From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.2 (2018-09-13) on dcvr.yhbt.net X-Spam-Level: X-Spam-Status: No, score=-3.9 required=3.0 tests=AWL,BAYES_00,DKIM_SIGNED, DKIM_VALID,DKIM_VALID_AU,HEADER_FROM_DIFFERENT_DOMAINS, MAILING_LIST_MULTI,RCVD_IN_MSPIKE_H4,RCVD_IN_MSPIKE_WL,SPF_HELO_PASS, SPF_PASS shortcircuit=no autolearn=unavailable autolearn_force=no version=3.4.2 Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by dcvr.yhbt.net (Postfix) with ESMTPS id D7E881F4B4 for ; Wed, 30 Dec 2020 20:15:53 +0000 (UTC) Received: from localhost ([::1]:51112 helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1kuhsv-0003Ft-1Q for normalperson@yhbt.net; Wed, 30 Dec 2020 15:15:53 -0500 Received: from eggs.gnu.org ([2001:470:142:3::10]:49912) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1kuhsW-0002zB-Dq for bug-gnulib@gnu.org; Wed, 30 Dec 2020 15:15:28 -0500 Received: from mail-qk1-x72c.google.com ([2607:f8b0:4864:20::72c]:45860) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_128_GCM_SHA256:128) (Exim 4.90_1) (envelope-from ) id 1kuhsQ-0001sK-Ou for bug-gnulib@gnu.org; Wed, 30 Dec 2020 15:15:28 -0500 Received: by mail-qk1-x72c.google.com with SMTP id b64so14827411qkc.12 for ; Wed, 30 Dec 2020 12:15:22 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=linaro.org; s=google; h=from:to:cc:subject:date:message-id:in-reply-to:references :mime-version:content-transfer-encoding; bh=rTAYWez9r91LR7R6P0A6703s+bfjK9PWnKTmhB+/S6M=; b=NLNgB0dZQsBjI8V/X3QwfDe60JsZQMqocTf9ewkg+bANi2Rw45811wOOcDxKnOIls9 WDLjh+KzPXIvUF8Z0OHIbjtCL9mgw3cpbwrxcY1cI1M++mc6ba6xlCejNLkAYUaf2yXv 81NFHWHh+Yz8BrYnlf3bfpm1ggWON3nETwaK+CVwhRPj4/WJkDHIHsypGM1WAf263WTO k1n6Wwq7SYP/ytzq4GAZL7svPZSXbrQ4jmm701Veo93NSrSIoM4C3PtOYPKPpiK3O3eo phw5o1KrStiLkHA3D16xaPqNthAMEKgL4CwYBz04QpRHyzrRxxsnRVQ2huMoiheTrUHR 4eDw== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:to:cc:subject:date:message-id:in-reply-to :references:mime-version:content-transfer-encoding; bh=rTAYWez9r91LR7R6P0A6703s+bfjK9PWnKTmhB+/S6M=; b=tHCSptVpAviljS1GscwnKVmUvtkXj89g8cs6lef1aRtJLfW/JzbcUzMHB3eck1R2Ol MlnywGLtEMkbA2cO/5+DvkfXvUWr5vzlAa981yEDQvpUR/CiG09kSXIt+g46pCYxrqk3 wsRuwBF0P8rsAi/enUNNFzfD43U1h7cDNfJ/ig+CDs9FZ2uxsoDukH2Bg/D/ooXq0enL ybGOjMWH21Wlto4q5/MVdJS7ygr2rgIq+aqUUjrV5us7shgH0c8Q1S+7pOidOzwtoUJh YcrF4+W1YarN19piGja+1MydCBz/iAxx7aMhFg/ipXSnOQ3JW9aAiz3oY6CuAoY8/W5N jmsQ== X-Gm-Message-State: AOAM53070jiXncvlEdRzCnyx1jrccp8TjrT0b1pO65LrVw3dQHtSA8pR 1kwCytNhP53pVEiJQtNN4LOJog== X-Google-Smtp-Source: ABdhPJzzei4aJxiQd6O08GU9gumKsYAyi5eW/fwAJc5GWKL+NCADAbG2l8egKqTwZKO3nuWFa4oiRQ== X-Received: by 2002:a37:a943:: with SMTP id s64mr55473837qke.207.1609359320093; Wed, 30 Dec 2020 12:15:20 -0800 (PST) Received: from localhost.localdomain ([177.194.48.209]) by smtp.googlemail.com with ESMTPSA id v145sm5949057qka.27.2020.12.30.12.15.17 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Wed, 30 Dec 2020 12:15:19 -0800 (PST) From: Adhemerval Zanella To: libc-alpha@sourceware.org, Paul Eggert Subject: [PATCH 5/5] posix: Sync fnmatch with gnulib Date: Wed, 30 Dec 2020 17:15:07 -0300 Message-Id: <20201230201507.2755086-5-adhemerval.zanella@linaro.org> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20201230201507.2755086-1-adhemerval.zanella@linaro.org> References: <20201230201507.2755086-1-adhemerval.zanella@linaro.org> MIME-Version: 1.0 Content-Transfer-Encoding: 8bit Received-SPF: pass client-ip=2607:f8b0:4864:20::72c; envelope-from=adhemerval.zanella@linaro.org; helo=mail-qk1-x72c.google.com X-Spam_score_int: -20 X-Spam_score: -2.1 X-Spam_bar: -- X-Spam_report: (-2.1 / 5.0 requ) BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_NONE=-0.0001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: bug-gnulib@gnu.org X-Mailman-Version: 2.1.23 Precedence: list List-Id: Gnulib discussion list List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Cc: bug-gnulib@gnu.org Errors-To: bug-gnulib-bounces+normalperson=yhbt.net@gnu.org Sender: "bug-gnulib" It sync with gnulib commit 43ee1a6bf with following changes: --- posix/fnmatch_loop.c +++ ../../gnulib/gnulib-lib/lib/fnmatch_loop.c @@ -978,12 +978,12 @@ bool no_leading_period, int flags, size_t alloca_used) { const CHAR *startp; - ssize_t level; + size_t level; struct patternlist { struct patternlist *next; CHAR malloced; - CHAR