30055979440

Committed 23 Jul 2026 05:57PM UTC coverage: 73.111% (+0.04%) from 73.067%

Build # 30055979440

Build Type

push

github

Committed by

yuwata

Commit Message

user-record: validate JSON shell fields with valid_shell()

The user record loader currently uses a generic filename-or-path check
for shell and fallbackShell. This allows values that homectl rejects,
including relative names, control characters, colons, and trailing
slashes.

Use valid_shell() for all three record locations and cover the top-level,
matching per-machine, and status fallback fields at the loader boundary.

Fixes #43066

Coverage Stats

35 of 36 new or added lines in 2 files covered. (97.22%)

514 existing lines in 41 files now uncovered.

346901 of 474487 relevant lines covered (73.11%)

1323376.8 hits per line

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

96.88

/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;

        assert(str);

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

        return unichar_console_width(c);
}

int unichar_console_width(char32_t c) {
        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) {
        assert(out);

        /* 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) {	107,879✔
17
18	if (ch >= 0x110000) /* End of unicode space */	107,879✔
19	return false;
20	if ((ch & 0xFFFFF800) == 0xD800) /* Reserved area for UTF-16 */	107,879✔
21	return false;
22	if ((ch >= 0xFDD0) && (ch <= 0xFDEF)) /* Reserved */	107,879✔
23	return false;
24	if ((ch & 0xFFFE) == 0xFFFE) /* BOM (Byte Order Mark) */	107,879✔
25	return false;	26✔
26
27	return true;
28	}
29
30	static bool unichar_is_control(char32_t ch) {	121,622,164✔
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')) \|\|	121,622,164✔
39	(0x7F <= ch && ch <= 0x9F);	121,621,983✔
40	}
41
42	/* count of characters used to encode one unicode char */
43	static size_t utf8_encoded_expected_len(uint8_t c) {	1,233,661,149✔
44	if (c < 0x80)	1,233,661,149✔
45	return 1;
46	if ((c & 0xe0) == 0xc0)	332,258✔
47	return 2;
48	if ((c & 0xf0) == 0xe0)	293,534✔
49	return 3;
50	if ((c & 0xf8) == 0xf0)	1,169✔
51	return 4;
52	if ((c & 0xfc) == 0xf8)	92✔
53	return 5;
54	if ((c & 0xfe) == 0xfc)	92✔
55	return 6;	1✔
56
57	return 0;
58	}
59
60	/* decode one unicode char */
61	int utf8_encoded_to_unichar(const char str, char32_t ret_unichar) {	135,304,215✔
62	char32_t unichar;	135,304,215✔
63	size_t len;	135,304,215✔
64
65	assert(str);	135,304,215✔
66	assert(ret_unichar);	135,304,215✔
67
68	len = utf8_encoded_expected_len(str[0]);	135,304,215✔
69
70	switch (len) {	135,304,215✔
71	case 1:	135,081,124✔
72	*ret_unichar = (char32_t)str[0];	135,081,124✔
73	return 1;	135,081,124✔
74	case 2:	28,201✔
75	unichar = str[0] & 0x1f;	28,201✔
76	break;	28,201✔
77	case 3:	194,249✔
78	unichar = (char32_t)str[0] & 0x0f;	194,249✔
79	break;	194,249✔
80	case 4:	640✔
81	unichar = (char32_t)str[0] & 0x07;	640✔
