From 025b4c410577a10692c608e7dbad712655abb858 Mon Sep 17 00:00:00 2001 From: Shawn Landden Date: Sat, 31 Oct 2015 20:39:15 -0700 Subject: [PATCH] utf8.[ch]: use char32_t and char16_t instead of int, int32_t, int16_t rework C11 utf8.[ch] to use char32_t instead of uint32_t when referring to unicode chars, to make things more expressive. --- TODO | 3 --- src/basic/escape.c | 12 ++++++------ src/basic/extract-word.c | 2 +- src/basic/json.c | 4 ++-- src/basic/string-util.c | 4 ++-- src/basic/utf8.c | 35 +++++++++++++++++++---------------- src/basic/utf8.h | 13 +++++++------ 7 files changed, 37 insertions(+), 36 deletions(-) diff --git a/TODO b/TODO index 8deb00b597..33b14e7b75 100644 --- a/TODO +++ b/TODO @@ -120,9 +120,6 @@ Features: * nspawn: as soon as networkd has a bus interface, hook up --network-interface=, --network-bridge= with networkd, to trigger netdev creation should an interface be missing -* rework C11 utf8.[ch] to use char32_t instead of uint32_t when referring - to unicode chars, to make things more expressive. - * "machinectl migrate" or similar to copy a container from or to a difference host, via ssh diff --git a/src/basic/escape.c b/src/basic/escape.c index add0d7795b..5cb8287b61 100644 --- a/src/basic/escape.c +++ b/src/basic/escape.c @@ -226,7 +226,7 @@ int cunescape_one(const char *p, size_t length, char *ret, uint32_t *ret_unicode int a[8]; unsigned i; - uint32_t c; + char32_t c; if (length != (size_t) -1 && length < 9) return -EINVAL; @@ -237,8 +237,8 @@ int cunescape_one(const char *p, size_t length, char *ret, uint32_t *ret_unicode return a[i]; } - c = ((uint32_t) a[0] << 28U) | ((uint32_t) a[1] << 24U) | ((uint32_t) a[2] << 20U) | ((uint32_t) a[3] << 16U) | - ((uint32_t) a[4] << 12U) | ((uint32_t) a[5] << 8U) | ((uint32_t) a[6] << 4U) | (uint32_t) a[7]; + c = (a[0] << 28U) | (a[1] << 24U) | (a[2] << 20U) | (a[3] << 16U) | + (a[4] << 12U) | (a[5] << 8U) | (a[6] << 4U) | (a[7] << 0U); /* Don't allow 0 chars */ if (c == 0) @@ -272,7 +272,7 @@ int cunescape_one(const char *p, size_t length, char *ret, uint32_t *ret_unicode case '7': { /* octal encoding */ int a, b, c; - uint32_t m; + char32_t m; if (length != (size_t) -1 && length < 3) return -EINVAL; @@ -294,7 +294,7 @@ int cunescape_one(const char *p, size_t length, char *ret, uint32_t *ret_unicode return -EINVAL; /* Don't allow bytes above 255 */ - m = ((uint32_t) a << 6U) | ((uint32_t) b << 3U) | (uint32_t) c; + m = (a << 6U) | (b << 3U) | (char32_t) c; if (m > 255) return -EINVAL; @@ -331,7 +331,7 @@ int cunescape_length_with_prefix(const char *s, size_t length, const char *prefi for (f = s, t = r + pl; f < s + length; f++) { size_t remaining; - uint32_t u; + char32_t u; char c; int k; diff --git a/src/basic/extract-word.c b/src/basic/extract-word.c index c0f9394fad..63dedf612c 100644 --- a/src/basic/extract-word.c +++ b/src/basic/extract-word.c @@ -99,7 +99,7 @@ int extract_first_word(const char **p, char **ret, const char *separators, Extra } if (flags & EXTRACT_CUNESCAPE) { - uint32_t u; + char32_t u; r = cunescape_one(*p, (size_t) -1, &c, &u); if (r < 0) { diff --git a/src/basic/json.c b/src/basic/json.c index 716705e5ff..a22bc865a7 100644 --- a/src/basic/json.c +++ b/src/basic/json.c @@ -319,7 +319,7 @@ static int json_parse_string(const char **p, char **ret) { else if (*c == 't') ch = '\t'; else if (*c == 'u') { - uint16_t x; + char16_t x; int r; r = unhex_ucs2(c + 1, &x); @@ -332,7 +332,7 @@ static int json_parse_string(const char **p, char **ret) { return -ENOMEM; if (!utf16_is_surrogate(x)) - n += utf8_encode_unichar(s + n, x); + n += utf8_encode_unichar(s + n, (char32_t) x); else if (utf16_is_trailing_surrogate(x)) return -EINVAL; else { diff --git a/src/basic/string-util.c b/src/basic/string-util.c index 63b9b79df9..4583a3ce2e 100644 --- a/src/basic/string-util.c +++ b/src/basic/string-util.c @@ -413,7 +413,7 @@ char *ellipsize_mem(const char *s, size_t old_length, size_t new_length, unsigne k = 0; for (i = s; k < x && i < s + old_length; i = utf8_next_char(i)) { - int c; + char32_t c; c = utf8_encoded_to_unichar(i); if (c < 0) @@ -425,7 +425,7 @@ char *ellipsize_mem(const char *s, size_t old_length, size_t new_length, unsigne x ++; for (j = s + old_length; k < new_length && j > i; ) { - int c; + char32_t c; j = utf8_prev_char(j); c = utf8_encoded_to_unichar(j); diff --git a/src/basic/utf8.c b/src/basic/utf8.c index 7600d99903..cc041d5b34 100644 --- a/src/basic/utf8.c +++ b/src/basic/utf8.c @@ -54,7 +54,7 @@ #include "utf8.h" #include "util.h" -bool unichar_is_valid(uint32_t ch) { +bool unichar_is_valid(char32_t ch) { if (ch >= 0x110000) /* End of unicode space */ return false; @@ -68,7 +68,7 @@ bool unichar_is_valid(uint32_t ch) { return true; } -static bool unichar_is_control(uint32_t ch) { +static bool unichar_is_control(char32_t ch) { /* 0 to ' '-1 is the C0 range. @@ -104,8 +104,9 @@ static int utf8_encoded_expected_len(const char *str) { } /* decode one unicode char */ -int utf8_encoded_to_unichar(const char *str) { - int unichar, len, i; +char32_t utf8_encoded_to_unichar(const char *str) { + char32_t unichar; + int len, i; assert(str); @@ -113,31 +114,31 @@ int utf8_encoded_to_unichar(const char *str) { switch (len) { case 1: - return (int)str[0]; + return (char32_t)str[0]; case 2: unichar = str[0] & 0x1f; break; case 3: - unichar = (int)str[0] & 0x0f; + unichar = (char32_t)str[0] & 0x0f; break; case 4: - unichar = (int)str[0] & 0x07; + unichar = (char32_t)str[0] & 0x07; break; case 5: - unichar = (int)str[0] & 0x03; + unichar = (char32_t)str[0] & 0x03; break; case 6: - unichar = (int)str[0] & 0x01; + unichar = (char32_t)str[0] & 0x01; break; default: return -EINVAL; } for (i = 1; i < len; i++) { - if (((int)str[i] & 0xc0) != 0x80) + if (((char32_t)str[i] & 0xc0) != 0x80) return -EINVAL; unichar <<= 6; - unichar |= (int)str[i] & 0x3f; + unichar |= (char32_t)str[i] & 0x3f; } return unichar; @@ -149,7 +150,8 @@ bool utf8_is_printable_newline(const char* str, size_t length, bool newline) { assert(str); for (p = str; length;) { - int encoded_len, val; + int encoded_len; + char32_t val; encoded_len = utf8_encoded_valid_unichar(p); if (encoded_len < 0 || @@ -277,7 +279,7 @@ char *ascii_is_valid(const char *str) { * Returns: The length in bytes that the UTF-8 representation does or would * occupy. */ -size_t utf8_encode_unichar(char *out_utf8, uint32_t g) { +size_t utf8_encode_unichar(char *out_utf8, char32_t g) { if (g < (1 << 7)) { if (out_utf8) @@ -321,7 +323,7 @@ char *utf16_to_utf8(const void *s, size_t length) { t = r; while (f < (const uint8_t*) s + length) { - uint16_t w1, w2; + char16_t w1, w2; /* see RFC 2781 section 2.2 */ @@ -329,7 +331,7 @@ char *utf16_to_utf8(const void *s, size_t length) { f += 2; if (!utf16_is_surrogate(w1)) { - t += utf8_encode_unichar(t, w1); + t += utf8_encode_unichar(t, (char32_t) w1); continue; } @@ -373,7 +375,8 @@ static int utf8_unichar_to_encoded_len(int unichar) { /* validate one encoded unicode char and return its length */ int utf8_encoded_valid_unichar(const char *str) { - int len, unichar, i; + int len, i; + char32_t unichar; assert(str); diff --git a/src/basic/utf8.h b/src/basic/utf8.h index e745649f06..20d38f12c8 100644 --- a/src/basic/utf8.h +++ b/src/basic/utf8.h @@ -22,12 +22,13 @@ ***/ #include +#include #include "macro.h" #define UTF8_REPLACEMENT_CHARACTER "\xef\xbf\xbd" -bool unichar_is_valid(uint32_t c); +bool unichar_is_valid(char32_t c); const char *utf8_is_valid(const char *s) _pure_; char *ascii_is_valid(const char *s) _pure_; @@ -38,20 +39,20 @@ bool utf8_is_printable_newline(const char* str, size_t length, bool newline) _pu char *utf8_escape_invalid(const char *s); char *utf8_escape_non_printable(const char *str); -size_t utf8_encode_unichar(char *out_utf8, uint32_t g); +size_t utf8_encode_unichar(char *out_utf8, char32_t g); char *utf16_to_utf8(const void *s, size_t length); int utf8_encoded_valid_unichar(const char *str); -int utf8_encoded_to_unichar(const char *str); +char32_t utf8_encoded_to_unichar(const char *str); -static inline bool utf16_is_surrogate(uint16_t c) { +static inline bool utf16_is_surrogate(char16_t c) { return (0xd800 <= c && c <= 0xdfff); } -static inline bool utf16_is_trailing_surrogate(uint16_t c) { +static inline bool utf16_is_trailing_surrogate(char16_t c) { return (0xdc00 <= c && c <= 0xdfff); } -static inline uint32_t utf16_surrogate_pair_to_unichar(uint16_t lead, uint16_t trail) { +static inline char32_t utf16_surrogate_pair_to_unichar(char16_t lead, char16_t trail) { return ((lead - 0xd800) << 10) + (trail - 0xdc00) + 0x10000; }