + 2;
+ defun_start_byte = CHAR_TO_BYTE (defun_start);
+ }
+ }
+ } while (defun_start < comment_end);
+
+ from_byte = CHAR_TO_BYTE (from);
UPDATE_SYNTAX_TABLE_FORWARD (from - 1);
}
- return from;
+ done:
+ *charpos_ptr = from;
+ *bytepos_ptr = from_byte;
+
+ return (from == comment_end) ? -1 : from;
}
\f
DEFUN ("syntax-table-p", Fsyntax_table_p, Ssyntax_table_p, 1, 1, 0,
(table)
Lisp_Object table;
{
+ int idx;
check_syntax_table (table);
current_buffer->syntax_table = table;
/* Indicate that this buffer now has a specified syntax table. */
- current_buffer->local_var_flags
- |= XFASTINT (buffer_local_flags.syntax_table);
+ idx = PER_BUFFER_VAR_IDX (syntax_table);
+ SET_PER_BUFFER_VALUE_P (current_buffer, idx, 1);
return table;
}
\f
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;
}
The second character of S is the matching parenthesis,\n\
used only if the first character is `(' or `)'.\n\
Any additional characters are flags.\n\
-Defined flags are the characters 1, 2, 3, 4, b, and p.\n\
+Defined flags are the characters 1, 2, 3, 4, b, p, and n.\n\
1 means CHAR is the start of a two-char comment start sequence.\n\
2 means CHAR is the second character of such a sequence.\n\
3 means CHAR is the start of a two-char comment end sequence.\n\
of a comment-start, or the first character of a comment-end sequence) using\n\
this flag:\n\
b means CHAR is part of comment sequence b.\n\
+ n means CHAR is part of a nestable comment sequence.\n\
\n\
p means CHAR is a prefix character for `backward-prefix-chars';\n\
such characters are treated as whitespace when they occur\n\
p = XSTRING (newentry)->data;
code = (enum syntaxcode) syntax_spec_code[*p++];
if (((int) code & 0377) == 0377)
- error ("invalid syntax description letter: %c", XINT (c));
+ error ("invalid syntax description letter: %c", p[-1]);
if (code == Sinherit)
{
if (*p)
{
int len;
- int character = STRING_CHAR_AND_LENGTH (p, XSTRING (newentry)->size - 1,
- len);
+ int character = (STRING_CHAR_AND_LENGTH
+ (p, STRING_BYTES (XSTRING (newentry)) - 1, len));
XSETINT (match, character);
if (XFASTINT (match) == ' ')
match = Qnil;
case 'b':
val |= 1 << 21;
break;
+
+ case 'n':
+ val |= 1 << 22;
+ break;
}
if (val < XVECTOR (Vsyntax_code_object)->size && NILP (match))
Lisp_Object value;
{
register enum syntaxcode code;
- char desc, match, start1, start2, end1, end2, prefix, comstyle;
+ char desc, start1, start2, end1, end2, prefix, comstyle, comnested;
char str[2];
Lisp_Object first, match_lisp;
return;
}
- first = XCONS (value)->car;
- match_lisp = XCONS (value)->cdr;
+ first = XCAR (value);
+ match_lisp = XCDR (value);
if (!INTEGERP (first) || !(NILP (match_lisp) || INTEGERP (match_lisp)))
{
end2 = (XINT (first) >> 19) & 1;
prefix = (XINT (first) >> 20) & 1;
comstyle = (XINT (first) >> 21) & 1;
+ comnested = (XINT (first) >> 22) & 1;
if ((int) code < 0 || (int) code >= (int) Smax)
{
insert ("p", 1);
if (comstyle)
insert ("b", 1);
+ if (comnested)
+ insert ("n", 1);
insert_string ("\twhich means: ");
insert_string (",\n\t is the second character of a comment-end sequence");
if (comstyle)
insert_string (" (comment style b)");
+ if (comnested)
+ insert_string (" (nestable)");
if (prefix)
insert_string (",\n\t is a prefix character for `backward-prefix-chars'");
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;
return 0;
}
UPDATE_SYNTAX_TABLE_FORWARD (from);
- ch0 = FETCH_CHAR (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;
{
if (from == end) break;
UPDATE_SYNTAX_TABLE_FORWARD (from);
- ch1 = FETCH_CHAR (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);
+ DEC_BOTH (from, from_byte);
UPDATE_SYNTAX_TABLE_BACKWARD (from);
- ch1 = FETCH_CHAR (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);
+ int temp_byte;
+
+ if (from == beg)
+ break;
+ temp_byte = dec_bytepos (from_byte);
UPDATE_SYNTAX_TABLE_BACKWARD (from);
- ch0 = FETCH_CHAR (temp_pos);
+ 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++;
DEFUN ("forward-word", Fforward_word, Sforward_word, 1, 1, "p",
"Move point forward ARG words (backward if ARG is negative).\n\
Normally returns t.\n\
-If an edge of the buffer is reached, point is left there\n\
-and nil is returned.")
+If an edge of the buffer or a field boundary is reached, point is left there\n\
+and the function returns nil. Field boundaries are not noticed if\n\
+`inhibit-field-text-motion' is non-nil.")
(count)
Lisp_Object count;
{
- int val;
+ int orig_val, val;
CHECK_NUMBER (count, 0);
- if (!(val = scan_words (PT, XINT (count))))
- {
- SET_PT (XINT (count) > 0 ? ZV : BEGV);
- return Qnil;
- }
+ val = orig_val = scan_words (PT, XINT (count));
+ if (! orig_val)
+ val = XINT (count) > 0 ? ZV : BEGV;
+
+ /* Avoid jumping out of an input field. */
+ val = XFASTINT (Fconstrain_to_field (make_number (val), make_number (PT),
+ Qt, Qnil, Qnil));
+
SET_PT (val);
- return Qt;
+ return val == orig_val ? Qt : Qnil;
}
\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\
+except that `]' is never special and `\\' quotes `^', `-' or `\\'\n\
+ (but not as the end of a range; quoting is never needed there).\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.")
