===================================================================
RCS file: /cvsrepo/anoncvs/cvs/src/usr.bin/tmux/utf8.c,v
retrieving revision 1.16
retrieving revision 1.17
diff -c -r1.16 -r1.17
*** src/usr.bin/tmux/utf8.c	2015/11/12 11:10:50	1.16
--- src/usr.bin/tmux/utf8.c	2015/11/12 12:19:57	1.17
***************
*** 1,4 ****
! /* $OpenBSD: utf8.c,v 1.16 2015/11/12 11:10:50 nicm Exp $ */
  
  /*
   * Copyright (c) 2008 Nicholas Marriott <nicm@users.sourceforge.net>
--- 1,4 ----
! /* $OpenBSD: utf8.c,v 1.17 2015/11/12 12:19:57 nicm Exp $ */
  
  /*
   * Copyright (c) 2008 Nicholas Marriott <nicm@users.sourceforge.net>
***************
*** 35,41 ****
  };
  
  /* Sorted, then repeatedly split in the middle to balance the tree. */
! struct utf8_width_entry utf8_width_table[] = {
  	{ 0x00b41, 0x00b44, 0, NULL, NULL },
  	{ 0x008e4, 0x00902, 0, NULL, NULL },
  	{ 0x006d6, 0x006dd, 0, NULL, NULL },
--- 35,41 ----
  };
  
  /* Sorted, then repeatedly split in the middle to balance the tree. */
! static struct utf8_width_entry utf8_width_table[] = {
  	{ 0x00b41, 0x00b44, 0, NULL, NULL },
  	{ 0x008e4, 0x00902, 0, NULL, NULL },
  	{ 0x006d6, 0x006dd, 0, NULL, NULL },
***************
*** 344,356 ****
  	{ 0xe0100, 0xe01ef, 0, NULL, NULL },
  	{ 0x100000, 0x10fffd, 0, NULL, NULL },
  };
  
! struct utf8_width_entry	*utf8_width_root = NULL;
  
- int	utf8_overlap(struct utf8_width_entry *, struct utf8_width_entry *);
- u_int	utf8_combine(const struct utf8_data *);
- u_int	utf8_width(const struct utf8_data *);
- 
  /* Set a single character. */
  void
  utf8_set(struct utf8_data *utf8data, u_char ch)
--- 344,353 ----
  	{ 0xe0100, 0xe01ef, 0, NULL, NULL },
  	{ 0x100000, 0x10fffd, 0, NULL, NULL },
  };
+ static struct utf8_width_entry	*utf8_width_root = NULL;
  
! static void	utf8_build(void);
  
  /* Set a single character. */
  void
  utf8_set(struct utf8_data *utf8data, u_char ch)
***************
*** 405,444 ****
  	if (utf8data->have != utf8data->size)
  		return (1);
  
! 	utf8data->width = utf8_width(utf8data);
  	return (0);
  }
  
- /* Check if two width tree entries overlap. */
- int
- utf8_overlap(struct utf8_width_entry *item1, struct utf8_width_entry *item2)
- {
- 	if (item1->first >= item2->first && item1->first <= item2->last)
- 		return (1);
- 	if (item1->last >= item2->first && item1->last <= item2->last)
- 		return (1);
- 	if (item2->first >= item1->first && item2->first <= item1->last)
- 		return (1);
- 	if (item2->last >= item1->first && item2->last <= item1->last)
- 		return (1);
- 	return (0);
- }
- 
  /* Build UTF-8 width tree. */
! void
  utf8_build(void)
  {
  	struct utf8_width_entry	**ptr, *item, *node;
! 	u_int			  i, j;
  
  	for (i = 0; i < nitems(utf8_width_table); i++) {
  		item = &utf8_width_table[i];
  
- 		for (j = 0; j < nitems(utf8_width_table); j++) {
- 			if (i != j && utf8_overlap(item, &utf8_width_table[j]))
- 				log_fatalx("utf8 overlap: %u %u", i, j);
- 		}
- 
  		ptr = &utf8_width_root;
  		while (*ptr != NULL) {
  			node = *ptr;
--- 402,421 ----
  	if (utf8data->have != utf8data->size)
  		return (1);
  
! 	utf8data->width = utf8_width(utf8_combine(utf8data));
  	return (0);
  }
  
  /* Build UTF-8 width tree. */
! static void
  utf8_build(void)
  {
  	struct utf8_width_entry	**ptr, *item, *node;
! 	u_int			  i;
  
  	for (i = 0; i < nitems(utf8_width_table); i++) {
  		item = &utf8_width_table[i];
  
  		ptr = &utf8_width_root;
  		while (*ptr != NULL) {
  			node = *ptr;
***************
*** 451,456 ****
--- 428,454 ----
  	}
  }
  
+ /* Lookup width of UTF-8 data in tree. */
+ u_int
+ utf8_width(u_int uc)
+ {
+ 	struct utf8_width_entry	*item;
+ 
+ 	if (utf8_width_root == NULL)
+ 		utf8_build();
+ 
+ 	item = utf8_width_root;
+ 	while (item != NULL) {
+ 		if (uc < item->first)
+ 			item = item->left;
+ 		else if (uc > item->last)
+ 			item = item->right;
+ 		else
+ 			return (item->width);
+ 	}
+ 	return (1);
+ }
+ 
  /* Combine UTF-8 into 32-bit Unicode. */
  u_int
  utf8_combine(const struct utf8_data *utf8data)
***************
*** 481,487 ****
  	return (value);
  }
  
! /* Split a UTF-8 character. */
  int
  utf8_split(u_int uc, struct utf8_data *utf8data)
  {
--- 479,485 ----
  	return (value);
  }
  
! /* Split 32-bit Unicode into UTF-8. */
  int
  utf8_split(u_int uc, struct utf8_data *utf8data)
  {
***************
*** 505,511 ****
  		utf8data->data[3] = 0x80 | (uc & 0x3f);
  	} else
  		return (-1);
! 	utf8data->width = utf8_width(utf8data);
  	return (0);
  }
  
--- 503,509 ----
  		utf8data->data[3] = 0x80 | (uc & 0x3f);
  	} else
  		return (-1);
! 	utf8data->width = utf8_width(uc);
  	return (0);
  }
  
***************
*** 519,545 ****
  		return (2);
  	}
  	ptr[0] = uc;
- 	return (1);
- }
- 
- /* Lookup width of UTF-8 data in tree. */
- u_int
- utf8_width(const struct utf8_data *utf8data)
- {
- 	struct utf8_width_entry	*item;
- 	u_int			 value;
- 
- 	value = utf8_combine(utf8data);
- 
- 	item = utf8_width_root;
- 	while (item != NULL) {
- 		if (value < item->first)
- 			item = item->left;
- 		else if (value > item->last)
- 			item = item->right;
- 		else
- 			return (item->width);
- 	}
  	return (1);
  }
  
--- 517,522 ----