15263807472

/* 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);

        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) {
        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;

        /* 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) {

        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) {	84,693✔
17
18	if (ch >= 0x110000) /* End of unicode space */	84,693✔
19	return false;
20	if ((ch & 0xFFFFF800) == 0xD800) /* Reserved area for UTF-16 */	84,693✔
21	return false;
22	if ((ch >= 0xFDD0) && (ch <= 0xFDEF)) /* Reserved */	84,693✔
23	return false;
24	if ((ch & 0xFFFE) == 0xFFFE) /* BOM (Byte Order Mark) */	84,693✔
25	return false;	6✔
26
27	return true;
28	}
29
30	static bool unichar_is_control(char32_t ch) {	101,703,697✔
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')) \|\|	101,703,697✔
39	(0x7F <= ch && ch <= 0x9F);	101,703,525✔
40	}
41
42	/* count of characters used to encode one unicode char */
43	static size_t utf8_encoded_expected_len(uint8_t c) {	633,129,351✔
44	if (c < 0x80)	633,129,351✔
45	return 1;
46	if ((c & 0xe0) == 0xc0)	271,959✔
47	return 2;
48	if ((c & 0xf0) == 0xe0)	245,367✔
49	return 3;
50	if ((c & 0xf8) == 0xf0)	1,027✔
51	return 4;
52	if ((c & 0xfc) == 0xf8)	64✔
53	return 5;
54	if ((c & 0xfe) == 0xfc)	64✔
UNCOV 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) {	107,657,781✔
62	char32_t unichar;	107,657,781✔
63	size_t len;	107,657,781✔
64
65	assert(str);	107,657,781✔
66
67	len = utf8_encoded_expected_len(str[0]);	107,657,781✔
68
69	switch (len) {	107,657,781✔
70	case 1:	107,471,774✔
71	*ret_unichar = (char32_t)str[0];	107,471,774✔
72	return 1;	107,471,774✔
73	case 2:	22,023✔
74	unichar = str[0] & 0x1f;	22,023✔
75	break;	22,023✔
76	case 3:	163,407✔
77	unichar = (char32_t)str[0] & 0x0f;	163,407✔
78	break;	163,407✔
79	case 4:	577✔
80	unichar = (char32_t)str[0] & 0x07;	577✔
81	break;	577✔
UNCOV 82	case 5:	×
UNCOV 83	unichar = (char32_t)str[0] & 0x03;	×
UNCOV 84	break;	×
UNCOV 85	case 6:	×
UNCOV 86	unichar = (char32_t)str[0] & 0x01;	×
87	break;	×
88	default:
89	return -EINVAL;
90	}
91
92	for (size_t i = 1; i < len; i++) {	536,567✔
93	if (((char32_t)str[i] & 0xc0) != 0x80)	350,566✔
94	return -EINVAL;
95
96	unichar <<= 6;	350,560✔
97	unichar \|= (char32_t)str[i] & 0x3f;	350,560✔
98	}
99
100	*ret_unichar = unichar;	186,001✔
101	return len;	186,001✔
102	}
103
104	bool utf8_is_printable_newline(const char* str, size_t length, bool allow_newline) {	3,319,290✔
105	assert(str);	3,319,290✔
106
107	for (const char *p = str; length > 0;) {	105,022,815✔
108	int encoded_len;	101,703,698✔
109	char32_t val;	101,703,698✔
110
111	encoded_len = utf8_encoded_valid_unichar(p, length);	101,703,698✔
112	if (encoded_len < 0)	101,703,698✔
113	return false;	173✔
114	assert(encoded_len > 0 && (size_t) encoded_len <= length);	101,703,697✔
115
116	if (utf8_encoded_to_unichar(p, &val) < 0 \|\|	101,703,697✔
117	unichar_is_control(val) \|\|	101,703,697✔
118	(!allow_newline && val == '\n'))	23,843,686✔
119	return false;
120
121	length -= encoded_len;	101,703,525✔
122	p += encoded_len;	101,703,525✔
123	}
124
125	return true;
126	}
127
128	char* utf8_is_valid_n(const char *str, size_t len_bytes) {	24,563,871✔
129	/* Check if the string is composed of valid utf8 characters. If length len_bytes is given, stop after
130	* len_bytes. Otherwise, stop at NUL. */
131
132	assert(str);	24,563,871✔
133
