27728979565

Committed 17 Jun 2026 12:11PM UTC coverage: 72.586% (-0.4%) from 73.001%

Build # 27728979565

Build Type

push

github

Committed by

web-flow

Commit Message

mkosi: fix suse build (#42637)

No longer exists since latest spec changes, as the binaries have moved
to another package so it doesn't get autogenerated anymore

Follow-up for db9d7265f

Coverage Stats

335884 of 462738 relevant lines covered (72.59%)

1397781.59 hits per line

Source File
Press 'n' to go to next uncovered line, 'b' for previous

95.58

/src/basic/utf8.c

/* SPDX-License-Identifier: LGPL-2.0-or-later */

/* Parts of this file are based on the GLIB utf8 validation functions. The original copyright follows.
 *
 * gutf8.c - Operations on UTF-8 strings.
 * Copyright (C) 1999 Tom Tromey
 * Copyright (C) 2000 Red Hat, Inc.
 */

#include "alloc-util.h"
#include "gunicode.h"
#include "hexdecoct.h"
#include "string-util.h"
#include "utf8.h"

bool unichar_is_valid(char32_t ch) {

        if (ch >= 0x110000) /* End of unicode space */
                return false;
        if ((ch & 0xFFFFF800) == 0xD800) /* Reserved area for UTF-16 */
                return false;
        if ((ch >= 0xFDD0) && (ch <= 0xFDEF)) /* Reserved */
                return false;
        if ((ch & 0xFFFE) == 0xFFFE) /* BOM (Byte Order Mark) */
                return false;

        return true;
}

static bool unichar_is_control(char32_t ch) {

        /*
          0 to ' '-1 is the C0 range.
          DEL=0x7F, and DEL+1 to 0x9F is C1 range.
          '\t' is in C0 range, but more or less harmless and commonly used.
        */

        return (ch < ' ' && !IN_SET(ch, '\t', '\n')) ||
                (0x7F <= ch && ch <= 0x9F);
}

/* count of characters used to encode one unicode char */
static size_t utf8_encoded_expected_len(uint8_t c) {
        if (c < 0x80)
                return 1;
        if ((c & 0xe0) == 0xc0)
                return 2;
        if ((c & 0xf0) == 0xe0)
                return 3;
        if ((c & 0xf8) == 0xf0)
                return 4;
        if ((c & 0xfc) == 0xf8)
                return 5;
        if ((c & 0xfe) == 0xfc)
                return 6;

        return 0;
}

/* decode one unicode char */
int utf8_encoded_to_unichar(const char *str, char32_t *ret_unichar) {
        char32_t unichar;
        size_t len;

        assert(str);
        assert(ret_unichar);

        len = utf8_encoded_expected_len(str[0]);

        switch (len) {
        case 1:
                *ret_unichar = (char32_t)str[0];
                return 1;
        case 2:
                unichar = str[0] & 0x1f;
                break;
        case 3:
                unichar = (char32_t)str[0] & 0x0f;
                break;
        case 4:
                unichar = (char32_t)str[0] & 0x07;
                break;
        case 5:
                unichar = (char32_t)str[0] & 0x03;
                break;
        case 6:
                unichar = (char32_t)str[0] & 0x01;
                break;
        default:
                return -EINVAL;
        }

        for (size_t i = 1; i < len; i++) {
                if (((char32_t)str[i] & 0xc0) != 0x80)
                        return -EINVAL;

                unichar <<= 6;
                unichar |= (char32_t)str[i] & 0x3f;
        }

        *ret_unichar = unichar;
        return len;
}

bool utf8_is_printable_newline(const char* str, size_t length, bool allow_newline) {
        assert(str);

        for (const char *p = str; length > 0;) {
                int encoded_len;
                char32_t val;

                encoded_len = utf8_encoded_valid_unichar(p, length);
                if (encoded_len < 0)
                        return false;
                assert(encoded_len > 0 && (size_t) encoded_len <= length);

                if (utf8_encoded_to_unichar(p, &val) < 0 ||
                    unichar_is_control(val) ||
                    (!allow_newline && val == '\n'))
                        return false;

                length -= encoded_len;
                p += encoded_len;
        }

        return true;
}

char* utf8_is_valid_n(const char *str, size_t len_bytes) {
        /* Check if the string is composed of valid utf8 characters. If length len_bytes is given, stop after
         * len_bytes. Otherwise, stop at NUL. */

        assert(str);

        for (size_t i = 0; len_bytes != SIZE_MAX ? i < len_bytes : str[i] != '\0'; ) {
                int len;

                if (_unlikely_(str[i] == '\0'))
                        return NULL; /* embedded NUL */

