}
oldi = i = count > 0 ? gl_state.forward_i : gl_state.backward_i;
- /* We are guarantied to be called with CHARPOS either in i,
+ /* We are guaranteed 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");
error ("Error in syntax_table logic for intervals <-");
/* Update the interval. */
i = update_interval (i, charpos);
- if (oldi->position != INTERVAL_LAST_POS (i))
+ if (!gl_state.left_ok || 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))
+ if (!gl_state.right_ok || i->position != INTERVAL_LAST_POS (oldi))
{
invalidate = 0;
gl_state.left_ok = 1; /* Invalidate the other end. */
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 (invalidate) /* Did not get to adjacent interval. */
+ { /* with the same table => */
+ /* invalidate the old range. */
if (count > 0)
{
gl_state.backward_i = i;
/* 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;
-}
+/* 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.
int comment_end_byte = from_byte;
int comstart_pos = 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;
register enum syntaxcode code;
int nesting = 1; /* current comment nesting */
int c;
+ int syntax = 0;
+
+ /* FIXME: A }} comment-ender style leads to incorrect behavior
+ in the case of {{ c }}} because we ignore the last two chars which are
+ assumed to be comment-enders although they aren't. */
/* 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;
+ int temp_byte, prev_syntax;
+ int com2start, com2end;
/* Move back and examine a character. */
DEC_BOTH (from, from_byte);
UPDATE_SYNTAX_TABLE_BACKWARD (from);
+ prev_syntax = syntax;
c = FETCH_CHAR (from_byte);
+ syntax = SYNTAX_WITH_FLAGS (c);
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)
- && prev_char_comend_first (from, from_byte))
- {
- code = Sendcomment;
- DEC_BOTH (from, from_byte);
- 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))
+ /* Check for 2-char comment markers. */
+ com2start = (SYNTAX_FLAGS_COMSTART_FIRST (syntax)
+ && SYNTAX_FLAGS_COMSTART_SECOND (prev_syntax)
+ && comstyle == SYNTAX_FLAGS_COMMENT_STYLE (prev_syntax)
+ && (SYNTAX_FLAGS_COMMENT_NESTED (prev_syntax)
+ || SYNTAX_FLAGS_COMMENT_NESTED (syntax)) == comnested);
+ com2end = (SYNTAX_FLAGS_COMEND_FIRST (syntax)
+ && SYNTAX_FLAGS_COMEND_SECOND (prev_syntax));
+
+ /* Nasty cases with overlapping 2-char comment markers:
+ - snmp-mode: -- c -- foo -- c --
+ --- c --
+ ------ c --
+ - c-mode: *||*
+ |* *|* *|
+ |*| |* |*|
+ /// */
+
+ /* If a 2-char comment sequence partly overlaps with another,
+ we don't try to be clever. */
+ if (from > stop && (com2end || com2start))
{
- 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);
+ int next = from, next_byte = from_byte, next_c, next_syntax;
+ DEC_BOTH (next, next_byte);
+ UPDATE_SYNTAX_TABLE_BACKWARD (next);
+ next_c = FETCH_CHAR (next_byte);
+ next_syntax = SYNTAX_WITH_FLAGS (next_c);
+ if (((com2start || comnested)
+ && SYNTAX_FLAGS_COMEND_SECOND (syntax)
+ && SYNTAX_FLAGS_COMEND_FIRST (next_syntax))
+ || ((com2end || comnested)
+ && SYNTAX_FLAGS_COMSTART_SECOND (syntax)
+ && comstyle == SYNTAX_FLAGS_COMMENT_STYLE (syntax)
+ && SYNTAX_FLAGS_COMSTART_FIRST (next_syntax)))
+ goto lossage;
+ /* UPDATE_SYNTAX_TABLE_FORWARD (next + 1); */
}
- else if (code == Scomment && comstyle != SYNTAX_COMMENT_STYLE (c))
- /* Ignore comment starters of a different style. */
+
+ if (com2start && comstart_pos == 0)
+ /* We're looking at a comment starter. But it might be a comment
+ ender as well (see snmp-mode). The first time we see one, we
+ need to consider it as a comment starter,
+ and the subsequent times as a comment ender. */
+ com2end = 0;
+
+ /* Turn a 2-char comment sequences into the appropriate syntax. */
+ if (com2end)
+ code = Sendcomment;
+ else if (com2start)
+ code = Scomment;
+ /* Ignore comment starters of a different style. */
+ else if (code == Scomment
+ && (comstyle != SYNTAX_FLAGS_COMMENT_STYLE (syntax)
+ || SYNTAX_FLAGS_COMMENT_NESTED (syntax) != comnested))
continue;
/* Ignore escaped characters, except comment-enders. */
break;
case Sendcomment:
- if (SYNTAX_COMMENT_STYLE (FETCH_CHAR (from_byte)) == comstyle)
+ if (SYNTAX_FLAGS_COMMENT_STYLE (syntax) == comstyle
+ && (SYNTAX_FLAGS_COMMENT_NESTED (prev_syntax)
+ || SYNTAX_FLAGS_COMMENT_NESTED (syntax)) == comnested)
/* This is the same style of comment ender as ours. */
{
if (comnested)
return Qnil;
}
-/* This comment supplies the doc string for modify-syntax-entry,
- for make-docfile to see. We cannot put this in the real DEFUN
- due to limits in the Unix cpp.
