summaryrefslogtreecommitdiff
path: root/lib/unistr/u8-mbtouc-unsafe.c
diff options
context:
space:
mode:
Diffstat (limited to 'lib/unistr/u8-mbtouc-unsafe.c')
-rw-r--r--lib/unistr/u8-mbtouc-unsafe.c140
1 files changed, 14 insertions, 126 deletions
diff --git a/lib/unistr/u8-mbtouc-unsafe.c b/lib/unistr/u8-mbtouc-unsafe.c
index 01d12dc4..679a686b 100644
--- a/lib/unistr/u8-mbtouc-unsafe.c
+++ b/lib/unistr/u8-mbtouc-unsafe.c
@@ -1,12 +1,21 @@
/* Look at first character in UTF-8 string.
- Copyright (C) 1999-2002, 2006-2007, 2009-2015 Free Software Foundation, Inc.
+ Copyright (C) 1999-2002, 2006-2007, 2009-2016 Free Software Foundation, Inc.
Written by Bruno Haible <bruno@clisp.org>, 2001.
- This program is free software: you can redistribute it and/or modify it
- under the terms of the GNU Lesser General Public License as published
- by the Free Software Foundation; either version 3 of the License, or
- (at your option) any later version.
+ This program is free software: you can redistribute it and/or
+ modify it under the terms of either:
+ * the GNU Lesser General Public License as published by the Free
+ Software Foundation; either version 3 of the License, or (at your
+ option) any later version.
+
+ or
+
+ * 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.
+
+ or both in parallel, as here.
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
@@ -44,17 +53,13 @@ u8_mbtouc_unsafe (ucs4_t *puc, const uint8_t *s, size_t n)
{
if (n >= 2)
{
-#if CONFIG_UNICODE_SAFETY
if ((s[1] ^ 0x80) < 0x40)
-#endif
{
*puc = ((unsigned int) (c & 0x1f) << 6)
| (unsigned int) (s[1] ^ 0x80);
return 2;
}
-#if CONFIG_UNICODE_SAFETY
/* invalid multibyte character */
-#endif
}
else
{
@@ -67,21 +72,18 @@ u8_mbtouc_unsafe (ucs4_t *puc, const uint8_t *s, size_t n)
{
if (n >= 3)
{
-#if CONFIG_UNICODE_SAFETY
if ((s[1] ^ 0x80) < 0x40)
{
if ((s[2] ^ 0x80) < 0x40)
{
if ((c >= 0xe1 || s[1] >= 0xa0)
&& (c != 0xed || s[1] < 0xa0))
-#endif
{
*puc = ((unsigned int) (c & 0x0f) << 12)
| ((unsigned int) (s[1] ^ 0x80) << 6)
| (unsigned int) (s[2] ^ 0x80);
return 3;
}
-#if CONFIG_UNICODE_SAFETY
/* invalid multibyte character */
*puc = 0xfffd;
return 3;
@@ -91,7 +93,6 @@ u8_mbtouc_unsafe (ucs4_t *puc, const uint8_t *s, size_t n)
return 2;
}
/* invalid multibyte character */
-#endif
}
else
{
@@ -107,7 +108,6 @@ u8_mbtouc_unsafe (ucs4_t *puc, const uint8_t *s, size_t n)
{
if (n >= 4)
{
-#if CONFIG_UNICODE_SAFETY
if ((s[1] ^ 0x80) < 0x40)
{
if ((s[2] ^ 0x80) < 0x40)
@@ -115,11 +115,8 @@ u8_mbtouc_unsafe (ucs4_t *puc, const uint8_t *s, size_t n)
if ((s[3] ^ 0x80) < 0x40)
{
if ((c >= 0xf1 || s[1] >= 0x90)
-#if 1