                len = utf8_encoded_valid_unichar(str + i,
                                                 len_bytes != SIZE_MAX ? len_bytes - i : SIZE_MAX);
                if (_unlikely_(len < 0))
                        return NULL; /* invalid character */

                i += len;
        }

        return (char*) str;
}

char* utf8_escape_invalid(const char *str) {
        char *p, *s;

        assert(str);

        p = s = malloc(strlen(str) * 4 + 1);
        if (!p)
                return NULL;

        while (*str) {
                int len;

                len = utf8_encoded_valid_unichar(str, SIZE_MAX);
                if (len > 0) {
                        s = mempcpy(s, str, len);
                        str += len;
                } else {
                        s = stpcpy(s, UTF8_REPLACEMENT_CHARACTER);
                        str += 1;
                }
        }

        *s = '\0';
        return str_realloc(p);
}

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;

        if (c == '\t')
                return 8; /* Assume a tab width of 8 */

        /* 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, bool force_ellipsis) {
        char *p, *s, *prev_s;
        size_t n = 0; /* estimated print width */

        assert(str);

        if (console_width == 0)
                return strdup("");

        size_t body = strlen(str) * 4;
        p = s = prev_s = malloc(body + STRLEN("…") + 1);
        if (!p)
                return NULL;

        for (;;) {
                int len;
                char *saved_s = s;

                if (!*str) { /* done! */
                        if (force_ellipsis)
                                goto truncation;
                        else
                                goto finish;
                }

                len = utf8_encoded_valid_unichar(str, SIZE_MAX);
                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 {
                                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;
                                        n += 4;
                                }
                        }
                } else {
                        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;
        }

 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, "…", strlen("…"));

 finish:
        *s = '\0';
        return str_realloc(p);
}

char* ascii_is_valid_n(const char *str, size_t len) {
        /* Check whether the string consists of valid ASCII bytes, i.e values between 1 and 127, inclusive.
         * Stops at len, or NUL byte if len is SIZE_MAX. */

        assert(str);

        for (size_t i = 0; len != SIZE_MAX ? i < len : str[i] != '\0'; i++)
                if ((unsigned char) str[i] >= 128 || str[i] == '\0')
                        return NULL;

        return (char*) str;
}

int utf8_to_ascii(const char *str, char replacement_char, char **ret) {
        /* Convert to a string that has only ASCII chars, replacing anything that is not ASCII
         * by replacement_char. */

        assert(str);
        assert(ret);

        _cleanup_free_ char *ans = new(char, strlen(str) + 1);
        if (!ans)
                return -ENOMEM;

        char *q = ans;

        for (const char *p = str; *p; q++) {
                int l;

                l = utf8_encoded_valid_unichar(p, SIZE_MAX);
                if (l < 0)  /* Non-UTF-8, let's not even try to propagate the garbage */
                        return l;

                if (l == 1)
                        *q = *p;
                else
                        /* non-ASCII, we need to replace it */
                        *q = replacement_char;

                p += l;
        }
        *q = '\0';

        *ret = TAKE_PTR(ans);
        return 0;
}

/**
 * utf8_encode_unichar() - Encode single UCS-4 character as UTF-8
 * @out_utf8: output buffer of at least 4 bytes or NULL
 * @g: UCS-4 character to encode
 *
 * This encodes a single UCS-4 character as UTF-8 and writes it into @out_utf8.
 * The length of the character is returned. It is not zero-terminated! If the
 * output buffer is NULL, only the length is returned.
 *
 * Returns: The length in bytes that the UTF-8 representation does or would
 *          occupy.
 */
size_t utf8_encode_unichar(char *out_utf8, char32_t g) {

        if (g < (1 << 7)) {
                if (out_utf8)
                        out_utf8[0] = g & 0x7f;
                return 1;
        } else if (g < (1 << 11)) {
                if (out_utf8) {
                        out_utf8[0] = 0xc0 | ((g >> 6) & 0x1f);
                        out_utf8[1] = 0x80 | (g & 0x3f);
                }
                return 2;
        } else if (g < (1 << 16)) {
                if (out_utf8) {
                        out_utf8[0] = 0xe0 | ((g >> 12) & 0x0f);
                        out_utf8[1] = 0x80 | ((g >> 6) & 0x3f);
                        out_utf8[2] = 0x80 | (g & 0x3f);
                }
                return 3;
        } else if (g < (1 << 21)) {
                if (out_utf8) {
                        out_utf8[0] = 0xf0 | ((g >> 18) & 0x07);
                        out_utf8[1] = 0x80 | ((g >> 12) & 0x3f);
                        out_utf8[2] = 0x80 | ((g >> 6) & 0x3f);
                        out_utf8[3] = 0x80 | (g & 0x3f);
                }
                return 4;
        }

