hyung-hwan/qse
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
31fc83fbeb
qse/qse/lib/cmn/str-fnmat.c

700 lines
16 KiB
C
Raw Blame History

/*
* $Id$
*
Copyright (c) 2006-2019 Chung, Hyung-Hwan. All rights reserved.
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions
are met:
1. Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
2. Redistributions in binary form must reproduce the above copyright
notice, this list of conditions and the following disclaimer in the
documentation and/or other materials provided with the distribution.
THIS SOFTWARE IS PROVIDED BY THE AUTHOR "AS IS" AND ANY EXPRESS OR
IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT,
INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*/
#include <qse/cmn/str.h>
#include <qse/cmn/chr.h>
/* ------------------------------------------------------------------------- */
/* | MBS VERSION | */
/* ------------------------------------------------------------------------- */
/* separator to use in a pattern.
* this also matches a backslash in non-unix OS where a blackslash
* is used as a path separator. MBS_SEPCHS defines OS path separators. */
#define MBS_SEP QSE_MT('/')
#define MBS_ESC QSE_MT('\\')
#define MBS_DOT QSE_MT('.')
#if defined(_WIN32) || defined(__OS2__) || defined(__DOS__)
# define MBS_ISSEPCH(c) ((c) == QSE_MT('/') || (c) == QSE_MT('\\'))
# define MBS_SEPCHS QSE_MT("/\\")
#else
# define MBS_ISSEPCH(c) ((c) == QSE_MT('/'))
# define MBS_SEPCHS QSE_MT("/")
#endif
static int __mbsxnfnmat (
const qse_mchar_t* str, qse_size_t slen,
const qse_mchar_t* ptn, qse_size_t plen, int flags, int no_first_period)
{
const qse_mchar_t* sp = str;
const qse_mchar_t* pp = ptn;
const qse_mchar_t* se = str + slen;
const qse_mchar_t* pe = ptn + plen;
qse_mchar_t sc, pc, pc2;
while (1)
{
if (pp < pe && *pp == MBS_ESC &&
!(flags & QSE_MBSFNMAT_NOESCAPE))
{
/* pattern is escaped and escaping is allowed. */
if ((++pp) >= pe)
{
/*
* the last character of the pattern is an MBS_ESC.
* matching is performed as if the end of the pattern is
* reached just without an MBS_ESC.
*/
if (sp < se) return 0;
return 1;
}
if (sp >= se) return 0; /* premature string termination */
sc = *sp; pc = *pp; /* pc is just a normal character */
if ((flags & QSE_MBSFNMAT_IGNORECASE) != 0)
{
/* make characters to lower-case */
sc = QSE_TOMLOWER(sc);
pc = QSE_TOMLOWER(pc);
}
if (sc != pc) return 0;
sp++; pp++;
continue;
}
if (pp >= pe)
{
/*
* the end of the pattern has been reached.
* the string must terminate too.
*/
return sp >= se;
}
if (sp >= se)
{
/* the string terminats prematurely */
while (pp < pe && *pp == QSE_MT('*')) pp++;
return pp >= pe;
}
sc = *sp; pc = *pp;
if (sc == MBS_DOT && (flags & QSE_MBSFNMAT_PERIOD))
{
/*
* a leading period in the staring must match
* a period in the pattern explicitly
*/
if ((!no_first_period && sp == str) ||
(MBS_ISSEPCH(sp[-1]) && (flags & QSE_MBSFNMAT_PATHNAME)))
{
if (pc != MBS_DOT) return 0;
sp++; pp++;
continue;
}
}
else if (MBS_ISSEPCH(sc) && (flags & QSE_MBSFNMAT_PATHNAME))
{
while (pc == QSE_MT('*'))
{
if ((++pp) >= pe) return 0;
pc = *pp;
}
/* a path separator must be matched explicitly */
if (pc != MBS_SEP) return 0;
sp++; pp++;
continue;
}
/* the handling of special pattern characters begins here */
if (pc == QSE_MT('?'))
