23966986054

Committed 03 Apr 2026 09:22PM UTC coverage: 72.107% (-0.3%) from 72.362%

Build # 23966986054

Build Type

push

github

Committed by

daandemeyer

Commit Message

fd-util: Add missing assert()

Coverage Stats

1 of 1 new or added line in 1 file covered. (100.0%)

7891 existing lines in 120 files now uncovered.

318256 of 441368 relevant lines covered (72.11%)

1186882.96 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("");

        p = s = prev_s = malloc(strlen(str) * 4 + 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. */

        _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) {	88,055✔
17
18	if (ch >= 0x110000) /* End of unicode space */	88,055✔
19	return false;
20	if ((ch & 0xFFFFF800) == 0xD800) /* Reserved area for UTF-16 */	88,055✔
21	return false;
22	if ((ch >= 0xFDD0) && (ch <= 0xFDEF)) /* Reserved */	88,055✔
23	return false;
24	if ((ch & 0xFFFE) == 0xFFFE) /* BOM (Byte Order Mark) */	88,055✔
25	return false;	6✔
26
27	return true;
28	}
29
30	static bool unichar_is_control(char32_t ch) {	120,845,390✔
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')) \|\|	120,845,390✔
39	(0x7F <= ch && ch <= 0x9F);	120,845,212✔
40	}
41
42	/* count of characters used to encode one unicode char */
43	static size_t utf8_encoded_expected_len(uint8_t c) {	1,061,899,823✔
44	if (c < 0x80)	1,061,899,823✔
45	return 1;
46	if ((c & 0xe0) == 0xc0)	283,558✔
47	return 2;
48	if ((c & 0xf0) == 0xe0)	256,443✔
49	return 3;
50	if ((c & 0xf8) == 0xf0)	854✔
51	return 4;
52	if ((c & 0xfc) == 0xf8)	64✔
53	return 5;
54	if ((c & 0xfe) == 0xfc)	64✔
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) {	131,729,027✔
62	char32_t unichar;	131,729,027✔
63	size_t len;	131,729,027✔
64
65	assert(str);	131,729,027✔
66	assert(ret_unichar);	131,729,027✔
67
68	len = utf8_encoded_expected_len(str[0]);	131,729,027✔
69
70	switch (len) {	131,729,027✔
71	case 1:	131,534,783✔
72	*ret_unichar = (char32_t)str[0];	131,534,783✔
73	return 1;	131,534,783✔
74	case 2:	22,364✔
75	unichar = str[0] & 0x1f;	22,364✔
76	break;	22,364✔
77	case 3:	171,381✔
78	unichar = (char32_t)str[0] & 0x0f;	171,381✔
79	break;	171,381✔
80	case 4:	499✔
81	unichar = (char32_t)str[0] & 0x07;	499✔
82	break;	499✔
83	case 5:	×
84	unichar = (char32_t)str[0] & 0x03;	×
85	break;	×
86	case 6:	×
87	unichar = (char32_t)str[0] & 0x01;	×
UNCOV 88	break;	×
89	default:
90	return -EINVAL;
91	}
92
93	for (size_t i = 1; i < len; i++) {	560,859✔
94	if (((char32_t)str[i] & 0xc0) != 0x80)	366,621✔
95	return -EINVAL;
96
97	unichar <<= 6;	366,615✔
98	unichar \|= (char32_t)str[i] & 0x3f;	366,615✔
99	}
100
101	*ret_unichar = unichar;	194,238✔
102	return len;	194,238✔
103	}
104
105	bool utf8_is_printable_newline(const char* str, size_t length, bool allow_newline) {	3,975,395✔
106	assert(str);	3,975,395✔
107
108	for (const char *p = str; length > 0;) {	124,820,607✔
109	int encoded_len;	120,845,391✔
110	char32_t val;	120,845,391✔
111
112	encoded_len = utf8_encoded_valid_unichar(p, length);	120,845,391✔
113	if (encoded_len < 0)	120,845,391✔
114	return false;	179✔
115	assert(encoded_len > 0 && (size_t) encoded_len <= length);	120,845,390✔
116
117	if (utf8_encoded_to_unichar(p, &val) < 0 \|\|	120,845,390✔
118	unichar_is_control(val) \|\|	120,845,390✔
119	(!allow_newline && val == '\n'))	25,179,598✔