        return 0;
}

char* utf16_to_utf8(const char16_t *s, size_t length /* bytes! */) {
        const uint8_t *f;
        char *r, *t;

        if (length == 0)
                return new0(char, 1);

        assert(s);

        if (length == SIZE_MAX) {
                length = char16_strlen(s);

                if (length > SIZE_MAX/2)
                        return NULL; /* overflow */

                length *= 2;
        }

        /* 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 > (SIZE_MAX - 1) / 2)
                return NULL; /* overflow */

        r = new(char, length * 2 + 1);
        if (!r)
                return NULL;

        f = (const uint8_t*) s;
        t = r;

        while (f + 1 < (const uint8_t*) s + length) {
                char16_t w1, w2;

                /* see RFC 2781 section 2.2 */

                w1 = f[1] << 8 | f[0];
                f += 2;

                if (!utf16_is_surrogate(w1)) {
                        t += utf8_encode_unichar(t, w1);
                        continue;
                }

                if (utf16_is_trailing_surrogate(w1))
                        continue; /* spurious trailing surrogate, ignore */

                if (f + 1 >= (const uint8_t*) s + length)
                        break;

                w2 = f[1] << 8 | f[0];
                f += 2;

                if (!utf16_is_trailing_surrogate(w2)) {
                        f -= 2;
                        continue; /* surrogate missing its trailing surrogate, ignore */
                }

                t += utf8_encode_unichar(t, utf16_surrogate_pair_to_unichar(w1, w2));
        }

        *t = 0;
        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;
        int r;

        if (length == 0)
                return new0(char16_t, 1);

        assert(s);

        if (length == SIZE_MAX)
                length = strlen(s);

        if (length > SIZE_MAX - 1)
                return NULL; /* overflow */

        n = new(char16_t, length + 1);
        if (!n)
                return NULL;

        p = n;

        for (size_t 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;
}

size_t char16_strsize(const char16_t *s) {
        POINTER_MAY_BE_NULL(s);

        return s ? (char16_strlen(s) + 1) * sizeof(*s) : 0;
}

/* expected size used to encode one unicode char */
static int utf8_unichar_to_encoded_len(char32_t unichar) {

        if (unichar < 0x80)
                return 1;
        if (unichar < 0x800)
                return 2;
        if (unichar < 0x10000)
                return 3;
        if (unichar < 0x200000)
                return 4;
        if (unichar < 0x4000000)
                return 5;

        return 6;
}

/* validate one encoded unicode char and return its length */
int utf8_encoded_valid_unichar(const char *str, size_t length /* bytes */) {
        char32_t unichar;
        size_t len;
        int r;

        assert(str);
        assert(length > 0);

        /* We read until NUL, at most length bytes. SIZE_MAX may be used to disable the length check. */

        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;

        /* check if expected encoded chars are available */
        for (size_t i = 0; i < len; i++)
                if ((str[i] & 0x80) != 0x80)
                        return -EINVAL;

        r = utf8_encoded_to_unichar(str, &unichar);
        if (r < 0)
                return r;

        /* check if encoded length matches encoded value */
        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 (int) len;
}

size_t utf8_n_codepoints(const char *str) {
        size_t n = 0;

        assert(str);

        /* Returns the number of UTF-8 codepoints in this string, or SIZE_MAX if the string is not valid UTF-8. */

        while (*str != 0) {
                int k;

                k = utf8_encoded_valid_unichar(str, SIZE_MAX);
                if (k < 0)
                        return SIZE_MAX;

                str += k;
                n++;
        }

        return n;
}

size_t utf8_console_width(const char *str) {
        POINTER_MAY_BE_NULL(str);

        if (isempty(str))
                return 0;

        /* Returns the approximate width a string will take on screen when printed on a character cell
         * terminal/console. */

        size_t n = 0;
        while (*str) {
                int w;

                w = utf8_char_console_width(str);
                if (w < 0)
                        return SIZE_MAX;

                n += w;
                str = utf8_next_char(str);
        }

        return n;
}

size_t utf8_last_length(const char *s, size_t n) {
        int r;

        assert(s);

        if (n == SIZE_MAX)
                n = strlen(s);

