#include "safe_str_lib.h"
#include "unw16ifcan.h"
#include "unw16ifcmb.h"
#include "unw16ifcmp.h"
#include "unw16ifexc.h"
#include "hangul.h"

Include dependency graph for wcsnorm_s.c:

Data Structures
struct	UNWIF_cc

Macros
#define	_UNICODE_MAX 0x10ffff

#define	CC_SEQ_SIZE 10

#define	CC_SEQ_STEP 5

Functions
bool	isExclusion (uint32_t uv)

bool	isSingleton (uint32_t uv)

bool	isNonStDecomp (uint32_t uv)

bool	isComp2nd (uint32_t uv)

static int	_decomp_canonical_s (wchar_t *dest, rsize_t dmax, uint32_t cp)

static int	_decomp_hangul_s (wchar_t *dest, rsize_t dmax, uint32_t cp)

EXPORT int	_decomp_s (wchar_t *restrict dest, rsize_t dmax, const uint32_t cp, const bool iscompat)

static int	_compare_cc (const void a, const void b)

static uint32_t	_composite_cp (uint32_t cp, uint32_t cp2)

static uint8_t	_combin_class (uint32_t cp)

EXPORT errno_t	wcsnorm_decompose_s (wchar_t restrict dest, rsize_t dmax, wchar_t restrict src, rsize_t *restrict lenp, bool iscompat)
	Converts the wide string to the canonical NFD normalization, as defined in the latest Unicode standard, latest 10.0. More...

EXPORT errno_t	wcsnorm_reorder_s (wchar_t restrict dest, rsize_t dmax, wchar_t restrict p, rsize_t len)
	Reorder all decomposed sequences in a wide string to NFD, as defined in the latest Unicode standard, latest 10.0. More...

EXPORT errno_t	wcsnorm_compose_s (wchar_t restrict dest, rsize_t dmax, wchar_t restrict p, rsize_t *restrict lenp, bool iscontig)
	Combine all decomposed sequences in a wide string to NFC, as defined in the latest Unicode standard, latest 10.0. More...

EXPORT errno_t	wcsnorm_s (wchar_t restrict dest, rsize_t dmax, wchar_t restrict src, wcsnorm_mode_t mode, rsize_t *restrict lenp)
	Converts the wide string to the canonical NFC or NFD normalization, as defined in the latest Unicode standard, latest 10.0. More...

Macro Definition Documentation

◆ _UNICODE_MAX

#define _UNICODE_MAX 0x10ffff

◆ CC_SEQ_SIZE

#define CC_SEQ_SIZE 10

◆ CC_SEQ_STEP

#define CC_SEQ_STEP 5

Function Documentation

◆ isExclusion()

bool isExclusion ( uint32_t uv )

◆ isSingleton()

bool isSingleton ( uint32_t uv )

◆ isNonStDecomp()

bool isNonStDecomp ( uint32_t uv )

◆ isComp2nd()

bool isComp2nd ( uint32_t uv )

◆ _decomp_canonical_s()

static int _decomp_canonical_s	(	wchar_t *	dest,
		rsize_t	dmax,
		uint32_t	cp
	)

static

◆ _decomp_hangul_s()

static int _decomp_hangul_s	(	wchar_t *	dest,
		rsize_t	dmax,
		uint32_t	cp
	)

static

◆ _decomp_s()

EXPORT int _decomp_s	(	wchar_t *restrict	dest,
		rsize_t	dmax,
		const uint32_t	cp,
		const bool	iscompat
	)

◆ _compare_cc()

static int _compare_cc	(	const void *	a,
		const void *	b
	)

static

◆ _composite_cp()

static uint32_t _composite_cp	(	uint32_t	cp,
		uint32_t	cp2
	)

static

◆ _combin_class()

static uint8_t _combin_class ( uint32_t cp )

static

◆ wcsnorm_decompose_s()

EXPORT errno_t wcsnorm_decompose_s	(	wchar_t *restrict	dest,
		rsize_t	dmax,
		wchar_t *restrict	src,
		rsize_t *restrict	lenp,
		bool	iscompat
	)

Converts the wide string to the canonical NFD normalization, as defined in the latest Unicode standard, latest 10.0.

The conversion stops at the first null or after dmax characters.

Composed characters are checked for the left-hand-size of the Decomposition_Mapping Unicode property, which means the codepoint will be normalized if the sequence is composed. This is equivalent to all 1963 combining mark characters, plus some remaining 869 non-mark and non-hangul normalizables. Hangul has some special normalization logic.

This function is used by wcsnorm_s() to do the argument checking, overlap checking and to do the first of three passes for NFC.

Parameters

[out]	dest	wide string to hold the result
[in]	dmax	maximum result buffer size
[in]	src	wide string
[out]	lenp	pointer to length of the result, may be NULL
[in]	iscompat	do NFKD, and not NFD (with `–enable-norm-compat`)

Precondition: dest and src shall not be null pointers.; dmax shall not equal zero and big enough for dest.; dmax shall not be greater than RSIZE_MAX_WSTR.

Returns: If there is a runtime-constraint violation, then if dest is not a null pointer and dmax is greater than zero and not greater than RSIZE_MAX_WSTR, then wcsnorm_s nulls dest.

