syntax.c

具有IDE功能的编辑器

/* editor syntax highlighting.

   Copyright (C) 1996-2000 the Free Software Foundation

   Authors: 1998 Paul Sheer

   This program is free software; you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
   the Free Software Foundation; either version 2 of the License, or
   (at your option) any later version.

   This program is distributed in the hope that it will be useful,
   but WITHOUT ANY WARRANTY; without even the implied warranty of
   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
   GNU General Public License for more details.

   You should have received a copy of the GNU General Public License
   along with this program; if not, write to the Free Software
   Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
   02111-1307, USA.
*/

#include <config.h>
#if defined(MIDNIGHT) || defined(GTK)
#include "edit.h"
#else
#include "coolwidget.h"
#endif
#if defined (HAVE_MAD) && ! defined (MIDNIGHT) && ! defined (GTK)
#include "mad.h"
#endif

/* bytes */
#define SYNTAX_MARKER_DENSITY 512

/*
   Mispelled words are flushed from the syntax highlighting rules
   when they have been around longer than
   TRANSIENT_WORD_TIME_OUT seconds. At a cursor rate of 30
   chars per second and say 3 chars + a space per word, we can
   accumulate 450 words absolute max with a value of 60. This is
   below this limit of 1024 words in a context.
 */
#define TRANSIENT_WORD_TIME_OUT 60

#define UNKNOWN_FORMAT "unknown"

#if !defined(MIDNIGHT) || defined(HAVE_SYNTAXH)

int option_syntax_highlighting = 1;
int option_auto_spellcheck = 1;

/* these three functions are called from the outside */
void edit_load_syntax (WEdit * edit, char **names, char *type);
void edit_free_syntax_rules (WEdit * edit);
void edit_get_syntax_color (WEdit * edit, long byte_index, int *fg, int *bg);

#ifdef HAVE_MAD
static void *mad_syntax_malloc (size_t x, char *file, int line)
#define syntax_malloc(x) mad_syntax_malloc (x, __FILE__, __LINE__)
#else
static void *syntax_malloc (size_t x)
#endif
{
    void *p;
#ifdef HAVE_MAD
    p = mad_alloc (x, file, line);
#else
    p = malloc (x);
#endif
    memset (p, 0, x);
    return p;
}

#define syntax_free(x) {if(x){free(x);(x)=0;}}

static long compare_word_to_right (WEdit * edit, long i, char *text, char *whole_left, char *whole_right, int line_start)
{
    unsigned char *p, *q;
    int c, d, j;
    if (!*text)
	return -1;
    c = edit_get_byte (edit, i - 1);
    if (line_start)
	if (c != '\n')
	    return -1;
    if (whole_left)
	if (strchr (whole_left, c))
	    return -1;
    for (p = (unsigned char *) text, q = p + strlen ((char *) p); (unsigned long) p < (unsigned long) q; p++, i++) {
	switch (*p) {
	case '\001':
	    p++;
	    for (;;) {
		c = edit_get_byte (edit, i);
		if (!*p)
		    if (whole_right)
			if (!strchr (whole_right, c))
			    break;
		if (c == *p)
		    break;
		if (c == '\n')
		    return -1;
		i++;
	    }
	    break;
	case '\002':
	    p++;
	    j = 0;
	    for (;;) {
		c = edit_get_byte (edit, i);
		if (c == *p) {
		    j = i;
		    if (*p == *text && !p[1])	/* handle eg '+' and @+@ keywords properly */
			break;
		}
		if (j && strchr ((char *) p + 1, c))	/* c exists further down, so it will get matched later */
		    break;
		if (c == '\n' || c == '\t' || c == ' ') {
		    if (!*p) {
			i--;
			break;
		    }
		    if (!j)
			return -1;
		    i = j;
		    break;
		}
		if (whole_right)
		    if (!strchr (whole_right, c)) {
			if (!*p) {
			    i--;
			    break;
			}
			if (!j)
			    return -1;
			i = j;
			break;
		    }
		i++;
	    }
	    break;
	case '\003':
	    p++;
	    c = -1;
	    for (;; i++) {
		d = c;
		c = edit_get_byte (edit, i);
		for (j = 0; p[j] != '\003'; j++)
		    if (c == p[j])
			goto found_char2;
		break;
	      found_char2:
		j = c;		/* dummy command */
	    }
	    i--;
	    while (*p != '\003')
		p++;
	    if (p[1] == d)
		i--;
	    break;
	case '\004':
	    p++;
	    c = edit_get_byte (edit, i);
	    for (; *p != '\004'; p++)
		if (c == *p)
		    goto found_char3;
	    return -1;
	  found_char3:
	    for (; *p != '\004'; p++);
	    break;
#if 0
/* especially for LaTeX */
	case '\005': {
                int b = 0;
	        p++;
	        for (;;) {
		    c = edit_get_byte (edit, i);
                    if (c == '\\') {
                        i += 2;
                        continue;
                    }
                    if (c == '{') {
                    }
		    if (c == *p)
		        break;
		    if (c == '\n')
		        return -1;
		    i++;
	        }
	        break;
            }
#endif
	default:
	    if (*p != edit_get_byte (edit, i))
		return -1;
	}
    }
    if (whole_right)
	if (strchr (whole_right, edit_get_byte (edit, i)))
	    return -1;
    return i;
}