        /* Determines length in bytes of last UTF-8 codepoint in string. If the string is empty, returns
         * zero. Treats invalid UTF-8 codepoints as 1 sized ones. */

        for (size_t last = 0;;) {
                if (n == 0)
                        return last;

                r = utf8_encoded_valid_unichar(s, n);
                if (r <= 0)
                        r = 1; /* treat invalid UTF-8 as byte-wide */

                s += r;
                n -= r;
                last = r;
        }
}

1	/* SPDX-License-Identifier: LGPL-2.0-or-later */
2
3	/* Parts of this file are based on the GLIB utf8 validation functions. The original copyright follows.
4	*
5	* gutf8.c - Operations on UTF-8 strings.
6	* Copyright (C) 1999 Tom Tromey
7	* Copyright (C) 2000 Red Hat, Inc.
8	*/
9
10	#include "alloc-util.h"
11	#include "gunicode.h"
12	#include "hexdecoct.h"
13	#include "string-util.h"
14	#include "utf8.h"
15
16	bool unichar_is_valid(char32_t ch) {	97,357✔
17
18	if (ch >= 0x110000) /* End of unicode space */	97,357✔
19	return false;
20	if ((ch & 0xFFFFF800) == 0xD800) /* Reserved area for UTF-16 */	97,357✔
21	return false;
22	if ((ch >= 0xFDD0) && (ch <= 0xFDEF)) /* Reserved */	97,357✔
23	return false;
24	if ((ch & 0xFFFE) == 0xFFFE) /* BOM (Byte Order Mark) */	97,357✔
25	return false;	26✔
26
27	return true;
28	}
29
30	static bool unichar_is_control(char32_t ch) {	120,744,702✔
31
32	/*
33	0 to ' '-1 is the C0 range.
34	DEL=0x7F, and DEL+1 to 0x9F is C1 range.
35	'\t' is in C0 range, but more or less harmless and commonly used.
36	*/
37
38	return (ch < ' ' && !IN_SET(ch, '\t', '\n')) \|\|	120,744,702✔
39	(0x7F <= ch && ch <= 0x9F);	120,744,521✔
40	}
41
42	/* count of characters used to encode one unicode char */
43	static size_t utf8_encoded_expected_len(uint8_t c) {	1,231,570,602✔
44	if (c < 0x80)	1,231,570,602✔
45	return 1;
46	if ((c & 0xe0) == 0xc0)	311,994✔
47	return 2;
48	if ((c & 0xf0) == 0xe0)	284,720✔
49	return 3;
50	if ((c & 0xf8) == 0xf0)	751✔
51	return 4;
52	if ((c & 0xfc) == 0xf8)	89✔
53	return 5;
54	if ((c & 0xfe) == 0xfc)	89✔
55	return 6;	×
56
57	return 0;
58	}
59
60	/* decode one unicode char */
61	int utf8_encoded_to_unichar(const char str, char32_t ret_unichar) {	133,640,148✔
62	char32_t unichar;	133,640,148✔
63	size_t len;	133,640,148✔
64
65	assert(str);	133,640,148✔
66	assert(ret_unichar);	133,640,148✔
67
68	len = utf8_encoded_expected_len(str[0]);	133,640,148✔
69
70	switch (len) {	133,640,148✔
71	case 1:	133,426,796✔
72	*ret_unichar = (char32_t)str[0];	133,426,796✔
73	return 1;	133,426,796✔
74	case 2:	22,456✔
75	unichar = str[0] & 0x1f;	22,456✔
76	break;	22,456✔
77	case 3:	190,461✔
78	unichar = (char32_t)str[0] & 0x0f;	190,461✔
79	break;	190,461✔
80	case 4:	435✔
81	unichar = (char32_t)str[0] & 0x07;	435✔
82	break;	435✔
83	case 5:	×
84	unichar = (char32_t)str[0] & 0x03;	×
85	break;	×
86	case 6:	×
87	unichar = (char32_t)str[0] & 0x01;	×
88	break;	×
89	default:
90	return -EINVAL;
91	}
92
93	for (size_t i = 1; i < len; i++) {	618,027✔
94	if (((char32_t)str[i] & 0xc0) != 0x80)	404,681✔
95	return -EINVAL;
96
97	unichar <<= 6;	404,675✔
98	unichar \|= (char32_t)str[i] & 0x3f;	404,675✔
99	}
100
101	*ret_unichar = unichar;	213,346✔
102	return len;	213,346✔
103	}
104
105	bool utf8_is_printable_newline(const char* str, size_t length, bool allow_newline) {	4,337,352✔
106	assert(str);	4,337,352✔
107
108	for (const char *p = str; length > 0;) {	125,081,873✔
109	int encoded_len;	120,744,703✔
