diff options
Diffstat (limited to 'wcsmbs')
| -rw-r--r-- | wcsmbs/Makefile | 1 | ||||
| -rw-r--r-- | wcsmbs/mbsnrtowcs.c | 143 | ||||
| -rw-r--r-- | wcsmbs/wchar.h | 27 | ||||
| -rw-r--r-- | wcsmbs/wcsnrtombs.c | 129 |
4 files changed, 294 insertions, 6 deletions
diff --git a/wcsmbs/Makefile b/wcsmbs/Makefile index f713892ae1..120f8690a9 100644 --- a/wcsmbs/Makefile +++ b/wcsmbs/Makefile @@ -29,6 +29,7 @@ routines := wcscat wcschr wcscmp wcscpy wcscspn wcsdup wcslen wcsncat \ wmemcmp wmemcpy wmemmove wmemset wcpcpy wcpncpy \ btowc wctob mbsinit \ mbrlen mbrtowc wcrtomb mbsrtowcs wcsrtombs \ + mbsnrtowcs wcsnrtombs \ wcstol wcstoul wcstoq wcstouq wcstod wcstold wcstof \ wcscoll wcsxfrm \ wcwidth wcswidth diff --git a/wcsmbs/mbsnrtowcs.c b/wcsmbs/mbsnrtowcs.c new file mode 100644 index 0000000000..27906957e9 --- /dev/null +++ b/wcsmbs/mbsnrtowcs.c @@ -0,0 +1,143 @@ +/* Copyright (C) 1996 Free Software Foundation, Inc. +This file is part of the GNU C Library. +Contributed by Ulrich Drepper <drepper@gnu.ai.mit.edu>, 1996. + +The GNU C Library is free software; you can redistribute it and/or +modify it under the terms of the GNU Library General Public License as +published by the Free Software Foundation; either version 2 of the +License, or (at your option) any later version. + +The GNU C Library is distributed in the hope that it will be useful, +but WITHOUT ANY WARRANTY; without even the implied warranty of +MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +Library General Public License for more details. + +You should have received a copy of the GNU Library General Public +License along with the GNU C Library; see the file COPYING.LIB. If +not, write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330, +Boston, MA 02111-1307, USA. */ + +#include <errno.h> +#include <wchar.h> + +#ifndef EILSEQ +#define EILSEQ EINVAL +#endif + + +/* We don't need the state really because we don't have shift states + to maintain between calls to this function. */ +static mbstate_t internal; + +/* This is a non-standard function but it is very useful in the + implementation of stdio because we have to deal with unterminated + buffers. At most NMC bytes will be converted. */ +size_t +__mbsnrtowcs (dst, src, nmc, len, ps) + wchar_t *dst; + const char **src; + size_t nmc; + size_t len; + mbstate_t *ps; +{ + size_t written = 0; + const char *run = *src; + const char *last = run + nmc; + + if (ps == NULL) + ps = &internal; + + if (dst == NULL) + /* The LEN parameter has to be ignored if we don't actually write + anything. */ + len = ~0; + + /* Copy all words. */ + while (written < len && run < last) + { + wchar_t value; + size_t count; + unsigned char byte = *run++; + + /* We expect a start of a new multibyte character. */ + if (byte < 0x80) + { + /* One byte sequence. */ + count = 0; + value = byte; + } + else if ((byte & 0xe0) == 0xc0) + { + count = 1; + value = byte & 0x1f; + } + else if ((byte & 0xf0) == 0xe0) + { + /* We expect three bytes. */ + count = 2; + value = byte & 0x0f; + } + else if ((byte & 0xf8) == 0xf0) + { + /* We expect four bytes. */ + count = 3; + value = byte & 0x07; + } + else if ((byte & 0xfc) == 0xf8) + { + /* We expect five bytes. */ + count = 4; + value = byte & 0x03; + } + else if ((byte & 0xfe) == 0xfc) + { + /* We expect six bytes. */ + count = 5; + value = byte & 0x01; + } + else + { + /* This is an illegal encoding. */ + errno = EILSEQ; + return (size_t) -1; + } + + /* Read the possible remaining bytes. */ + while (count-- > 0) + { + byte = *run++; + + if ((byte & 0xc0) != 0x80) + { + /* This is an illegal encoding. */ + errno = EILSEQ; + return (size_t) -1; + } + + value <<= 6; + value |= byte & 0x3f; + } + + /* Store value is required. */ + if (dst != NULL) + *dst++ = value; + + /* The whole sequence is read. Check whether end of string is + reached. */ + if (value == L'\0') + { + /* Found the end of the string. */ + *src = NULL; + return written; + } + + /* Increment counter of produced words. */ + ++written; + } + + /* Store address of next byte to process. */ + *src = run; + + return written; +} +weak_alias (__mbsnrtowcs, mbsnrtowcs) diff --git a/wcsmbs/wchar.h b/wcsmbs/wchar.h index 806bafa655..0346364b44 100644 --- a/wcsmbs/wchar.h +++ b/wcsmbs/wchar.h @@ -131,7 +131,7 @@ extern wchar_t *wmemcpy __P ((wchar_t *__s1, __const wchar_t *__s2, /* Copy N bytes of SRC to DEST, guaranteeing correct behavior for overlapping strings. */ extern wchar_t *wmemmove __P ((wchar_t *__s1, __const wchar_t *__s2, - size_t __N)); + size_t __n)); /* Set N bytes of S to C. */ extern wchar_t *wmemset __P ((wchar_t *__s, wchar_t __c, size_t __n)); @@ -168,8 +168,8 @@ extern __inline size_t mbrlen (__const char *s, size_t n, mbstate_t *ps) { return ps != NULL ? mbrtowc (NULL, s, n, ps) : __mbrlen (s, n, NULL); } #endif -/* Write wide character representation of