rework the bracket searching code to handle multibyte bracket characters

git-svn-id: svn://svn.savannah.gnu.org/nano/trunk/nano@3258 35c25a1d-7b9e-4130-9fde-d3aeb78583b8

ChangeLog

@@ -7,6 +7,11 @@ CVS code -
 	  atom in the default regex string for quotestr, according to
 	  ASCII.  Changes to main(), nano.1, nanorc.5, and
 	  nanorc.sample. (DLR)
+	- Rework the bracket searching code to handle multibyte bracket
+	  characters.  New functions mbstrpbrk() and mbrevstrpbrk();
+	  changes to find_statusbar_bracket_match(),
+	  do_statusbar_find_bracket(), find_bracket_match(), and
+	  do_find_bracket(). (DLR)
 - chars.c:
   mbstrchr()
 	- Make parameter c const. (DLR)

src/chars.c

@@ -813,27 +813,7 @@ size_t mbstrnlen(const char *s, size_t maxlen)
 	return strnlen(s, maxlen);
 }
 
-#ifndef NANO_TINY
+#if !defined(NANO_TINY) || !defined(DISABLE_JUSTIFY)
-/* This function is equivalent to strpbrk(), except in that it scans the
- * string in reverse, starting at rev_start. */
-char *revstrpbrk(const char *s, const char *accept, const char
-	*rev_start)
-{
-    assert(s != NULL && accept != NULL && rev_start != NULL);
-
-    for (; rev_start >= s; rev_start--) {
-	const char *q = (*rev_start == '\0') ? NULL : strchr(accept,
-		*rev_start);
-
-	if (q != NULL)
-	    return (char *)rev_start;
-    }
-
-    return NULL;
-}
-#endif /* !NANO_TINY */
-
-#ifndef DISABLE_JUSTIFY
 /* This function is equivalent to strchr() for multibyte strings. */
 char *mbstrchr(const char *s, const char *c)
 {
@@ -879,8 +859,75 @@ char *mbstrchr(const char *s, const char *c)
 #endif
 	return strchr(s, *c);
 }
+#endif /* !NANO_TINY || !DISABLE_JUSTIFY */
 
-#ifdef ENABLE_NANORC
+#ifndef NANO_TINY
+/* This function is equivalent to strpbrk() for multibyte strings. */
+char *mbstrpbrk(const char *s, const char *accept)
+{
+    assert(s != NULL && accept != NULL);
+
+#ifdef ENABLE_UTF8
+    if (ISSET(USE_UTF8)) {
+	while (*s != '\0') {
+	    if (mbstrchr(accept, s) != NULL)
+		return (char *)s;
+
+	    s += move_mbright(s, 0);
+	}
+
+	return NULL;
+    } else
+#endif
+	return strpbrk(s, accept);
+}
+
+/* This function is equivalent to strpbrk(), except in that it scans the
+ * string in reverse, starting at rev_start. */
+char *revstrpbrk(const char *s, const char *accept, const char
+	*rev_start)
+{
+    assert(s != NULL && accept != NULL && rev_start != NULL);
+
+    for (; rev_start >= s; rev_start--) {
+	const char *q = (*rev_start == '\0') ? NULL : strchr(accept,
+		*rev_start);
+
+	if (q != NULL)
+	    return (char *)rev_start;
+    }
+
+    return NULL;
+}
+
+/* This function is equivalent to strpbrk() for multibyte strings,
+ * except in that it scans the string in reverse, starting at
+ * rev_start. */
+char *mbrevstrpbrk(const char *s, const char *accept, const char
+	*rev_start)
+{
+    assert(s != NULL && accept != NULL && rev_start != NULL);
+
+#ifdef ENABLE_UTF8
+    if (ISSET(USE_UTF8)) {
+	while (rev_start >= s) {
+	    const char *q = (*rev_start == '\0') ? NULL :
+		mbstrchr(accept, rev_start);
+
+	    if (q != NULL)
+		return (char *)rev_start;
+
+	    rev_start = s + move_mbleft(s, rev_start - s);
+	}
+
+	return NULL;
+    } else
+#endif
+	return revstrpbrk(s, accept, rev_start);
+}
+#endif /* !NANO_TINY */
+
+#if !defined(DISABLE_JUSTIFY) && defined(ENABLE_NANORC)
 /* Return TRUE if the string s contains one or more blank characters,
  * and FALSE otherwise. */
 bool has_blank_chars(const char *s)
@@ -926,8 +973,7 @@ bool has_blank_mbchars(const char *s)
 #endif
 	return has_blank_chars(s);
 }
-#endif /* ENABLE_NANORC */
+#endif /* !DISABLE_JUSTIFY && ENABLE_NANORC */
-#endif /* !DISABLE_JUSTIFY */
 
