X-Git-Url: http://git.ipfire.org/?a=blobdiff_plain;f=src%2Fbasic%2Futf8.c;h=afc24700dd77a786d8cf9392916dece1b3785c3a;hb=4c6d51390c66ccbf647c2943657e8bb693312f38;hp=4da9a405cb85914c13457f5a6a4c0b6358499353;hpb=4ff183d419f9ec271ae311dae0bc01b4bb204b50;p=thirdparty%2Fsystemd.git
diff --git a/src/basic/utf8.c b/src/basic/utf8.c
index 4da9a405cb8..afc24700dd7 100644
--- a/src/basic/utf8.c
+++ b/src/basic/utf8.c
@@ -1,23 +1,4 @@
/* SPDX-License-Identifier: LGPL-2.1+ */
-/***
- This file is part of systemd.
-
- Copyright 2008-2011 Kay Sievers
- Copyright 2012 Lennart Poettering
-
- systemd 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 2.1 of the License, or
- (at your option) any later version.
-
- systemd 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
- Lesser General Public License for more details.
-
- You should have received a copy of the GNU Lesser General Public License
- along with systemd; If not, see .
-***/
/* Parts of this file are based on the GLIB utf8 validation functions. The
* original license text follows. */
@@ -48,8 +29,10 @@
#include
#include "alloc-util.h"
+#include "gunicode.h"
#include "hexdecoct.h"
#include "macro.h"
+#include "string-util.h"
#include "utf8.h"
bool unichar_is_valid(char32_t ch) {
@@ -79,12 +62,7 @@ static bool unichar_is_control(char32_t ch) {
}
/* count of characters used to encode one unicode char */
-static int utf8_encoded_expected_len(const char *str) {
- unsigned char c;
-
- assert(str);
-
- c = (unsigned char) str[0];
+static size_t utf8_encoded_expected_len(uint8_t c) {
if (c < 0x80)
return 1;
if ((c & 0xe0) == 0xc0)
@@ -104,11 +82,11 @@ static int utf8_encoded_expected_len(const char *str) {
/* decode one unicode char */
int utf8_encoded_to_unichar(const char *str, char32_t *ret_unichar) {
char32_t unichar;
- int len, i;
+ size_t len, i;
assert(str);
- len = utf8_encoded_expected_len(str);
+ len = utf8_encoded_expected_len(str[0]);
switch (len) {
case 1:
@@ -136,6 +114,7 @@ int utf8_encoded_to_unichar(const char *str, char32_t *ret_unichar) {
for (i = 1; i < len; i++) {
if (((char32_t)str[i] & 0xc0) != 0x80)
return -EINVAL;
+
unichar <<= 6;
unichar |= (char32_t)str[i] & 0x3f;
}
@@ -150,14 +129,14 @@ bool utf8_is_printable_newline(const char* str, size_t length, bool newline) {
assert(str);
- for (p = str; length;) {
+ for (p = str; length > 0;) {
int encoded_len, r;
char32_t val;
- encoded_len = utf8_encoded_valid_unichar(p);
- if (encoded_len < 0 ||
- (size_t) encoded_len > length)
+ encoded_len = utf8_encoded_valid_unichar(p, length);
+ if (encoded_len < 0)
return false;
+ assert(encoded_len > 0 && (size_t) encoded_len <= length);
r = utf8_encoded_to_unichar(p, &val);
if (r < 0 ||
@@ -172,22 +151,23 @@ bool utf8_is_printable_newline(const char* str, size_t length, bool newline) {
return true;
}
-const char *utf8_is_valid(const char *str) {
- const uint8_t *p;
+char *utf8_is_valid(const char *str) {
+ const char *p;
assert(str);
- for (p = (const uint8_t*) str; *p; ) {
+ p = str;
+ while (*p) {
int len;
- len = utf8_encoded_valid_unichar((const char *)p);
+ len = utf8_encoded_valid_unichar(p, (size_t) -1);
if (len < 0)
return NULL;
p += len;
}
- return str;
+ return (char*) str;
}
char *utf8_escape_invalid(const char *str) {
@@ -202,7 +182,7 @@ char *utf8_escape_invalid(const char *str) {
while (*str) {
int len;
- len = utf8_encoded_valid_unichar(str);
+ len = utf8_encoded_valid_unichar(str, (size_t) -1);
if (len > 0) {
s = mempcpy(s, str, len);
str += len;
@@ -213,52 +193,102 @@ char *utf8_escape_invalid(const char *str) {
}
*s = '\0';
-
+ (void) str_realloc(&p);