82	break;	640✔
83	case 5:	×
84	unichar = (char32_t)str[0] & 0x03;	×
85	break;	×
86	case 6:	1✔
87	unichar = (char32_t)str[0] & 0x01;	1✔
88	break;	1✔
89	default:
90	return -EINVAL;
91	}
92
93	for (size_t i = 1; i < len; i++) {	641,696✔
94	if (((char32_t)str[i] & 0xc0) != 0x80)	418,614✔
95	return -EINVAL;
96
97	unichar <<= 6;	418,605✔
98	unichar \|= (char32_t)str[i] & 0x3f;	418,605✔
99	}
100
101	*ret_unichar = unichar;	223,082✔
102	return len;	223,082✔
103	}
104
105	bool utf8_is_printable_newline(const char* str, size_t length, bool allow_newline) {	4,381,055✔
106	assert(str);	4,381,055✔
107
108	for (const char *p = str; length > 0;) {	126,003,038✔
109	int encoded_len;	121,622,165✔
110	char32_t val;	121,622,165✔
111
112	encoded_len = utf8_encoded_valid_unichar(p, length);	121,622,165✔
113	if (encoded_len < 0)	121,622,165✔
114	return false;	182✔
115	assert(encoded_len > 0 && (size_t) encoded_len <= length);	121,622,164✔
116
117	if (utf8_encoded_to_unichar(p, &val) < 0 \|\|	121,622,164✔
118	unichar_is_control(val) \|\|	121,622,164✔
119	(!allow_newline && val == '\n'))	22,274,232✔
120	return false;
121
122	length -= encoded_len;	121,621,983✔
123	p += encoded_len;	121,621,983✔
124	}
125
126	return true;
127	}
128
129	char* utf8_is_valid_n(const char *str, size_t len_bytes) {	56,756,889✔
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);	56,756,889✔
134
135	for (size_t i = 0; len_bytes != SIZE_MAX ? i < len_bytes : str[i] != '\0'; ) {	983,583,772✔
136	int len;	926,828,024✔
137
138	if (_unlikely_(str[i] == '\0'))	926,828,024✔
139	return NULL; /* embedded NUL */
140
141	len = utf8_encoded_valid_unichar(str + i,	926,828,021✔
142	len_bytes != SIZE_MAX ? len_bytes - i : SIZE_MAX);
143	if (_unlikely_(len < 0))	926,828,021✔
144	return NULL; /* invalid character */
145
146	i += len;	926,826,883✔
147	}
148
149	return (char*) str;
150	}
151
152	char* utf8_escape_invalid(const char *str) {	15,074✔
153	char p, s;	15,074✔
154
155	assert(str);	15,074✔
156
157	p = s = malloc(strlen(str) * 4 + 1);	15,074✔
158	if (!p)	15,074✔
159	return NULL;
160
161	while (*str) {	325,967✔
162	int len;	310,893✔
163
164	len = utf8_encoded_valid_unichar(str, SIZE_MAX);	310,893✔
165	if (len > 0) {	310,893✔
166	s = mempcpy(s, str, len);	310,841✔
167	str += len;	310,841✔
168	} else {
169	s = stpcpy(s, UTF8_REPLACEMENT_CHARACTER);	52✔
170	str += 1;	52✔
171	}
172	}
173
174	*s = '\0';	15,074✔
175	return str_realloc(p);	15,074✔
176	}
177
178	int utf8_char_console_width(const char *str) {	7,482,933✔
179	char32_t c;	7,482,933✔
180	int r;	7,482,933✔
181
182	assert(str);	7,482,933✔
183
184	r = utf8_encoded_to_unichar(str, &c);	7,482,933✔
185	if (r < 0)	7,482,933✔
186	return r;	7,482,933✔
187
188	return unichar_console_width(c);	7,482,932✔
189	}
190
191	int unichar_console_width(char32_t c) {	10,910,089✔
192	if (c == '\t')	10,910,089✔
193	return 8; /* Assume a tab width of 8 */
194
195	/* TODO: we should detect combining characters */