{
/* match any single character */
sp++; pp++;
}
else if (pc == QSE_MT('*'))
{
/* match zero or more characters */
/* compact asterisks */
do { pp++; } while (pp < pe && *pp == QSE_MT('*'));
if (pp >= pe)
{
/*
* if the last character in the pattern is an asterisk,
* the string should not have any directory separators
* when QSE_MBSFNMAT_PATHNAME is set.
*/
if (flags & QSE_MBSFNMAT_PATHNAME)
{
if (qse_mbsxpbrk(sp, se-sp, MBS_SEPCHS) != QSE_NULL) return 0;
}
return 1;
}
else
{
do
{
if (__mbsxnfnmat(sp, se - sp, pp, pe - pp, flags, 1))
{
return 1;
}
if (MBS_ISSEPCH(*sp) &&
(flags & QSE_MBSFNMAT_PATHNAME)) break;
sp++;
}
while (sp < se);
return 0;
}
}
else if (pc == QSE_MT('['))
{
/* match range */
int negate = 0;
int matched = 0;
if ((++pp) >= pe) return 0;
if (*pp == QSE_MT('!')) { negate = 1; pp++; }
while (pp < pe && *pp != QSE_MT(']'))
{
if (*pp == QSE_MT('['))
{
qse_size_t pl = pe - pp;
if (pl >= 10)
{
if (qse_mbszcmp(pp, QSE_MT("[:xdigit:]"), 10) == 0)
{
matched = QSE_ISMXDIGIT(sc);
pp += 10; continue;
}
}
if (pl >= 9)
{
/* TODO: binary search or something more efficient ? */
if (qse_mbszcmp(pp, QSE_MT("[:upper:]"), 9) == 0)
{
matched = QSE_ISMUPPER(sc);
pp += 9; continue;
}
else if (qse_mbszcmp(pp, QSE_MT("[:lower:]"), 9) == 0)
{
matched = QSE_ISMLOWER(sc);
pp += 9; continue;
}
else if (qse_mbszcmp(pp, QSE_MT("[:alpha:]"), 9) == 0)
{
matched = QSE_ISMALPHA(sc);
pp += 9; continue;
}
else if (qse_mbszcmp(pp, QSE_MT("[:digit:]"), 9) == 0)
{
matched = QSE_ISMDIGIT(sc);
pp += 9; continue;
}
else if (qse_mbszcmp(pp, QSE_MT("[:alnum:]"), 9) == 0)
{
matched = QSE_ISMALNUM(sc);
pp += 9; continue;
}
else if (qse_mbszcmp(pp, QSE_MT("[:space:]"), 9) == 0)
{
matched = QSE_ISMSPACE(sc);
pp += 9; continue;
}
else if (qse_mbszcmp(pp, QSE_MT("[:print:]"), 9) == 0)
{
matched = QSE_ISMPRINT(sc);
pp += 9; continue;
}
else if (qse_mbszcmp(pp, QSE_MT("[:graph:]"), 9) == 0)
{
matched = QSE_ISMGRAPH(sc);
pp += 9; continue;
}
else if (qse_mbszcmp(pp, QSE_MT("[:cntrl:]"), 9) == 0)
{
matched = QSE_ISMCNTRL(sc);
pp += 9; continue;
}
else if (qse_mbszcmp(pp, QSE_MT("[:punct:]"), 9) == 0)
{
matched = QSE_ISMPUNCT(sc);
pp += 9; continue;
}
}
/*
* characters in an invalid class name are
* just treated as normal characters
*/
}
if (*pp == MBS_ESC &&
!(flags & QSE_MBSFNMAT_NOESCAPE)) pp++;
else if (*pp == QSE_MT(']')) break;
if (pp >= pe) break;
pc = *pp;
if ((flags & QSE_MBSFNMAT_IGNORECASE) != 0)
{
sc = QSE_TOMLOWER(sc);
pc = QSE_TOMLOWER(pc);
}
if (pp + 1 < pe && pp[1] == QSE_MT('-'))
{
pp += 2; /* move the a character next to a dash */
if (pp >= pe)
{
if (sc >= pc) matched = 1;
break;
}
if (*pp == MBS_ESC &&