120	return false;
121
122	length -= encoded_len;	120,845,212✔
123	p += encoded_len;	120,845,212✔
124	}
125
126	return true;
127	}
128
129	char* utf8_is_valid_n(const char *str, size_t len_bytes) {	45,839,710✔
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);	45,839,710✔
134
135	for (size_t i = 0; len_bytes != SIZE_MAX ? i < len_bytes : str[i] != '\0'; ) {	820,593,275✔
136	int len;	774,754,694✔
137
138	if (_unlikely_(str[i] == '\0'))	774,754,694✔
139	return NULL; /* embedded NUL */
140
141	len = utf8_encoded_valid_unichar(str + i,	774,754,691✔
142	len_bytes != SIZE_MAX ? len_bytes - i : SIZE_MAX);
143	if (_unlikely_(len < 0))	774,754,691✔
144	return NULL; /* invalid character */
145
146	i += len;	774,753,565✔
147	}
148
149	return (char*) str;
150	}
151
152	char* utf8_escape_invalid(const char *str) {	12,337✔
153	char p, s;	12,337✔
154
155	assert(str);	12,337✔
156
157	p = s = malloc(strlen(str) * 4 + 1);	12,337✔
158	if (!p)	12,337✔
159	return NULL;
160
161	while (*str) {	287,214✔
162	int len;	274,877✔
163
164	len = utf8_encoded_valid_unichar(str, SIZE_MAX);	274,877✔
165	if (len > 0) {	274,877✔
166	s = mempcpy(s, str, len);	274,861✔
167	str += len;	274,861✔
168	} else {
169	s = stpcpy(s, UTF8_REPLACEMENT_CHARACTER);	16✔
170	str += 1;	16✔
171	}
172	}
173
174	*s = '\0';	12,337✔
175	return str_realloc(p);	12,337✔
176	}
177
178	int utf8_char_console_width(const char *str) {	8,520,195✔
179	char32_t c;	8,520,195✔
180	int r;	8,520,195✔
181
182	r = utf8_encoded_to_unichar(str, &c);	8,520,195✔
183	if (r < 0)	8,520,195✔
184	return r;	8,520,195✔
185
186	if (c == '\t')	8,520,194✔
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;	8,478,681✔
192	}
193
194	char* utf8_escape_non_printable_full(const char *str, size_t console_width, bool force_ellipsis) {	1,327✔
195	char p, s, *prev_s;	1,327✔
196	size_t n = 0; /* estimated print width */	1,327✔
197
198	assert(str);	1,327✔
199
200	if (console_width == 0)	1,327✔
201	return strdup("");	9✔
202
203	p = s = prev_s = malloc(strlen(str) * 4 + 1);	1,318✔
204	if (!p)	1,318✔
205	return NULL;
206
207	for (;;) {	52,259✔
208	int len;	52,259✔
209	char *saved_s = s;	52,259✔
210
211	if (!str) { / done! */	52,259✔
212	if (force_ellipsis)	1,244✔
213	goto truncation;	73✔
214	else
215	goto finish;	1,171✔
216	}
217
218	len = utf8_encoded_valid_unichar(str, SIZE_MAX);	51,015✔
219	if (len > 0) {	51,015✔
220	if (utf8_is_printable(str, len)) {	50,919✔
221	int w;	50,756✔
222
223	w = utf8_char_console_width(str);	50,756✔
224	assert(w >= 0);	50,756✔
225	if (n + w > console_width)	50,756✔
226	goto truncation;	34✔
227
228	s = mempcpy(s, str, len);	50,722✔
229	str += len;	50,722✔
230	n += w;	50,722✔
231
232	} else {
233	for (; len > 0; len--) {	288✔
234	if (n + 4 > console_width)	163✔
235	goto truncation;	38✔
236
237	*(s++) = '\\';	125✔
238	*(s++) = 'x';	125✔
239	(s++) = hexchar((int) str >> 4);	125✔
240	(s++) = hexchar((int) str);	125✔
241
242	str += 1;	125✔
243	n += 4;	125✔
244	}
245	}
246	} else {
247	if (n + 1 > console_width)	96✔
248	goto truncation;	2✔
249
250	s = mempcpy(s, UTF8_REPLACEMENT_CHARACTER, strlen(UTF8_REPLACEMENT_CHARACTER));	94✔
251	str += 1;	94✔
252	n += 1;	94✔
253	}
254
255	prev_s = saved_s;
256	}
257
258	truncation:	147✔
259	/* Try to go back one if we don't have enough space for the ellipsis */