return p;
}
-char *utf8_escape_non_printable(const char *str) {
- char *p, *s;
+static int utf8_char_console_width(const char *str) {
+ char32_t c;
+ int r;
+
+ r = utf8_encoded_to_unichar(str, &c);
+ if (r < 0)
+ return r;
+
+ /* TODO: we should detect combining characters */
+
+ return unichar_iswide(c) ? 2 : 1;
+}
+
+char *utf8_escape_non_printable_full(const char *str, size_t console_width) {
+ char *p, *s, *prev_s;
+ size_t n = 0; /* estimated print width */
assert(str);
- p = s = malloc(strlen(str) * 4 + 1);
+ if (console_width == 0)
+ return strdup("");
+
+ p = s = prev_s = malloc(strlen(str) * 4 + 1);
if (!p)
return NULL;
- while (*str) {
+ for (;;) {
int len;
+ char *saved_s = s;
- len = utf8_encoded_valid_unichar(str);
+ if (!*str) /* done! */
+ goto finish;
+
+ len = utf8_encoded_valid_unichar(str, (size_t) -1);
if (len > 0) {
if (utf8_is_printable(str, len)) {
+ int w;
+
+ w = utf8_char_console_width(str);
+ assert(w >= 0);
+ if (n + w > console_width)
+ goto truncation;
+
s = mempcpy(s, str, len);
str += len;
+ n += w;
+
} else {
- while (len > 0) {
+ for (; len > 0; len--) {
+ if (n + 4 > console_width)
+ goto truncation;
+
*(s++) = '\\';
*(s++) = 'x';
*(s++) = hexchar((int) *str >> 4);
*(s++) = hexchar((int) *str);
str += 1;
- len--;
+ n += 4;
}
}
} else {
- s = stpcpy(s, UTF8_REPLACEMENT_CHARACTER);
+ if (n + 1 > console_width)
+ goto truncation;
+
+ s = mempcpy(s, UTF8_REPLACEMENT_CHARACTER, strlen(UTF8_REPLACEMENT_CHARACTER));
str += 1;
+ n += 1;
}
+
+ prev_s = saved_s;
}
- *s = '\0';
+ truncation:
+ /* Try to go back one if we don't have enough space for the ellipsis */
+ if (n + 1 >= console_width)
+ s = prev_s;
+
+ s = mempcpy(s, "â¦", strlen("â¦"));
+ finish:
+ *s = '\0';
+ (void) str_realloc(&p);
return p;
}
char *ascii_is_valid(const char *str) {
const char *p;
+ /* Check whether the string consists of valid ASCII bytes,
+ * i.e values between 0 and 127, inclusive. */
+
assert(str);
for (p = str; *p; p++)
@@ -268,6 +298,21 @@ char *ascii_is_valid(const char *str) {
return (char*) str;
}
+char *ascii_is_valid_n(const char *str, size_t len) {
+ size_t i;
+
+ /* Very similar to ascii_is_valid(), but checks exactly len
+ * bytes and rejects any NULs in that range. */
+
+ assert(str);
+
+ for (i = 0; i < len; i++)
+ if ((unsigned char) str[i] >= 128 || str[i] == 0)
+ return NULL;
+
+ return (char*) str;
+}
+
/**
* utf8_encode_unichar() - Encode single UCS-4 character as UTF-8
* @out_utf8: output buffer of at least 4 bytes or NULL
@@ -312,18 +357,25 @@ size_t utf8_encode_unichar(char *out_utf8, char32_t g) {
return 0;
}
-char *utf16_to_utf8(const void *s, size_t length) {
+char *utf16_to_utf8(const char16_t *s, size_t length /* bytes! */) {
const uint8_t *f;
char *r, *t;
- r = new(char, (length * 4 + 1) / 2 + 1);
+ assert(s);
+
+ /* Input length is in bytes, i.e. the shortest possible character takes 2 bytes. Each unicode character may
+ * take up to 4 bytes in UTF-8. Let's also account for a trailing NUL byte. */
+ if (length * 2 < length)
+ return NULL; /* overflow */
+
+ r = new(char, length * 2 + 1);
if (!r)
return NULL;
- f = s;
+ f = (const uint8_t*) s;
t = r;
- while (f < (const uint8_t*) s + length) {
+ while (f + 1 < (const uint8_t*) s + length) {
char16_t w1, w2;
/* see RFC 2781 section 2.2 */
@@ -333,13 +385,13 @@ char *utf16_to_utf8(const void *s, size_t length) {
if (!utf16_is_surrogate(w1)) {
t += utf8_encode_unichar(t, w1);
-
continue;
}
if (utf16_is_trailing_surrogate(w1))
- continue;
- else if (f >= (const uint8_t*) s + length)
+ continue; /* spurious trailing surrogate, ignore */
+