196
197	return unichar_iswide(c) ? 2 : 1;	10,851,741✔
198	}
199
200	char* utf8_escape_non_printable_full(const char *str, size_t console_width, bool force_ellipsis) {	4,284✔
201	char p, s, *prev_s;	4,284✔
202	size_t n = 0; /* estimated print width */	4,284✔
203
204	assert(str);	4,284✔
205
206	if (console_width == 0)	4,284✔
207	return strdup("");	9✔
208
209	size_t body = strlen(str) * 4;	4,275✔
210	p = s = prev_s = malloc(body + STRLEN("…") + 1);	4,275✔
211	if (!p)	4,275✔
212	return NULL;
213
214	for (;;) {	424,697✔
215	int len;	214,486✔
216	char *saved_s = s;	214,486✔
217
218	if (!str) { / done! */	214,486✔
219	if (force_ellipsis)	4,201✔
220	goto truncation;	68✔
221	else
222	goto finish;	4,133✔
223	}
224
225	len = utf8_encoded_valid_unichar(str, SIZE_MAX);	210,285✔
226	if (len > 0) {	210,285✔
227	if (utf8_is_printable(str, len)) {	210,189✔
228	int w;	210,023✔
229
230	w = utf8_char_console_width(str);	210,023✔
231	assert(w >= 0);	210,023✔
232	if (n + w > console_width)	210,023✔
233	goto truncation;	34✔
234
235	s = mempcpy(s, str, len);	209,989✔
236	str += len;	209,989✔
237	n += w;	209,989✔
238
239	} else {
240	for (; len > 0; len--) {	294✔
241	if (n + 4 > console_width)	166✔
242	goto truncation;	38✔
243
244	*(s++) = '\\';	128✔
245	*(s++) = 'x';	128✔
246	(s++) = hexchar((int) str >> 4);	128✔
247	(s++) = hexchar((int) str);	128✔
248
249	str += 1;	128✔
250	n += 4;	128✔
251	}
252	}
253	} else {
254	if (n + 1 > console_width)	96✔
255	goto truncation;	2✔
256
257	s = mempcpy(s, UTF8_REPLACEMENT_CHARACTER, strlen(UTF8_REPLACEMENT_CHARACTER));	94✔
258	str += 1;	94✔
259	n += 1;	94✔
260	}
261
262	prev_s = saved_s;	210,211✔
263	}
264
265	truncation:	142✔
266	/* Try to go back one if we don't have enough space for the ellipsis */
267	if (n + 1 > console_width)	142✔
268	s = prev_s;	82✔
269
270	s = mempcpy(s, "…", strlen("…"));	142✔
271
272	finish:	4,275✔
273	*s = '\0';	4,275✔
274	return str_realloc(p);	4,275✔
275	}
276
277	char* ascii_is_valid_n(const char *str, size_t len) {	109,183✔
278	/* Check whether the string consists of valid ASCII bytes, i.e values between 1 and 127, inclusive.
279	* Stops at len, or NUL byte if len is SIZE_MAX. */
280
281	assert(str);	109,183✔
282
283	for (size_t i = 0; len != SIZE_MAX ? i < len : str[i] != '\0'; i++)	2,689,404✔
284	if ((unsigned char) str[i] >= 128 \|\| str[i] == '\0')	2,583,484✔
285	return NULL;
286
287	return (char*) str;
288	}
289
290	int utf8_to_ascii(const char str, char replacement_char, char *ret) {	51✔
291	/* Convert to a string that has only ASCII chars, replacing anything that is not ASCII
292	* by replacement_char. */
293
294	assert(str);	51✔
295	assert(ret);	51✔
296
297	_cleanup_free_ char *ans = new(char, strlen(str) + 1);	102✔
298	if (!ans)	51✔
299	return -ENOMEM;
300
301	char *q = ans;
302
303	for (const char p = str; p; q++) {	393✔
304	int l;	344✔
305
306	l = utf8_encoded_valid_unichar(p, SIZE_MAX);	344✔