260	if (n + 1 > console_width)	147✔
261	s = prev_s;	88✔
262
263	s = mempcpy(s, "…", strlen("…"));	147✔
264
265	finish:	1,318✔
266	*s = '\0';	1,318✔
267	return str_realloc(p);	1,318✔
268	}
269
270	char* ascii_is_valid_n(const char *str, size_t len) {	65,784✔
271	/* Check whether the string consists of valid ASCII bytes, i.e values between 1 and 127, inclusive.
272	* Stops at len, or NUL byte if len is SIZE_MAX. */
273
274	assert(str);	65,784✔
275
276	for (size_t i = 0; len != SIZE_MAX ? i < len : str[i] != '\0'; i++)	1,681,148✔
277	if ((unsigned char) str[i] >= 128 \|\| str[i] == '\0')	1,618,765✔
278	return NULL;
279
280	return (char*) str;
281	}
282
283	int utf8_to_ascii(const char str, char replacement_char, char *ret) {	49✔
284	/* Convert to a string that has only ASCII chars, replacing anything that is not ASCII
285	* by replacement_char. */
286
287	_cleanup_free_ char *ans = new(char, strlen(str) + 1);	98✔
288	if (!ans)	49✔
289	return -ENOMEM;
290
291	char *q = ans;
292
293	for (const char p = str; p; q++) {	385✔
294	int l;	338✔
295
296	l = utf8_encoded_valid_unichar(p, SIZE_MAX);	338✔
297	if (l < 0) /* Non-UTF-8, let's not even try to propagate the garbage */	338✔
298	return l;
299
300	if (l == 1)	336✔
301	q = p;	319✔
302	else
303	/* non-ASCII, we need to replace it */
304	*q = replacement_char;	17✔
305
306	p += l;	336✔
307	}
308	*q = '\0';	47✔
309
310	*ret = TAKE_PTR(ans);	47✔
311	return 0;	47✔
312	}
313
314	/**
315	* utf8_encode_unichar() - Encode single UCS-4 character as UTF-8
316	* @out_utf8: output buffer of at least 4 bytes or NULL
317	* @g: UCS-4 character to encode
318	*
319	* This encodes a single UCS-4 character as UTF-8 and writes it into @out_utf8.
320	* The length of the character is returned. It is not zero-terminated! If the
321	* output buffer is NULL, only the length is returned.
322	*
323	* Returns: The length in bytes that the UTF-8 representation does or would
324	* occupy.
325	*/
326	size_t utf8_encode_unichar(char *out_utf8, char32_t g) {	76,175✔
327
328	if (g < (1 << 7)) {	76,175✔
329	if (out_utf8)	76,051✔
330	out_utf8[0] = g & 0x7f;	76,051✔
331	return 1;	76,051✔
332	} else if (g < (1 << 11)) {	124✔
333	if (out_utf8) {	17✔
334	out_utf8[0] = 0xc0 \| ((g >> 6) & 0x1f);	17✔
335	out_utf8[1] = 0x80 \| (g & 0x3f);	17✔
336	}
337	return 2;	17✔
338	} else if (g < (1 << 16)) {	107✔
339	if (out_utf8) {	96✔
340	out_utf8[0] = 0xe0 \| ((g >> 12) & 0x0f);	96✔
341	out_utf8[1] = 0x80 \| ((g >> 6) & 0x3f);	96✔
342	out_utf8[2] = 0x80 \| (g & 0x3f);	96✔
343	}
344	return 3;	96✔
345	} else if (g < (1 << 21)) {	11✔
346	if (out_utf8) {	11✔
347	out_utf8[0] = 0xf0 \| ((g >> 18) & 0x07);	11✔
348	out_utf8[1] = 0x80 \| ((g >> 12) & 0x3f);	11✔
349	out_utf8[2] = 0x80 \| ((g >> 6) & 0x3f);	11✔
350	out_utf8[3] = 0x80 \| (g & 0x3f);	11✔
351	}
352	return 4;	11✔
353	}
354
355	return 0;
356	}
357
358	char* utf16_to_utf8(const char16_t s, size_t length / bytes! */) {	2,380✔
359	const uint8_t *f;	2,380✔
360	char r, t;	2,380✔
361
362	if (length == 0)	2,380✔
UNCOV 363	return new0(char, 1);	×
364
365	assert(s);	2,380✔
366
367	if (length == SIZE_MAX) {	2,380✔
368	length = char16_strlen(s);	6✔
369
370	if (length > SIZE_MAX/2)	6✔
371	return NULL; /* overflow */
372
373	length *= 2;	6✔
374	}
375