str[]; + CHAR str[FLEXIBLE_ARRAY_MEMBER]; } *list = NULL; struct patternlist **lastp = &list; size_t pattern_len = STRLEN (pattern); @@ -994,7 +994,7 @@ /* Parse the pattern. Store the individual parts in the list. */ level = 0; - for (startp = p = pattern + 1; level >= 0; ++p) + for (startp = p = pattern + 1; ; ++p) if (*p == L_('\0')) { /* This is an invalid pattern. */ @@ -1065,6 +1065,7 @@ *lastp = newp; \ lastp = &newp->next NEW_PATTERN; + break; } } else if (*p == L_('|')) Because otherwise it triggers some glibc regressions: FAIL: posix/tst-fnmatch FAIL: posix/tst-fnmatch-mem FAIL: posix/tst-fnmatch3 $ cat posix/tst-fnmatch.out [...] 537: fnmatch ("x?y", "x/y/z", FNM_PATHNAME|FNM_LEADING_DIR) = FNM_NOMATCH Didn't expect signal from child: got `Aborted' $ cat posix/tst-fnmatch3.out Didn't expect signal from child: got `Aborted' It is due the loops returns early and hits the assert below. Checked on x86_64-linux-gnu. --- posix/fnmatch.c | 548 +++++------- posix/fnmatch_loop.c | 1977 +++++++++++++++++++++--------------------- 2 files changed, 1208 insertions(+), 1317 deletions(-) diff --git a/posix/fnmatch.c b/posix/fnmatch.c index 315584f900..c80809e783 100644 --- a/posix/fnmatch.c +++ b/posix/fnmatch.c @@ -15,38 +15,29 @@ License along with the GNU C Library; if not, see . */ -#if HAVE_CONFIG_H -# include +#ifndef _LIBC +# include #endif /* Enable GNU extensions in fnmatch.h. */ #ifndef _GNU_SOURCE -# define _GNU_SOURCE 1 +# define _GNU_SOURCE 1 #endif +#include + #include #include -#include #include #include - -#if defined STDC_HEADERS || defined _LIBC -# include -#endif - -#ifdef _LIBC +#include +#if defined _LIBC || HAVE_ALLOCA # include -#else -# define alloca_account(size., var) alloca (size) -#endif - -/* For platform which support the ISO C amendement 1 functionality we - support user defined character classes. */ -#if defined _LIBC || (defined HAVE_WCTYPE_H && defined HAVE_WCHAR_H) -/* Solaris 2.5 has a bug: must be included before . */ -# include -# include #endif +#include +#include +#include +#include /* We need some of the locale data (the collation sequence information) but there is no interface to get this information in general. Therefore @@ -57,199 +48,125 @@ # include # define CONCAT(a,b) __CONCAT(a,b) +# define btowc __btowc +# define iswctype __iswctype # define mbsrtowcs __mbsrtowcs +# define mempcpy __mempcpy +# define strnlen __strnlen +# define towlower __towlower +# define wcscat __wcscat +# define wcslen __wcslen +# define wctype __wctype +# define wmemchr __wmemchr +# define wmempcpy __wmempcpy # define fnmatch __fnmatch extern int fnmatch (const char *pattern, const char *string, int flags); #endif -/* We often have to test for FNM_FILE_NAME and FNM_PERIOD being both set. */ -#define NO_LEADING_PERIOD(flags) \ - ((flags & (FNM_FILE_NAME | FNM_PERIOD)) == (FNM_FILE_NAME | FNM_PERIOD)) - -/* Comment out all this code if we are using the GNU C Library, and are not - actually compiling the library itself. This code is part of the GNU C - Library, but also included in many other GNU distributions. Compiling - and linking in this code is a waste when using the GNU C library - (especially if it is a shared library). Rather than having every GNU - program understand `configure --with-gnu-libc' and omit the object files, - it is simpler to just do this in the source for each such file. */ +#ifdef _LIBC +# if (__GNUC__ >= 7) || (__clang_major__ >= 10) +# define FALLTHROUGH __attribute__ ((__fallthrough__)) +# else +# define FALLTHROUGH ((void) 0) +# endif +#else +# include "attribute.h" +#endif -#if defined _LIBC || !defined __GNU_LIBRARY__ +#include +#include +#ifdef _LIBC +typedef ptrdiff_t idx_t; +#else +# include "idx.h" +#endif -# if defined STDC_HEADERS || !defined isascii -# define ISASCII(c) 1 -# else -# define ISASCII(c) isascii(c) -# endif +/* We often have to test for FNM_FILE_NAME and FNM_PERIOD being both set. */ +#define NO_LEADING_PERIOD(flags) \ + ((flags & (FNM_FILE_NAME | FNM_PERIOD)) == (FNM_FILE_NAME | FNM_PERIOD)) -# ifdef isblank -# define ISBLANK(c) (ISASCII (c) && isblank (c)) +#ifndef _LIBC +# if HAVE_ALLOCA +/* The OS usually guarantees only one guard page at the bottom of the stack, + and a page size can be as small as 4096 bytes. So we cannot safely + allocate anything larger