110	char32_t val;	120,744,703✔
111
112	encoded_len = utf8_encoded_valid_unichar(p, length);	120,744,703✔
113	if (encoded_len < 0)	120,744,703✔
114	return false;	182✔
115	assert(encoded_len > 0 && (size_t) encoded_len <= length);	120,744,702✔
116
117	if (utf8_encoded_to_unichar(p, &val) < 0 \|\|	120,744,702✔
118	unichar_is_control(val) \|\|	120,744,702✔
119	(!allow_newline && val == '\n'))	22,006,007✔
120	return false;
121
122	length -= encoded_len;	120,744,521✔
123	p += encoded_len;	120,744,521✔
124	}
125
126	return true;
127	}
128
129	char* utf8_is_valid_n(const char *str, size_t len_bytes) {	55,451,055✔
130	/* Check if the string is composed of valid utf8 characters. If length len_bytes is given, stop after
131	* len_bytes. Otherwise, stop at NUL. */
132
133	assert(str);	55,451,055✔
134
135	for (size_t i = 0; len_bytes != SIZE_MAX ? i < len_bytes : str[i] != '\0'; ) {	982,223,724✔
136	int len;	926,773,807✔
137
138	if (_unlikely_(str[i] == '\0'))	926,773,807✔
139	return NULL; /* embedded NUL */
140
141	len = utf8_encoded_valid_unichar(str + i,	926,773,804✔
142	len_bytes != SIZE_MAX ? len_bytes - i : SIZE_MAX);
143	if (_unlikely_(len < 0))	926,773,805✔
144	return NULL; /* invalid character */
145
146	i += len;	926,772,669✔
147	}
148
149	return (char*) str;
150	}
151
152	char* utf8_escape_invalid(const char *str) {	15,836✔
153	char p, s;	15,836✔
154
155	assert(str);	15,836✔
156
157	p = s = malloc(strlen(str) * 4 + 1);	15,836✔
158	if (!p)	15,836✔
159	return NULL;
160
161	while (*str) {	348,172✔
162	int len;	332,336✔
163
164	len = utf8_encoded_valid_unichar(str, SIZE_MAX);	332,336✔
165	if (len > 0) {	332,336✔
166	s = mempcpy(s, str, len);	332,284✔
167	str += len;	332,284✔
168	} else {
169	s = stpcpy(s, UTF8_REPLACEMENT_CHARACTER);	52✔
170	str += 1;	52✔
171	}
172	}
173
174	*s = '\0';	15,836✔
175	return str_realloc(p);	15,836✔
176	}
177
178	int utf8_char_console_width(const char *str) {	10,245,007✔
179	char32_t c;	10,245,007✔
180	int r;	10,245,007✔
181
182	r = utf8_encoded_to_unichar(str, &c);	10,245,007✔
183	if (r < 0)	10,245,007✔
184	return r;	10,245,007✔
185
186	if (c == '\t')	10,245,006✔
187	return 8; /* Assume a tab width of 8 */
188
189	/* TODO: we should detect combining characters */
190
191	return unichar_iswide(c) ? 2 : 1;	10,191,361✔
192	}
193
194	char* utf8_escape_non_printable_full(const char *str, size_t console_width, bool force_ellipsis) {	4,083✔
195	char p, s, *prev_s;	4,083✔
196	size_t n = 0; /* estimated print width */	4,083✔
197
198	assert(str);	4,083✔
199
200	if (console_width == 0)	4,083✔
201	return strdup("");	9✔
202
203	size_t body = strlen(str) * 4;	4,074✔
204	p = s = prev_s = malloc(body + STRLEN("…") + 1);	4,074✔
205	if (!p)	4,074✔
206	return NULL;
207
208	for (;;) {	395,028✔
209	int len;	199,551✔
210	char *saved_s = s;	199,551✔
211
212	if (!str) { / done! */	199,551✔
213	if (force_ellipsis)	4,000✔
214	goto truncation;	68✔
215	else
216	goto finish;	3,932✔
217	}
218
219	len = utf8_encoded_valid_unichar(str, SIZE_MAX);	195,551✔
220	if (len > 0) {	195,551✔
221	if (utf8_is_printable(str, len)) {	195,455✔
222	int w;	195,289✔
223
224	w = utf8_char_console_width(str);	195,289✔
225	assert(w >= 0);	195,289✔
226	if (n + w > console_width)	195,289✔
227	goto truncation;	34✔
228
229	s = mempcpy(s, str, len);	195,255✔
230	str += len;	195,255✔
231	n += w;	195,255✔
232
233	} else {
234	for (; len > 0; len--) {	294✔
235	if (n + 4 > console_width)	166✔
236	goto truncation;	38✔
237
238	*(s++) = '\\';	128✔