134	for (size_t i = 0; len_bytes != SIZE_MAX ? i < len_bytes : str[i] != '\0'; ) {	422,712,025✔
135	int len;	398,149,283✔
136
137	if (_unlikely_(str[i] == '\0'))	398,149,283✔
138	return NULL; /* embedded NUL */
139
140	len = utf8_encoded_valid_unichar(str + i,	398,149,280✔
141	len_bytes != SIZE_MAX ? len_bytes - i : SIZE_MAX);
142	if (_unlikely_(len < 0))	398,149,280✔
143	return NULL; /* invalid character */
144
145	i += len;	398,148,154✔
146	}
147
148	return (char*) str;
149	}
150
151	char* utf8_escape_invalid(const char *str) {	2,568✔
152	char p, s;	2,568✔
153
154	assert(str);	2,568✔
155
156	p = s = malloc(strlen(str) * 4 + 1);	2,568✔
157	if (!p)	2,568✔
158	return NULL;
159
160	while (*str) {	59,855✔
161	int len;	57,287✔
162
163	len = utf8_encoded_valid_unichar(str, SIZE_MAX);	57,287✔
164	if (len > 0) {	57,287✔
165	s = mempcpy(s, str, len);	57,271✔
166	str += len;	57,271✔
167	} else {
168	s = stpcpy(s, UTF8_REPLACEMENT_CHARACTER);	16✔
169	str += 1;	16✔
170	}
171	}
172
173	*s = '\0';	2,568✔
174	return str_realloc(p);	2,568✔
175	}
176
177	int utf8_char_console_width(const char *str) {	4,470,313✔
178	char32_t c;	4,470,313✔
179	int r;	4,470,313✔
180
181	r = utf8_encoded_to_unichar(str, &c);	4,470,313✔
182	if (r < 0)	4,470,313✔
183	return r;	4,470,313✔
184
185	if (c == '\t')	4,470,312✔
186	return 8; /* Assume a tab width of 8 */
187
188	/* TODO: we should detect combining characters */
189
190	return unichar_iswide(c) ? 2 : 1;	4,459,749✔
191	}
192
193	char* utf8_escape_non_printable_full(const char *str, size_t console_width, bool force_ellipsis) {	551✔
194	char p, s, *prev_s;	551✔
195	size_t n = 0; /* estimated print width */	551✔
196
197	assert(str);	551✔
198
199	if (console_width == 0)	551✔
200	return strdup("");	9✔
201
202	p = s = prev_s = malloc(strlen(str) * 4 + 1);	542✔
203	if (!p)	542✔
204	return NULL;
205
206	for (;;) {	20,681✔
207	int len;	20,681✔
208	char *saved_s = s;	20,681✔
209
210	if (!str) { / done! */	20,681✔
211	if (force_ellipsis)	468✔
212	goto truncation;	73✔
213	else
214	goto finish;	395✔
215	}
216
217	len = utf8_encoded_valid_unichar(str, SIZE_MAX);	20,213✔
218	if (len > 0) {	20,213✔
219	if (utf8_is_printable(str, len)) {	20,117✔
220	int w;	19,954✔
221
222	w = utf8_char_console_width(str);	19,954✔
223	assert(w >= 0);	19,954✔
224	if (n + w > console_width)	19,954✔
225	goto truncation;	34✔
226
227	s = mempcpy(s, str, len);	19,920✔
228	str += len;	19,920✔
229	n += w;	19,920✔
230
231	} else {
232	for (; len > 0; len--) {	288✔
233	if (n + 4 > console_width)	163✔
234	goto truncation;	38✔
235
236	*(s++) = '\\';	125✔
237	*(s++) = 'x';	125✔
238	(s++) = hexchar((int) str >> 4);	125✔
239	(s++) = hexchar((int) str);	125✔
240
241	str += 1;	125✔
242	n += 4;	125✔
243	}
244	}
245	} else {
246	if (n + 1 > console_width)	96✔
247	goto truncation;	2✔
248
249	s = mempcpy(s, UTF8_REPLACEMENT_CHARACTER, strlen(UTF8_REPLACEMENT_CHARACTER));	94✔
250	str += 1;	94✔
251	n += 1;	94✔
252	}
253
254	prev_s = saved_s;
255	}
256
257	truncation:	147✔
258	/* Try to go back one if we don't have enough space for the ellipsis */
259	if (n + 1 > console_width)	147✔
260	s = prev_s;	88✔
261
262	s = mempcpy(s, "…", strlen("…"));	147✔
263
264	finish:	542✔
265	*s = '\0';	542✔
266	return str_realloc(p);	542✔
267	}
268
269	char* ascii_is_valid_n(const char *str, size_t len) {	48,968✔
270	/* Check whether the string consists of valid ASCII bytes, i.e values between 1 and 127, inclusive.