than 4096 bytes. Also care for the possibility + of a few compiler-allocated temporary stack slots. */ +# define __libc_use_alloca(n) ((n) < 4032) # else -# define ISBLANK(c) ((c) == ' ' || (c) == '\t') -# endif -# ifdef isgraph -# define ISGRAPH(c) (ISASCII (c) && isgraph (c)) -# else -# define ISGRAPH(c) (ISASCII (c) && isprint (c) && !isspace (c)) +/* Just use malloc. */ +# define __libc_use_alloca(n) false +# undef alloca +# define alloca(n) malloc (n) # endif +# define alloca_account(size, avar) ((avar) += (size), alloca (size)) +#endif -# define ISPRINT(c) (ISASCII (c) && isprint (c)) -# define ISDIGIT(c) (ISASCII (c) && isdigit (c)) -# define ISALNUM(c) (ISASCII (c) && isalnum (c)) -# define ISALPHA(c) (ISASCII (c) && isalpha (c)) -# define ISCNTRL(c) (ISASCII (c) && iscntrl (c)) -# define ISLOWER(c) (ISASCII (c) && islower (c)) -# define ISPUNCT(c) (ISASCII (c) && ispunct (c)) -# define ISSPACE(c) (ISASCII (c) && isspace (c)) -# define ISUPPER(c) (ISASCII (c) && isupper (c)) -# define ISXDIGIT(c) (ISASCII (c) && isxdigit (c)) - -# define STREQ(s1, s2) ((strcmp (s1, s2) == 0)) - -# if defined _LIBC || (defined HAVE_WCTYPE_H && defined HAVE_WCHAR_H) -/* The GNU C library provides support for user-defined character classes - and the functions from ISO C amendement 1. */ -# ifdef CHARCLASS_NAME_MAX -# define CHAR_CLASS_MAX_LENGTH CHARCLASS_NAME_MAX -# else +/* Provide support for user-defined character classes, based on the functions + from ISO C 90 amendment 1. */ +#ifdef CHARCLASS_NAME_MAX +# define CHAR_CLASS_MAX_LENGTH CHARCLASS_NAME_MAX +#else /* This shouldn't happen but some implementation might still have this problem. Use a reasonable default value. */ -# define CHAR_CLASS_MAX_LENGTH 256 -# endif - -# ifdef _LIBC -# define IS_CHAR_CLASS(string) __wctype (string) -# else -# define IS_CHAR_CLASS(string) wctype (string) -# endif - -# ifdef _LIBC -# define ISWCTYPE(WC, WT) __iswctype (WC, WT) -# else -# define ISWCTYPE(WC, WT) iswctype (WC, WT) -# endif - -# if (HAVE_MBSTATE_T && HAVE_MBSRTOWCS) || _LIBC -/* In this case we are implementing the multibyte character handling. */ -# define HANDLE_MULTIBYTE 1 -# endif +# define CHAR_CLASS_MAX_LENGTH 256 +#endif -# else -# define CHAR_CLASS_MAX_LENGTH 6 /* Namely, `xdigit'. */ - -# define IS_CHAR_CLASS(string) \ - (STREQ (string, "alpha") || STREQ (string, "upper") \ - || STREQ (string, "lower") || STREQ (string, "digit") \ - || STREQ (string, "alnum") || STREQ (string, "xdigit") \ - || STREQ (string, "space") || STREQ (string, "print") \ - || STREQ (string, "punct") || STREQ (string, "graph") \ - || STREQ (string, "cntrl") || STREQ (string, "blank")) -# endif +#define IS_CHAR_CLASS(string) wctype (string) /* Avoid depending on library functions or files whose names are inconsistent. */ -# if !defined _LIBC && !defined getenv -extern char *getenv (); -# endif - -# ifndef errno -extern int errno; -# endif - /* Global variable. */ static int posixly_correct; -/* This function doesn't exist on most systems. */ - -# if !defined HAVE___STRCHRNUL && !defined _LIBC -static char * -__strchrnul (const char *s, int c) -{ - char *result = strchr (s, c); - if (result == NULL) - result = strchr (s, '\0'); - return result; -} -# endif - -# if HANDLE_MULTIBYTE && !defined HAVE___STRCHRNUL && !defined _LIBC -static wchar_t * -__wcschrnul (const wchar_t *s, wint_t c) -{ - wchar_t *result = wcschr (s, c); - if (result == NULL) - result = wcschr (s, '\0'); - return result; -} -# endif - /* Note that this evaluates C many times. */ -# ifdef _LIBC -# define FOLD(c) ((flags & FNM_CASEFOLD) ? tolower (c) : (c)) -# else -# define FOLD(c) ((flags & FNM_CASEFOLD) && ISUPPER (c) ? tolower (c) : (c)) -# endif -# define CHAR char -# define UCHAR unsigned char -# define INT int -# define FCT internal_fnmatch -# define EXT ext_match -# define END end_pattern -# define STRUCT fnmatch_struct -# define L(CS) CS -# ifdef _LIBC -# define BTOWC(C) __btowc (C) -# else -# define BTOWC(C) btowc (C) -# endif -# define STRLEN(S) strlen (S) -# define STRCAT(D, S) strcat (D, S) -# define MEMPCPY(D, S, N) __mempcpy (D, S, N) -# define MEMCHR(S, C, N) memchr (S, C, N) -# define STRCOLL(S1, S2) strcoll (S1, S2) -# define WIDE_CHAR_VERSION 0 +#define FOLD(c) ((flags & FNM_CASEFOLD) ? tolower (c) : (c)) +#define CHAR