 #ifdef ENABLE_UTF8
 /* Return TRUE if wc is valid Unicode, and FALSE otherwise. */

src/prompt.c

@@ -3,7 +3,7 @@
  *   prompt.c                                                             *
  *                                                                        *
  *   Copyright (C) 1999-2004 Chris Allegretta                             *
- *   Copyright (C) 2005 David Lawrence Ramsey                             *
+ *   Copyright (C) 2005-2006 David Lawrence Ramsey                        *
  *   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, or (at your option)  *
@@ -672,7 +672,7 @@ bool find_statusbar_bracket_match(bool reverse, const char
 {
     const char *rev_start = NULL, *found = NULL;
 
-    assert(strlen(bracket_set) == 2);
+    assert(mbstrlen(bracket_set) == 2);
 
     /* rev_start might end up 1 character before the start or after the
      * end of the line.  This won't be a problem because we'll skip over
@@ -687,8 +687,8 @@ bool find_statusbar_bracket_match(bool reverse, const char
 	 * is found. */
 	found = ((rev_start > answer && *(rev_start - 1) == '\0') ||
 		rev_start < answer) ? NULL : (reverse ?
-		revstrpbrk(answer, bracket_set, rev_start) :
+		mbrevstrpbrk(answer, bracket_set, rev_start) :
-		strpbrk(rev_start, bracket_set));
+		mbstrpbrk(rev_start, bracket_set));
 
 	/* We've found a potential match. */
 	if (found != NULL)
@@ -711,48 +711,81 @@ bool find_statusbar_bracket_match(bool reverse, const char
 void do_statusbar_find_bracket(void)
 {
     size_t statusbar_x_save, pww_save;
-    const char *bracket_list = "()<>[]{}";
+    const char *bracket_list = "(<[{)>]}";
 	/* The list of brackets we can find matches to. */
-    const char *pos;
+    const char *ch;
 	/* The location in bracket_list of the bracket at the current
 	 * cursor position. */
-    char ch;
+    int ch_len;
-	/* The bracket at the current cursor position. */
+	/* The length of ch in bytes. */
-    char wanted_ch;
+    const char *wanted_ch;
-	/* The bracket complementing the bracket at the current cursor
+	/* The location in bracket_list of the bracket complementing the
-	 * position. */
+	 * bracket at the current cursor position. */
-    char bracket_set[3];
+    int wanted_ch_len;
+	/* The length of wanted_ch in bytes. */
+    char *bracket_set;
 	/* The pair of characters in ch and wanted_ch. */
+    size_t bracket_halflist;
+	/* The number of characters in one half of bracket_list. */
     size_t count = 1;
 	/* The initial bracket count. */
     bool reverse;
 	/* The direction we search. */
+    char *found_ch;
+	/* The character we find. */
 
-    assert(strlen(bracket_list) % 2 == 0);
+    assert(mbstrlen(bracket_list) % 2 == 0);
 
-    ch = answer[statusbar_x];
+    ch = answer + statusbar_x;
 
-    if (ch == '\0' || (pos = strchr(bracket_list, ch)) == NULL)
+    if (ch == '\0' || (ch = mbstrchr(bracket_list, ch)) == NULL)
 	return;
 
     /* Save where we are. */
     statusbar_x_save = statusbar_x;
     pww_save = statusbar_pww;
 
-    /* If we're on an opening bracket, we want to search forwards for a
+    /* If we're on an opening bracket, which must be in the first half
-     * closing bracket, and if we're on a closing bracket, we want to
+     * of bracket_list, we want to search forwards for a closing
-     * search backwards for an opening bracket. */
+     * bracket.  If we're on a closing bracket, which must be in the
-    reverse = ((pos - bracket_list) % 2 != 0);
+     * second half of bracket_list, we want to search backwards for an
-    wanted_ch = reverse ? *(pos - 1) : *(pos + 1);
+     * opening bracket. */
-    bracket_set[0] = ch;
+    bracket_halflist = mbstrlen(bracket_list) / 2;
-    bracket_set[1] = wanted_ch;
+    reverse = ((ch - bracket_list) > bracket_halflist);
-    bracket_set[2] = '\0';
+
+    /* If we're on an opening bracket, set wanted_ch to the character
+     * that's bracket_halflist characters after ch.  If we're on a
+     * closing bracket, set wanted_ch to the character that's
+     * bracket_halflist characters before ch. */
+    wanted_ch = ch;
+
+    while (bracket_halflist > 0) {
+	if (reverse)
+	    wanted_ch = bracket_list + move_mbleft(bracket_list,
+		wanted_ch - bracket_list);
+	else
+	    wanted_ch += move_mbright(wanted_ch, 0);
+
+	bracket_halflist--;
+    }
+
+    ch_len = parse_mbchar(ch, NULL, NULL);
+    wanted_ch_len = parse_mbchar(wanted_ch, NULL, NULL);
+
+    /* Fill bracket_set in with the values of ch and wanted_ch. */
+    bracket_set = charalloc((mb_cur_max() * 2) + 1);
+    strncpy(bracket_set, ch, ch_len);
+    strncpy(bracket_set + ch_len, wanted_ch, wanted_ch_len);
+    null_at(&bracket_set, ch_len + wanted_ch_len);
+
+    found_ch = charalloc(mb_cur_max() + 1);
 