239	*(s++) = 'x';	128✔
240	(s++) = hexchar((int) str >> 4);	128✔
241	(s++) = hexchar((int) str);	128✔
242
243	str += 1;	128✔
244	n += 4;	128✔
245	}
246	}
247	} else {
248	if (n + 1 > console_width)	96✔
249	goto truncation;	2✔
250
251	s = mempcpy(s, UTF8_REPLACEMENT_CHARACTER, strlen(UTF8_REPLACEMENT_CHARACTER));	94✔
252	str += 1;	94✔
253	n += 1;	94✔
254	}
255
256	prev_s = saved_s;	195,477✔
257	}
258
259	truncation:	142✔
260	/* Try to go back one if we don't have enough space for the ellipsis */
261	if (n + 1 > console_width)	142✔
262	s = prev_s;	82✔
263
264	s = mempcpy(s, "…", strlen("…"));	142✔
265
266	finish:	4,074✔
267	*s = '\0';	4,074✔
268	return str_realloc(p);	4,074✔
269	}
270
271	char* ascii_is_valid_n(const char *str, size_t len) {	98,846✔
272	/* Check whether the string consists of valid ASCII bytes, i.e values between 1 and 127, inclusive.
273	* Stops at len, or NUL byte if len is SIZE_MAX. */
274
275	assert(str);	98,846✔
276
277	for (size_t i = 0; len != SIZE_MAX ? i < len : str[i] != '\0'; i++)	2,445,020✔
278	if ((unsigned char) str[i] >= 128 \|\| str[i] == '\0')	2,349,580✔
279	return NULL;
280
281	return (char*) str;
282	}
283
284	int utf8_to_ascii(const char str, char replacement_char, char *ret) {	50✔
285	/* Convert to a string that has only ASCII chars, replacing anything that is not ASCII
286	* by replacement_char. */
287
288	assert(str);	50✔
289	assert(ret);	50✔
290
291	_cleanup_free_ char *ans = new(char, strlen(str) + 1);	100✔
292	if (!ans)	50✔
293	return -ENOMEM;
294
295	char *q = ans;
296
297	for (const char p = str; p; q++) {	389✔
298	int l;	341✔
299
300	l = utf8_encoded_valid_unichar(p, SIZE_MAX);	341✔
301	if (l < 0) /* Non-UTF-8, let's not even try to propagate the garbage */	341✔
302	return l;
303
304	if (l == 1)	339✔
305	q = p;	322✔
306	else
307	/* non-ASCII, we need to replace it */
308	*q = replacement_char;	17✔
309
310	p += l;	339✔
311	}
312	*q = '\0';	48✔
313
314	*ret = TAKE_PTR(ans);	48✔
315	return 0;	48✔
316	}
317
318	/**
319	* utf8_encode_unichar() - Encode single UCS-4 character as UTF-8
320	* @out_utf8: output buffer of at least 4 bytes or NULL
321	* @g: UCS-4 character to encode
322	*
323	* This encodes a single UCS-4 character as UTF-8 and writes it into @out_utf8.
324	* The length of the character is returned. It is not zero-terminated! If the
325	* output buffer is NULL, only the length is returned.
326	*
327	* Returns: The length in bytes that the UTF-8 representation does or would
328	* occupy.
329	*/
330	size_t utf8_encode_unichar(char *out_utf8, char32_t g) {	81,950✔
331
332	if (g < (1 << 7)) {	81,950✔
333	if (out_utf8)	81,826✔
334	out_utf8[0] = g & 0x7f;	81,826✔
335	return 1;
336	} else if (g < (1 << 11)) {	124✔
337	if (out_utf8) {	17✔
338	out_utf8[0] = 0xc0 \| ((g >> 6) & 0x1f);	17✔
339	out_utf8[1] = 0x80 \| (g & 0x3f);	17✔
340	}
341	return 2;
342	} else if (g < (1 << 16)) {	107✔
343	if (out_utf8) {	96✔
344	out_utf8[0] = 0xe0 \| ((g >> 12) & 0x0f);	96✔
345	out_utf8[1] = 0x80 \| ((g >> 6) & 0x3f);	96✔
346	out_utf8[2] = 0x80 \| (g & 0x3f);	96✔
347	}
348	return 3;
349	} else if (g < (1 << 21)) {	11✔
350	if (out_utf8) {	11✔
351	out_utf8[0] = 0xf0 \| ((g >> 18) & 0x07);	11✔
352	out_utf8[1] = 0x80 \| ((g >> 12) & 0x3f);	11✔
353	out_utf8[2] = 0x80 \| ((g >> 6) & 0x3f);	11✔
354	out_utf8[3] = 0x80 \| (g & 0x3f);	11✔
355	}
356	return 4;
357	}
358
359	return 0;
360	}
361
362	char* utf16_to_utf8(const char16_t s, size_t length / bytes! */) {	2,510✔