271	* Stops at len, or NUL byte if len is SIZE_MAX. */
272
273	assert(str);	48,968✔
274
275	for (size_t i = 0; len != SIZE_MAX ? i < len : str[i] != '\0'; i++)	1,304,526✔
276	if ((unsigned char) str[i] >= 128 \|\| str[i] == '\0')	1,258,904✔
277	return NULL;
278
279	return (char*) str;
280	}
281
282	int utf8_to_ascii(const char str, char replacement_char, char *ret) {	34✔
283	/* Convert to a string that has only ASCII chars, replacing anything that is not ASCII
284	* by replacement_char. */
285
286	_cleanup_free_ char *ans = new(char, strlen(str) + 1);	68✔
287	if (!ans)	34✔
288	return -ENOMEM;
289
290	char *q = ans;
291
292	for (const char p = str; p; q++) {	211✔
293	int l;	179✔
294
295	l = utf8_encoded_valid_unichar(p, SIZE_MAX);	179✔
296	if (l < 0) /* Non-UTF-8, let's not even try to propagate the garbage */	179✔
297	return l;
298
299	if (l == 1)	177✔
300	q = p;	160✔
301	else
302	/* non-ASCII, we need to replace it */
303	*q = replacement_char;	17✔
304
305	p += l;	177✔
306	}
307	*q = '\0';	32✔
308
309	*ret = TAKE_PTR(ans);	32✔
310	return 0;	32✔
311	}
312
313	/**
314	* utf8_encode_unichar() - Encode single UCS-4 character as UTF-8
315	* @out_utf8: output buffer of at least 4 bytes or NULL
316	* @g: UCS-4 character to encode
317	*
318	* This encodes a single UCS-4 character as UTF-8 and writes it into @out_utf8.
319	* The length of the character is returned. It is not zero-terminated! If the
320	* output buffer is NULL, only the length is returned.
321	*
322	* Returns: The length in bytes that the UTF-8 representation does or would
323	* occupy.
324	*/
325	size_t utf8_encode_unichar(char *out_utf8, char32_t g) {	68,024✔
326
327	if (g < (1 << 7)) {	68,024✔
328	if (out_utf8)	67,900✔
329	out_utf8[0] = g & 0x7f;	67,900✔
330	return 1;	67,900✔
331	} else if (g < (1 << 11)) {	124✔
332	if (out_utf8) {	17✔
333	out_utf8[0] = 0xc0 \| ((g >> 6) & 0x1f);	17✔
334	out_utf8[1] = 0x80 \| (g & 0x3f);	17✔
335	}
336	return 2;	17✔
337	} else if (g < (1 << 16)) {	107✔
338	if (out_utf8) {	96✔
339	out_utf8[0] = 0xe0 \| ((g >> 12) & 0x0f);	96✔
340	out_utf8[1] = 0x80 \| ((g >> 6) & 0x3f);	96✔
341	out_utf8[2] = 0x80 \| (g & 0x3f);	96✔
342	}
343	return 3;	96✔
344	} else if (g < (1 << 21)) {	11✔
345	if (out_utf8) {	11✔
346	out_utf8[0] = 0xf0 \| ((g >> 18) & 0x07);	11✔
347	out_utf8[1] = 0x80 \| ((g >> 12) & 0x3f);	11✔
348	out_utf8[2] = 0x80 \| ((g >> 6) & 0x3f);	11✔
349	out_utf8[3] = 0x80 \| (g & 0x3f);	11✔
350	}
351	return 4;	11✔
352	}
353
354	return 0;
355	}
356
357	char* utf16_to_utf8(const char16_t s, size_t length / bytes! */) {	1,917✔
358	const uint8_t *f;	1,917✔
359	char r, t;	1,917✔
360
361	if (length == 0)	1,917✔
UNCOV 362	return new0(char, 1);	×
363
364	assert(s);	1,917✔
365
366	if (length == SIZE_MAX) {	1,917✔
367	length = char16_strlen(s);	6✔
368
369	if (length > SIZE_MAX/2)	6✔
370	return NULL; /* overflow */
371
372	length *= 2;	6✔
373	}
374
375	/* Input length is in bytes, i.e. the shortest possible character takes 2 bytes. Each unicode character may
376	* take up to 4 bytes in UTF-8. Let's also account for a trailing NUL byte. */
377	if (length > (SIZE_MAX - 1) / 2)	1,917✔
