From 54af55b990eda5a6a0140a3cded8094b42c0c3b7 Mon Sep 17 00:00:00 2001 From: Nicolas Pitre Date: Wed, 9 Apr 2025 21:13:59 -0400 Subject: [PATCH] vt: create ucs_recompose.c using gen_ucs_recompose.py This provides ucs_recompose() to recompose two Unicode characters into a single character if possible. This is needed for the VT to properly display decomposed UTF8 sequences. Note: scripts/checkpatch.pl complains about "... exceeds 100 columns". Please ignore. Signed-off-by: Nicolas Pitre Link: https://lore.kernel.org/r/20250410011839.64418-8-nico@fluxnic.net Signed-off-by: Greg Kroah-Hartman --- drivers/tty/vt/Makefile | 2 +- drivers/tty/vt/ucs_recompose.c | 170 +++++++++++++++++++++++++++++++++ include/linux/consolemap.h | 6 ++ 3 files changed, 177 insertions(+), 1 deletion(-) create mode 100644 drivers/tty/vt/ucs_recompose.c diff --git a/drivers/tty/vt/Makefile b/drivers/tty/vt/Makefile index bee69277bbc3..a63f6c9438da 100644 --- a/drivers/tty/vt/Makefile +++ b/drivers/tty/vt/Makefile @@ -8,7 +8,7 @@ obj-$(CONFIG_VT) += vt_ioctl.o vc_screen.o \ selection.o keyboard.o \ vt.o defkeymap.o obj-$(CONFIG_CONSOLE_TRANSLATIONS) += consolemap.o consolemap_deftbl.o \ - ucs_width.o + ucs_width.o ucs_recompose.o # Files generated that shall be removed upon make clean clean-files := consolemap_deftbl.c defkeymap.c diff --git a/drivers/tty/vt/ucs_recompose.c b/drivers/tty/vt/ucs_recompose.c new file mode 100644 index 000000000000..5c30c989def3 --- /dev/null +++ b/drivers/tty/vt/ucs_recompose.c @@ -0,0 +1,170 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * ucs_recompose.c - Unicode character recomposition + * + * Auto-generated by gen_ucs_recompose.py + * + * Unicode Version: 16.0.0 + * + * This file contains a table with most commonly used Latin, Greek, and + * Cyrillic recomposition pairs only (71 entries). To generate a table with + * all possible recomposition pairs from the Unicode BMP (1000 entries) + * instead, run: + * + * python gen_ucs_recompose.py --full + */ + +#include +#include +#include +#include + +/* + * Structure for recomposition pairs. + * First element is the base character, second is the combining mark, + * third is the recomposed character. + * Using uint16_t to save space since all values are within BMP range. + */ +struct recomposition { + uint16_t base; + uint16_t combining; + uint16_t recomposed; +}; + +/* + * Table of most commonly used Latin, Greek, and Cyrillic recomposition pairs only + * Sorted by base character and then combining character for binary search + */ +static const struct recomposition recomposition_table[] = { + { 0x0041, 0x0300, 0x00C0 }, /* LATIN CAPITAL LETTER A + COMBINING GRAVE ACCENT = LATIN CAPITAL LETTER A WITH GRAVE */ + { 