376	/* Input length is in bytes, i.e. the shortest possible character takes 2 bytes. Each unicode character may
377	* take up to 4 bytes in UTF-8. Let's also account for a trailing NUL byte. */
378	if (length > (SIZE_MAX - 1) / 2)	2,380✔
379	return NULL; /* overflow */
380
381	r = new(char, length * 2 + 1);	2,380✔
382	if (!r)	2,380✔
383	return NULL;
384
385	f = (const uint8_t*) s;
386	t = r;
387
388	while (f + 1 < (const uint8_t*) s + length) {	78,361✔
389	char16_t w1, w2;	75,981✔
390
391	/* see RFC 2781 section 2.2 */
392
393	w1 = f[1] << 8 \| f[0];	75,981✔
394	f += 2;	75,981✔
395
396	if (!utf16_is_surrogate(w1)) {	75,981✔
397	t += utf8_encode_unichar(t, w1);	75,974✔
398	continue;	75,974✔
399	}
400
401	if (utf16_is_trailing_surrogate(w1))	7✔
402	continue; /* spurious trailing surrogate, ignore */	1✔
403
404	if (f + 1 >= (const uint8_t*) s + length)	6✔
405	break;
406
407	w2 = f[1] << 8 \| f[0];	6✔
408	f += 2;	6✔
409
410	if (!utf16_is_trailing_surrogate(w2)) {	6✔
411	f -= 2;	1✔
412	continue; /* surrogate missing its trailing surrogate, ignore */	1✔
413	}
414
415	t += utf8_encode_unichar(t, utf16_surrogate_pair_to_unichar(w1, w2));	5✔
416	}
417
418	*t = 0;	2,380✔
419	return r;	2,380✔
420	}
421
422	size_t utf16_encode_unichar(char16_t *out, char32_t c) {	16✔
423
424	/* Note that this encodes as little-endian. */
425
426	switch (c) {	16✔
427
428	case 0 ... 0xd7ffU:	12✔
429	case 0xe000U ... 0xffffU:
430	out[0] = htole16(c);	12✔
431	return 1;	12✔
432
433	case 0x10000U ... 0x10ffffU:	4✔
434	c -= 0x10000U;	4✔
435	out[0] = htole16((c >> 10) + 0xd800U);	4✔
436	out[1] = htole16((c & 0x3ffU) + 0xdc00U);	4✔
437	return 2;	4✔
438
439	default: /* A surrogate (invalid) */
440	return 0;
441	}
442	}
443
444	char16_t utf8_to_utf16(const char s, size_t length) {	367✔
445	char16_t n, p;	367✔
446	int r;	367✔
447
448	if (length == 0)	367✔
UNCOV 449	return new0(char16_t, 1);	×
450
451	assert(s);	367✔
452
453	if (length == SIZE_MAX)	367✔
454	length = strlen(s);	366✔
455
456	if (length > SIZE_MAX - 1)	366✔
457	return NULL; /* overflow */
458
459	n = new(char16_t, length + 1);	367✔
460	if (!n)	367✔
461	return NULL;
462
463	p = n;
464
465	for (size_t i = 0; i < length;) {	5,963✔
466	char32_t unichar;	5,596✔
467	size_t e;	5,596✔
468
469	e = utf8_encoded_expected_len(s[i]);	5,596✔
470	if (e <= 1) /* Invalid and single byte characters are copied as they are */	5,596✔
471	goto copy;	5,580✔
472
473	if (i + e > length) /* sequence longer than input buffer, then copy as-is */	16✔
UNCOV 474	goto copy;	×
475
476	r = utf8_encoded_to_unichar(s + i, &unichar);	16✔
477	if (r < 0) /* sequence invalid, then copy as-is */	16✔
UNCOV 478	goto copy;	×
479
480	p += utf16_encode_unichar(p, unichar);	16✔
481	i += e;	16✔
482	continue;	16✔
483
484	copy:	5,580✔
485	*(p++) = htole16(s[i++]);	5,580✔
486	}
487
488	*p = 0;	367✔
489	return n;	367✔
490	}
491
492	size_t char16_strlen(const char16_t *s) {	631✔
493	size_t n = 0;	631✔
494
495	assert(s);	631✔
496
497	while (*s != 0)	9,408✔
498	n++, s++;	8,777✔
499
500	return n;	631✔
501	}
502
503	size_t char16_strsize(const char16_t *s) {	3✔
504	POINTER_MAY_BE_NULL(s);	3✔
505
506	return s ? (char16_strlen(s) + 1) * sizeof(*s) : 0;	3✔
507	}
508
509	/* expected size used to encode one unicode char */