378	return NULL; /* overflow */
379
380	r = new(char, length * 2 + 1);	1,917✔
381	if (!r)	1,917✔
382	return NULL;
383
384	f = (const uint8_t*) s;
385	t = r;
386
387	while (f + 1 < (const uint8_t*) s + length) {	69,752✔
388	char16_t w1, w2;	67,835✔
389
390	/* see RFC 2781 section 2.2 */
391
392	w1 = f[1] << 8 \| f[0];	67,835✔
393	f += 2;	67,835✔
394
395	if (!utf16_is_surrogate(w1)) {	67,835✔
396	t += utf8_encode_unichar(t, w1);	67,828✔
397	continue;	67,828✔
398	}
399
400	if (utf16_is_trailing_surrogate(w1))	7✔
401	continue; /* spurious trailing surrogate, ignore */	1✔
402
403	if (f + 1 >= (const uint8_t*) s + length)	6✔
404	break;
405
406	w2 = f[1] << 8 \| f[0];	6✔
407	f += 2;	6✔
408
409	if (!utf16_is_trailing_surrogate(w2)) {	6✔
410	f -= 2;	1✔
411	continue; /* surrogate missing its trailing surrogate, ignore */	1✔
412	}
413
414	t += utf8_encode_unichar(t, utf16_surrogate_pair_to_unichar(w1, w2));	5✔
415	}
416
417	*t = 0;	1,917✔
418	return r;	1,917✔
419	}
420
421	size_t utf16_encode_unichar(char16_t *out, char32_t c) {	16✔
422
423	/* Note that this encodes as little-endian. */
424
425	switch (c) {	16✔
426
427	case 0 ... 0xd7ffU:	12✔
428	case 0xe000U ... 0xffffU:
429	out[0] = htole16(c);	12✔
430	return 1;	12✔
431
432	case 0x10000U ... 0x10ffffU:	4✔
433	c -= 0x10000U;	4✔
434	out[0] = htole16((c >> 10) + 0xd800U);	4✔
435	out[1] = htole16((c & 0x3ffU) + 0xdc00U);	4✔
436	return 2;	4✔
437
438	default: /* A surrogate (invalid) */
439	return 0;
440	}
441	}
442
443	char16_t utf8_to_utf16(const char s, size_t length) {	139✔
444	char16_t n, p;	139✔
445	int r;	139✔
446
447	if (length == 0)	139✔
UNCOV 448	return new0(char16_t, 1);	×
449
450	assert(s);	139✔
451
452	if (length == SIZE_MAX)	139✔
453	length = strlen(s);	138✔
454
455	if (length > SIZE_MAX - 1)	138✔
456	return NULL; /* overflow */
457
458	n = new(char16_t, length + 1);	139✔
459	if (!n)	139✔
460	return NULL;
461
462	p = n;
463
464	for (size_t i = 0; i < length;) {	1,373✔
465	char32_t unichar;	1,234✔
466	size_t e;	1,234✔
467
468	e = utf8_encoded_expected_len(s[i]);	1,234✔
469	if (e <= 1) /* Invalid and single byte characters are copied as they are */	1,234✔
470	goto copy;	1,218✔
471
472	if (i + e > length) /* sequence longer than input buffer, then copy as-is */	16✔
UNCOV 473	goto copy;	×
474
475	r = utf8_encoded_to_unichar(s + i, &unichar);	16✔
476	if (r < 0) /* sequence invalid, then copy as-is */	16✔
UNCOV 477	goto copy;	×
478
479	p += utf16_encode_unichar(p, unichar);	16✔
480	i += e;	16✔
481	continue;	16✔
482
483	copy:	1,218✔
484	*(p++) = htole16(s[i++]);	1,218✔
485	}
486
487	*p = 0;	139✔
488	return n;	139✔
489	}
490
491	size_t char16_strlen(const char16_t *s) {	403✔
492	size_t n = 0;	403✔
493
494	assert(s);	403✔
495
496	while (*s != 0)	4,818✔
497	n++, s++;	4,415✔
498
499	return n;	403✔
500	}
501
502	size_t char16_strsize(const char16_t *s) {	3✔
503	return s ? (char16_strlen(s) + 1) * sizeof(*s) : 0;	3✔
504	}
505
506	/* expected size used to encode one unicode char */
507	static int utf8_unichar_to_encoded_len(char32_t unichar) {	84,683✔
508
509	if (unichar < 0x80)	84,683✔
510	return 1;