&& (c < 0xf4 || (c == 0xf4 && s[1] < 0x90))
-#endif
)
-#endif
{
*puc = ((unsigned int) (c & 0x07) << 18)
| ((unsigned int) (s[1] ^ 0x80) << 12)
@@ -127,7 +124,6 @@ u8_mbtouc_unsafe (ucs4_t *puc, const uint8_t *s, size_t n)
| (unsigned int) (s[3] ^ 0x80);
return 4;
}
-#if CONFIG_UNICODE_SAFETY
/* invalid multibyte character */
*puc = 0xfffd;
return 4;
@@ -141,7 +137,6 @@ u8_mbtouc_unsafe (ucs4_t *puc, const uint8_t *s, size_t n)
return 2;
}
/* invalid multibyte character */
-#endif
}
else
{
@@ -155,113 +150,6 @@ u8_mbtouc_unsafe (ucs4_t *puc, const uint8_t *s, size_t n)
return 3;
}
}
-#if 0
- else if (c < 0xfc)
- {
- if (n >= 5)
- {
-#if CONFIG_UNICODE_SAFETY
- if ((s[1] ^ 0x80) < 0x40)
- {
- if ((s[2] ^ 0x80) < 0x40)
- {
- if ((s[3] ^ 0x80) < 0x40)
- {
- if ((s[4] ^ 0x80) < 0x40)
- {
- if (c >= 0xf9 || s[1] >= 0x88)
-#endif
- {
- *puc = ((unsigned int) (c & 0x03) << 24)
- | ((unsigned int) (s[1] ^ 0x80) << 18)
- | ((unsigned int) (s[2] ^ 0x80) << 12)
- | ((unsigned int) (s[3] ^ 0x80) << 6)
- | (unsigned int) (s[4] ^ 0x80);
- return 5;
- }
-#if CONFIG_UNICODE_SAFETY
- /* invalid multibyte character */
- *puc = 0xfffd;
- return 5;
- }
- /* invalid multibyte character */
- *puc = 0xfffd;
- return 4;
- }
- /* invalid multibyte character */
- *puc = 0xfffd;
- return 3;
- }
- /* invalid multibyte character */
- return 2;
- }
- /* invalid multibyte character */
-#endif
- }
- else
- {
- /* incomplete multibyte character */
- *puc = 0xfffd;
- return n;
- }
- }
- else if (c < 0xfe)
- {
- if (n >= 6)
- {
-#if CONFIG_UNICODE_SAFETY
- if ((s[1] ^ 0x80) < 0x40)
- {
- if ((s[2] ^ 0x80) < 0x40)
- {
- if ((s[3] ^ 0x80) < 0x40)
- {
- if ((s[4] ^ 0x80) < 0x40)
- {
- if ((s[5] ^ 0x80) < 0x40)
- {
- if (c >= 0xfd || s[1] >= 0x84)
-#endif
- {
- *puc = ((unsigned int) (c & 0x01) << 30)
- | ((unsigned int) (s[1] ^ 0x80) << 24)
- | ((unsigned int) (s[2] ^ 0x80) << 18)
- | ((unsigned int) (s[3] ^ 0x80) << 12)
- | ((unsigned int) (s[4] ^ 0x80) << 6)
- | (unsigned int) (s[5] ^ 0x80);
- return 6;
- }
-#if CONFIG_UNICODE_SAFETY
- /* invalid multibyte character */
- *puc = 0xfffd;
- return 6;
- }
- /* invalid multibyte character */
- *puc = 0xfffd;
- return 5;
- }
- /* invalid multibyte character */
- *puc = 0xfffd;
- return 4;
- }
- /* invalid multibyte character */
- *puc = 0xfffd;
- return 3;
- }
- /* invalid multibyte character */
- return 2;
- }
- /* invalid multibyte character */
-#endif
- }
- else
- {
- /* incomplete multibyte character */
- *puc = 0xfffd;
- return n;
- }
- }
-#endif
}
/* invalid multibyte character */
*puc = 0xfffd;
generated by cgit v1.2.3 (git 2.25.1) at 2024-11-11 03:14:06 +0000