return skip_chars (0, 1, syntax, lim);
}
-Lisp_Object
+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];
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 *) alloca (XSTRING (string)->size * (sizeof (int)) * 2);
+ int *char_ranges;
int n_char_ranges = 0;
int negate = 0;
- register int i;
+ register int i, i_byte;
int multibyte = !NILP (current_buffer->enable_multibyte_characters);
+ int string_multibyte;
+ int size_byte;
+ unsigned char *str;
+ int len;
CHECK_STRING (string, 0);
+ char_ranges = (int *) alloca (XSTRING (string)->size * (sizeof (int)) * 2);
+ string_multibyte = STRING_MULTIBYTE (string);
+ str = XSTRING (string)->data;
+ size_byte = STRING_BYTES (XSTRING (string));
+
+ /* Adjust the multibyteness of the string to that of the buffer. */
+ if (multibyte != string_multibyte)
+ {
+ int nbytes;
+
+ if (multibyte)
+ nbytes = count_size_as_multibyte (XSTRING (string)->data,
+ XSTRING (string)->size);
+ else
+ nbytes = XSTRING (string)->size;
+ if (nbytes != size_byte)
+ {
+ str = (unsigned char *) alloca (nbytes);
+ copy_text (XSTRING (string)->data, str, size_byte,
+ string_multibyte, multibyte);
+ size_byte = nbytes;
+ }
+ }
if (NILP (lim))
XSETINT (lim, forwardp ? ZV : BEGV);
else
- CHECK_NUMBER_COERCE_MARKER (lim, 1);
+ CHECK_NUMBER_COERCE_MARKER (lim, 0);
/* In any case, don't allow scan outside bounds of buffer. */
- /* jla turned this off, for no known reason.
- bfox turned the ZV part on, and rms turned the
- BEGV part back on. */
if (XINT (lim) > ZV)
XSETFASTINT (lim, ZV);
if (XINT (lim) < BEGV)
XSETFASTINT (lim, BEGV);
- p = XSTRING (string)->data;
- pend = p + XSTRING (string)->size;
bzero (fastmap, sizeof fastmap);
- if (p != pend && *p == '^')
+ i_byte = 0;
+
+ if (i_byte < size_byte
+ && XSTRING (string)->data[0] == '^')
{
- negate = 1; p++;
+ negate = 1; 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 (p != pend)
+ while (i_byte < size_byte)
{
- c = *p;
- if (multibyte)
- {
- ch = STRING_CHAR (p, pend - p);
- p += BYTES_BY_CHAR_HEAD (*p);
- }
- else
- {
- ch = c;
- p++;
- }
+ int c_leading_code = str[i_byte];
+
+ c = STRING_CHAR_AND_LENGTH (str + i_byte, size_byte - i_byte, len);
+ i_byte += len;
+
if (syntaxp)
- fastmap[syntax_spec_code[c]] = 1;
+ fastmap[syntax_spec_code[c & 0377]] = 1;
else
{
if (c == '\\')
{
- if (p == pend) break;
- c = *p++;
+ if (i_byte == size_byte)
+ break;
+
+ c_leading_code = str[i_byte];
+ c = STRING_CHAR_AND_LENGTH (str+i_byte, size_byte-i_byte, len);
+ i_byte += len;
}
- if (p != pend && *p == '-')
+ if (i_byte < size_byte
+ && str[i_byte] == '-')
{
- unsigned int ch2;
+ unsigned int c2, c2_leading_code;
- p++;
- if (p == pend) break;
- if (SINGLE_BYTE_CHAR_P (ch))
- while (c <= *p)
- {
- fastmap[c] = 1;
- c++;
- }
- else
+ /* Skip over the dash. */
+ i_byte++;
+
+ if (i_byte == size_byte)
+ break;
+
+ /* Get the end of the range. */
+ c2_leading_code = str[i_byte];
+ c2 =STRING_CHAR_AND_LENGTH (str+i_byte, size_byte-i_byte, len);
+ i_byte += len;
+
+ if (SINGLE_BYTE_CHAR_P (c))
{
- fastmap[c] = 1; /* C is the base leading-code. */
- ch2 = STRING_CHAR (p, pend - p);
- if (ch <= ch2)
- char_ranges[n_char_ranges++] = ch,
- char_ranges[n_char_ranges++] = ch2;
+ if (! SINGLE_BYTE_CHAR_P (c2))
+ {
+ /* Handle a range such as \177-\377 in multibyte
+ mode. Split that into two ranges, the low
+ one ending at 0237, and the high one starting
+ at the smallest character in the charset of
+ C2 and ending at C2. */
+ int charset = CHAR_CHARSET (c2);
+ int c1 = MAKE_CHAR (charset, 0, 0);
+
+ fastmap[c2_leading_code] = 1;
+ char_ranges[n_char_ranges++] = c1;
+ char_ranges[n_char_ranges++] = c2;
+ c2 = 0237;
+ }
+ while (c <= c2)
+ {
+ fastmap[c] = 1;
+ c++;
+ }
+ }
+ else if (! SINGLE_BYTE_CHAR_P (c2))
+ {
+ if (c_leading_code != c2_leading_code)
+ error ("Invalid character range: %s",
+ XSTRING (string)->data);
+ if (c <= c2)
+ {
+ fastmap[c_leading_code] = 1;
+ char_ranges[n_char_ranges++] = c;
+ char_ranges[n_char_ranges++] = c2;
+ }
}
- p += multibyte ? BYTES_BY_CHAR_HEAD (*p) : 1;
}
else
{
- fastmap[c] = 1;
- if (!SINGLE_BYTE_CHAR_P (ch))
+ if (SINGLE_BYTE_CHAR_P (c))
+ fastmap[c] = 1;
+ else
{
- char_ranges[n_char_ranges++] = ch;
- char_ranges[n_char_ranges++] = ch;