char +#define UCHAR unsigned char +#define INT int +#define FCT internal_fnmatch +#define EXT ext_match +#define END end_pattern +#define STRUCT fnmatch_struct +#define L_(CS) CS +#define BTOWC(C) btowc (C) +#define STRLEN(S) strlen (S) +#define STRCAT(D, S) strcat (D, S) +#define MEMPCPY(D, S, N) mempcpy (D, S, N) +#define MEMCHR(S, C, N) memchr (S, C, N) +#define WIDE_CHAR_VERSION 0 +#ifdef _LIBC # include # define FINDIDX findidx -# include "fnmatch_loop.c" - - -# if HANDLE_MULTIBYTE -/* Note that this evaluates C many times. */ -# ifdef _LIBC -# define FOLD(c) ((flags & FNM_CASEFOLD) ? __towlower (c) : (c)) -# else -# define FOLD(c) ((flags & FNM_CASEFOLD) && ISUPPER (c) ? towlower (c) : (c)) -# endif -# define CHAR wchar_t -# define UCHAR wint_t -# define INT wint_t -# define FCT internal_fnwmatch -# define EXT ext_wmatch -# define END end_wpattern -# define STRUCT fnwmatch_struct -# define L(CS) L##CS -# define BTOWC(C) (C) -# define STRLEN(S) __wcslen (S) -# define STRCAT(D, S) __wcscat (D, S) -# define MEMPCPY(D, S, N) __wmempcpy (D, S, N) -# define MEMCHR(S, C, N) __wmemchr (S, C, N) -# define STRCOLL(S1, S2) wcscoll (S1, S2) -# ifdef _LIBC -# define WMEMCMP(S1, S2, N) __wmemcmp (S1, S2, N) -# else -# define WMEMCMP(S1, S2, N) wmemcmp (S1, S2, N) -# endif -# define WIDE_CHAR_VERSION 1 +#endif +#include "fnmatch_loop.c" + + +#define FOLD(c) ((flags & FNM_CASEFOLD) ? towlower (c) : (c)) +#define CHAR wchar_t +#define UCHAR wint_t +#define INT wint_t +#define FCT internal_fnwmatch +#define EXT ext_wmatch +#define END end_wpattern +#define L_(CS) L##CS +#define BTOWC(C) (C) +#define STRLEN(S) wcslen (S) +#define STRCAT(D, S) wcscat (D, S) +#define MEMPCPY(D, S, N) wmempcpy (D, S, N) +#define MEMCHR(S, C, N) wmemchr (S, C, N) +#define WIDE_CHAR_VERSION 1 +#ifdef _LIBC /* Change the name the header defines so it doesn't conflict with the version included above. */ -# define findidx findidxwc -# include -# undef findidx -# define FINDIDX findidxwc +# define findidx findidxwc +# include +# undef findidx +# define FINDIDX findidxwc +#endif -# undef IS_CHAR_CLASS +#undef IS_CHAR_CLASS /* We have to convert the wide character string in a multibyte string. But we know that the character class names consist of alphanumeric characters from the portable character set, and since the wide character encoding @@ -265,42 +182,42 @@ is_char_class (const wchar_t *wcs) do { /* Test for a printable character from the portable character set. */ -# ifdef _LIBC +#ifdef _LIBC if (*wcs < 0x20 || *wcs > 0x7e - || *wcs == 0x24 || *wcs == 0x40 || *wcs == 0x60) - return (wctype_t) 0; -# else + || *wcs == 0x24 || *wcs == 0x40 || *wcs == 0x60) + return (wctype_t) 0; +#else switch (*wcs) - { - case L' ': case L'!': case L'"': case L'#': case L'%': - case L'&': case L'\'': case L'(': case L')': case L'*': - case L'+': case L',': case L'-': case L'.': case L'/': - case L'0': case L'1': case L'2': case L'3': case L'4': - case L'5': case L'6': case L'7': case L'8': case L'9': - case L':': case L';': case L'<': case L'=': case L'>': - case L'?': - case L'A': case L'B': case L'C': case L'D': case L'E': - case L'F': case L'G': case L'H': case L'I': case L'J': - case L'K': case L'L': case L'M': case L'N': case L'O': - case L'P': case L'Q': case L'R': case L'S': case L'T': - case L'U': case L'V': case L'W': case L'X': case L'Y': - case L'Z': - case L'[': case L'\\': case L']': case L'^': case L'_': - case L'a': case L'b': case L'c': case L'd': case L'e': - case L'f': case L'g': case L'h': case L'i': case L'j': - case L'k': case L'l': case L'm': case L'n': case L'o': - case L'p': case L'q': case L'r': case L's': case L't': - case L'u': case L'v': case L'w': case L'x': case L'y': - case L'z': case L'{': case L'|': case L'}': case L'~': - break; - default: - return (wctype_t) 0; - } -# endif + { + case L' ': case L'!': case L'"': case L'#': case L'%': + case L'&': case L'\'': case L'(': case L')': case L'*': + case L'+': case L',': case L'-': case L'.': case L'/': + case L'0': case L'1': case L'2': case L'3': case L'4': + case L'5': case L'6': case L'7': case L'8': case L'9': + case L':': case L';': case L'<': case L'=': case L'>': + case L'?': + case L'A': case L'B': case L'C': case L'D': case L'E': + case L'F': case L'G': case L'H': case L'I': case L'J': + case