multibyte chracter string SRC - to DST. */ +/* Write wide character representation of multibyte character string + SRC to DST. */ extern size_t mbsrtowcs __P ((wchar_t *__dst, __const char **__src, size_t __len, mbstate_t *__ps)); @@ -180,6 +180,21 @@ extern size_t wcsrtombs __P ((char *__dst, __const wchar_t **__src, #ifdef __USE_GNU +/* Write wide character representation of at most NMC bytes of the + multibyte character string SRC to DST. */ +extern size_t __mbsnrtowcs __P ((wchar_t *__dst, __const char **__src, + size_t __nmc, size_t __len, mbstate_t *__ps)); +extern size_t mbsnrtowcs __P ((wchar_t *__dst, __const char **__src, + size_t __nmc, size_t __len, mbstate_t *__ps)); + +/* Write multibyte character representation of at most NWC characters + from the wide character string SRC to DST. */ +extern size_t __wcsnrtombs __P ((char *__dst, __const wchar_t **__src, + size_t __nwc, size_t __len, mbstate_t *__ps)); +extern size_t wcsnrtombs __P ((char *__dst, __const wchar_t **__src, + size_t __nwc, size_t __len, mbstate_t *__ps)); + + /* The following functions are extensions found in X/Open CAE. */ /* Determine number of column positions required for C. */ @@ -229,11 +244,11 @@ extern unsigned long long int wcstouq __P ((__const wchar_t *__nptr, /* The internal entry points for `wcstoX' take an extra flag argument saying whether or not to parse locale-dependent number grouping. */ extern double __wcstod_internal __P ((__const wchar_t *__nptr, - wchar_t **_endptr, int __group)); + wchar_t **__endptr, int __group)); extern float __wcstof_internal __P ((__const wchar_t *__nptr, - wchar_t **_endptr, int __group)); + wchar_t **__endptr, int __group)); extern __long_double_t __wcstold_internal __P ((__const wchar_t *__nptr, - wchar_t **_endptr, + wchar_t **__endptr, int __group)); extern long int __wcstol_internal __P ((__const wchar_t *__nptr, diff --git a/wcsmbs/wcsnrtombs.c b/wcsmbs/wcsnrtombs.c new file mode 100644 index 0000000000..fb50a11732 --- /dev/null +++ b/wcsmbs/wcsnrtombs.c @@ -0,0 +1,129 @@ +/* Copyright (C) 1996 Free Software Foundation, Inc. +This file is part of the GNU C Library. +Contributed by Ulrich Drepper <drepper@gnu.ai.mit.edu>, 1996. + +The GNU C Library is free software; you can redistribute it and/or +modify it under the terms of the GNU Library General Public License as +published by the Free Software Foundation; either version 2 of the +License, or (at your option) any later version. + +The GNU C Library is distributed in the hope that it will be useful, +but WITHOUT ANY WARRANTY; without even the implied warranty of +MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +Library General Public License for more details. + +You should have received a copy of the GNU Library General Public +License along with the GNU C Library; see the file COPYING.LIB. If +not, write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330, +Boston, MA 02111-1307, USA. */ + +#include <errno.h> +#include <wchar.h> + +#ifndef EILSEQ +#define EILSEQ EINVAL +#endif + + +static const wchar_t encoding_mask[] = +{ + ~0x7ff, ~0xffff, ~0x1fffff, ~0x3ffffff +}; + +static const unsigned char encoding_byte[] = +{ + 0xc0, 0xe0, 0xf0, 0xf8, 0xfc +}; + +/* We don't need the state really because we don't have shift states + to maintain between calls to this function. */ +static mbstate_t internal; + +/* This is a non-standard function but it is very useful in the + implementation of stdio because we have to deal with unterminated + buffers. At most NWC wide character will be converted. */ +size_t +__wcsnrtombs (dst, src, nwc, len, ps) + char *dst; + const wchar_t **src; + size_t nwc; + size_t len; + mbstate_t *ps; +{ + size_t written = 0; + const wchar_t *run = *src; + + if (ps == NULL) + ps = &internal; + + if (dst == NULL) + /* The LEN parameter has to be ignored if we don't actually write + anything. */ + len = ~0; + + while (written < len && nwc-- > 0) + { + wchar_t wc = *run++; + + if (wc < 0 || wc > 0x7fffffff) + { + /* This is no correct ISO 10646 character. */ + errno = EILSEQ; + return (size_t) -1; + } + + if (wc == L'\0') + { + /* Found the end. */ + if (dst != NULL) + *dst = '\0'; + *src = NULL; + return written; + } + else if (wc < 0x80) + { + /* It's an one byte sequence. */ + if (dst != NULL) + *dst++ = (char) wc; + ++written; + } + else + { + size_t step; + + for (step = 2; step < 6; ++step) + if ((wc & encoding_mask[step - 2]) == 0) + break; + + if (written + step >= len) + /* Too long. */ + break; + + if (dst != NULL) + { + size_t cnt = step; + + dst[0] = encoding_byte[cnt - 2]; + + --cnt; + do + { + dst[cnt] = 0x80 | (wc & 0x3f); + wc >>= 6; + } + while (--cnt > 0); + dst[0] |= wc; + + dst += step; + } + + written += step; + } + } + + /* Store position of first unprocessed word. */ + *src = run; + + return written; +} +weak_alias (__wcsnrtombs, wcsnrtombs) |