commit bash-20151023 snapshot

This commit is contained in:
Chet Ramey
2015-11-02 10:49:23 -05:00
parent 33723c8402
commit 07357ec296
64 changed files with 6405 additions and 8823 deletions
-241
View File
@@ -16,132 +16,6 @@
/* Written by Bruno Haible <bruno@clisp.org>. */
/* A multibyte character is a short subsequence of a char* string,
representing a single wide character.
We use multibyte characters instead of wide characters because of
the following goals:
1) correct multibyte handling, i.e. operate according to the LC_CTYPE
locale,
2) ease of maintenance, i.e. the maintainer needs not know all details
of the ISO C 99 standard,
3) don't fail grossly if the input is not in the encoding set by the
locale, because often different encodings are in use in the same
countries (ISO-8859-1/UTF-8, EUC-JP/Shift_JIS, ...),
4) fast in the case of ASCII characters,
5) portability, i.e. don't make unportable assumptions about wchar_t.
Multibyte characters are only accessed through the mb* macros.
mb_ptr (mbc)
return a pointer to the beginning of the multibyte sequence.
mb_len (mbc)
returns the number of bytes occupied by the multibyte sequence.
Always > 0.
mb_iseq (mbc, sc)
returns true if mbc is the standard ASCII character sc.
mb_isnul (mbc)
returns true if mbc is the nul character.
mb_cmp (mbc1, mbc2)
returns a positive, zero, or negative value depending on whether mbc1
sorts after, same or before mbc2.
mb_casecmp (mbc1, mbc2)
returns a positive, zero, or negative value depending on whether mbc1
sorts after, same or before mbc2, modulo upper/lowercase conversion.
mb_equal (mbc1, mbc2)
returns true if mbc1 and mbc2 are equal.
mb_caseequal (mbc1, mbc2)
returns true if mbc1 and mbc2 are equal modulo upper/lowercase conversion.
mb_isalnum (mbc)
returns true if mbc is alphanumeric.
mb_isalpha (mbc)
returns true if mbc is alphabetic.
mb_isascii(mbc)
returns true if mbc is plain ASCII.
mb_isblank (mbc)
returns true if mbc is a blank.
mb_iscntrl (mbc)
returns true if mbc is a control character.
mb_isdigit (mbc)
returns true if mbc is a decimal digit.
mb_isgraph (mbc)
returns true if mbc is a graphic character.
mb_islower (mbc)
returns true if mbc is lowercase.
mb_isprint (mbc)
returns true if mbc is a printable character.
mb_ispunct (mbc)
returns true if mbc is a punctuation character.
mb_isspace (mbc)
returns true if mbc is a space character.
mb_isupper (mbc)
returns true if mbc is uppercase.
mb_isxdigit (mbc)
returns true if mbc is a hexadecimal digit.
mb_width (mbc)
returns the number of columns on the output device occupied by mbc.
Always >= 0.
mb_putc (mbc, stream)
outputs mbc on stream, a byte oriented FILE stream opened for output.
mb_setascii (&mbc, sc)
assigns the standard ASCII character sc to mbc.
mb_copy (&destmbc, &srcmbc)
copies srcmbc to destmbc.
Here are the function prototypes of the macros.
typedef int bool;
extern const char * mb_ptr (const mbchar_t mbc);
extern size_t mb_len (const mbchar_t mbc);
extern bool mb_iseq (const mbchar_t mbc, char sc);
extern bool mb_isnul (const mbchar_t mbc);
extern int mb_cmp (const mbchar_t mbc1, const mbchar_t mbc2);
extern int mb_casecmp (const mbchar_t mbc1, const mbchar_t mbc2);
extern bool mb_equal (const mbchar_t mbc1, const mbchar_t mbc2);
extern bool mb_caseequal (const mbchar_t mbc1, const mbchar_t mbc2);
extern bool mb_isalnum (const mbchar_t mbc);
extern bool mb_isalpha (const mbchar_t mbc);
extern bool mb_isascii (const mbchar_t mbc);
extern bool mb_isblank (const mbchar_t mbc);
extern bool mb_iscntrl (const mbchar_t mbc);
extern bool mb_isdigit (const mbchar_t mbc);
extern bool mb_isgraph (const mbchar_t mbc);
extern bool mb_islower (const mbchar_t mbc);
extern bool mb_isprint (const mbchar_t mbc);
extern bool mb_ispunct (const mbchar_t mbc);
extern bool mb_isspace (const mbchar_t mbc);
extern bool mb_isupper (const mbchar_t mbc);
extern bool mb_isxdigit (const mbchar_t mbc);
extern int mb_width (const mbchar_t mbc);
extern void mb_putc (const mbchar_t mbc, FILE *stream);
extern void mb_setascii (mbchar_t *new, char sc);
extern void mb_copy (mbchar_t *new, const mbchar_t *old);