510	static int utf8_unichar_to_encoded_len(char32_t unichar) {	88,045✔
511
512	if (unichar < 0x80)	88,045✔
513	return 1;
514	if (unichar < 0x800)	88,045✔
515	return 2;
516	if (unichar < 0x10000)	83,621✔
517	return 3;
518	if (unichar < 0x200000)	285✔
519	return 4;
UNCOV 520	if (unichar < 0x4000000)	×
UNCOV 521	return 5;	×
522
523	return 6;
524	}
525
526	/* validate one encoded unicode char and return its length */
527	int utf8_encoded_valid_unichar(const char str, size_t length / bytes */) {	930,165,200✔
528	char32_t unichar;	930,165,200✔
529	size_t len;	930,165,200✔
530	int r;	930,165,200✔
531
532	assert(str);	930,165,200✔
533	assert(length > 0);	930,165,200✔
534
535	/* We read until NUL, at most length bytes. SIZE_MAX may be used to disable the length check. */
536
537	len = utf8_encoded_expected_len(str[0]);	930,165,200✔
538	if (len == 0)	930,165,200✔
539	return -EINVAL;	930,165,200✔
540
541	/* Do we have a truncated multi-byte character? */
542	if (len > length)	930,165,136✔
543	return -EINVAL;
544
545	/* ascii is valid */
546	if (len == 1)	930,165,128✔
547	return 1;
548
549	/* check if expected encoded chars are available */
550	for (size_t i = 0; i < len; i++)	350,867✔
551	if ((str[i] & 0x80) != 0x80)	262,817✔
552	return -EINVAL;
553
554	r = utf8_encoded_to_unichar(str, &unichar);	88,050✔
555	if (r < 0)	88,050✔
556	return r;
557
558	/* check if encoded length matches encoded value */
559	if (utf8_unichar_to_encoded_len(unichar) != (int) len)	88,045✔
560	return -EINVAL;
561
562	/* check if value has valid range */
563	if (!unichar_is_valid(unichar))	88,045✔
564	return -EINVAL;	6✔
565
566	return (int) len;
567	}
568
569	size_t utf8_n_codepoints(const char *str) {	6✔
570	size_t n = 0;	6✔
571
572	assert(str);	6✔
573
574	/* Returns the number of UTF-8 codepoints in this string, or SIZE_MAX if the string is not valid UTF-8. */
575
576	while (*str != 0) {	34✔
577	int k;	29✔
578
579	k = utf8_encoded_valid_unichar(str, SIZE_MAX);	29✔
580	if (k < 0)	29✔
581	return SIZE_MAX;
582
583	str += k;	28✔
584	n++;	28✔
585	}
586
587	return n;
588	}
589
590	size_t utf8_console_width(const char *str) {	473,332✔
591	POINTER_MAY_BE_NULL(str);	473,332✔
592
593	if (isempty(str))	473,332✔
594	return 0;
595
596	/* Returns the approximate width a string will take on screen when printed on a character cell
597	* terminal/console. */
598
599	size_t n = 0;
600	while (*str) {	6,575,086✔
601	int w;	6,120,915✔
602
603	w = utf8_char_console_width(str);	6,120,915✔
604	if (w < 0)	6,120,915✔
605	return SIZE_MAX;
606
607	n += w;	6,120,914✔
608	str = utf8_next_char(str);	6,120,914✔
609	}
610
611	return n;
612	}
613
614	size_t utf8_last_length(const char *s, size_t n) {	9✔
615	int r;	9✔
616
617	assert(s);	9✔
618
619	if (n == SIZE_MAX)	9✔
620	n = strlen(s);	7✔
621
622	/* Determines length in bytes of last UTF-8 codepoint in string. If the string is empty, returns
623	* zero. Treats invalid UTF-8 codepoints as 1 sized ones. */
624
625	for (size_t last = 0;;) {	19✔
626	if (n == 0)	28✔
627	return last;	9✔
628
629	r = utf8_encoded_valid_unichar(s, n);	19✔
630	if (r <= 0)	19✔
UNCOV 631	r = 1; /* treat invalid UTF-8 as byte-wide */	×
632
633	s += r;	19✔
634	n -= r;	19✔
635	last = r;	19✔
636	}
637	}

systemd / systemd / 23966986054

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