307	if (l < 0) /* Non-UTF-8, let's not even try to propagate the garbage */	344✔
308	return l;
309
310	if (l == 1)	342✔
311	q = p;	325✔
312	else
313	/* non-ASCII, we need to replace it */
314	*q = replacement_char;	17✔
315
316	p += l;	342✔
317	}
318	*q = '\0';	49✔
319
320	*ret = TAKE_PTR(ans);	49✔
321	return 0;	49✔
322	}
323
324	/**
325	* utf8_encode_unichar() - Encode single UCS-4 character as UTF-8
326	* @out_utf8: output buffer of at least 4 bytes or NULL
327	* @g: UCS-4 character to encode
328	*
329	* This encodes a single UCS-4 character as UTF-8 and writes it into @out_utf8.
330	* The length of the character is returned. It is not zero-terminated! If the
331	* output buffer is NULL, only the length is returned.
332	*
333	* Returns: The length in bytes that the UTF-8 representation does or would
334	* occupy.
335	*/
336	size_t utf8_encode_unichar(char *out_utf8, char32_t g) {	84,481✔
337
338	if (g < (1 << 7)) {	84,481✔
339	if (out_utf8)	84,354✔
340	out_utf8[0] = g & 0x7f;	84,354✔
341	return 1;
342	} else if (g < (1 << 11)) {	127✔
343	if (out_utf8) {	17✔
344	out_utf8[0] = 0xc0 \| ((g >> 6) & 0x1f);	17✔
345	out_utf8[1] = 0x80 \| (g & 0x3f);	17✔
346	}
347	return 2;
348	} else if (g < (1 << 16)) {	110✔
349	if (out_utf8) {	99✔
350	out_utf8[0] = 0xe0 \| ((g >> 12) & 0x0f);	99✔
351	out_utf8[1] = 0x80 \| ((g >> 6) & 0x3f);	99✔
352	out_utf8[2] = 0x80 \| (g & 0x3f);	99✔
353	}
354	return 3;
355	} else if (g < (1 << 21)) {	11✔
356	if (out_utf8) {	11✔
357	out_utf8[0] = 0xf0 \| ((g >> 18) & 0x07);	11✔
358	out_utf8[1] = 0x80 \| ((g >> 12) & 0x3f);	11✔
359	out_utf8[2] = 0x80 \| ((g >> 6) & 0x3f);	11✔
360	out_utf8[3] = 0x80 \| (g & 0x3f);	11✔
361	}
362	return 4;
363	}
364
365	return 0;
366	}
367
368	char* utf16_to_utf8(const char16_t s, size_t length / bytes! */) {	2,606✔
369	const uint8_t *f;	2,606✔
370	char r, t;	2,606✔
371
372	if (length == 0)	2,606✔
UNCOV 373	return new0(char, 1);	×
374
375	assert(s);	2,606✔
376
377	if (length == SIZE_MAX) {	2,606✔
378	length = char16_strlen(s);	6✔
379
380	if (length > SIZE_MAX/2)	6✔
381	return NULL; /* overflow */
382
383	length *= 2;	6✔
384	}
385
386	/* Input length is in bytes, i.e. the shortest possible character takes 2 bytes. Each unicode character may
387	* take up to 4 bytes in UTF-8. Let's also account for a trailing NUL byte. */
388	if (length > (SIZE_MAX - 1) / 2)	2,606✔
389	return NULL; /* overflow */
390
391	r = new(char, length * 2 + 1);	2,606✔
392	if (!r)	2,606✔
393	return NULL;
394
395	f = (const uint8_t*) s;
396	t = r;
397
398	while (f + 1 < (const uint8_t*) s + length) {	86,849✔
399	char16_t w1, w2;	84,243✔
400
401	/* see RFC 2781 section 2.2 */
402
403	w1 = f[1] << 8 \| f[0];	84,243✔
404	f += 2;	84,243✔
405
406	if (!utf16_is_surrogate(w1)) {	84,243✔
407	t += utf8_encode_unichar(t, w1);	84,236✔
408	continue;	84,236✔
409	}
410
411	if (utf16_is_trailing_surrogate(w1))	7✔