!(flags & QSE_MBSFNMAT_NOESCAPE))
{
if ((++pp) >= pe)
{
if (sc >= pc) matched = 1;
break;
}
}
else if (*pp == QSE_MT(']'))
{
if (sc >= pc) matched = 1;
break;
}
pc2 = *pp;
if ((flags & QSE_MBSFNMAT_IGNORECASE) != 0)
pc2 = QSE_TOMLOWER(pc2);
if (sc >= pc && sc <= pc2) matched = 1;
pp++;
}
else
{
if (sc == pc) matched = 1;
pp++;
}
}
if (negate) matched = !matched;
if (!matched) return 0;
sp++; if (pp < pe) pp++;
}
else
{
/* a normal character */
if ((flags & QSE_MBSFNMAT_IGNORECASE) != 0)
{
sc = QSE_TOMLOWER(sc);
pc = QSE_TOMLOWER(pc);
}
if (sc != pc) return 0;
sp++; pp++;
}
}
/* will never reach here. but make some immature compilers happy... */
return 0;
}
int qse_mbsfnmat (const qse_mchar_t* str, const qse_mchar_t* ptn, int flags)
{
return __mbsxnfnmat (
str, qse_mbslen(str), ptn, qse_mbslen(ptn), flags, 0);
}
int qse_mbsxfnmat (
const qse_mchar_t* str, qse_size_t slen, const qse_mchar_t* ptn, int flags)
{
return __mbsxnfnmat (str, slen, ptn, qse_mbslen(ptn), flags, 0);
}
int qse_mbsnfnmat (
const qse_mchar_t* str, const qse_mchar_t* ptn, qse_size_t plen, int flags)
{
return __mbsxnfnmat (str, qse_mbslen(str), ptn, plen, flags, 0);
}
int qse_mbsxnfnmat (
const qse_mchar_t* str, qse_size_t slen,
const qse_mchar_t* ptn, qse_size_t plen, int flags)
{
return __mbsxnfnmat (str, slen, ptn, plen, flags, 0);
}
/* ------------------------------------------------------------------------- */
/* | WCS VERSION | */
/* ------------------------------------------------------------------------- */
/* separator to use in a pattern.
* this also matches a backslash in non-unix OS where a blackslash
* is used as a path separator. WCS_SEPCHS defines OS path separators. */
#define WCS_SEP QSE_WT('/')
#define WCS_ESC QSE_WT('\\')
#define WCS_DOT QSE_WT('.')
#if defined(_WIN32) || defined(__OS2__) || defined(__DOS__)
# define WCS_ISSEPCH(c) ((c) == QSE_WT('/') || (c) == QSE_WT('\\'))
# define WCS_SEPCHS QSE_WT("/\\")
#else
# define WCS_ISSEPCH(c) ((c) == QSE_WT('/'))
# define WCS_SEPCHS QSE_WT("/")
#endif
static int __wcsxnfnmat (
const qse_wchar_t* str, qse_size_t slen,
const qse_wchar_t* ptn, qse_size_t plen, int flags, int no_first_period)
{
const qse_wchar_t* sp = str;
const qse_wchar_t* pp = ptn;
const qse_wchar_t* se = str + slen;
const qse_wchar_t* pe = ptn + plen;
qse_wchar_t sc, pc, pc2;
while (1)
{
if (pp < pe && *pp == WCS_ESC &&
!(flags & QSE_WCSFNMAT_NOESCAPE))
{
/* pattern is escaped and escaping is allowed. */
if ((++pp) >= pe)
{
/*
* the last character of the pattern is an WCS_ESC.
* matching is performed as if the end of the pattern is
* reached just without an WCS_ESC.