363	const uint8_t *f;	2,510✔
364	char r, t;	2,510✔
365
366	if (length == 0)	2,510✔
367	return new0(char, 1);	×
368
369	assert(s);	2,510✔
370
371	if (length == SIZE_MAX) {	2,510✔
372	length = char16_strlen(s);	6✔
373
374	if (length > SIZE_MAX/2)	6✔
375	return NULL; /* overflow */
376
377	length *= 2;	6✔
378	}
379
380	/* Input length is in bytes, i.e. the shortest possible character takes 2 bytes. Each unicode character may
381	* take up to 4 bytes in UTF-8. Let's also account for a trailing NUL byte. */
382	if (length > (SIZE_MAX - 1) / 2)	2,510✔
383	return NULL; /* overflow */
384
385	r = new(char, length * 2 + 1);	2,510✔
386	if (!r)	2,510✔
387	return NULL;
388
389	f = (const uint8_t*) s;
390	t = r;
391
392	while (f + 1 < (const uint8_t*) s + length) {	84,230✔
393	char16_t w1, w2;	81,720✔
394
395	/* see RFC 2781 section 2.2 */
396
397	w1 = f[1] << 8 \| f[0];	81,720✔
398	f += 2;	81,720✔
399
400	if (!utf16_is_surrogate(w1)) {	81,720✔
401	t += utf8_encode_unichar(t, w1);	81,713✔
402	continue;	81,713✔
403	}
404
405	if (utf16_is_trailing_surrogate(w1))	7✔
406	continue; /* spurious trailing surrogate, ignore */	1✔
407
408	if (f + 1 >= (const uint8_t*) s + length)	6✔
409	break;
410
411	w2 = f[1] << 8 \| f[0];	6✔
412	f += 2;	6✔
413
414	if (!utf16_is_trailing_surrogate(w2)) {	6✔
415	f -= 2;	1✔
416	continue; /* surrogate missing its trailing surrogate, ignore */	1✔
417	}
418
419	t += utf8_encode_unichar(t, utf16_surrogate_pair_to_unichar(w1, w2));	5✔
420	}
421
422	*t = 0;	2,510✔
423	return r;	2,510✔
424	}
425
426	size_t utf16_encode_unichar(char16_t *out, char32_t c) {	16✔
427
428	/* Note that this encodes as little-endian. */
429
430	switch (c) {	16✔
431
432	case 0 ... 0xd7ffU:	12✔
433	case 0xe000U ... 0xffffU:
434	out[0] = htole16(c);	12✔
435	return 1;	12✔
436
437	case 0x10000U ... 0x10ffffU:	4✔
438	c -= 0x10000U;	4✔
439	out[0] = htole16((c >> 10) + 0xd800U);	4✔
440	out[1] = htole16((c & 0x3ffU) + 0xdc00U);	4✔
441	return 2;	4✔
442
443	default: /* A surrogate (invalid) */
444	return 0;
445	}
446	}
447
448	char16_t utf8_to_utf16(const char s, size_t length) {	413✔
449	char16_t n, p;	413✔
450	int r;	413✔
451
452	if (length == 0)	413✔
453	return new0(char16_t, 1);	×
454
455	assert(s);	413✔
456
457	if (length == SIZE_MAX)	413✔
458	length = strlen(s);	412✔
459
460	if (length > SIZE_MAX - 1)	412✔
461	return NULL; /* overflow */
462
463	n = new(char16_t, length + 1);	413✔
464	if (!n)	413✔
465	return NULL;
466
467	p = n;
468
469	for (size_t i = 0; i < length;) {	6,832✔
470	char32_t unichar;	6,419✔
471	size_t e;	6,419✔
472
473	e = utf8_encoded_expected_len(s[i]);	6,419✔
474	if (e <= 1) /* Invalid and single byte characters are copied as they are */	6,419✔
475	goto copy;	6,403✔
476
477	if (i + e > length) /* sequence longer than input buffer, then copy as-is */	16✔
478	goto copy;	×
479
480	r = utf8_encoded_to_unichar(s + i, &unichar);	16✔
481	if (r < 0) /* sequence invalid, then copy as-is */	16✔
482	goto copy;	×
483
484	p += utf16_encode_unichar(p, unichar);	16✔
485	i += e;	16✔
486	continue;	16✔
487
488	copy:	6,403✔
489	*(p++) = htole16(s[i++]);	6,403✔
490	}
491
492	*p = 0;	413✔
493	return n;	413✔
494	}
495
496	size_t char16_strlen(const char16_t *s) {	677✔
497	size_t n = 0;	677✔
498
499	assert(s);	677✔
500
501	while (*s != 0)	10,277✔
502	n++, s++;	9,600✔
503
504	return n;	677✔
505	}
506
507	size_t char16_strsize(const char16_t *s) {	3✔
508	POINTER_MAY_BE_NULL(s);	3✔
509
510	return s ? (char16_strlen(s) + 1) * sizeof(*s) : 0;	3✔