+ fastmap[c_leading_code] = 1;
+ char_ranges[n_char_ranges++] = c;
+ char_ranges[n_char_ranges++] = c;
}
}
}
{
int start_point = PT;
int pos = PT;
+ int pos_byte = PT_BYTE;
immediate_quit = 1;
if (syntaxp)
{
if (multibyte)
{
- while (pos < XINT (lim)
- && fastmap[(int) SYNTAX (FETCH_CHAR (pos))])
- {
- INC_POS (pos);
- UPDATE_SYNTAX_TABLE_FORWARD (pos);
- }
+ 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))
{
- int savepos = pos;
- DEC_POS (pos);
+ 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))])
+ if (!fastmap[(int) SYNTAX (FETCH_CHAR (pos_byte))])
{
- pos = savepos;
+ pos++;
+ pos_byte = savepos;
break;
}
}
}
else
{
- while (pos > XINT (lim))
- {
- pos--;
- UPDATE_SYNTAX_TABLE_BACKWARD (pos);
- if (!fastmap[(int) SYNTAX (FETCH_BYTE (pos))])
- {
- pos++;
+ if (pos > XINT (lim))
+ while (fastmap[(int) SYNTAX (FETCH_BYTE (pos - 1))])
+ {
+ pos--;
+ if (pos <= XINT (lim))
break;
- }
- }
+ UPDATE_SYNTAX_TABLE_BACKWARD (pos - 1);
+ }
}
}
}
if (forwardp)
{
if (multibyte)
- while (pos < XINT (lim) && fastmap[(c = FETCH_BYTE (pos))])
+ while (pos < XINT (lim) && fastmap[(c = FETCH_BYTE (pos_byte))])
{
- if (!BASE_LEADING_CODE_P (c))
- pos++;
- else if (n_char_ranges)
+ /* If we are looking at a multibyte character, we
+ must look up the character in the table
+ CHAR_RANGES. If there's no data in the table,
+ that character is not what we want to skip. */
+ if (BASE_LEADING_CODE_P (c)
+ && (c = FETCH_MULTIBYTE_CHAR (pos_byte),
+ ! SINGLE_BYTE_CHAR_P (c)))
{
- /* We much check CHAR_RANGES for a multibyte
- character. */
- ch = FETCH_MULTIBYTE_CHAR (pos);
+ /* The following code do the right thing even if
+ n_char_ranges is zero (i.e. no data in
+ CHAR_RANGES). */
for (i = 0; i < n_char_ranges; i += 2)
- if ((ch >= char_ranges[i] && ch <= char_ranges[i + 1]))
+ if (c >= char_ranges[i] && c <= char_ranges[i + 1])
break;
if (!(negate ^ (i < n_char_ranges)))
break;
-
- INC_POS (pos);
- }
- else
- {
- if (!negate) break;
- INC_POS (pos);
}
+ INC_BOTH (pos, pos_byte);
}
else
while (pos < XINT (lim) && fastmap[FETCH_BYTE (pos)])
if (multibyte)
while (pos > XINT (lim))
{
- int savepos = pos;
- DEC_POS (pos);
- if (fastmap[(c = FETCH_BYTE (pos))])
- {
- 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);
- 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 = savepos;
- break;
- }
- }
- else
- if (!negate)
- {
- pos = savepos;
- break;
- }
- }
- else
+ int prev_pos_byte = pos_byte;
+
+ DEC_POS (prev_pos_byte);
+ if (!fastmap[(c = FETCH_BYTE (prev_pos_byte))])
+ break;
+
+ /* See the comment in the previous similar code. */
+ if (BASE_LEADING_CODE_P (c)
+ && (c = FETCH_MULTIBYTE_CHAR (prev_pos_byte),
+ ! SINGLE_BYTE_CHAR_P (c)))
{
- pos = savepos;
- break;
+ for (i = 0; i < n_char_ranges; i += 2)
+ if (c >= char_ranges[i] && c <= char_ranges[i + 1])
+ break;
+ if (!(negate ^ (i < n_char_ranges)))
+ break;
}
+ pos--;
+ pos_byte = prev_pos_byte;
}
else
while (pos > XINT (lim) && fastmap[FETCH_BYTE (pos - 1)])
}
}
+#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
- SET_PT (pos);
+ if (! multibyte)
+ pos_byte = pos;
+
+ SET_PT_BOTH (pos, pos_byte);
immediate_quit = 0;
return make_number (PT - start_point);
}
}
\f
+/* Jump over a comment, assuming we are at the beginning of one.
+ FROM is the current position.
+ FROM_BYTE is the bytepos corresponding to FROM.
+ Do not move past STOP (a charpos).
+ The comment over which we have to jump is of style STYLE
+ (either SYNTAX_COMMENT_STYLE(foo) or ST_COMMENT_STYLE).
+ NESTING should be positive to indicate the nesting at the beginning
+ for nested comments and should be zero or negative else.
+ ST_COMMENT_STYLE cannot be nested.
+ PREV_SYNTAX is the SYNTAX_WITH_FLAGS of the previous character
+ (or 0 If the search cannot start in the middle of a two-character).
+
+ If successful, return 1 and store the charpos of the comment's end
+ into *CHARPOS_PTR and the corresponding bytepos into *BYTEPOS_PTR.
+ Else, return 0 and store the charpos STOP into *CHARPOS_PTR, the
+ corresponding bytepos into *BYTEPOS_PTR and the current nesting
+ (as defined for state.incomment) in *INCOMMENT_PTR.
+
+ The comment end is the last character of the comment rather than the
+ character just after the comment.