#define XXX							\
	    if (*s < '\005' || *s == (unsigned char) c)		\
		goto done;					\
	    s++;

static inline char *xx_strchr (const unsigned char *s, int c)
{
  repeat:
    XXX XXX XXX XXX XXX XXX XXX XXX;
    XXX XXX XXX XXX XXX XXX XXX XXX;
    goto repeat;
  done:
    return (char *) s;
}

static inline struct syntax_rule apply_rules_going_right (WEdit * edit, long i, struct syntax_rule rule)
{
    struct context_rule *r;
    int contextchanged = 0, c;
    int found_right = 0, found_left = 0, keyword_foundleft = 0, keyword_foundright = 0;
    int is_end;
    long end = 0;
    struct syntax_rule _rule = rule;
    if (!(c = edit_get_byte (edit, i)))
	return rule;
    is_end = (rule.end == (unsigned char) i);
/* check to turn off a keyword */
    if (_rule.keyword) {
	struct key_word *k;
	k = edit->rules[_rule.context]->keyword[_rule.keyword];
	if (edit_get_byte (edit, i - 1) == '\n')
	    _rule.keyword = 0;
	if (is_end) {
	    _rule.keyword = 0;
	    keyword_foundleft = 1;
	}
    }
/* check to turn off a context */
    if (_rule.context && !_rule.keyword) {
	long e;
	r = edit->rules[_rule.context];
	if (r->first_right == c && !(rule.border & RULE_ON_RIGHT_BORDER) && (e = compare_word_to_right (edit, i, r->right, r->whole_word_chars_left, r->whole_word_chars_right, r->line_start_right)) > 0) {
	    _rule.end = e;
	    found_right = 1;
	    _rule.border = RULE_ON_RIGHT_BORDER;
	    if (r->between_delimiters)
		_rule.context = 0;
	} else if (is_end && rule.border & RULE_ON_RIGHT_BORDER) {
/* always turn off a context at 4 */
	    found_left = 1;
	    _rule.border = 0;
	    if (!keyword_foundleft)
		_rule.context = 0;
	} else if (is_end && rule.border & RULE_ON_LEFT_BORDER) {
/* never turn off a context at 2 */
	    found_left = 1;
	    _rule.border = 0;
	}
    }
/* check to turn on a keyword */
    if (!_rule.keyword) {
	char *p;
	p = (r = edit->rules[_rule.context])->keyword_first_chars;
	while (*(p = xx_strchr ((unsigned char *) p + 1, c))) {
	    struct key_word *k;
	    int count;
	    long e;
	    count = (unsigned long) p - (unsigned long) r->keyword_first_chars;
	    k = r->keyword[count];
	    e = compare_word_to_right (edit, i, k->keyword, k->whole_word_chars_left, k->whole_word_chars_right, k->line_start);
	    if (e > 0) {
		end = e;
		_rule.end = e;
		_rule.keyword = count;
		keyword_foundright = 1;
		break;
	    }
	}
    }
/* check to turn on a context */
    if (!_rule.context) {
	if (!found_left && is_end) {
	    if (rule.border & RULE_ON_RIGHT_BORDER) {
		_rule.border = 0;
		_rule.context = 0;
		contextchanged = 1;
		_rule.keyword = 0;
	    } else if (rule.border & RULE_ON_LEFT_BORDER) {
		r = edit->rules[_rule._context];
		_rule.border = 0;
		if (r->between_delimiters) {
		    long e;
		    _rule.context = _rule._context;
		    contextchanged = 1;
		    _rule.keyword = 0;
		    if (r->first_right == c && (e = compare_word_to_right (edit, i, r->right, r->whole_word_chars_left, r->whole_word_chars_right, r->line_start_right)) >= end) {
			_rule.end = e;
			found_right = 1;
			_rule.border = RULE_ON_RIGHT_BORDER;
			_rule.context = 0;
		    }
		}
	    }
	}
	if (!found_right) {
	    int count;
	    struct context_rule **rules = edit->rules;
	    for (count = 1; rules[count]; count++) {
		r = rules[count];
		if (r->first_left == c) {
		    long e;
		    e = compare_word_to_right (edit, i, r->left, r->whole_word_chars_left, r->whole_word_chars_right, r->line_start_left);
		    if (e >= end && (!_rule.keyword || keyword_foundright)) {
			_rule.end = e;
			found_right = 1;
			_rule.border = RULE_ON_LEFT_BORDER;
			_rule._context = count;
			if (!r->between_delimiters)
			    if (!_rule.keyword)
				_rule.context = count;
			break;
		    }
		}
	    }
	}
    }
/* check again to turn on a keyword if the context switched */
    if (contextchanged && !_rule.keyword) {
	char *p;
	p = (r = edit->rules[_rule.context])->keyword_first_chars;
	while (*(p = xx_strchr ((unsigned char *) p + 1, c))) {
	    struct key_word *k;
	    int count;
	    long e;
	    count = (unsigned long) p - (unsigned long) r->keyword_first_chars;
	    k = r->keyword[count];
	    e = compare_word_to_right (edit, i, k->keyword, k->whole_word_chars_left, k->whole_word_chars_right, k->line_start);
	    if (e > 0) {
		_rule.end = e;
		_rule.keyword = count;
		break;
	    }
	}
    }
    return _rule;
}

