Enlightenment CVS committal
Author : dj2
Project : e17
Module : libs/ewl
Dir : e17/libs/ewl/src/lib
Modified Files:
ewl_text.c
Log Message:
- patch from pfritz to make the utf8 stuff a bit smarter
===================================================================
RCS file: /cvs/e/e17/libs/ewl/src/lib/ewl_text.c,v
retrieving revision 1.127
retrieving revision 1.128
diff -u -3 -r1.127 -r1.128
--- ewl_text.c 14 Sep 2006 05:35:38 -0000 1.127
+++ ewl_text.c 23 Sep 2006 21:12:03 -0000 1.128
@@ -18,6 +18,21 @@
*/
static Ecore_Hash *context_hash = NULL;
+static const char ewl_text_trailing_bytes[256] = {
+ 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+ 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+ 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+ 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+ 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+ 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
+ 2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2, 2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,
+ 3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3, 4,4,4,4,4,4,4,4,5,5,5,5,6,6,6,6
+};
+
+/* returns length of next utf-8 sequence */
+#define EWL_TEXT_CHAR_BYTE_LEN(s) \
+ (ewl_text_trailing_bytes[(unsigned int)(unsigned char)((s)[0])])
+
static void ewl_text_context_cb_free(void *data);
static void ewl_text_context_print(Ewl_Text_Context *tx, const char *indent);
static char *ewl_text_context_name_get(Ewl_Text_Context *tx,
@@ -68,8 +83,11 @@
unsigned int byte_len,
unsigned int *char_idx,
unsigned int *char_len);
-static unsigned int ewl_text_char_length_get(const char *text);
-static char *ewl_text_text_next_char(const char *text, unsigned int *idx);
+static char *ewl_text_text_next_char(const char *text,
+ unsigned int *idx);
+static char *ewl_text_text_utf8_validate(const char *text,
+ unsigned int *char_len,
+ unsigned int *byte_len);
/**
@@ -331,9 +349,9 @@
}
else
{
- evas_textblock_cursor_char_geometry_get(cursor,
+ evas_textblock_cursor_char_geometry_get(cursor,
&cx, &cy, &cw, &ch);
- if (tx > (cx + ((cw + 1) >> 1)))
+ if (tx > (cx + ((cw + 1) >> 1)))
char_idx ++;
}
@@ -450,16 +468,15 @@
int char_len = 0;
int byte_len = 0;
unsigned int byte_idx;
+ char *valid_text = NULL;
DENTER_FUNCTION(DLEVEL_STABLE);
DCHECK_PARAM_PTR("t", t);
DCHECK_TYPE("t", t, EWL_TEXT_TYPE);
- if (text)
- {
- byte_len = strlen(text);
- char_len = ewl_text_char_length_get(text);
- }
+ if (text)
+ valid_text = ewl_text_text_utf8_validate(text, &char_len,
+ &byte_len);
/* Limit the index to be within safe boundaries */
if (char_idx > t->length.chars + 1)
@@ -497,7 +514,8 @@
t->text + byte_idx,
t->length.bytes -
byte_idx);
- memcpy(t->text + byte_idx, text, byte_len);
+ memcpy(t->text + byte_idx, valid_text, byte_len);
+ FREE(valid_text);
t->length.chars += char_len;
t->length.bytes += byte_len;
t->text[t->length.bytes] = '\0';
@@ -2348,7 +2366,7 @@
DCHECK_PARAM_PTR("t", t);
DCHECK_TYPE("t", t, EWL_TEXT_TYPE);
- child = ewl_text_tree_node_get(t->formatting.tree, char_idx, TRUE); /*
XXX TRUE or FALSE? */
+ child = ewl_text_tree_node_get(t->formatting.tree, char_idx, TRUE);
parent = child->parent;
while (parent)
{
@@ -2422,7 +2440,7 @@
DCHECK_TYPE("t", t, EWL_TEXT_TYPE);
child = ewl_text_tree_node_in_bytes_get(t->formatting.tree,
- byte_idx, TRUE); /* XXX TRUE or FALSE?
*/
+ byte_idx, TRUE);
parent = child->parent;
while (parent)
{
@@ -2479,67 +2497,144 @@
DLEAVE_FUNCTION(DLEVEL_STABLE);
}
-/* Counts the number of characters in the given piece of text. Assume the
- * text is utf8 so take that into account when counting. */
-static unsigned int
-ewl_text_char_length_get(const char *text)
+/*
+ * This function checks if a given character is a utf character.
+ * It only checks the first character in the string.