412	continue; /* spurious trailing surrogate, ignore */	1✔
413
414	if (f + 1 >= (const uint8_t*) s + length)	6✔
415	break;
416
417	w2 = f[1] << 8 \| f[0];	6✔
418	f += 2;	6✔
419
420	if (!utf16_is_trailing_surrogate(w2)) {	6✔
421	f -= 2;	1✔
422	continue; /* surrogate missing its trailing surrogate, ignore */	1✔
423	}
424
425	t += utf8_encode_unichar(t, utf16_surrogate_pair_to_unichar(w1, w2));	5✔
426	}
427
428	*t = 0;	2,606✔
429	return r;	2,606✔
430	}
431
432	size_t utf16_encode_unichar(char16_t *out, char32_t c) {	16✔
433	assert(out);	16✔
434
435	/* Note that this encodes as little-endian. */
436
437	switch (c) {	16✔
438
439	case 0 ... 0xd7ffU:	12✔
440	case 0xe000U ... 0xffffU:
441	out[0] = htole16(c);	12✔
442	return 1;	12✔
443
444	case 0x10000U ... 0x10ffffU:	4✔
445	c -= 0x10000U;	4✔
446	out[0] = htole16((c >> 10) + 0xd800U);	4✔
447	out[1] = htole16((c & 0x3ffU) + 0xdc00U);	4✔
448	return 2;	4✔
449
450	default: /* A surrogate (invalid) */
451	return 0;
452	}
453	}
454
455	char16_t utf8_to_utf16(const char s, size_t length) {	691✔
456	char16_t n, p;	691✔
457	int r;	691✔
458
459	if (length == 0)	691✔
UNCOV 460	return new0(char16_t, 1);	×
461
462	assert(s);	691✔
463
464	if (length == SIZE_MAX)	691✔
465	length = strlen(s);	690✔
466
467	if (length > SIZE_MAX - 1)	690✔
468	return NULL; /* overflow */
469
470	n = new(char16_t, length + 1);	691✔
471	if (!n)	691✔
472	return NULL;
473
474	p = n;
475
476	for (size_t i = 0; i < length;) {	12,223✔
477	char32_t unichar;	11,532✔
478	size_t e;	11,532✔
479
480	e = utf8_encoded_expected_len(s[i]);	11,532✔
481	if (e <= 1) /* Invalid and single byte characters are copied as they are */	11,532✔
482	goto copy;	11,516✔
483
484	if (i + e > length) /* sequence longer than input buffer, then copy as-is */	16✔
UNCOV 485	goto copy;	×
486
487	r = utf8_encoded_to_unichar(s + i, &unichar);	16✔
488	if (r < 0) /* sequence invalid, then copy as-is */	16✔
UNCOV 489	goto copy;	×
490
491	p += utf16_encode_unichar(p, unichar);	16✔
492	i += e;	16✔
493	continue;	16✔
494
495	copy:	11,516✔
496	*(p++) = htole16(s[i++]);	11,516✔
497	}
498
499	*p = 0;	691✔
500	return n;	691✔
501	}
502
503	size_t char16_strlen(const char16_t *s) {	955✔
504	size_t n = 0;	955✔
505
506	assert(s);	955✔
507
508	while (*s != 0)	15,668✔
509	n++, s++;	14,713✔
510
511	return n;	955✔
512	}
513
514	size_t char16_strsize(const char16_t *s) {	3✔
515	POINTER_MAY_BE_NULL(s);	3✔
516
517	return s ? (char16_strlen(s) + 1) * sizeof(*s) : 0;	3✔
518	}
519
520	/* expected size used to encode one unicode char */
521	static int utf8_unichar_to_encoded_len(char32_t unichar) {	107,869✔
522
523	if (unichar < 0x80)	107,869✔
524	return 1;
525	if (unichar < 0x800)	107,869✔
526	return 2;
527	if (unichar < 0x10000)	97,674✔
528	return 3;
529	if (unichar < 0x200000)	430✔
530	return 4;
UNCOV 531	if (unichar < 0x4000000)	×
UNCOV 532	return 5;	×
533
534	return 6;
535	}
536
537	/* validate one encoded unicode char and return its length */