*/
if (sp < se) return 0;
return 1;
}
if (sp >= se) return 0; /* premature string termination */
sc = *sp; pc = *pp; /* pc is just a normal character */
if ((flags & QSE_WCSFNMAT_IGNORECASE) != 0)
{
/* make characters to lower-case */
sc = QSE_TOWLOWER(sc);
pc = QSE_TOWLOWER(pc);
}
if (sc != pc) return 0;
sp++; pp++;
continue;
}
if (pp >= pe)
{
/*
* the end of the pattern has been reached.
* the string must terminate too.
*/
return sp >= se;
}
if (sp >= se)
{
/* the string terminats prematurely */
while (pp < pe && *pp == QSE_WT('*')) pp++;
return pp >= pe;
}
sc = *sp; pc = *pp;
if (sc == WCS_DOT && (flags & QSE_WCSFNMAT_PERIOD))
{
/*
* a leading period in the staring must match
* a period in the pattern explicitly
*/
if ((!no_first_period && sp == str) ||
(WCS_ISSEPCH(sp[-1]) && (flags & QSE_WCSFNMAT_PATHNAME)))
{
if (pc != WCS_DOT) return 0;
sp++; pp++;
continue;
}
}
else if (WCS_ISSEPCH(sc) && (flags & QSE_WCSFNMAT_PATHNAME))
{
while (pc == QSE_WT('*'))
{
if ((++pp) >= pe) return 0;
pc = *pp;
}
/* a path separator must be matched explicitly */
if (pc != WCS_SEP) return 0;
sp++; pp++;
continue;
}
/* the handling of special pattern characters begins here */
if (pc == QSE_WT('?'))
{
/* match any single character */
sp++; pp++;
}
else if (pc == QSE_WT('*'))
{
/* match zero or more characters */
/* compact asterisks */
do { pp++; } while (pp < pe && *pp == QSE_WT('*'));
if (pp >= pe)
{
/*
* if the last character in the pattern is an asterisk,
* the string should not have any directory separators
* when QSE_WCSFNMAT_PATHNAME is set.
*/
if (flags & QSE_WCSFNMAT_PATHNAME)
{
if (qse_wcsxpbrk(sp, se-sp, WCS_SEPCHS) != QSE_NULL) return 0;
}
return 1;
}
else
{
do
{
if (__wcsxnfnmat(sp, se - sp, pp, pe - pp, flags, 1))
{
return 1;
}
if (WCS_ISSEPCH(*sp) &&
(flags & QSE_WCSFNMAT_PATHNAME)) break;
sp++;
}
while (sp < se);
return 0;
}
}
else if (pc == QSE_WT('['))
{
/* match range */
int negate = 0;
int matched = 0;
if ((++pp) >= pe) return 0;
if (*pp == QSE_WT('!')) { negate = 1; pp++; }
while (pp < pe && *pp != QSE_WT(']'))
{
if (*pp == QSE_WT('['))
{
qse_size_t pl = pe - pp;
if (pl >= 10)
{
if (qse_wcszcmp(pp, QSE_WT("[:xdigit:]"), 10) == 0)
{
matched = QSE_ISWXDIGIT(sc);
pp += 10; continue;
}
}
if (pl >= 9)
{
if (qse_wcszcmp(pp, QSE_WT("[:upper:]"), 9) == 0)
{
matched = QSE_ISWUPPER(sc);
pp += 9; continue;
}
else if (qse_wcszcmp(pp, QSE_WT("[:lower:]"), 9) == 0)
{
matched = QSE_ISWLOWER(sc);
pp += 9; continue;
}
else if (qse_wcszcmp(pp, QSE_WT("[:alpha:]"), 9) == 0)
{
matched = QSE_ISWALPHA(sc);
pp += 9; continue;
}