511	if (unichar < 0x800)	84,683✔
512	return 2;
513	if (unichar < 0x10000)	80,441✔
514	return 3;
515	if (unichar < 0x200000)	380✔
516	return 4;
UNCOV 517	if (unichar < 0x4000000)	×
UNCOV 518	return 5;	×
519
520	return 6;
521	}
522
523	/* validate one encoded unicode char and return its length */
524	int utf8_encoded_valid_unichar(const char str, size_t length / bytes */) {	525,470,336✔
525	char32_t unichar;	525,470,336✔
526	size_t len;	525,470,336✔
527	int r;	525,470,336✔
528
529	assert(str);	525,470,336✔
530	assert(length > 0);	525,470,336✔
531
532	/* We read until NUL, at most length bytes. SIZE_MAX may be used to disable the length check. */
533
534	len = utf8_encoded_expected_len(str[0]);	525,470,336✔
535	if (len == 0)	525,470,336✔
536	return -EINVAL;	525,470,336✔
537
538	/* Do we have a truncated multi-byte character? */
539	if (len > length)	525,470,272✔
540	return -EINVAL;
541
542	/* ascii is valid */
543	if (len == 1)	525,470,264✔
544	return 1;
545
546	/* check if expected encoded chars are available */
547	for (size_t i = 0; i < len; i++)	337,696✔
548	if ((str[i] & 0x80) != 0x80)	253,008✔
549	return -EINVAL;
550
551	r = utf8_encoded_to_unichar(str, &unichar);	84,688✔
552	if (r < 0)	84,688✔
553	return r;
554
555	/* check if encoded length matches encoded value */
556	if (utf8_unichar_to_encoded_len(unichar) != (int) len)	84,683✔
557	return -EINVAL;
558
559	/* check if value has valid range */
560	if (!unichar_is_valid(unichar))	84,683✔
561	return -EINVAL;	6✔
562
563	return (int) len;
564	}
565
566	size_t utf8_n_codepoints(const char *str) {	6✔
567	size_t n = 0;	6✔
568
569	/* Returns the number of UTF-8 codepoints in this string, or SIZE_MAX if the string is not valid UTF-8. */
570
571	while (*str != 0) {	34✔
572	int k;	29✔
573
574	k = utf8_encoded_valid_unichar(str, SIZE_MAX);	29✔
575	if (k < 0)	29✔
576	return SIZE_MAX;
577
578	str += k;	28✔
579	n++;	28✔
580	}
581
582	return n;
583	}
584
585	size_t utf8_console_width(const char *str) {	300,606✔
586
587	if (isempty(str))	300,606✔
588	return 0;
589
590	/* Returns the approximate width a string will take on screen when printed on a character cell
591	* terminal/console. */
592
593	size_t n = 0;
594	while (*str) {	4,110,264✔
595	int w;	3,822,391✔
596
597	w = utf8_char_console_width(str);	3,822,391✔
598	if (w < 0)	3,822,391✔
599	return SIZE_MAX;
600
601	n += w;	3,822,390✔
602	str = utf8_next_char(str);	3,822,390✔
603	}
604
605	return n;
606	}
607
608	size_t utf8_last_length(const char *s, size_t n) {	9✔
609	int r;	9✔
610
611	assert(s);	9✔
612
613	if (n == SIZE_MAX)	9✔
614	n = strlen(s);	7✔
615
616	/* Determines length in bytes of last UTF-8 codepoint in string. If the string is empty, returns
617	* zero. Treats invalid UTF-8 codepoints as 1 sized ones. */
618
619	for (size_t last = 0;;) {	19✔
620	if (n == 0)	28✔
621	return last;	9✔
622
623	r = utf8_encoded_valid_unichar(s, n);	19✔
624	if (r <= 0)	19✔
UNCOV 625	r = 1; /* treat invalid UTF-8 as byte-wide */	×
626
627	s += r;	19✔
628	n -= r;	19✔
629	last = r;	19✔
630	}
631	}

systemd / systemd / 15263807472

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