     while (TRUE) {
 	if (find_statusbar_bracket_match(reverse, bracket_set)) {
 	    /* If we found an identical bracket, increment count.  If we
 	     * found a complementary bracket, decrement it. */
-	    count += (answer[statusbar_x] == ch) ? 1 : -1;
+	    parse_mbchar(answer + statusbar_x, found_ch, NULL);
+	    count += (strncmp(found_ch, ch, ch_len) == 0) ? 1 : -1;
 
 	    /* If count is zero, we've found a matching bracket.  Update
 	     * the statusbar prompt and get out. */
@@ -769,6 +802,10 @@ void do_statusbar_find_bracket(void)
 	    break;
 	}
     }
+
+    /* Clean up. */
+    free(bracket_set);
+    free(found_ch);
 }
 #endif /* !NANO_TINY */
 

src/proto.h

@@ -196,17 +196,20 @@ size_t mbstrlen(const char *s);
 size_t nstrnlen(const char *s, size_t maxlen);
 #endif
 size_t mbstrnlen(const char *s, size_t maxlen);
+#if !defined(NANO_TINY) || !defined(DISABLE_JUSTIFY)
+char *mbstrchr(const char *s, const char *c);
+#endif
 #ifndef NANO_TINY
+char *mbstrpbrk(const char *s, const char *accept);
 char *revstrpbrk(const char *s, const char *accept, const char
 	*rev_start);
+char *mbrevstrpbrk(const char *s, const char *accept, const char
+	*rev_start);
 #endif
-#ifndef DISABLE_JUSTIFY
+#if !defined(DISABLE_JUSTIFY) && defined(ENABLE_NANORC)
-char *mbstrchr(const char *s, const char *c);
-#ifdef ENABLE_NANORC
 bool has_blank_chars(const char *s);
 bool has_blank_mbchars(const char *s);
 #endif
-#endif
 #ifdef ENABLE_UTF8
 bool is_valid_unicode(wchar_t wc);
 #endif

src/search.c

@@ -3,7 +3,7 @@
  *   search.c                                                             *
  *                                                                        *
  *   Copyright (C) 1999-2004 Chris Allegretta                             *
- *   Copyright (C) 2005 David Lawrence Ramsey                             *
+ *   Copyright (C) 2005-2006 David Lawrence Ramsey                        *
  *   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, or (at your option)  *
@@ -276,9 +276,9 @@ bool findnextstr(
     filestruct *fileptr = openfile->current;
     const char *rev_start = NULL, *found = NULL;
     size_t found_len;
-	/* The length of the match we found. */
+	/* The length of the match we find. */
     size_t current_x_find = 0;
-	/* The location in the current line of the match we found. */
+	/* The location in the current line of the match we find. */
     ssize_t current_y_find = openfile->current_y;
 
     /* rev_start might end up 1 character before the start or after the
@@ -1076,7 +1076,7 @@ bool find_bracket_match(bool reverse, const char *bracket_set)
     const char *rev_start = NULL, *found = NULL;
     ssize_t current_y_find = openfile->current_y;
 
-    assert(strlen(bracket_set) == 2);
+    assert(mbstrlen(bracket_set) == 2);
 
     /* rev_start might end up 1 character before the start or after the
      * end of the line.  This won't be a problem because we'll skip over
@@ -1091,8 +1091,8 @@ bool find_bracket_match(bool reverse, const char *bracket_set)
     while (TRUE) {
 	found = ((rev_start > fileptr->data && *(rev_start - 1) ==
 		'\0') || rev_start < fileptr->data) ? NULL : (reverse ?
-		revstrpbrk(fileptr->data, bracket_set, rev_start) :
+		mbrevstrpbrk(fileptr->data, bracket_set, rev_start) :
-		strpbrk(rev_start, bracket_set));
+		mbstrpbrk(rev_start, bracket_set));
 