-
-DEFUN ("modify-syntax-entry", foo, bar, 2, 3, 0,
- "Set syntax for character CHAR according to string S.\n\
-The syntax is changed only for table TABLE, which defaults to\n\
- the current buffer's syntax table.\n\
-The first character of S should be one of the following:\n\
- Space or - whitespace syntax. w word constituent.\n\
- _ symbol constituent. . punctuation.\n\
- ( open-parenthesis. ) close-parenthesis.\n\
- \" string quote. \\ escape.\n\
- $ paired delimiter. ' expression quote or prefix operator.\n\
- < comment starter. > comment ender.\n\
- / character-quote. @ inherit from `standard-syntax-table'.\n\
-\n\
-Only single-character comment start and end sequences are represented thus.\n\
-Two-character sequences are represented as described below.\n\
-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, 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\
- 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\
-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\
-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\
- between expressions.")
- (char, s, table)
-*/
-
-DEFUN ("modify-syntax-entry", Fmodify_syntax_entry, Smodify_syntax_entry, 2, 3,
- /* I really don't know why this is interactive
- help-form should at least be made useful whilst reading the second arg
- */
- "cSet syntax for character: \nsSet syntax for %s to: ",
- 0 /* See immediately above */)
- (c, newentry, syntax_table)
- Lisp_Object c, newentry, syntax_table;
+DEFUN ("string-to-syntax", Fstring_to_syntax, Sstring_to_syntax, 1, 1, 0,
+ "Convert a syntax specification STRING into syntax cell form.\n\
+STRING should be a string as it is allowed as argument of\n\
+`modify-syntax-entry'. Value is the equivalent cons cell\n\
+\(CODE . MATCHING-CHAR) that can be used as value of a `syntax-table'\n\
+text property.")
+ (string)
+ Lisp_Object string;
{
register unsigned char *p;
register enum syntaxcode code;
int val;
Lisp_Object match;
- CHECK_NUMBER (c, 0);
- CHECK_STRING (newentry, 1);
-
- if (NILP (syntax_table))
- syntax_table = current_buffer->syntax_table;
- else
- check_syntax_table (syntax_table);
+ CHECK_STRING (string, 0);
- p = XSTRING (newentry)->data;
+ p = XSTRING (string)->data;
code = (enum syntaxcode) syntax_spec_code[*p++];
if (((int) code & 0377) == 0377)
error ("invalid syntax description letter: %c", p[-1]);
if (code == Sinherit)
- {
- SET_RAW_SYNTAX_ENTRY (syntax_table, XINT (c), Qnil);
- return Qnil;
- }
+ return Qnil;
if (*p)
{
int len;
int character = (STRING_CHAR_AND_LENGTH
- (p, STRING_BYTES (XSTRING (newentry)) - 1, len));
+ (p, STRING_BYTES (XSTRING (string)) - 1, len));
XSETINT (match, character);
if (XFASTINT (match) == ' ')
match = Qnil;
}
if (val < XVECTOR (Vsyntax_code_object)->size && NILP (match))
- newentry = XVECTOR (Vsyntax_code_object)->contents[val];
+ return XVECTOR (Vsyntax_code_object)->contents[val];
else
/* 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, XINT (c), newentry);
+ return Fcons (make_number (val), match);
+}
+
+/* This comment supplies the doc string for modify-syntax-entry,
+ for make-docfile to see. We cannot put this in the real DEFUN
+ due to limits in the Unix cpp.