L'K': case L'L': case L'M': case L'N': case L'O': + case L'P': case L'Q': case L'R': case L'S': case L'T': + case L'U': case L'V': case L'W': case L'X': case L'Y': + case L'Z': + case L'[': case L'\\': case L']': case L'^': case L'_': + case L'a': case L'b': case L'c': case L'd': case L'e': + case L'f': case L'g': case L'h': case L'i': case L'j': + case L'k': case L'l': case L'm': case L'n': case L'o': + case L'p': case L'q': case L'r': case L's': case L't': + case L'u': case L'v': case L'w': case L'x': case L'y': + case L'z': case L'{': case L'|': case L'}': case L'~': + break; + default: + return (wctype_t) 0; + } +#endif /* Avoid overrunning the buffer. */ if (cp == s + CHAR_CLASS_MAX_LENGTH) - return (wctype_t) 0; + return (wctype_t) 0; *cp++ = (char) *wcs++; } @@ -308,23 +225,17 @@ is_char_class (const wchar_t *wcs) *cp = '\0'; -# ifdef _LIBC - return __wctype (s); -# else return wctype (s); -# endif } -# define IS_CHAR_CLASS(string) is_char_class (string) +#define IS_CHAR_CLASS(string) is_char_class (string) -# include "fnmatch_loop.c" -# endif +#include "fnmatch_loop.c" int fnmatch (const char *pattern, const char *string, int flags) { -# if HANDLE_MULTIBYTE - if (__builtin_expect (MB_CUR_MAX, 1) != 1) + if (__glibc_unlikely (MB_CUR_MAX != 1)) { mbstate_t ps; size_t n; @@ -338,126 +249,113 @@ fnmatch (const char *pattern, const char *string, int flags) /* Convert the strings into wide characters. */ memset (&ps, '\0', sizeof (ps)); p = pattern; -#ifdef _LIBC - n = __strnlen (pattern, 1024); -#else - n = strlen (pattern); -#endif + n = strnlen (pattern, 1024); if (__glibc_likely (n < 1024)) - { - wpattern = (wchar_t *) alloca_account ((n + 1) * sizeof (wchar_t), - alloca_used); - n = mbsrtowcs (wpattern, &p, n + 1, &ps); - if (__glibc_unlikely (n == (size_t) -1)) - /* Something wrong. - XXX Do we have to set `errno' to something which mbsrtows hasn't - already done? */ - return -1; - if (p) - { - memset (&ps, '\0', sizeof (ps)); - goto prepare_wpattern; - } - } + { + wpattern = (wchar_t *) alloca_account ((n + 1) * sizeof (wchar_t), + alloca_used); + n = mbsrtowcs (wpattern, &p, n + 1, &ps); + if (__glibc_unlikely (n == (size_t) -1)) + /* Something wrong. + XXX Do we have to set 'errno' to something which mbsrtows hasn't + already done? */ + return -1; + if (p) + { + memset (&ps, '\0', sizeof (ps)); + goto prepare_wpattern; + } + } else - { - prepare_wpattern: - n = mbsrtowcs (NULL, &pattern, 0, &ps); - if (__glibc_unlikely (n == (size_t) -1)) - /* Something wrong. - XXX Do we have to set `errno' to something which mbsrtows hasn't - already done? */ - return -1; - if (__glibc_unlikely (n >= (size_t) -1 / sizeof (wchar_t))) - { - __set_errno (ENOMEM); - return -2; - } - wpattern_malloc = wpattern - = (wchar_t *) malloc ((n + 1) * sizeof (wchar_t)); - assert (mbsinit (&ps)); - if (wpattern == NULL) - return -2; - (void) mbsrtowcs (wpattern, &pattern, n + 1, &ps); - } + { + prepare_wpattern: + n = mbsrtowcs (NULL, &pattern, 0, &ps); + if (__glibc_unlikely (n == (size_t) -1)) + /* Something wrong. + XXX Do we have to set 'errno' to something which mbsrtows hasn't + already done? */ + return -1; + if (__glibc_unlikely (n >= (size_t) -1 / sizeof (wchar_t))) + { + __set_errno (ENOMEM); + return -2; + } + wpattern_malloc = wpattern + = (wchar_t *) malloc ((n + 1) * sizeof (wchar_t)); + assert (mbsinit (&ps)); + if (wpattern == NULL) + return -2; + (void) mbsrtowcs (wpattern, &pattern, n + 1, &ps); + } assert (mbsinit (&ps)); -#ifdef _LIBC - n = __strnlen (string, 1024); -#else - n = strlen (string); -#endif + n = strnlen (string, 1024); p = string; if (__glibc_likely (n < 1024)) - { - wstring = (wchar_t *) alloca_account ((n + 1) * sizeof (wchar_t), - alloca_used); - n = mbsrtowcs (wstring, &p, n + 1, &ps); - if (__glibc_unlikely (n == (size_t) -1)) - { - /* Something wrong. - XXX Do we have to set `errno' to something which - mbsrtows hasn't already done? */ - free_return: - free (wpattern_malloc); - return -1; - } - if (p) - { - memset (&ps, '\0', sizeof (ps)); - goto prepare_wstring; - } - } + { + wstring = (wchar_t *) alloca_account ((n + 1) * sizeof (wchar_t), + alloca_used); + n = mbsrtowcs (wstring, &p, n + 1, &ps); + if (__glibc_unlikely (n == (size_t) -1)) + { + /* Something wrong. + XXX Do we have to set 'errno' to