+ */
+static int
+ewl_text_char_is_legal_utf8(const char *c)
{
- unsigned int length = 0, idx;
- const char *t;
+ unsigned const char *t;
DENTER_FUNCTION(DLEVEL_STABLE);
+ DCHECK_PARAM_PTR_RET("c", c, FALSE);
- if (!text || (strlen(text) == 0))
- DRETURN_INT(length, DLEVEL_STABLE);
-
- t = text;
- while ((t = ewl_text_text_next_char(t, &idx)))
- length ++;
+ t = c;
+ if (!t) DRETURN_INT(FALSE, DLEVEL_STABLE);
+
+ if (t[0] < 0x80)
+ {
+ /*
+ * this a noraml 7-bit ASCII character
+ * -> legal utf8
+ */
+ DRETURN_INT(TRUE, DLEVEL_STABLE);
+ }
- DRETURN_INT(length, DLEVEL_STABLE);
+ switch (EWL_TEXT_CHAR_BYTE_LEN(t))
+ {
+ case 2:
+ /* 2 byte */
+ if ((t[1] & 0xc0) != 0x80)
+ DRETURN_INT(FALSE, DLEVEL_STABLE);
+ break;
+
+ case 3:
+ /* 3 byte */
+ if (((t[1] & 0xc0) != 0x80)
+ || ((t[2] & 0xc0) != 0x80))
+ DRETURN_INT(FALSE, DLEVEL_STABLE);
+ break;
+
+ case 4:
+ /* 4 byte */
+ if (((t[1] & 0xc0) != 0x80)
+ || ((t[2] & 0xc0) != 0x80)
+ || ((t[3] & 0xc0) != 0x80))
+ DRETURN_INT(FALSE, DLEVEL_STABLE);
+ break;
+
+ default:
+ /*
+ * this is actually:
+ * case 1:
+ * We already checked if it is a 7-bit ASCII
character,
+ * so anything else with the length of 1 byte is
not
+ * a valid utf8 character
+ * case 5: case 6:
+ * Although a character sequences of the length 5
or 6
+ * is possible it is not a legal utf8 character
+ */
+ return FALSE;
+ }
+
+ DRETURN_INT(TRUE, DLEVEL_STABLE);
}
-/* This is stolen from evas_common_font_utf8_get_next() */
+/*
+ * This function return the next character of a utf string.
+ * The text pointer should point on the leading byte of the
+ * current character, otherwise it will return the adress of
+ * the next byte.
+ */
static char *
ewl_text_text_next_char(const char *text, unsigned int *idx)
{
- unsigned char d, d2, d3, d4;
-
+ int len;
+
DENTER_FUNCTION(DLEVEL_STABLE);
+ DCHECK_PARAM_PTR_RET("text", text, NULL);
- *idx = 0;
+ len = EWL_TEXT_CHAR_BYTE_LEN(text);
+ if (idx) *idx = len;
- if (!text || (text[0] == '\0')) DRETURN_PTR(NULL, DLEVEL_STABLE);
+ DRETURN_PTR(text + len, DLEVEL_STABLE);
+}
- d = text[(*idx)++];
- if (!d) DRETURN_PTR(text + *idx, DLEVEL_STABLE); /* error .. */
+/*
+ * This function valdiates a a given utf-string and return a copy
+ * of it. Should the string contain illegal bytes, it will replace
+ * them with a question mark. This function doesn't check if the
+ * correspondending unicode exists for the single character nor
+ * if the font provides it.
+ */
+static char *
+ewl_text_text_utf8_validate(const char *text, unsigned int *char_len,
+ unsigned int *byte_len)
+{
+ char *t, *new_t;
+ unsigned int idx;
+ unsigned int c_len;
- if (d < 0x80)
- DRETURN_PTR(text + *idx, DLEVEL_STABLE);
+ DENTER_FUNCTION(DLEVEL_STABLE);
+ DCHECK_PARAM_PTR_RET("text", text, NULL);
+
+ new_t = t = strdup(text);
+ c_len = 0;
- if ((d & 0xe0) == 0xc0)
- {
- /* 2 byte */
- if (((d2 = text[(*idx)++]) & 0xc0) != 0x80)
- DRETURN_PTR(text + *idx, DLEVEL_STABLE); /* error .. */
- }
- else if ((d & 0xf0) == 0xe0)
+ while (*t)
{
- /* 3 byte */
- if ((((d2 = text[(*idx)++]) & 0xc0) != 0x80)
- || (((d3 = text[(*idx)++]) & 0xc0) != 0x80))
- DRETURN_PTR(text + *idx, DLEVEL_STABLE); /* error .. */
- }
- else
- {
- /* 4 byte */
- if ((((d2 = text[(*idx)++]) & 0xc0) != 0x80)
- || (((d3 = text[(*idx)++]) & 0xc0) != 0x80)
- || (((d4 = text[(*idx)++]) & 0xc0) != 0x80))
- DRETURN_PTR(text + *idx, DLEVEL_STABLE); /* error .. */
+ if (ewl_text_char_is_legal_utf8(t))
+ {
+ /*
+ * the current character is valid utf-character
+ * so we can jump to the next character
+ */
+ t = ewl_text_text_next_char(t, &idx);
+ }
+ else
+ {
+ /*
+ * oops, we found a illegal utf-character, or better
+ * something else. Replace this byte and hope
+ * the next one will be better :)
+ */
+ *t = '?';
+ t++;
+
+ printf("found a non utf8 character\n");
+ }
+ c_len++;
}
-
- DRETURN_PTR(text + *idx, DLEVEL_STABLE);
+
+ /*
+ * Well this is just a by-product, so we can use it
+ * without doing this loop again
+ */
+ if (char_len) *char_len = c_len;
+ if (byte_len) *byte_len = t - new_t;
+
+ DRETURN_PTR(new_t, DLEVEL_STABLE);
}
static void
@@ -2597,7 +2692,7 @@
txt = ewl_text_text_next_char(tmp, &idx);
}
- else if (*tmp == '\r' && *(ewl_text_text_next_char(tmp, &idx))
== '\n')
+ else if (*tmp == '\r' && *(tmp + 1) == '\n')
{
*tmp = '\0';
if (*txt) evas_textblock_cursor_text_append(cursor,
txt);
-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys -- and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
enlightenment-cvs mailing list
enlightenment-cvs@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-cvs