else if (qse_wcszcmp(pp, QSE_WT("[:digit:]"), 9) == 0)
{
matched = QSE_ISWDIGIT(sc);
pp += 9; continue;
}
else if (qse_wcszcmp(pp, QSE_WT("[:alnum:]"), 9) == 0)
{
matched = QSE_ISWALNUM(sc);
pp += 9; continue;
}
else if (qse_wcszcmp(pp, QSE_WT("[:space:]"), 9) == 0)
{
matched = QSE_ISWSPACE(sc);
pp += 9; continue;
}
else if (qse_wcszcmp(pp, QSE_WT("[:print:]"), 9) == 0)
{
matched = QSE_ISWPRINT(sc);
pp += 9; continue;
}
else if (qse_wcszcmp(pp, QSE_WT("[:graph:]"), 9) == 0)
{
matched = QSE_ISWGRAPH(sc);
pp += 9; continue;
}
else if (qse_wcszcmp(pp, QSE_WT("[:cntrl:]"), 9) == 0)
{
matched = QSE_ISWCNTRL(sc);
pp += 9; continue;
}
else if (qse_wcszcmp(pp, QSE_WT("[:punct:]"), 9) == 0)
{
matched = QSE_ISWPUNCT(sc);
pp += 9; continue;
}
}
/*
* characters in an invalid class name are
* just treated as normal characters
*/
}
if (*pp == WCS_ESC &&
!(flags & QSE_WCSFNMAT_NOESCAPE)) pp++;
else if (*pp == QSE_WT(']')) break;
if (pp >= pe) break;
pc = *pp;
if ((flags & QSE_WCSFNMAT_IGNORECASE) != 0)
{
sc = QSE_TOWLOWER(sc);
pc = QSE_TOWLOWER(pc);
}
if (pp + 1 < pe && pp[1] == QSE_WT('-'))
{
pp += 2; /* move the a character next to a dash */
if (pp >= pe)
{
if (sc >= pc) matched = 1;
break;
}
if (*pp == WCS_ESC &&
!(flags & QSE_WCSFNMAT_NOESCAPE))
{
if ((++pp) >= pe)
{
if (sc >= pc) matched = 1;
break;
}
}
else if (*pp == QSE_WT(']'))
{
if (sc >= pc) matched = 1;
break;
}
pc2 = *pp;
if ((flags & QSE_WCSFNMAT_IGNORECASE) != 0)
pc2 = QSE_TOWLOWER(pc2);
if (sc >= pc && sc <= pc2) matched = 1;
pp++;
}
else
{
if (sc == pc) matched = 1;
pp++;
}
}
if (negate) matched = !matched;
if (!matched) return 0;
sp++; if (pp < pe) pp++;
}
else
{
/* a normal character */
if ((flags & QSE_WCSFNMAT_IGNORECASE) != 0)
{
sc = QSE_TOWLOWER(sc);
pc = QSE_TOWLOWER(pc);
}
if (sc != pc) return 0;
sp++; pp++;
}
}
/* will never reach here. but make some immature compilers happy... */
return 0;
}
int qse_wcsfnmat (const qse_wchar_t* str, const qse_wchar_t* ptn, int flags)
{
return __wcsxnfnmat (
str, qse_wcslen(str), ptn, qse_wcslen(ptn), flags, 0);
}
int qse_wcsxfnmat (
const qse_wchar_t* str, qse_size_t slen, const qse_wchar_t* ptn, int flags)
{
return __wcsxnfnmat (str, slen, ptn, qse_wcslen(ptn), flags, 0);
}
int qse_wcsnfnmat (
const qse_wchar_t* str, const qse_wchar_t* ptn, qse_size_t plen, int flags)
{
return __wcsxnfnmat (str, qse_wcslen(str), ptn, plen, flags, 0);
}
int qse_wcsxnfnmat (
const qse_wchar_t* str, qse_size_t slen,
const qse_wchar_t* ptn, qse_size_t plen, int flags)
{
return __wcsxnfnmat (str, slen, ptn, plen, flags, 0);
}
Reference in New Issue View Git Blame Copy Permalink