+
+ Global syntax data is assumed to initially be valid for FROM and
+ remains valid for forward search starting at the returned position. */
+
+static int
+forw_comment (from, from_byte, stop, nesting, style, prev_syntax,
+ charpos_ptr, bytepos_ptr, incomment_ptr)
+ int from, from_byte, stop;
+ int nesting, style, prev_syntax;
+ int *charpos_ptr, *bytepos_ptr, *incomment_ptr;
+{
+ register int c, c1;
+ register enum syntaxcode code;
+ register int syntax;
+
+ if (nesting <= 0) nesting = -1;
+
+ /* Enter the loop in the middle so that we find
+ a 2-char comment ender if we start in the middle of it. */
+ syntax = prev_syntax;
+ if (syntax != 0) goto forw_incomment;
+
+ while (1)
+ {
+ if (from == stop)
+ {
+ *incomment_ptr = nesting;
+ *charpos_ptr = from;
+ *bytepos_ptr = from_byte;
+ return 0;
+ }
+ c = FETCH_CHAR (from_byte);
+ syntax = SYNTAX_WITH_FLAGS (c);
+ code = syntax & 0xff;
+ if (code == Sendcomment
+ && SYNTAX_FLAGS_COMMENT_STYLE (syntax) == style
+ && --nesting <= 0)
+ /* we have encountered a comment end of the same style
+ as the comment sequence which began this comment
+ section */
+ break;
+ if (code == Scomment_fence
+ && style == ST_COMMENT_STYLE)
+ /* we have encountered a comment end of the same style
+ as the comment sequence which began this comment
+ section. */
+ break;
+ if (nesting > 0
+ && code == Scomment
+ && SYNTAX_FLAGS_COMMENT_STYLE (syntax) == style)
+ /* we have encountered a nested comment of the same style
+ as the comment sequence which began this comment section */
+ nesting++;
+ INC_BOTH (from, from_byte);
+ UPDATE_SYNTAX_TABLE_FORWARD (from);
+
+ forw_incomment:
+ if (from < stop && SYNTAX_FLAGS_COMEND_FIRST (syntax)
+ && SYNTAX_FLAGS_COMMENT_STYLE (syntax) == style
+ && (c1 = FETCH_CHAR (from_byte),
+ SYNTAX_COMEND_SECOND (c1)))
+ {
+ if (--nesting <= 0)
+ /* we have encountered a comment end of the same style
+ as the comment sequence which began this comment
+ section */
+ break;
+ else
+ {
+ INC_BOTH (from, from_byte);
+ UPDATE_SYNTAX_TABLE_FORWARD (from);
+ }
+ }
+ if (nesting > 0
+ && from < stop
+ && SYNTAX_FLAGS_COMSTART_FIRST (syntax)
+ && (c1 = FETCH_CHAR (from_byte),
+ SYNTAX_COMMENT_STYLE (c1) == style
+ && SYNTAX_COMSTART_SECOND (c1)))
+ /* we have encountered a nested comment of the same style
+ as the comment sequence which began this comment
+ section */
+ {
+ INC_BOTH (from, from_byte);
+ UPDATE_SYNTAX_TABLE_FORWARD (from);
+ nesting++;
+ }
+ }
+ *charpos_ptr = from;
+ *bytepos_ptr = from_byte;
+ return 1;
+}
+
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 comstyle = 0; /* style of comment encountered */
+ int comnested = 0; /* whether the comment is nestable or not */
int found;
int count1;
- int temp_pos;
+ int out_charpos, out_bytepos;
+ int dummy;
CHECK_NUMBER (count, 0);
count1 = XINT (count);
QUIT;
from = PT;
+ from_byte = PT_BYTE;
SETUP_SYNTAX_TABLE (from, count1);
while (count1 > 0)
{
do
{
+ int comstart_first;
+
if (from == stop)
{
- SET_PT (from);
+ SET_PT_BOTH (from, from_byte);
immediate_quit = 0;
return Qnil;
}
- UPDATE_SYNTAX_TABLE_FORWARD (from);
- c = FETCH_CHAR (from);
+ c = FETCH_CHAR (from_byte);
code = SYNTAX (c);
- INC_POS (from);
- comstyle = 0;
- if (from < stop && SYNTAX_COMSTART_FIRST (c)
- && (c1 = FETCH_CHAR (from),
+ comstart_first = SYNTAX_COMSTART_FIRST (c);
+ comnested = SYNTAX_COMMENT_NESTED (c);
+ comstyle = SYNTAX_COMMENT_STYLE (c);
+ INC_BOTH (from, from_byte);
+ UPDATE_SYNTAX_TABLE_FORWARD (from);
+ if (from < stop && comstart_first
+ && (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);
+ comnested = comnested || SYNTAX_COMMENT_NESTED (c1);
+ INC_BOTH (from, from_byte);
+ UPDATE_SYNTAX_TABLE_FORWARD (from);
}
+ /* FIXME: here we ignore 2-char endcomments while we don't
+ when going backwards. */
}
while (code == Swhitespace || code == Sendcomment);
- if (code != Scomment && code != Scomment_fence)
+
+ if (code == Scomment_fence)
+ comstyle = ST_COMMENT_STYLE;
+ else if (code != Scomment)
{
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. */
- while (1)
+ found = forw_comment (from, from_byte, stop, comnested, comstyle, 0,
+ &out_charpos, &out_bytepos, &dummy);
+ from = out_charpos; from_byte = out_bytepos;
+ if (!found)
{
- if (from == stop)
- {
- immediate_quit = 0;
- SET_PT (from);
- return Qnil;
- }
- UPDATE_SYNTAX_TABLE_FORWARD (from);
- c = FETCH_CHAR (from);
- INC_POS (from);
- 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),
- 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; }
+ immediate_quit = 0;
+ SET_PT_BOTH (from, from_byte);
+ return Qnil;
}
+ INC_BOTH (from, from_byte);
+ UPDATE_SYNTAX_TABLE_FORWARD (from);
/* We have skipped one comment. */
count1--;
}
while (count1 < 0)
{
- while (from > stop)
+ while (1)
{
- int quoted;
+ int quoted, comstart_second;
+
+ 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);
+ /* char_quoted does UPDATE_SYNTAX_TABLE_BACKWARD (from). */
+ quoted = char_quoted (from, from_byte);
if (quoted)
{
- DEC_POS (from);
- goto leave; /* ????? XXXXX */
+ DEC_BOTH (from, from_byte);
+ goto leave;
}
- UPDATE_SYNTAX_TABLE_BACKWARD (from);
- c = FETCH_CHAR (from);
+ c = FETCH_CHAR (from_byte);
code = SYNTAX (c);
comstyle = 0;
+ comnested = SYNTAX_COMMENT_NESTED (c);
if (code == Sendcomment)
comstyle = SYNTAX_COMMENT_STYLE (c);
- temp_pos = from;
- DEC_POS (temp_pos);
+ comstart_second = SYNTAX_COMSTART_SECOND (c);