something which + mbsrtows hasn't already done? */ + free_return: + free (wpattern_malloc); + return -1; + } + if (p) + { + memset (&ps, '\0', sizeof (ps)); + goto prepare_wstring; + } + } else - { - prepare_wstring: - n = mbsrtowcs (NULL, &string, 0, &ps); - if (__glibc_unlikely (n == (size_t) -1)) - /* Something wrong. - XXX Do we have to set `errno' to something which mbsrtows hasn't - already done? */ - goto free_return; - if (__glibc_unlikely (n >= (size_t) -1 / sizeof (wchar_t))) - { - free (wpattern_malloc); - __set_errno (ENOMEM); - return -2; - } - - wstring_malloc = wstring - = (wchar_t *) malloc ((n + 1) * sizeof (wchar_t)); - if (wstring == NULL) - { - free (wpattern_malloc); - return -2; - } - assert (mbsinit (&ps)); - (void) mbsrtowcs (wstring, &string, n + 1, &ps); - } + { + prepare_wstring: + n = mbsrtowcs (NULL, &string, 0, &ps); + if (__glibc_unlikely (n == (size_t) -1)) + /* Something wrong. + XXX Do we have to set 'errno' to something which mbsrtows hasn't + already done? */ + goto free_return; + if (__glibc_unlikely (n >= (size_t) -1 / sizeof (wchar_t))) + { + free (wpattern_malloc); + __set_errno (ENOMEM); + return -2; + } + + wstring_malloc = wstring + = (wchar_t *) malloc ((n + 1) * sizeof (wchar_t)); + if (wstring == NULL) + { + free (wpattern_malloc); + return -2; + } + assert (mbsinit (&ps)); + (void) mbsrtowcs (wstring, &string, n + 1, &ps); + } int res = internal_fnwmatch (wpattern, wstring, wstring + n, - flags & FNM_PERIOD, flags, NULL, - alloca_used); + flags & FNM_PERIOD, flags, NULL, + alloca_used); free (wstring_malloc); free (wpattern_malloc); return res; } -# endif /* mbstate_t and mbsrtowcs or _LIBC. */ return internal_fnmatch (pattern, string, string + strlen (string), - flags & FNM_PERIOD, flags, NULL, 0); + flags & FNM_PERIOD, flags, NULL, 0); } -# ifdef _LIBC -# undef fnmatch +#undef fnmatch versioned_symbol (libc, __fnmatch, fnmatch, GLIBC_2_2_3); -# if SHLIB_COMPAT(libc, GLIBC_2_0, GLIBC_2_2_3) +#if SHLIB_COMPAT(libc, GLIBC_2_0, GLIBC_2_2_3) strong_alias (__fnmatch, __fnmatch_old) compat_symbol (libc, __fnmatch_old, fnmatch, GLIBC_2_0); -# endif +#endif libc_hidden_ver (__fnmatch, fnmatch) -# endif - -#endif /* _LIBC or not __GNU_LIBRARY__. */ diff --git a/posix/fnmatch_loop.c b/posix/fnmatch_loop.c index 0f890d4782..2804eae6b1 100644 --- a/posix/fnmatch_loop.c +++ b/posix/fnmatch_loop.c @@ -15,28 +15,30 @@ License along with the GNU C Library; if not, see . */ -#include +#ifdef _LIBC +# include +#endif struct STRUCT { const CHAR *pattern; const CHAR *string; - int no_leading_period; + bool no_leading_period; }; -/* Match STRING against the filename pattern PATTERN, returning zero if +/* Match STRING against the file name pattern PATTERN, returning zero if it matches, nonzero if not. */ static int FCT (const CHAR *pattern, const CHAR *string, - const CHAR *string_end, int no_leading_period, int flags, - struct STRUCT *ends, size_t alloca_used); + const CHAR *string_end, bool no_leading_period, int flags, + struct STRUCT *ends, size_t alloca_used); static int EXT (INT opt, const CHAR *pattern, const CHAR *string, - const CHAR *string_end, int no_leading_period, int flags, - size_t alloca_used); + const CHAR *string_end, bool no_leading_period, int flags, + size_t alloca_used); static const CHAR *END (const CHAR *patternp); static int FCT (const CHAR *pattern, const CHAR *string, const CHAR *string_end, - int no_leading_period, int flags, struct STRUCT *ends, size_t alloca_used) + bool no_leading_period, int flags, struct STRUCT *ends, size_t alloca_used) { const CHAR *p = pattern, *n = string; UCHAR c; @@ -50,882 +52,868 @@ FCT (const CHAR *pattern, const CHAR *string, const CHAR *string_end, # endif #endif - while ((c = *p++) != L('\0')) + while ((c = *p++) != L_('\0')) { - int new_no_leading_period = 0; + bool new_no_leading_period = false; c = FOLD (c); switch (c) - { - case L('?'): - if (__builtin_expect (flags & FNM_EXTMATCH, 0) && *p == '(') - { - int res = EXT (c, p, n, string_end, no_leading_period, - flags, alloca_used); - if (res != -1) - return res; - } - - if (n == string_end) - return FNM_NOMATCH; - else if (*n == L('/') && (flags & FNM_FILE_NAME)) - return FNM_NOMATCH; - else if (*n == L('.') && no_leading_period) - return FNM_NOMATCH; - break; - - case L('\\'): - if (!