static struct syntax_rule edit_get_rule (WEdit * edit, long byte_index)
{
    long i;
    if (edit->syntax_invalidate) {
	struct _syntax_marker *s;
	while (edit->syntax_marker && edit->syntax_marker->offset >= edit->last_get_rule) {
	    s = edit->syntax_marker->next;
	    syntax_free (edit->syntax_marker);
	    edit->syntax_marker = s;
	}
	if (edit->syntax_marker) {
	    edit->last_get_rule = edit->syntax_marker->offset;
	    edit->rule = edit->syntax_marker->rule;
	} else {
	    edit->last_get_rule = -1;
	    memset (&edit->rule, 0, sizeof (edit->rule));
	}
	edit->syntax_invalidate = 0;
    }
    if (byte_index > edit->last_get_rule) {
	for (i = edit->last_get_rule + 1; i <= byte_index; i++) {
	    edit->rule = apply_rules_going_right (edit, i, edit->rule);
	    if (i > (edit->syntax_marker ? edit->syntax_marker->offset + SYNTAX_MARKER_DENSITY : SYNTAX_MARKER_DENSITY)) {
		struct _syntax_marker *s;
		s = edit->syntax_marker;
		edit->syntax_marker = syntax_malloc (sizeof (struct _syntax_marker));
		edit->syntax_marker->next = s;
		edit->syntax_marker->offset = i;
		edit->syntax_marker->rule = edit->rule;
	    }
	}
    } else if (byte_index < edit->last_get_rule) {
	struct _syntax_marker *s;
	for (;;) {
	    if (!edit->syntax_marker) {
		memset (&edit->rule, 0, sizeof (edit->rule));
		for (i = -1; i <= byte_index; i++)
		    edit->rule = apply_rules_going_right (edit, i, edit->rule);
		break;
	    }
	    if (byte_index >= edit->syntax_marker->offset) {
		edit->rule = edit->syntax_marker->rule;
		for (i = edit->syntax_marker->offset + 1; i <= byte_index; i++)
		    edit->rule = apply_rules_going_right (edit, i, edit->rule);
		break;
	    }
	    s = edit->syntax_marker->next;
	    syntax_free (edit->syntax_marker);
	    edit->syntax_marker = s;
	}
    }
    edit->last_get_rule = byte_index;
    return edit->rule;
}

static void translate_rule_to_color (WEdit * edit, struct syntax_rule rule, int *fg, int *bg)
{
    struct key_word *k;
    k = edit->rules[rule.context]->keyword[rule.keyword];
    *bg = k->bg;
    *fg = k->fg;
}

void edit_get_syntax_color (WEdit * edit, long byte_index, int *fg, int *bg)
{
    if (edit->rules && byte_index < edit->last_byte && option_syntax_highlighting) {
	translate_rule_to_color (edit, edit_get_rule (edit, byte_index), fg, bg);
    } else {
#ifdef MIDNIGHT
	*fg = EDITOR_NORMAL_COLOR;
#else
	*fg = NO_COLOR;
	*bg = NO_COLOR;
#endif
    }
}


/*
   Returns 0 on error/eof or a count of the number of bytes read
   including the newline. Result must be free'd.
 */
#ifdef HAVE_MAD
static int mad_read_one_line (char **line, FILE * f, char *file, int line_)
#define read_one_line(a,b) mad_read_one_line(a,b,__FILE__,__LINE__)
#else
static int read_one_line (char **line, FILE * f)
#endif
{
    char *p;
    int len = 256, c, r = 0, i = 0;
#ifdef HAVE_MAD
    p = mad_syntax_malloc (len, file, line_);
#else
    p = syntax_malloc (len);
#endif
    for (;;) {
	c = fgetc (f);
	if (c == -1) {
	    if (errno == EINTR)
		continue;
	    r = 0;
	    break;
	} else if (c == '\n') {
	    r = i + 1;		/* extra 1 for the newline just read */
	    break;
	} else {
	    if (i >= len - 1) {
		char *q;
		q = syntax_malloc (len * 2);
		memcpy (q, p, len);
		syntax_free (p);
		p = q;
		len *= 2;
	    }
	    p[i++] = c;
	}
    }
    p[i] = 0;
    *line = p;
    return r;
}

static char *strdup_convert (char *s)
{
    char *r, *p;
    p = r = (char *) strdup (s);
    while (*s) {
	switch (*s) {
	case '\\':
	    s++;
	    switch (*s) {