*/
#ifndef _SHMBCHAR_H
#define _SHMBCHAR_H 1
@@ -158,121 +32,6 @@
#include <wchar.h>
#include <wctype.h>
#define MBCHAR_BUF_SIZE 24
struct mbchar
{
const char *ptr; /* pointer to current character */
size_t bytes; /* number of bytes of current character, > 0 */
int wc_valid; /* true if wc is a valid wide character */
wchar_t wc; /* if wc_valid: the current character */
char buf[MBCHAR_BUF_SIZE]; /* room for the bytes, used for file input only */
};
/* EOF (not a real character) is represented with bytes = 0 and
wc_valid = false. */
typedef struct mbchar mbchar_t;
/* Access the current character. */
#define mb_ptr(mbc) ((mbc).ptr)
#define mb_len(mbc) ((mbc).bytes)
/* Comparison of characters. */
#define mb_iseq(mbc, sc) ((mbc).wc_valid && (mbc).wc == (sc))
#define mb_isnul(mbc) ((mbc).wc_valid && (mbc).wc == 0)
#define mb_cmp(mbc1, mbc2) \
((mbc1).wc_valid \
? ((mbc2).wc_valid \
? (int) (mbc1).wc - (int) (mbc2).wc \
: -1) \
: ((mbc2).wc_valid \
? 1 \
: (mbc1).bytes == (mbc2).bytes \
? memcmp ((mbc1).ptr, (mbc2).ptr, (mbc1).bytes) \
: (mbc1).bytes < (mbc2).bytes \
? (memcmp ((mbc1).ptr, (mbc2).ptr, (mbc1).bytes) > 0 ? 1 : -1) \
: (memcmp ((mbc1).ptr, (mbc2).ptr, (mbc2).bytes) >= 0 ? 1 : -1)))
#define mb_casecmp(mbc1, mbc2) \
((mbc1).wc_valid \
? ((mbc2).wc_valid \
? (int) towlower ((mbc1).wc) - (int) towlower ((mbc2).wc) \
: -1) \
: ((mbc2).wc_valid \
? 1 \
: (mbc1).bytes == (mbc2).bytes \
? memcmp ((mbc1).ptr, (mbc2).ptr, (mbc1).bytes) \
: (mbc1).bytes < (mbc2).bytes \
? (memcmp ((mbc1).ptr, (mbc2).ptr, (mbc1).bytes) > 0 ? 1 : -1) \
: (memcmp ((mbc1).ptr, (mbc2).ptr, (mbc2).bytes) >= 0 ? 1 : -1)))
#define mb_equal(mbc1, mbc2) \
((mbc1).wc_valid && (mbc2).wc_valid \
? (mbc1).wc == (mbc2).wc \
: (mbc1).bytes == (mbc2).bytes \
&& memcmp ((mbc1).ptr, (mbc2).ptr, (mbc1).bytes) == 0)
#define mb_caseequal(mbc1, mbc2) \
((mbc1).wc_valid && (mbc2).wc_valid \
? towlower ((mbc1).wc) == towlower ((mbc2).wc) \
: (mbc1).bytes == (mbc2).bytes \
&& memcmp ((mbc1).ptr, (mbc2).ptr, (mbc1).bytes) == 0)
/* <ctype.h>, <wctype.h> classification. */
#define mb_isascii(mbc) \
((mbc).wc_valid && (mbc).wc >= 0 && (mbc).wc <= 127)
#define mb_isalnum(mbc) ((mbc).wc_valid && iswalnum ((mbc).wc))
#define mb_isalpha(mbc) ((mbc).wc_valid && iswalpha ((mbc).wc))
#define mb_isblank(mbc) ((mbc).wc_valid && iswblank ((mbc).wc))
#define mb_iscntrl(mbc) ((mbc).wc_valid && iswcntrl ((mbc).wc))
#define mb_isdigit(mbc) ((mbc).wc_valid && iswdigit ((mbc).wc))
#define mb_isgraph(mbc) ((mbc).wc_valid && iswgraph ((mbc).wc))
#define mb_islower(mbc) ((mbc).wc_valid && iswlower ((mbc).wc))
#define mb_isprint(mbc) ((mbc).wc_valid && iswprint ((mbc).wc))
#define mb_ispunct(mbc) ((mbc).wc_valid && iswpunct ((mbc).wc))
#define mb_isspace(mbc) ((mbc).wc_valid && iswspace ((mbc).wc))
#define mb_isupper(mbc) ((mbc).wc_valid && iswupper ((mbc).wc))
#define mb_isxdigit(mbc) ((mbc).wc_valid && iswxdigit ((mbc).wc))
/* Extra <wchar.h> function. */
/* Unprintable characters appear as a small box of width 1. */
#define MB_UNPRINTABLE_WIDTH 1
static inline int
mb_width_aux (wint_t wc)
{
int w = wcwidth (wc);
/* For unprintable characters, arbitrarily return 0 for control characters
and MB_UNPRINTABLE_WIDTH otherwise. */
return (w >= 0 ? w : iswcntrl (wc) ? 0 : MB_UNPRINTABLE_WIDTH);
}
#define mb_width(mbc) \
((mbc).wc_valid ? mb_width_aux ((mbc).wc) : MB_UNPRINTABLE_WIDTH)
/* Output. */
#define mb_putc(mbc, stream) fwrite ((mbc).ptr, 1, (mbc).bytes, (stream))
/* Assignment. */
#define mb_setascii(mbc, sc) \
((mbc)->ptr = (mbc)->buf, (mbc)->bytes = 1, (mbc)->wc_valid = 1, \
(mbc)->wc = (mbc)->buf[0] = (sc))
/* Copying a character. */
static inline void
mb_copy (mbchar_t *new_mbc, const mbchar_t *old_mbc)
{
if (old_mbc->ptr == &old_mbc->buf[0])
{
memcpy (&new_mbc->buf[0], &old_mbc->buf[0], old_mbc->bytes);
new_mbc->ptr = &new_mbc->buf[0];
}
else
new_mbc->ptr = old_mbc->ptr;
new_mbc->bytes = old_mbc->bytes;
if ((new_mbc->wc_valid = old_mbc->wc_valid))
new_mbc->wc = old_mbc->wc;
}
/* is_basic(c) tests whether the single-byte character c is in the
ISO C "basic character set".