if (from > stop && SYNTAX_COMEND_SECOND (c)
- && (c1 = FETCH_CHAR (temp_pos),
- SYNTAX_COMEND_FIRST (c1))
- && !char_quoted (temp_pos))
+ && prev_char_comend_first (from, from_byte)
+ && !char_quoted (from - 1, dec_bytepos (from_byte)))
{
/* We must record the comment style encountered so that
later, we can match only the proper comment begin
sequence of the same style. */
+ DEC_BOTH (from, from_byte);
code = Sendcomment;
+ /* Calling char_quoted, above, set up global syntax position
+ at the new value of FROM. */
+ c1 = FETCH_CHAR (from_byte);
comstyle = SYNTAX_COMMENT_STYLE (c1);
- from = temp_pos;
+ comnested = comnested || SYNTAX_COMMENT_NESTED (c1);
}
- if (from > stop && SYNTAX_COMSTART_SECOND (c)
- && (c1 = FETCH_CHAR (temp_pos),
- SYNTAX_COMSTART_FIRST (c1))
- && !char_quoted (temp_pos))
+ if (from > stop && comstart_second
+ && prev_char_comstart_first (from, from_byte)
+ && !char_quoted (from - 1, dec_bytepos (from_byte)))
{
- /* 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;
- from = temp_pos;
+ DEC_BOTH (from, from_byte);
}
if (code == Scomment_fence)
{
/* Skip until first preceding unquoted comment_fence. */
- int found = 0, ini = from;
+ int found = 0, ini = from, ini_byte = from_byte;
- while (--from != stop)
+ while (1)
{
+ DEC_BOTH (from, from_byte);
+ if (from == stop)
+ break;
UPDATE_SYNTAX_TABLE_BACKWARD (from);
- c = FETCH_CHAR (from);
- if (SYNTAX (c) == Scomment_fence && !char_quoted (from))
+ c = FETCH_CHAR (from_byte);
+ if (SYNTAX (c) == Scomment_fence
+ && !char_quoted (from, from_byte))
{
found = 1;
break;
if (found == 0)
{
from = ini; /* Set point to ini + 1. */
+ from_byte = ini_byte;
goto leave;
}
}
else if (code == Sendcomment)
{
-#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, comnested, comstyle,
+ &out_charpos, &out_bytepos);
+ if (found == -1)
{
- if (from != stop) DEC_POS (from);
- while (1)
- {
- 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;
- }
- break;
+#if 0 /* cc-mode (and maybe others) relies on the bogus behavior. */
+ /* Failure: we should go back to the end of this
+ not-quite-endcomment. */
+ if (SYNTAX(c) != code)
+ /* It was a two-char Sendcomment. */
+ INC_BOTH (from, from_byte);
+ goto leave;
+#endif
}
-#endif /* 0 */
- found = back_comment (from, stop, comstyle);
- if (found != -1) from = found;
-#if 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);
-
- UPDATE_SYNTAX_TABLE_BACKWARD (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.comstr_start;
- else
- /* We can't grok this as a comment; scan it normally. */
- from = comment_end;
- }
- }
-#endif /* 0 */
- /* We have skipped one comment. */
+ else
+ /* We have skipped one comment. */
+ from = out_charpos, from_byte = out_bytepos;
break;
}
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
-Lisp_Object
+/* Return syntax code of character C if C is a single byte character
+ or `multibyte_symbol_p' is zero. Otherwise, retrun Ssymbol. */
+
+#define SYNTAX_WITH_MULTIBYTE_CHECK(c) \
+ ((SINGLE_BYTE_CHAR_P (c) || !multibyte_symbol_p) \
+ ? SYNTAX (c) : Ssymbol)
+
+static Lisp_Object
scan_lists (from, count, depth, sexpflag)
register int from;
int count, depth, sexpflag;
register enum syntaxcode code, temp_code;
int min_depth = depth; /* Err out if depth gets less than this. */
int comstyle = 0; /* style of comment encountered */
+ int comnested = 0; /* whether the comment is nestable or not */
int temp_pos;
int last_good = from;
int found;
+ int from_byte;
+ int out_bytepos, out_charpos;
+ int temp, dummy;
+ int multibyte_symbol_p = sexpflag && multibyte_syntax_as_symbol;
if (depth > 0) min_depth = 0;
+ if (from > ZV) from = ZV;
+ if (from < BEGV) from = BEGV;
+
+ from_byte = CHAR_TO_BYTE (from);
+
immediate_quit = 1;
QUIT;
{
while (from < stop)
{
+ int comstart_first, prefix;
UPDATE_SYNTAX_TABLE_FORWARD (from);
- c = FETCH_CHAR (from);
- code = SYNTAX (c);
+ c = FETCH_CHAR (from_byte);
+ code = SYNTAX_WITH_MULTIBYTE_CHECK (c);
+ comstart_first = SYNTAX_COMSTART_FIRST (c);
+ comnested = SYNTAX_COMMENT_NESTED (c);
+ comstyle = SYNTAX_COMMENT_STYLE (c);
+ prefix = SYNTAX_PREFIX (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))
+ if (from < stop && comstart_first
+ && SYNTAX_COMSTART_SECOND (FETCH_CHAR (from_byte))
&& parse_sexp_ignore_comments)
{
/* we have encountered a comment start sequence and we
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);
+ c1 = FETCH_CHAR (from_byte);
+ comstyle = SYNTAX_COMMENT_STYLE (c1);
+ comnested = comnested || SYNTAX_COMMENT_NESTED (c1);
+ INC_BOTH (from, from_byte);
+ UPDATE_SYNTAX_TABLE_FORWARD (from);
}
- UPDATE_SYNTAX_TABLE_FORWARD (from);
- if (SYNTAX_PREFIX (c))
+ if (prefix)
continue;
switch (SWITCH_ENUM_CAST (code))
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:
while (from < stop)
{
UPDATE_SYNTAX_TABLE_FORWARD (from);
- switch (SWITCH_ENUM_CAST (SYNTAX (FETCH_CHAR (from))))
+
+ /* Some compilers can't handle this inside the switch. */
+ c = FETCH_CHAR (from_byte);
+ temp = SYNTAX_WITH_MULTIBYTE_CHECK (c);
+ switch (temp)
{
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:
+ comstyle = ST_COMMENT_STYLE;
+ /* FALLTHROUGH */
+ case Scomment:
if (!parse_sexp_ignore_comments) break;
- while (1)
+ UPDATE_SYNTAX_TABLE_FORWARD (from);
+ found = forw_comment (from, from_byte, stop,
+ comnested, comstyle, 0,
+ &out_charpos, &out_bytepos, &dummy);
+ from = out_charpos, from_byte = out_bytepos;