0x0041, 0x0301, 0x00C1 }, /* LATIN CAPITAL LETTER A + COMBINING ACUTE ACCENT = LATIN CAPITAL LETTER A WITH ACUTE */ + { 0x0041, 0x0302, 0x00C2 }, /* LATIN CAPITAL LETTER A + COMBINING CIRCUMFLEX ACCENT = LATIN CAPITAL LETTER A WITH CIRCUMFLEX */ + { 0x0041, 0x0303, 0x00C3 }, /* LATIN CAPITAL LETTER A + COMBINING TILDE = LATIN CAPITAL LETTER A WITH TILDE */ + { 0x0041, 0x0308, 0x00C4 }, /* LATIN CAPITAL LETTER A + COMBINING DIAERESIS = LATIN CAPITAL LETTER A WITH DIAERESIS */ + { 0x0041, 0x030A, 0x00C5 }, /* LATIN CAPITAL LETTER A + COMBINING RING ABOVE = LATIN CAPITAL LETTER A WITH RING ABOVE */ + { 0x0043, 0x0327, 0x00C7 }, /* LATIN CAPITAL LETTER C + COMBINING CEDILLA = LATIN CAPITAL LETTER C WITH CEDILLA */ + { 0x0045, 0x0300, 0x00C8 }, /* LATIN CAPITAL LETTER E + COMBINING GRAVE ACCENT = LATIN CAPITAL LETTER E WITH GRAVE */ + { 0x0045, 0x0301, 0x00C9 }, /* LATIN CAPITAL LETTER E + COMBINING ACUTE ACCENT = LATIN CAPITAL LETTER E WITH ACUTE */ + { 0x0045, 0x0302, 0x00CA }, /* LATIN CAPITAL LETTER E + COMBINING CIRCUMFLEX ACCENT = LATIN CAPITAL LETTER E WITH CIRCUMFLEX */ + { 0x0045, 0x0308, 0x00CB }, /* LATIN CAPITAL LETTER E + COMBINING DIAERESIS = LATIN CAPITAL LETTER E WITH DIAERESIS */ + { 0x0049, 0x0300, 0x00CC }, /* LATIN CAPITAL LETTER I + COMBINING GRAVE ACCENT = LATIN CAPITAL LETTER I WITH GRAVE */ + { 0x0049, 0x0301, 0x00CD }, /* LATIN CAPITAL LETTER I + COMBINING ACUTE ACCENT = LATIN CAPITAL LETTER I WITH ACUTE */ + { 0x0049, 0x0302, 0x00CE }, /* LATIN CAPITAL LETTER I + COMBINING CIRCUMFLEX ACCENT = LATIN CAPITAL LETTER I WITH CIRCUMFLEX */ + { 0x0049, 0x0308, 0x00CF }, /* LATIN CAPITAL LETTER I + COMBINING DIAERESIS = LATIN CAPITAL LETTER I WITH DIAERESIS */ + { 0x004E, 0x0303, 0x00D1 }, /* LATIN CAPITAL LETTER N + COMBINING TILDE = LATIN CAPITAL LETTER N WITH TILDE */ + { 0x004F, 0x0300, 0x00D2 }, /* LATIN CAPITAL LETTER O + COMBINING GRAVE ACCENT = LATIN CAPITAL LETTER O WITH GRAVE */ + { 0x004F, 0x0301, 0x00D3 }, /* LATIN CAPITAL LETTER O + COMBINING ACUTE ACCENT = LATIN CAPITAL LETTER O WITH ACUTE */ + { 0x004F, 0x0302, 0x00D4 }, /* LATIN CAPITAL LETTER O + COMBINING CIRCUMFLEX ACCENT = LATIN CAPITAL LETTER O WITH CIRCUMFLEX */ + { 0x004F, 0x0303, 0x00D5 }, /* LATIN CAPITAL LETTER O + COMBINING TILDE = LATIN CAPITAL LETTER O WITH TILDE */ + { 0x004F, 0x0308, 0x00D6 }, /* LATIN CAPITAL LETTER O + COMBINING DIAERESIS = LATIN CAPITAL LETTER O WITH DIAERESIS */ + { 0x0055, 0x0300, 0x00D9 }, /* LATIN CAPITAL LETTER U + COMBINING GRAVE ACCENT = LATIN CAPITAL LETTER U WITH GRAVE */ + { 0x0055, 0x0301, 0x00DA }, /* LATIN CAPITAL LETTER U + COMBINING ACUTE ACCENT = LATIN CAPITAL LETTER U WITH ACUTE */ + { 0x0055, 0x0302, 0x00DB }, /* LATIN CAPITAL LETTER U + COMBINING CIRCUMFLEX ACCENT = LATIN CAPITAL LETTER U WITH CIRCUMFLEX */ + { 0x0055, 0x0308, 0x00DC }, /* LATIN CAPITAL LETTER U + COMBINING DIAERESIS = LATIN CAPITAL LETTER U WITH DIAERESIS */ + { 0x0059, 0x0301, 0x00DD }, /* LATIN CAPITAL LETTER Y + COMBINING ACUTE ACCENT = LATIN CAPITAL LETTER Y WITH ACUTE */ + { 0x0061, 0x0300, 0x00E0 }, /* LATIN SMALL LETTER A + COMBINING GRAVE ACCENT = LATIN SMALL LETTER A WITH GRAVE */ + { 0x0061, 0x0301, 0x00E1 }, /* LATIN SMALL LETTER A + COMBINING ACUTE ACCENT = LATIN SMALL LETTER A WITH ACUTE */ + { 0x0061, 0x0302, 0x00E2 }, /* LATIN SMALL LETTER A + COMBINING CIRCUMFLEX ACCENT = LATIN SMALL LETTER A WITH CIRCUMFLEX */ + { 0x0061, 0x0303, 0x00E3 }, /* LATIN SMALL LETTER A + COMBINING TILDE = LATIN SMALL LETTER A WITH TILDE */ + { 0x0061, 0x0308, 0x00E4 }, /* LATIN SMALL LETTER A + COMBINING DIAERESIS = LATIN SMALL LETTER A WITH DIAERESIS */ + { 0x0061, 0x030A, 0x00E5 }, /* LATIN SMALL LETTER A + COMBINING RING ABOVE = LATIN SMALL LETTER A WITH RING ABOVE */ + { 0x0063, 0x0327, 0x00E7 }, /* LATIN SMALL LETTER C + COMBINING CEDILLA = LATIN SMALL LETTER C WITH CEDILLA */ + { 0x0065, 0x0300, 0x00E8 }, /* LATIN SMALL LETTER E + COMBINING GRAVE ACCENT = LATIN SMALL LETTER E WITH GRAVE */ + { 0x0065, 0x0301, 0x00E9 }, /* LATIN SMALL LETTER E + COMBINING ACUTE ACCENT = LATIN SMALL LETTER E WITH ACUTE */ + { 0x0065, 0x0302, 0x00EA }, /* LATIN SMALL LETTER E + COMBINING CIRCUMFLEX ACCENT = LATIN SMALL LETTER E WITH CIRCUMFLEX */ + { 0x0065, 0x0308, 0x00EB }, /* LATIN SMALL LETTER E + COMBINING DIAERESIS = LATIN SMALL LETTER E WITH DIAERESIS */ + { 0x0069, 0x0300, 0x00EC }, /* LATIN SMALL LETTER I + COMBINING GRAVE ACCENT = LATIN SMALL LETTER I WITH GRAVE */ + { 0x0069, 0x0301, 0x00ED }, /* LATIN SMALL LETTER I + COMBINING ACUTE ACCENT = LATIN SMALL LETTER I WITH ACUTE */ + { 0x0069, 0x0302, 0x00EE }, /* LATIN SMALL LETTER I + COMBINING CIRCUMFLEX ACCENT = LATIN SMALL LETTER I WITH CIRCUMFLEX */ + { 0x0069, 0x0308, 0x00EF }, /* LATIN SMALL LETTER I + COMBINING DIAERESIS = LATIN SMALL LETTER I WITH DIAERESIS */ + { 0x006E, 0x0303, 0x00F1 }, /* LATIN SMALL LETTER N + COMBINING TILDE = LATIN SMALL LETTER N WITH TILDE */ + { 0x006F, 0x0300, 0x00F2 }, /* LATIN SMALL LETTER O + COMBINING GRAVE ACCENT = LATIN SMALL LETTER O WITH GRAVE */ + { 0x006F, 0x0301, 0x00F3 }, /* LATIN SMALL LETTER O + COMBINING ACUTE ACCENT = LATIN SMALL LETTER O WITH ACUTE */ + { 0x006F, 0x0302, 0x00F4 }, /* LATIN SMALL LETTER O + COMBINING CIRCUMFLEX ACCENT = LATIN SMALL LETTER O WITH CIRCUMFLEX */ + { 0x006F, 0x0303, 0x00F5 }, /* LATIN SMALL LETTER O + COMBINING TILDE = LATIN SMALL LETTER O WITH TILDE */ + { 0x006F, 0x0308, 0x00F6 }, /* LATIN SMALL LETTER O + COMBINING