(flags & FNM_NOESCAPE)) - { - c = *p++; - if (c == L('\0')) - /* Trailing \ loses. */ - return FNM_NOMATCH; - c = FOLD (c); - } - if (n == string_end || FOLD ((UCHAR) *n) != c) - return FNM_NOMATCH; - break; - - case L('*'): - if (__builtin_expect (flags & FNM_EXTMATCH, 0) && *p == '(') - { - int res = EXT (c, p, n, string_end, no_leading_period, - flags, alloca_used); - if (res != -1) - return res; - } - else if (ends != NULL) - { - ends->pattern = p - 1; - ends->string = n; - ends->no_leading_period = no_leading_period; - return 0; - } - - if (n != string_end && *n == L('.') && no_leading_period) - return FNM_NOMATCH; - - for (c = *p++; c == L('?') || c == L('*'); c = *p++) - { - if (*p == L('(') && (flags & FNM_EXTMATCH) != 0) - { - const CHAR *endp = END (p); - if (endp != p) - { - /* This is a pattern. Skip over it. */ - p = endp; - continue; - } - } - - if (c == L('?')) - { - /* A ? needs to match one character. */ - if (n == string_end) - /* There isn't another character; no match. */ - return FNM_NOMATCH; - else if (*n == L('/') - && __builtin_expect (flags & FNM_FILE_NAME, 0)) - /* A slash does not match a wildcard under - FNM_FILE_NAME. */ - return FNM_NOMATCH; - else - /* One character of the string is consumed in matching - this ? wildcard, so *??? won't match if there are - less than three characters. */ - ++n; - } - } - - if (c == L('\0')) - /* The wildcard(s) is/are the last element of the pattern. - If the name is a file name and contains another slash - this means it cannot match, unless the FNM_LEADING_DIR - flag is set. */ - { - int result = (flags & FNM_FILE_NAME) == 0 ? 0 : FNM_NOMATCH; - - if (flags & FNM_FILE_NAME) - { - if (flags & FNM_LEADING_DIR) - result = 0; - else - { - if (MEMCHR (n, L('/'), string_end - n) == NULL) - result = 0; - } - } - - return result; - } - else - { - const CHAR *endp; - struct STRUCT end; - - end.pattern = NULL; - endp = MEMCHR (n, (flags & FNM_FILE_NAME) ? L('/') : L('\0'), - string_end - n); - if (endp == NULL) - endp = string_end; - - if (c == L('[') - || (__builtin_expect (flags & FNM_EXTMATCH, 0) != 0 - && (c == L('@') || c == L('+') || c == L('!')) - && *p == L('('))) - { - int flags2 = ((flags & FNM_FILE_NAME) - ? flags : (flags & ~FNM_PERIOD)); - - for (--p; n < endp; ++n, no_leading_period = 0) - if (FCT (p, n, string_end, no_leading_period, flags2, - &end, alloca_used) == 0) - goto found; - } - else if (c == L('/') && (flags & FNM_FILE_NAME)) - { - while (n < string_end && *n != L('/')) - ++n; - if (n < string_end && *n == L('/') - && (FCT (p, n + 1, string_end, flags & FNM_PERIOD, flags, - NULL, alloca_used) == 0)) - return 0; - } - else - { - int flags2 = ((flags & FNM_FILE_NAME) - ? flags : (flags & ~FNM_PERIOD)); - - if (c == L('\\') && !(flags & FNM_NOESCAPE)) - c = *p; - c = FOLD (c); - for (--p; n < endp; ++n, no_leading_period = 0) - if (FOLD ((UCHAR) *n) == c - && (FCT (p, n, string_end, no_leading_period, flags2, - &end, alloca_used) == 0)) - { - found: - if (end.pattern == NULL) - return 0; - break; - } - if (end.pattern != NULL) - { - p = end.pattern; - n = end.string; - no_leading_period = end.no_leading_period; - continue; - } - } - } - - /* If we come here no match is possible with the wildcard. */ - return FNM_NOMATCH; - - case L('['): - { - /* Nonzero if the sense of the character class is inverted. */ - const CHAR *p_init = p; - const CHAR *n_init = n; - int not; - CHAR cold; - UCHAR fn; - - if (posixly_correct == 0) - posixly_correct = getenv ("POSIXLY_CORRECT") != NULL ? 1 : -1; - - if (n == string_end) - return FNM_NOMATCH; - - if (*n == L('.') && no_leading_period) - return FNM_NOMATCH; - - if (*n == L('/') && (flags & FNM_FILE_NAME)) - /* `/' cannot be matched. */ - return FNM_NOMATCH; - - not = (*p == L('!') || (posixly_correct < 0 && *p == L('^'))); - if (not) - ++p; - - fn = FOLD ((UCHAR) *n); - - c = *p++; - for (;;) - { - if (!