+ if (!found)
{
- if (from == stop)
- {
- if (depth == 0)
- goto done;
- goto lose;
- }
- UPDATE_SYNTAX_TABLE_FORWARD (from);
- c = FETCH_CHAR (from);
- 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);
- if (from < stop && SYNTAX_COMEND_FIRST (c)
- && SYNTAX_COMEND_SECOND (FETCH_CHAR (from))
- && 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; }
+ if (depth == 0)
+ goto done;
+ goto lose;
}
+ INC_BOTH (from, from_byte);
+ UPDATE_SYNTAX_TABLE_FORWARD (from);
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;
case Sstring:
case Sstring_fence:
- temp_pos = from;
- DEC_POS (temp_pos);
+ temp_pos = dec_bytepos (from_byte);
stringterm = FETCH_CHAR (temp_pos);
while (1)
{
if (from >= stop) goto lose;
UPDATE_SYNTAX_TABLE_FORWARD (from);
+ c = FETCH_CHAR (from_byte);
if (code == Sstring
- ? (FETCH_CHAR (from) == stringterm)
- : SYNTAX (FETCH_CHAR (from)) == Sstring_fence)
+ ? c == stringterm
+ : SYNTAX_WITH_MULTIBYTE_CHECK (c) == Sstring_fence)
break;
- switch (SWITCH_ENUM_CAST (SYNTAX (FETCH_CHAR (from))))
+
+ /* Some compilers can't handle this inside the switch. */
+ temp = SYNTAX_WITH_MULTIBYTE_CHECK (c);
+ switch (temp)
{
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 (from > stop)
{
- DEC_POS (from);
+ DEC_BOTH (from, from_byte);
UPDATE_SYNTAX_TABLE_BACKWARD (from);
- if (quoted = char_quoted (from))
- {
- DEC_POS (from);
- UPDATE_SYNTAX_TABLE_BACKWARD (from);
- }
- c = FETCH_CHAR (from);
- code = SYNTAX (c);
+ c = FETCH_CHAR (from_byte);
+ code = SYNTAX_WITH_MULTIBYTE_CHECK (c);
if (depth == min_depth)
last_good = from;
comstyle = 0;
+ comnested = SYNTAX_COMMENT_NESTED (c);
if (code == Sendcomment)
comstyle = SYNTAX_COMMENT_STYLE (c);
- temp_pos = from;
- DEC_POS (temp_pos);
if (from > stop && SYNTAX_COMEND_SECOND (c)
- && (c1 = FETCH_CHAR (temp_pos), SYNTAX_COMEND_FIRST (c1))
- && !char_quoted (temp_pos)
+ && prev_char_comend_first (from, from_byte)
&& parse_sexp_ignore_comments)
{
- /* we must record the comment style encountered so that
+ /* We must record the comment style encountered so that
later, we can match only the proper comment begin
- sequence of the same style */
+ sequence of the same style. */
+ DEC_BOTH (from, from_byte);
+ UPDATE_SYNTAX_TABLE_BACKWARD (from);
code = Sendcomment;
+ c1 = FETCH_CHAR (from_byte);
comstyle = SYNTAX_COMMENT_STYLE (c1);
- from = temp_pos;
+ comnested = comnested || SYNTAX_COMMENT_NESTED (c1);
}
- if (SYNTAX_PREFIX (c))
+ /* Quoting turns anything except a comment-ender
+ into a word character. Note that this cannot be true
+ if we decremented FROM in the if-statement above. */
+ 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. */
while (from > stop)
{
- temp_pos = from;
- DEC_POS (temp_pos);
- UPDATE_SYNTAX_TABLE_BACKWARD (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_WITH_MULTIBYTE_CHECK (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);
- UPDATE_SYNTAX_TABLE_BACKWARD (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 = SYNTAX_WITH_MULTIBYTE_CHECK (c1);
if (! (quoted || temp_code == Sword
|| 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);
- UPDATE_SYNTAX_TABLE_BACKWARD (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. */
- {
- 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;
- }
-#endif /* 0 */
- found = back_comment (from, stop, comstyle);
- if (found != -1) from = found;
-#if 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.comstr_start;
- else
- /* We can't grok this as a comment; scan it normally. */
- from = comment_end;
- }
- }
-#endif /* 0 */
+ found = back_comment (from, from_byte, stop, comnested, comstyle,
+ &out_charpos, &out_bytepos);
+ /* FIXME: if found == -1, then it really wasn't a comment-end.
+ For single-char Sendcomment, we can't do much about it apart
+ from skipping the char.
+ For 2-char endcomments, we could try again, taking both
+ chars as separate entities, but it's a lot of trouble
+ for very little gain, so we don't bother either. -sm */
+ if (found != -1)
+ from = out_charpos, from_byte = out_bytepos;
break;
case Scomment_fence:
case Sstring_fence:
while (1)
{
- DEC_POS (from);
+ DEC_BOTH (from, from_byte);
if (from == stop) goto lose;
UPDATE_SYNTAX_TABLE_BACKWARD (from);
- if (!char_quoted (from)
- && SYNTAX (FETCH_CHAR (from)) == code)
+ if (!char_quoted (from, from_byte)
+ && (c = FETCH_CHAR (from_byte),
+ SYNTAX_WITH_MULTIBYTE_CHECK (c) == code))
break;
}
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);
- UPDATE_SYNTAX_TABLE_BACKWARD (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;
}
()
{
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;
- if (pos > beg)
+ if (pos <= beg)
{
- SETUP_SYNTAX_TABLE (pos, -1);
+ SET_PT_BOTH (opoint, opoint_byte);
+
+ return Qnil;
}
- DEC_POS (temp_pos);
- while (pos > beg && !char_quoted (temp_pos)
+ SETUP_SYNTAX_TABLE (pos, -1);
+
+ DEC_BOTH (pos, pos_byte);
+
+ while (!char_quoted (pos, pos_byte)
/* Previous statement updates syntax table. */
- && ((c = FETCH_CHAR (temp_pos), SYNTAX (c) == Squote)
+ && ((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;
struct lisp_parse_state state;
register enum syntaxcode code;
+ int c1;
+ int comnested;
struct level { int last, prev; };
struct level levelstart[100];
register struct level *curlevel = levelstart;
struct level *endlevel = levelstart + 100;
- int prev;
register int depth; /* Paren depth of current scanning location.