+ if (f + 1 >= (const uint8_t*) s + length)
break;
w2 = f[1] << 8 | f[0];
@@ -347,7 +399,7 @@ char *utf16_to_utf8(const void *s, size_t length) {
if (!utf16_is_trailing_surrogate(w2)) {
f -= 2;
- continue;
+ continue; /* surrogate missing its trailing surrogate, ignore */
}
t += utf8_encode_unichar(t, utf16_surrogate_pair_to_unichar(w1, w2));
@@ -357,6 +409,79 @@ char *utf16_to_utf8(const void *s, size_t length) {
return r;
}
+size_t utf16_encode_unichar(char16_t *out, char32_t c) {
+
+ /* Note that this encodes as little-endian. */
+
+ switch (c) {
+
+ case 0 ... 0xd7ffU:
+ case 0xe000U ... 0xffffU:
+ out[0] = htole16(c);
+ return 1;
+
+ case 0x10000U ... 0x10ffffU:
+ c -= 0x10000U;
+ out[0] = htole16((c >> 10) + 0xd800U);
+ out[1] = htole16((c & 0x3ffU) + 0xdc00U);
+ return 2;
+
+ default: /* A surrogate (invalid) */
+ return 0;
+ }
+}
+
+char16_t *utf8_to_utf16(const char *s, size_t length) {
+ char16_t *n, *p;
+ size_t i;
+ int r;
+
+ assert(s);
+
+ n = new(char16_t, length + 1);
+ if (!n)
+ return NULL;
+
+ p = n;
+
+ for (i = 0; i < length;) {
+ char32_t unichar;
+ size_t e;
+
+ e = utf8_encoded_expected_len(s[i]);
+ if (e <= 1) /* Invalid and single byte characters are copied as they are */
+ goto copy;
+
+ if (i + e > length) /* sequence longer than input buffer, then copy as-is */
+ goto copy;
+
+ r = utf8_encoded_to_unichar(s + i, &unichar);
+ if (r < 0) /* sequence invalid, then copy as-is */
+ goto copy;
+
+ p += utf16_encode_unichar(p, unichar);
+ i += e;
+ continue;
+
+ copy:
+ *(p++) = htole16(s[i++]);
+ }
+
+ *p = 0;
+ return n;
+}
+
+size_t char16_strlen(const char16_t *s) {
+ size_t n = 0;
+
+ assert(s);
+
+ while (*s != 0)
+ n++, s++;
+
+ return n;
+}
+
/* expected size used to encode one unicode char */
static int utf8_unichar_to_encoded_len(char32_t unichar) {
@@ -375,16 +500,24 @@ static int utf8_unichar_to_encoded_len(char32_t unichar) {
}
/* validate one encoded unicode char and return its length */
-int utf8_encoded_valid_unichar(const char *str) {
- int len, i, r;
+int utf8_encoded_valid_unichar(const char *str, size_t length /* bytes */) {
char32_t unichar;
+ size_t len, i;
+ int r;
assert(str);
+ assert(length > 0);
+
+ /* We read until NUL, at most length bytes. (size_t) -1 may be used to disable the length check. */
- len = utf8_encoded_expected_len(str);
+ len = utf8_encoded_expected_len(str[0]);
if (len == 0)
return -EINVAL;
+ /* Do we have a truncated multi-byte character? */
+ if (len > length)
+ return -EINVAL;
+
/* ascii is valid */
if (len == 1)
return 1;
@@ -399,12 +532,51 @@ int utf8_encoded_valid_unichar(const char *str) {
return r;
/* check if encoded length matches encoded value */
- if (utf8_unichar_to_encoded_len(unichar) != len)
+ if (utf8_unichar_to_encoded_len(unichar) != (int) len)
return -EINVAL;
/* check if value has valid range */
if (!unichar_is_valid(unichar))
return -EINVAL;
- return len;
+ return (int) len;
+}
+
+size_t utf8_n_codepoints(const char *str) {
+ size_t n = 0;
+
+ /* Returns the number of UTF-8 codepoints in this string, or (size_t) -1 if the string is not valid UTF-8. */
+
+ while (*str != 0) {
+ int k;
+
+ k = utf8_encoded_valid_unichar(str, (size_t) -1);
+ if (k < 0)
+ return (size_t) -1;
+
+ str += k;
+ n++;
+ }
+
+ return n;
+}
+
+size_t utf8_console_width(const char *str) {
+ size_t n = 0;
+
+ /* Returns the approximate width a string will take on screen when printed on a character cell
+ * terminal/console. */
+
+ while (*str) {
+ int w;
+
+ w = utf8_char_console_width(str);
+ if (w < 0)
+ return (size_t) -1;
+
+ n += w;
+ str = utf8_next_char(str);
+ }
+
+ return n;
}