511	}
512
513	/* expected size used to encode one unicode char */
514	static int utf8_unichar_to_encoded_len(char32_t unichar) {	97,347✔
515
516	if (unichar < 0x80)	97,347✔
517	return 1;
518	if (unichar < 0x800)	97,347✔
519	return 2;
520	if (unichar < 0x10000)	92,857✔
521	return 3;
522	if (unichar < 0x200000)	221✔
523	return 4;
524	if (unichar < 0x4000000)	×
525	return 5;	×
526
527	return 6;
528	}
529
530	/* validate one encoded unicode char and return its length */
531	int utf8_encoded_valid_unichar(const char str, size_t length / bytes */) {	1,097,924,033✔
532	char32_t unichar;	1,097,924,033✔
533	size_t len;	1,097,924,033✔
534	int r;	1,097,924,033✔
535
536	assert(str);	1,097,924,033✔
537	assert(length > 0);	1,097,924,033✔
538
539	/* We read until NUL, at most length bytes. SIZE_MAX may be used to disable the length check. */
540
541	len = utf8_encoded_expected_len(str[0]);	1,097,924,033✔
542	if (len == 0)	1,097,924,033✔
543	return -EINVAL;	1,097,924,033✔
544
545	/* Do we have a truncated multi-byte character? */
546	if (len > length)	1,097,923,944✔
547	return -EINVAL;
548
549	/* ascii is valid */
550	if (len == 1)	1,097,923,936✔
551	return 1;
552
553	/* check if expected encoded chars are available */
554	for (size_t i = 0; i < len; i++)	387,947✔
555	if ((str[i] & 0x80) != 0x80)	290,595✔
556	return -EINVAL;
557
558	r = utf8_encoded_to_unichar(str, &unichar);	97,352✔
559	if (r < 0)	97,352✔
560	return r;
561
562	/* check if encoded length matches encoded value */
563	if (utf8_unichar_to_encoded_len(unichar) != (int) len)	97,347✔
564	return -EINVAL;
565
566	/* check if value has valid range */
567	if (!unichar_is_valid(unichar))	97,347✔
568	return -EINVAL;	26✔
569
570	return (int) len;
571	}
572
573	size_t utf8_n_codepoints(const char *str) {	6✔
574	size_t n = 0;	6✔
575
576	assert(str);	6✔
577
578	/* Returns the number of UTF-8 codepoints in this string, or SIZE_MAX if the string is not valid UTF-8. */
579
580	while (*str != 0) {	34✔
581	int k;	29✔
582
583	k = utf8_encoded_valid_unichar(str, SIZE_MAX);	29✔
584	if (k < 0)	29✔
585	return SIZE_MAX;
586
587	str += k;	28✔
588	n++;	28✔
589	}
590
591	return n;
592	}
593
594	size_t utf8_console_width(const char *str) {	540,011✔
595	POINTER_MAY_BE_NULL(str);	540,011✔
596
597	if (isempty(str))	540,011✔
598	return 0;
599
600	/* Returns the approximate width a string will take on screen when printed on a character cell
601	* terminal/console. */
602
603	size_t n = 0;
604	while (*str) {	7,487,176✔
605	int w;	6,968,778✔
606
607	w = utf8_char_console_width(str);	6,968,778✔
608	if (w < 0)	6,968,778✔
609	return SIZE_MAX;
610
611	n += w;	6,968,777✔
612	str = utf8_next_char(str);	6,968,777✔
613	}
614
615	return n;
616	}
617
618	size_t utf8_last_length(const char *s, size_t n) {	9✔
619	int r;	9✔
620
621	assert(s);	9✔
622
623	if (n == SIZE_MAX)	9✔
624	n = strlen(s);	7✔
625
626	/* Determines length in bytes of last UTF-8 codepoint in string. If the string is empty, returns
627	* zero. Treats invalid UTF-8 codepoints as 1 sized ones. */
628
629	for (size_t last = 0;;) {	9✔
630	if (n == 0)	28✔
631	return last;	9✔
632
633	r = utf8_encoded_valid_unichar(s, n);	19✔
634	if (r <= 0)	19✔
635	r = 1; /* treat invalid UTF-8 as byte-wide */	×
636
637	s += r;	19✔
638	n -= r;	19✔
639	last = r;	19✔
640	}
641	}

systemd / systemd / 27728979565

Source File Press 'n' to go to next uncovered line, 'b' for previous

Source File
Press 'n' to go to next uncovered line, 'b' for previous