level - levelstart equals this except
when the depth becomes negative. */
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;
+ int found;
+ int out_bytepos, out_charpos;
+ int temp;
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;
- SETUP_SYNTAX_TABLE (from, 1);
-
if (NILP (oldstate))
{
depth = 0;
oldstate = Fcdr (oldstate);
tem = Fcar (oldstate);
/* Check whether we are inside string_fence-style string: */
- state.instring = ( !NILP (tem)
- ? ( INTEGERP (tem) ? XINT (tem) : ST_STRING_STYLE)
- : -1);
+ state.instring = (!NILP (tem)
+ ? (INTEGERP (tem) ? XINT (tem) : ST_STRING_STYLE)
+ : -1);
oldstate = Fcdr (oldstate);
tem = Fcar (oldstate);
- state.incomment = !NILP (tem);
+ state.incomment = (!NILP (tem)
+ ? (INTEGERP (tem) ? XINT (tem) : -1)
+ : 0);
oldstate = Fcdr (oldstate);
tem = Fcar (oldstate);
start_quoted = !NILP (tem);
- /* if the eight element of the list is nil, we are in comment
+ /* if the eighth element of the list is nil, we are in comment
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) ? 0 : ( EQ (tem, Qsyntax_table)
- ? ST_COMMENT_STYLE : 1 );
+ 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) ;
+ oldstate = Fcdr (oldstate);
+ tem = Fcar (oldstate);
+ while (!NILP (tem)) /* >= second enclosing sexps. */
+ {
+ /* curlevel++->last ran into compiler bug on Apollo */
+ curlevel->last = XINT (Fcar (tem));
+ if (++curlevel == endlevel)
+ curlevel--; /* error ("Nesting too deep for parser"); */
+ curlevel->prev = -1;
+ curlevel->last = -1;
+ tem = Fcdr (tem);
+ }
}
state.quoted = 0;
mindepth = depth;
curlevel->prev = -1;
curlevel->last = -1;
+ SETUP_SYNTAX_TABLE (prev_from, 1);
+ prev_from_syntax = SYNTAX_WITH_FLAGS (FETCH_CHAR (prev_from_byte));
+ UPDATE_SYNTAX_TABLE_FORWARD (from);
+
/* Enter the loop at a place appropriate for initial state. */
- if (state.incomment) goto startincomment;
+ if (state.incomment)
+ goto startincomment;
if (state.instring >= 0)
{
nofence = state.instring != ST_STRING_STYLE;
- if (start_quoted) goto startquotedinstring;
+ if (start_quoted)
+ goto startquotedinstring;
goto startinstring;
}
- if (start_quoted) goto startquoted;
+ else if (start_quoted)
+ goto startquoted;
+
+#if 0 /* This seems to be redundant with the identical code above. */
+ SETUP_SYNTAX_TABLE (prev_from, 1);
+ prev_from_syntax = SYNTAX_WITH_FLAGS (FETCH_CHAR (prev_from_byte));
+ UPDATE_SYNTAX_TABLE_FORWARD (from);
+#endif
while (from < end)
{
- UPDATE_SYNTAX_TABLE_FORWARD (from);
- code = SYNTAX (FETCH_CHAR (from));
INC_FROM;
+ code = prev_from_syntax & 0xff;
if (code == Scomment)
- state.comstr_start = prev_from;
+ {
+ state.comstyle = SYNTAX_FLAGS_COMMENT_STYLE (prev_from_syntax);
+ state.incomment = (SYNTAX_FLAGS_COMMENT_NESTED (prev_from_syntax) ?
+ 1 : -1);
+ 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)));
+ state.comstyle = ST_COMMENT_STYLE;
+ state.incomment = -1;
state.comstr_start = prev_from;
- if (code != Scomment_fence) INC_FROM;
code = Scomment;
}
else if (from < end)
- if (SYNTAX_COMSTART_FIRST (FETCH_CHAR (prev_from)))
- if (SYNTAX_COMSTART_SECOND (FETCH_CHAR (from)))
- /* Duplicate code to avoid a very complex if-expression
+ if (SYNTAX_FLAGS_COMSTART_FIRST (prev_from_syntax))
+ if (c1 = FETCH_CHAR (from_byte),
+ SYNTAX_COMSTART_SECOND (c1))
+ /* 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)));
+ state.comstyle = SYNTAX_COMMENT_STYLE (FETCH_CHAR (from_byte));
+ comnested = SYNTAX_FLAGS_COMMENT_NESTED (prev_from_syntax);
+ comnested = comnested || SYNTAX_COMMENT_NESTED (c1);
+ state.incomment = comnested ? 1 : -1;
state.comstr_start = prev_from;
- if (code != Scomment_fence) INC_FROM;
+ 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)
{
- UPDATE_SYNTAX_TABLE_FORWARD (from);
- switch (SWITCH_ENUM_CAST (SYNTAX (FETCH_CHAR (from))))
+ /* Some compilers can't handle this inside the switch. */
+ temp = SYNTAX (FETCH_CHAR (from_byte));
+ switch (temp)
{
case Scharquote:
case Sescape:
curlevel->prev = curlevel->last;
break;
- startincomment:
- 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 || boundary_stop) goto done;
- commentloop:
- while (1)
- {
- if (from == end) goto done;
- UPDATE_SYNTAX_TABLE_FORWARD (from);
- prev = FETCH_CHAR (from);
- 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)
- /* Only terminate the comment section if the end-comment
- sequence of the same style as the start sequence has
- been encountered. */
- { break; }
- }
- INC_FROM;
+ startincomment:
+ /* The (from == BEGV) test was to enter the loop in the middle so
+ that we find a 2-char comment ender even if we start in the
+ middle of it. We don't want to do that if we're just at the
+ beginning of the comment (think of (*) ... (*)). */
+ found = forw_comment (from, from_byte, end,
+ state.incomment, state.comstyle,
+ (from == BEGV || from < state.comstr_start + 3)
+ ? 0 : prev_from_syntax,
+ &out_charpos, &out_bytepos, &state.incomment);
+ from = out_charpos; from_byte = out_bytepos;
+ /* Beware! prev_from and friends are invalid now.