DIAERESIS = LATIN SMALL LETTER O WITH DIAERESIS */ + { 0x0075, 0x0300, 0x00F9 }, /* LATIN SMALL LETTER U + COMBINING GRAVE ACCENT = LATIN SMALL LETTER U WITH GRAVE */ + { 0x0075, 0x0301, 0x00FA }, /* LATIN SMALL LETTER U + COMBINING ACUTE ACCENT = LATIN SMALL LETTER U WITH ACUTE */ + { 0x0075, 0x0302, 0x00FB }, /* LATIN SMALL LETTER U + COMBINING CIRCUMFLEX ACCENT = LATIN SMALL LETTER U WITH CIRCUMFLEX */ + { 0x0075, 0x0308, 0x00FC }, /* LATIN SMALL LETTER U + COMBINING DIAERESIS = LATIN SMALL LETTER U WITH DIAERESIS */ + { 0x0079, 0x0301, 0x00FD }, /* LATIN SMALL LETTER Y + COMBINING ACUTE ACCENT = LATIN SMALL LETTER Y WITH ACUTE */ + { 0x0079, 0x0308, 0x00FF }, /* LATIN SMALL LETTER Y + COMBINING DIAERESIS = LATIN SMALL LETTER Y WITH DIAERESIS */ + { 0x0391, 0x0301, 0x0386 }, /* GREEK CAPITAL LETTER ALPHA + COMBINING ACUTE ACCENT = GREEK CAPITAL LETTER ALPHA WITH TONOS */ + { 0x0395, 0x0301, 0x0388 }, /* GREEK CAPITAL LETTER EPSILON + COMBINING ACUTE ACCENT = GREEK CAPITAL LETTER EPSILON WITH TONOS */ + { 0x0397, 0x0301, 0x0389 }, /* GREEK CAPITAL LETTER ETA + COMBINING ACUTE ACCENT = GREEK CAPITAL LETTER ETA WITH TONOS */ + { 0x0399, 0x0301, 0x038A }, /* GREEK CAPITAL LETTER IOTA + COMBINING ACUTE ACCENT = GREEK CAPITAL LETTER IOTA WITH TONOS */ + { 0x039F, 0x0301, 0x038C }, /* GREEK CAPITAL LETTER OMICRON + COMBINING ACUTE ACCENT = GREEK CAPITAL LETTER OMICRON WITH TONOS */ + { 0x03A5, 0x0301, 0x038E }, /* GREEK CAPITAL LETTER UPSILON + COMBINING ACUTE ACCENT = GREEK CAPITAL LETTER UPSILON WITH TONOS */ + { 0x03A9, 0x0301, 0x038F }, /* GREEK CAPITAL LETTER OMEGA + COMBINING ACUTE ACCENT = GREEK CAPITAL LETTER OMEGA WITH TONOS */ + { 0x03B1, 0x0301, 0x03AC }, /* GREEK SMALL LETTER ALPHA + COMBINING ACUTE ACCENT = GREEK SMALL LETTER ALPHA WITH TONOS */ + { 0x03B5, 0x0301, 0x03AD }, /* GREEK SMALL LETTER EPSILON + COMBINING ACUTE ACCENT = GREEK SMALL LETTER EPSILON WITH TONOS */ + { 0x03B7, 0x0301, 0x03AE }, /* GREEK SMALL LETTER ETA + COMBINING ACUTE ACCENT = GREEK SMALL LETTER ETA WITH TONOS */ + { 0x03B9, 0x0301, 0x03AF }, /* GREEK SMALL LETTER IOTA + COMBINING ACUTE ACCENT = GREEK SMALL LETTER IOTA WITH TONOS */ + { 0x03BF, 0x0301, 0x03CC }, /* GREEK SMALL LETTER OMICRON + COMBINING ACUTE ACCENT = GREEK SMALL LETTER OMICRON WITH TONOS */ + { 0x03C5, 0x0301, 0x03CD }, /* GREEK SMALL LETTER UPSILON + COMBINING ACUTE ACCENT = GREEK SMALL LETTER UPSILON WITH TONOS */ + { 0x03C9, 0x0301, 0x03CE }, /* GREEK SMALL LETTER OMEGA + COMBINING ACUTE ACCENT = GREEK SMALL LETTER OMEGA WITH TONOS */ + { 0x0418, 0x0306, 0x0419 }, /* CYRILLIC CAPITAL LETTER I + COMBINING BREVE = CYRILLIC CAPITAL LETTER SHORT I */ + { 0x0423, 0x0306, 0x040E }, /* CYRILLIC CAPITAL LETTER U + COMBINING BREVE = CYRILLIC CAPITAL LETTER SHORT U */ + { 0x0438, 0x0306, 0x0439 }, /* CYRILLIC SMALL LETTER I + COMBINING BREVE = CYRILLIC SMALL LETTER SHORT I */ + { 0x0443, 0x0306, 0x045E }, /* CYRILLIC SMALL LETTER U + COMBINING BREVE = CYRILLIC SMALL LETTER SHORT U */ +}; + +/* + * Boundary values for quick rejection + * These are calculated by analyzing the table during generation + */ +#define MIN_BASE_CHAR 0x0041 +#define MAX_BASE_CHAR 0x0443 +#define MIN_COMBINING_CHAR 0x0300 +#define MAX_COMBINING_CHAR 0x0327 + +struct compare_key { + uint16_t base; + uint16_t combining; +}; + +static int recomposition_compare(const void *key, const void *element) +{ + const struct compare_key *search_key = key; + const struct recomposition *table_entry = element; + + /* Compare base character first */ + if (search_key->base < table_entry->base) + return -1; + if (search_key->base > table_entry->base) + return 1; + + /* Base characters match, now compare combining character */ + if (search_key->combining < table_entry->combining) + return -1; + if (search_key->combining > table_entry->combining) + return 1; + + /* Both match */ + return 0; +} + +/** + * Attempt to recompose two Unicode characters into a single character. + * + * @param previous: Previous Unicode code point (UCS-4) + * @param current: Current Unicode code point (UCS-4) + * Return: Recomposed Unicode code point, or 0 if no recomposition is possible + */ +uint32_t ucs_recompose(uint32_t base, uint32_t combining) +{ + /* Check if characters are within the range of our table */ + if (base < MIN_BASE_CHAR || base > MAX_BASE_CHAR || + combining < MIN_COMBINING_CHAR || combining > MAX_COMBINING_CHAR) + return 0; + + struct compare_key key = { base, combining }; + + struct recomposition *result = + __inline_bsearch(&key, recomposition_table, + ARRAY_SIZE(recomposition_table), + sizeof(*recomposition_table), + recomposition_compare); + + return result ? result->recomposed : 0; +} diff --git a/include/linux/consolemap.h b/include/linux/consolemap.h index b3a911866662..4d3a34c288e5 100644 --- a/include/linux/consolemap.h +++ b/include/linux/consolemap.h @@ -30,6 +30,7 @@ int conv_uni_to_8bit(u32 uni); void console_map_init(void); bool ucs_is_double_width(uint32_t cp); bool ucs_is_zero_width(uint32_t cp); +uint32_t ucs_recompose(uint32_t base, uint32_t combining); #else static inline u16 inverse_translate(const struct vc_data *conp, u16 glyph, bool use_unicode) @@ -69,6 +70,11 @@ static inline bool ucs_is_zero_width(uint32_t cp) { return false; } + +static inline uint32_t ucs_recompose(uint32_t base, uint32_t combining) +{ + return 0; +} #endif /* CONFIG_CONSOLE_TRANSLATIONS */ #endif /* __LINUX_CONSOLEMAP_H__ */ -- 2.25.1