 	/* We've found a potential match. */
 	if (found != NULL)
@@ -1130,28 +1130,34 @@ void do_find_bracket(void)
 {
     filestruct *current_save;
     size_t current_x_save, pww_save;
-    const char *bracket_list = "()<>[]{}";
+    const char *bracket_list = "(<[{)>]}";
 	/* The list of brackets we can find matches to. */
-    const char *pos;
+    const char *ch;
 	/* The location in bracket_list of the bracket at the current
 	 * cursor position. */
-    char ch;
+    int ch_len;
-	/* The bracket at the current cursor position. */
+	/* The length of ch in bytes. */
-    char wanted_ch;
+    const char *wanted_ch;
-	/* The bracket complementing the bracket at the current cursor
+	/* The location in bracket_list of the bracket complementing the
-	 * position. */
+	 * bracket at the current cursor position. */
-    char bracket_set[3];
+    int wanted_ch_len;
+	/* The length of wanted_ch in bytes. */
+    char *bracket_set;
 	/* The pair of characters in ch and wanted_ch. */
+    size_t bracket_halflist;
+	/* The number of characters in one half of bracket_list. */
     size_t count = 1;
 	/* The initial bracket count. */
     bool reverse;
 	/* The direction we search. */
+    char *found_ch;
+	/* The character we find. */
 
-    assert(strlen(bracket_list) % 2 == 0);
+    assert(mbstrlen(bracket_list) % 2 == 0);
 
-    ch = openfile->current->data[openfile->current_x];
+    ch = openfile->current->data + openfile->current_x;
 
-    if (ch == '\0' || (pos = strchr(bracket_list, ch)) == NULL) {
+    if (ch == '\0' || (ch = mbstrchr(bracket_list, ch)) == NULL) {
 	statusbar(_("Not a bracket"));
 	return;
     }
@@ -1161,21 +1167,48 @@ void do_find_bracket(void)
     current_x_save = openfile->current_x;
     pww_save = openfile->placewewant;
 
-    /* If we're on an opening bracket, we want to search forwards for a
+    /* If we're on an opening bracket, which must be in the first half
-     * closing bracket, and if we're on a closing bracket, we want to
+     * of bracket_list, we want to search forwards for a closing
-     * search backwards for an opening bracket. */
+     * bracket.  If we're on a closing bracket, which must be in the
-    reverse = ((pos - bracket_list) % 2 != 0);
+     * second half of bracket_list, we want to search backwards for an
-    wanted_ch = reverse ? *(pos - 1) : *(pos + 1);
+     * opening bracket. */
-    bracket_set[0] = ch;
+    bracket_halflist = mbstrlen(bracket_list) / 2;
-    bracket_set[1] = wanted_ch;
+    reverse = ((ch - bracket_list) > bracket_halflist);
-    bracket_set[2] = '\0';
+
+    /* If we're on an opening bracket, set wanted_ch to the character
+     * that's bracket_halflist characters after ch.  If we're on a
+     * closing bracket, set wanted_ch to the character that's
+     * bracket_halflist characters before ch. */
+    wanted_ch = ch;
+
+    while (bracket_halflist > 0) {
+	if (reverse)
+	    wanted_ch = bracket_list + move_mbleft(bracket_list,
+		wanted_ch - bracket_list);
+	else
+	    wanted_ch += move_mbright(wanted_ch, 0);
+
+	bracket_halflist--;
+    }
+
+    ch_len = parse_mbchar(ch, NULL, NULL);
+    wanted_ch_len = parse_mbchar(wanted_ch, NULL, NULL);
+
+    /* Fill bracket_set in with the values of ch and wanted_ch. */
+    bracket_set = charalloc((mb_cur_max() * 2) + 1);
+    strncpy(bracket_set, ch, ch_len);
+    strncpy(bracket_set + ch_len, wanted_ch, wanted_ch_len);
+    null_at(&bracket_set, ch_len + wanted_ch_len);
+
+    found_ch = charalloc(mb_cur_max() + 1);
 
     while (TRUE) {
 	if (find_bracket_match(reverse, bracket_set)) {
 	    /* If we found an identical bracket, increment count.  If we
 	     * found a complementary bracket, decrement it. */
-	    count += (openfile->current->data[openfile->current_x] ==
+	    parse_mbchar(openfile->current->data + openfile->current_x,
-		ch) ? 1 : -1;
+		found_ch, NULL);
+	    count += (strncmp(found_ch, ch, ch_len) == 0) ? 1 : -1;
 
 	    /* If count is zero, we've found a matching bracket.  Update
 	     * the screen and get out. */
@@ -1193,6 +1226,10 @@ void do_find_bracket(void)
 	    break;
 	}
     }
+
+    /* Clean up. */
+    free(bracket_set);
+    free(found_ch);
 }
 
 #ifdef ENABLE_NANORC