+DEFUN ("modify-syntax-entry", foo, bar, 2, 3, 0,
+ "Set syntax for character CHAR according to string S.\n\
+The syntax is changed only for table TABLE, which defaults to\n\
+ the current buffer's syntax table.\n\
+The first character of S should be one of the following:\n\
+ Space or - whitespace syntax. w word constituent.\n\
+ _ symbol constituent. . punctuation.\n\
+ ( open-parenthesis. ) close-parenthesis.\n\
+ \" string quote. \\ escape.\n\
+ $ paired delimiter. ' expression quote or prefix operator.\n\
+ < comment starter. > comment ender.\n\
+ / character-quote. @ inherit from `standard-syntax-table'.\n\
+ | generic string fence. ! generic comment fence.\n\
+\n\
+Only single-character comment start and end sequences are represented thus.\n\
+Two-character sequences are represented as described below.\n\
+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, 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\
+ 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\
+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\
+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\
+ between expressions.")
+ (char, s, table)
+*/
+
+DEFUN ("modify-syntax-entry", Fmodify_syntax_entry, Smodify_syntax_entry, 2, 3,
+ /* I really don't know why this is interactive
+ help-form should at least be made useful whilst reading the second arg
+ */
+ "cSet syntax for character: \nsSet syntax for %s to: ",
+ 0 /* See immediately above */)
+ (c, newentry, syntax_table)
+ Lisp_Object c, newentry, syntax_table;
+{
+ CHECK_NUMBER (c, 0);
+
+ if (NILP (syntax_table))
+ syntax_table = current_buffer->syntax_table;
+ else
+ check_syntax_table (syntax_table);
+
+ SET_RAW_SYNTAX_ENTRY (syntax_table, XINT (c), Fstring_to_syntax (newentry));
return Qnil;
}
\f
case Sclose:
insert_string ("close"); break;
case Squote:
- insert_string ("quote"); break;
+ insert_string ("prefix"); break;
case Sstring:
insert_string ("string"); break;
case Smath:
insert_string ("comment"); break;
case Sendcomment:
insert_string ("endcomment"); break;
+ case Sinherit:
+ insert_string ("inherit"); break;
+ case Scomment_fence:
+ insert_string ("comment fence"); break;
+ case Sstring_fence:
+ insert_string ("string fence"); break;
default:
insert_string ("invalid");
return;
Lisp_Object string, lim;
{
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, set the
{
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. */
+ /* Handle a range starting with a character of
+ less than 256, and ending with a character of
+ not less than 256. Split that into two
+ ranges, the low one ending at 0377, 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);
char_ranges[n_char_ranges++] = c1;
char_ranges[n_char_ranges++] = c2;
- c2 = 0237;
+ c2 = 0377;
}
while (c <= c2)
{
code = syntax & 0xff;
if (code == Sendcomment
&& SYNTAX_FLAGS_COMMENT_STYLE (syntax) == style
- && --nesting <= 0)
+ && (SYNTAX_FLAGS_COMMENT_NESTED (syntax) ?