538	int utf8_encoded_valid_unichar(const char str, size_t length / bytes */) {	1,098,345,402✔
539	char32_t unichar;	1,098,345,402✔
540	size_t len;	1,098,345,402✔
541	int r;	1,098,345,402✔
542
543	assert(str);	1,098,345,402✔
544	assert(length > 0);	1,098,345,402✔
545
546	/* We read until NUL, at most length bytes. SIZE_MAX may be used to disable the length check. */
547
548	len = utf8_encoded_expected_len(str[0]);	1,098,345,402✔
549	if (len == 0)	1,098,345,402✔
550	return -EINVAL;	1,098,345,402✔
551
552	/* Do we have a truncated multi-byte character? */
553	if (len > length)	1,098,345,311✔
554	return -EINVAL;
555
556	/* ascii is valid */
557	if (len == 1)	1,098,345,302✔
558	return 1;
559
560	/* check if expected encoded chars are available */
561	for (size_t i = 0; i < len; i++)	424,539✔
562	if ((str[i] & 0x80) != 0x80)	316,665✔
563	return -EINVAL;
564
565	r = utf8_encoded_to_unichar(str, &unichar);	107,874✔
566	if (r < 0)	107,874✔
567	return r;
568
569	/* check if encoded length matches encoded value */
570	if (utf8_unichar_to_encoded_len(unichar) != (int) len)	107,869✔
571	return -EINVAL;
572
573	/* check if value has valid range */
574	if (!unichar_is_valid(unichar))	107,869✔
575	return -EINVAL;	26✔
576
577	return (int) len;
578	}
579
580	size_t utf8_n_codepoints(const char *str) {	6✔
581	size_t n = 0;	6✔
582
583	assert(str);	6✔
584
585	/* Returns the number of UTF-8 codepoints in this string, or SIZE_MAX if the string is not valid UTF-8. */
586
587	while (*str != 0) {	34✔
588	int k;	29✔
589
590	k = utf8_encoded_valid_unichar(str, SIZE_MAX);	29✔
591	if (k < 0)	29✔
592	return SIZE_MAX;
593
594	str += k;	28✔
595	n++;	28✔
596	}
597
598	return n;
599	}
600
601	size_t utf8_console_width(const char *str) {	565,617✔
602	POINTER_MAY_BE_NULL(str);	565,617✔
603
604	if (isempty(str))	565,617✔
605	return 0;
606
607	/* Returns the approximate width a string will take on screen when printed on a character cell
608	* terminal/console. */
609
610	size_t n = 0;
611	while (*str) {	7,814,112✔
612	int w;	7,272,910✔
613
614	w = utf8_char_console_width(str);	7,272,910✔
615	if (w < 0)	7,272,910✔
616	return SIZE_MAX;
617
618	n += w;	7,272,909✔
619	str = utf8_next_char(str);	7,272,909✔
620	}
621
622	return n;
623	}
624
625	size_t utf8_last_length(const char *s, size_t n) {	9✔
626	int r;	9✔
627
628	assert(s);	9✔
629
630	if (n == SIZE_MAX)	9✔
631	n = strlen(s);	7✔
632
633	/* Determines length in bytes of last UTF-8 codepoint in string. If the string is empty, returns
634	* zero. Treats invalid UTF-8 codepoints as 1 sized ones. */
635
636	for (size_t last = 0;;) {	9✔
637	if (n == 0)	28✔
638	return last;	9✔
639
640	r = utf8_encoded_valid_unichar(s, n);	19✔
641	if (r <= 0)	19✔
UNCOV 642	r = 1; /* treat invalid UTF-8 as byte-wide */	×
643
644	s += r;	19✔
645	n -= r;	19✔
646	last = r;	19✔
647	}
648	}

systemd / systemd / 30055979440

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