Return values

EOK	on success
ESNULLP	when dest or src is NULL pointer
ESZEROL	when dmax = 0
ESLEMIN	when dmax < 5 or 19 with a compat mode
ESLEMAX	when dmax > RSIZE_MAX_WSTR
ESOVRLP	when buffers overlap
ESNOSPC	when dmax too small for the result buffer
EOF	on some normalization error

See also: wcsfc_s(), wcsnorm_s(), wcsnorm_compose_s(), wcsnorm_reorder_s(), ICU, gnulib/libunistring, utf8proc

◆ wcsnorm_reorder_s()

EXPORT errno_t wcsnorm_reorder_s	(	wchar_t *restrict	dest,
		rsize_t	dmax,
		wchar_t *restrict	p,
		rsize_t	len
	)

Reorder all decomposed sequences in a wide string to NFD, as defined in the latest Unicode standard, latest 10.0.

The conversion stops at the first null or after dmax characters.

Parameters

[out]	dest	wide string to hold the result
[in]	dmax	maximum result buffer size
[in]	p	wide string to be converted
[in]	len	length of p

Precondition: dest and p shall not be null pointers.; dmax shall not equal zero and big enough for dest.; dmax shall not be greater than RSIZE_MAX_WSTR.

Returns: If there is a runtime-constraint violation, then if dest is not a null pointer and dmax is greater than zero and not greater than RSIZE_MAX_WSTR, then wcsnorm_reorder_s nulls dest.

Return values

EOK	on success
ESNOSPC	when dmax too small for the result buffer
EOF	on some normalization error

See also: wcsnorm_s(), wcsnorm_decompose_s(), ICU, gnulib/libunistring, utf8proc

◆ wcsnorm_compose_s()

EXPORT errno_t wcsnorm_compose_s	(	wchar_t *restrict	dest,
		rsize_t	dmax,
		wchar_t *restrict	p,
		rsize_t *restrict	lenp,
		bool	iscontig
	)

Combine all decomposed sequences in a wide string to NFC, as defined in the latest Unicode standard, latest 10.0.

The conversion stops at the first null or after dmax characters.

Parameters

[out]	dest	wide string to hold the result
[in]	dmax	maximum result buffer size
[in]	p	wide string to be converted
[out]	lenp	pointer to length of p and the result length.
[in]	iscontig	if true, the result will only be a fast FCC

Precondition: dest, p and lenp shall not be null pointers.; dmax shall not equal zero and big enough for dest.; dmax shall not be greater than RSIZE_MAX_WSTR.

Returns: If there is a runtime-constraint violation, then if dest is not a null pointer and dmax is greater than zero and not greater than RSIZE_MAX_WSTR, then wcsnorm_reorder_s nulls dest.

Return values

EOK	on success
ESNOSPC	when dmax too small for the result buffer
EOF	on some normalization error

See also: wcsnorm_s(), wcsnorm_decompose_s(), ICU, gnulib/libunistring, utf8proc

◆ wcsnorm_s()

EXPORT errno_t wcsnorm_s	(	wchar_t *restrict	dest,
		rsize_t	dmax,
		wchar_t *restrict	src,
		wcsnorm_mode_t	mode,
		rsize_t *restrict	lenp
	)

Converts the wide string to the canonical NFC or NFD normalization, as defined in the latest Unicode standard, latest 10.0.

The conversion stops at the first null or after dmax characters.

Decomposed characters are checked for the left-hand-size and then right-hand-side of the Decomposition_Mapping Unicode property, which means the codepoint will be normalized if the sequence is composed or decomposed (NFD or NFKD). This is equivalent to all 1963 combining mark characters, plus some remaining 869 non-mark and non-hangul normalizables. Hangul has some special normalization logic.

The compat tables for NFKC or NFKD are too large for a libc, and mostly unused. As default we only provide the smaller canonical conversions, but it can be enabled with –enable-norm-compat. The compat modes also don't roundtrip.

Parameters

[out]	dest	wide string to hold the result
[in]	dmax	maximum length of string
[in]	src	wide string
[in]	mode	convert to nfc or just nfd. experimentally to fast modes FCD or FCC. optionally to compat modes NFKD, NFKC with `–enable-norm-compat`

See also: enum wcsnorm_mode.

Parameters

[out] lenp pointer to length of the result, may be NULL

Precondition: dest and src shall not be null pointers.; dmax shall not equal zero and big enough for dest.; dmax shall not be greater than RSIZE_MAX_WSTR.

Returns: If there is a runtime-constraint violation, then if dest is not a null pointer and dmax is greater than zero and not greater than RSIZE_MAX_WSTR, then wcsnorm_s nulls dest.

Return values

EOK	on success
ESNULLP	when dest or src is NULL pointer
ESZEROL	when dmax = 0
ESLEMIN	when dmax < 5
ESLEMAX	when dmax > RSIZE_MAX_WSTR
ESOVRLP	when buffers overlap
ESNOSPC	when dmax too small for the result buffer
EOF	any other normalization error

See also: wcsfc_s(), ICU, gnulib/libunistring, utf8proc

Data Structures

Macros

Functions

Macro Definition Documentation

◆ _UNICODE_MAX

◆ CC_SEQ_SIZE

◆ CC_SEQ_STEP

Function Documentation

◆ isExclusion()

◆ isSingleton()

◆ isNonStDecomp()

◆ isComp2nd()

◆ _decomp_canonical_s()

◆ _decomp_hangul_s()

◆ _decomp_s()

◆ _compare_cc()

◆ _composite_cp()

◆ _combin_class()

◆ wcsnorm_decompose_s()

◆ wcsnorm_reorder_s()

◆ wcsnorm_compose_s()

◆ wcsnorm_s()