+ (nesting > 0 && --nesting == 0) : nesting < 0))
/* 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_NESTED (syntax)
&& 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 */
if (from < stop && SYNTAX_FLAGS_COMEND_FIRST (syntax)
&& SYNTAX_FLAGS_COMMENT_STYLE (syntax) == style
&& (c1 = FETCH_CHAR (from_byte),
- SYNTAX_COMEND_SECOND (c1)))
+ SYNTAX_COMEND_SECOND (c1))
+ && ((SYNTAX_FLAGS_COMMENT_NESTED (syntax) ||
+ SYNTAX_COMMENT_NESTED (c1)) ? nesting > 0 : nesting < 0))
{
if (--nesting <= 0)
/* we have encountered a comment end of the same style
&& SYNTAX_FLAGS_COMSTART_FIRST (syntax)
&& (c1 = FETCH_CHAR (from_byte),
SYNTAX_COMMENT_STYLE (c1) == style
- && SYNTAX_COMSTART_SECOND (c1)))
+ && SYNTAX_COMSTART_SECOND (c1))
+ && (SYNTAX_FLAGS_COMMENT_NESTED (syntax) ||
+ SYNTAX_COMMENT_NESTED (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);
}
- /* FIXME: here we ignore 2-char endcomments while we don't
- when going backwards. */
}
- while (code == Swhitespace || code == Sendcomment);
+ while (code == Swhitespace || (code == Sendcomment && c == '\n'));
if (code == Scomment_fence)
comstyle = ST_COMMENT_STYLE;
{
while (1)
{
- int quoted, comstart_second;
+ int quoted;
if (from <= stop)
{
comnested = SYNTAX_COMMENT_NESTED (c);
if (code == Sendcomment)
comstyle = SYNTAX_COMMENT_STYLE (c);
- comstart_second = SYNTAX_COMSTART_SECOND (c);
if (from > stop && SYNTAX_COMEND_SECOND (c)
&& prev_char_comend_first (from, from_byte)
&& !char_quoted (from - 1, dec_bytepos (from_byte)))
comstyle = SYNTAX_COMMENT_STYLE (c1);
comnested = comnested || SYNTAX_COMMENT_NESTED (c1);
}
- if (from > stop && comstart_second
- && prev_char_comstart_first (from, from_byte)
- && !char_quoted (from - 1, dec_bytepos (from_byte)))
- {
- code = Scomment;
- DEC_BOTH (from, from_byte);
- }
if (code == Scomment_fence)
{
&out_charpos, &out_bytepos);
if (found == -1)
{
-#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
+ if (c == '\n')
+ /* This end-of-line is not an end-of-comment.
+ Treat it like a whitespace.
+ CC-mode (and maybe others) relies on this behavior. */
+ ;
+ else
+ {
+ /* 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;
+ }
}
else
- /* We have skipped one comment. */
- from = out_charpos, from_byte = out_bytepos;
- break;
+ {
+ /* We have skipped one comment. */
+ from = out_charpos, from_byte = out_bytepos;
+ break;
+ }
}
- else if (code != Swhitespace && code != Scomment)
+ else if (code != Swhitespace)
{
leave:
immediate_quit = 0;
}
\f
/* Return syntax code of character C if C is a single byte character
- or `multibyte_symbol_p' is zero. Otherwise, retrun Ssymbol. */
+ or `multibyte_symbol_p' is zero. Otherwise, return Ssymbol. */
#define SYNTAX_WITH_MULTIBYTE_CHECK(c) \
((SINGLE_BYTE_CHAR_P (c) || !multibyte_symbol_p) \
if (from >= stop) goto lose;
UPDATE_SYNTAX_TABLE_FORWARD (from);
c = FETCH_CHAR (from_byte);
- if (code == Sstring
- ? c == stringterm
+ if (code == Sstring
+ ? (c == stringterm
+ && SYNTAX_WITH_MULTIBYTE_CHECK (c) == Sstring)
: SYNTAX_WITH_MULTIBYTE_CHECK (c) == Sstring_fence)
break;
temp_pos--;
UPDATE_SYNTAX_TABLE_BACKWARD (from - 1);
if (!char_quoted (from - 1, temp_pos)
- && stringterm == FETCH_CHAR (temp_pos))
+ && stringterm == (c = FETCH_CHAR (temp_pos))
+ && SYNTAX_WITH_MULTIBYTE_CHECK (c) == Sstring)
break;
DEC_BOTH (from, from_byte);
}
Qchar_table_extra_slots = intern ("char-table-extra-slots");
/* Create objects which can be shared among syntax tables. */
- Vsyntax_code_object = Fmake_vector (make_number (13), Qnil);
+ Vsyntax_code_object = Fmake_vector (make_number (Smax), Qnil);
for (i = 0; i < XVECTOR (Vsyntax_code_object)->size; i++)
XVECTOR (Vsyntax_code_object)->contents[i]
= Fcons (make_number (i), Qnil);
defsubr (&Sset_syntax_table);
defsubr (&Schar_syntax);
defsubr (&Smatching_paren);
+ defsubr (&Sstring_to_syntax);
defsubr (&Smodify_syntax_entry);
defsubr (&Sdescribe_syntax);