+ Luckily, the `done' doesn't use them and the INC_FROM
+ sets them to a sane value without looking at them. */
+ if (!found) goto done;
+ INC_FROM;
state.incomment = 0;
state.comstyle = 0; /* reset the comment style */
if (boundary_stop) goto done;
/* curlevel++->last ran into compiler bug on Apollo */
curlevel->last = prev_from;
if (++curlevel == endlevel)
- error ("Nesting too deep for parser");
+ curlevel--; /* error ("Nesting too deep for parser"); */
curlevel->prev = -1;
curlevel->last = -1;
if (targetdepth == depth) goto done;
if (stopbefore) goto stop; /* this arg means stop at sexp start */
curlevel->last = prev_from;
state.instring = (code == Sstring
- ? (FETCH_CHAR (prev_from))
+ ? (FETCH_CHAR (prev_from_byte))
: ST_STRING_STYLE);
if (boundary_stop) goto done;
startinstring:
{
- nofence = state.instring != ST_STRING_STYLE;
+ nofence = state.instring != ST_STRING_STYLE;
- while (1)
+ while (1)
+ {
+ int c;
+
+ if (from >= end) goto done;
+ c = FETCH_CHAR (from_byte);
+ /* Some compilers can't handle this inside the switch. */
+ temp = SYNTAX (c);
+
+ /* Check TEMP here so that if the char has
+ a syntax-table property which says it is NOT
+ a string character, it does not end the string. */
+ if (nofence && c == state.instring && temp == Sstring)
+ break;
+
+ switch (temp)
{
- int c;
-
- if (from >= end) goto done;
- c = FETCH_CHAR (from);
- if (nofence && c == state.instring) break;
- UPDATE_SYNTAX_TABLE_FORWARD (from);
- 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;
+ 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;
state.prevlevelstart
= (curlevel == levelstart) ? -1 : (curlevel - 1)->last;
state.location = from;
+ state.levelstarts = Qnil;
+ while (--curlevel >= levelstart)
+ state.levelstarts = Fcons (make_number (curlevel->last),
+ state.levelstarts);
immediate_quit = 0;
*stateptr = state;
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 nine elements describing final state of parsing:\n\
+Value is a list of ten 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\
- or t if the string should be terminated by an explicit\n\
- `syntax-table' property.)\n\
- 4. t if inside a comment.\n\
+ or t if the string should be terminated by a generic string delimiter.)\n\
+ 4. nil if outside a comment, t if inside a non-nestable comment, \n\
+ else an integer (the current comment nesting).\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'; `syntax-table' if given by an explicit\n\
- `syntax-table' property.\n\
- 8. character address of start of last comment or string; nil if none.\n\
+ 7. t if in a comment of style b; symbol `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\
+ 9. Intermediate data for continuation of parsing (subject to change).\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 arg COMMENTSTOP non-nil means stop at the start of a comment. If\n\
-it is `syntax-table', stop after the start of a comment or a string, or\n\
-after end of a comment or a string.")
+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 symbol `syntax-table', stop after the start of a comment or a\n\
+ string, 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)
? 0 : (EQ (commentstop, Qsyntax_table) ? -1 : 1)));
Fcons (state.instring >= 0
? (state.instring == ST_STRING_STYLE
? Qt : make_number (state.instring)) : Qnil,
- Fcons (state.incomment ? Qt : Qnil,
+ Fcons (state.incomment < 0 ? Qt :
+ (state.incomment == 0 ? Qnil :
+ make_number (state.incomment)),
Fcons (state.quoted ? Qt : Qnil,
- Fcons (make_number (state.mindepth),
- Fcons (state.comstyle
- ? (state.comstyle == ST_COMMENT_STYLE
- ? Qsyntax_table : Qt) : Qnil,
- Fcons (state.comstr_start != -1 ? make_number (state.comstr_start) : Qnil,
- Qnil)))))))));
+ Fcons (make_number (state.mindepth),
+ Fcons ((state.comstyle
+ ? (state.comstyle == ST_COMMENT_STYLE
+ ? Qsyntax_table : Qt) :
+ Qnil),
+ Fcons (((state.incomment
+ || (state.instring >= 0))
+ ? make_number (state.comstr_start)
+ : Qnil),
+ Fcons (state.levelstarts, 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);
c = ".,;:?!#@~^'`"[i];
SET_RAW_SYNTAX_ENTRY (Vstandard_syntax_table, c, temp);
}
+
+ /* All multibyte characters have syntax `word' by default. */
+ temp = XVECTOR (Vsyntax_code_object)->contents[(int) Sword];
+ for (i = CHAR_TABLE_SINGLE_BYTE_SLOTS; i < CHAR_TABLE_ORDINARY_SLOTS; i++)
+ XCHAR_TABLE (Vstandard_syntax_table)->contents[i] = temp;
}
+void
syms_of_syntax ()
{
Qsyntax_table_p = intern ("syntax-table-p");
Qscan_error = intern ("scan-error");
staticpro (&Qscan_error);
Fput (Qscan_error, Qerror_conditions,
- Fcons (Qerror, Qnil));
+ Fcons (Qscan_error, Fcons (Qerror, Qnil)));
Fput (Qscan_error, Qerror_message,
build_string ("Scan error"));
DEFVAR_BOOL ("words-include-escapes", &words_include_escapes,
"Non-nil means `forward-word', etc., should treat escape chars part of words.");
+ DEFVAR_BOOL ("multibyte-syntax-as-symbol", &multibyte_syntax_as_symbol,
+ "Non-nil means `scan-sexps' treats all multibyte characters as symbol.");
+ multibyte_syntax_as_symbol = 0;
+
+ DEFVAR_BOOL ("open-paren-in-column-0-is-defun-start",
+ &open_paren_in_column_0_is_defun_start,
+ "Non-nil means an open paren in column 0 denotes the start of a defun.");
+ open_paren_in_column_0_is_defun_start = 1;
+
defsubr (&Ssyntax_table_p);
defsubr (&Ssyntax_table);
defsubr (&Sstandard_syntax_table);