(flags & FNM_NOESCAPE) && c == L('\\')) - { - if (*p == L('\0')) - return FNM_NOMATCH; - c = FOLD ((UCHAR) *p); - ++p; - - goto normal_bracket; - } - else if (c == L('[') && *p == L(':')) - { - /* Leave room for the null. */ - CHAR str[CHAR_CLASS_MAX_LENGTH + 1]; - size_t c1 = 0; -#if defined _LIBC || (defined HAVE_WCTYPE_H && defined HAVE_WCHAR_H) - wctype_t wt; -#endif - const CHAR *startp = p; - - for (;;) - { - if (c1 == CHAR_CLASS_MAX_LENGTH) - /* The name is too long and therefore the pattern - is ill-formed. */ - return FNM_NOMATCH; - - c = *++p; - if (c == L(':') && p[1] == L(']')) - { - p += 2; - break; - } - if (c < L('a') || c >= L('z')) - { - /* This cannot possibly be a character class name. - Match it as a normal range. */ - p = startp; - c = L('['); - goto normal_bracket; - } - str[c1++] = c; - } - str[c1] = L('\0'); - -#if defined _LIBC || (defined HAVE_WCTYPE_H && defined HAVE_WCHAR_H) - wt = IS_CHAR_CLASS (str); - if (wt == 0) - /* Invalid character class name. */ - return FNM_NOMATCH; - -# if defined _LIBC && ! WIDE_CHAR_VERSION - /* The following code is glibc specific but does - there a good job in speeding up the code since - we can avoid the btowc() call. */ - if (_ISCTYPE ((UCHAR) *n, wt)) - goto matched; -# else - if (ISWCTYPE (BTOWC ((UCHAR) *n), wt)) - goto matched; -# endif + { + case L_('?'): + if (__glibc_unlikely (flags & FNM_EXTMATCH) && *p == '(') + { + int res = EXT (c, p, n, string_end, no_leading_period, + flags, alloca_used); + if (res != -1) + return res; + } + + if (n == string_end) + return FNM_NOMATCH; + else if (*n == L_('/') && (flags & FNM_FILE_NAME)) + return FNM_NOMATCH; + else if (*n == L_('.') && no_leading_period) + return FNM_NOMATCH; + break; + + case L_('\\'): + if (!(flags & FNM_NOESCAPE)) + { + c = *p++; + if (c == L_('\0')) + /* Trailing \ loses. */ + return FNM_NOMATCH; + c = FOLD (c); + } + if (n == string_end || FOLD ((UCHAR) *n) != c) + return FNM_NOMATCH; + break; + + case L_('*'): + if (__glibc_unlikely (flags & FNM_EXTMATCH) && *p == '(') + { + int res = EXT (c, p, n, string_end, no_leading_period, + flags, alloca_used); + if (res != -1) + return res; + } + else if (ends != NULL) + { + ends->pattern = p - 1; + ends->string = n; + ends->no_leading_period = no_leading_period; + return 0; + } + + if (n != string_end && *n == L_('.') && no_leading_period) + return FNM_NOMATCH; + + for (c = *p++; c == L_('?') || c == L_('*'); c = *p++) + { + if (*p == L_('(') && (flags & FNM_EXTMATCH) != 0) + { + const CHAR *endp = END (p); + if (endp != p) + { + /* This is a pattern. Skip over it. */ + p = endp; + continue; + } + } + + if (c == L_('?')) + { + /* A ? needs to match one character. */ + if (n == string_end) + /* There isn't another character; no match. */ + return FNM_NOMATCH; + else if (*n == L_('/') + && __glibc_unlikely (flags & FNM_FILE_NAME)) + /* A slash does not match a wildcard under + FNM_FILE_NAME. */ + return FNM_NOMATCH; + else + /* One character of the string is consumed in matching + this ? wildcard, so *??? won't match if there are + less than three characters. */ + ++n; + } + } + + if (c == L_('\0')) + /* The wildcard(s) is/are the last element of the pattern. + If the name is a file name and contains another slash + this means it cannot match, unless the FNM_LEADING_DIR + flag is set. */ + { + int result = (flags & FNM_FILE_NAME) == 0 ? 0 : FNM_NOMATCH; + + if (flags & FNM_FILE_NAME) + { + if (flags & FNM_LEADING_DIR) + result = 0; + else + { + if (MEMCHR (n, L_('/'), string_end - n) == NULL) + result = 0; + } + } + + return result; + } + else + { + const CHAR *endp; + struct STRUCT end; + + end.pattern = NULL; + endp = MEMCHR (n, (flags & FNM_FILE_NAME) ? L_('/') : L_('\0'), + string_end - n); + if (endp == NULL) + endp = string_end; + + if (c == L_('[') + || (__glibc_unlikely (flags & FNM_EXTMATCH) + && (c == L_('@') || c == L_('+') || c == L_('!')) + && *p == L_('('))) + { + int flags2 = ((flags & FNM_FILE_NAME) + ? flags : (flags & ~FNM_PERIOD)); + + for (--p; n < endp; ++n, no_leading_period = false) + if (FCT (p, n, string_end, no_leading_period, flags2, + &end, alloca_used) == 0) + goto found; + } + else if (c == L_('/') && (flags & FNM_FILE_NAME)) + { + while (n < string_end && *n != L_('/')) + ++n; + if (n < string_end && *n == L_('/') + && (FCT (p, n + 1, string_end, flags & FNM_PERIOD, flags, + NULL, alloca_used) == 0)) + return 0; + } + else + { + int flags2 = ((flags & FNM_FILE_NAME) + ? flags : (flags & ~FNM_PERIOD)); + + if (c == L_('\\') && !(flags & FNM_NOESCAPE)) + c = *p; + c = FOLD (c); + for (--p; n < endp; ++n, no_leading_period = false) + if (FOLD ((UCHAR) *n) == c + && (FCT (p, n, string_end, no_leading_period, flags2, + &end, alloca_used) == 0)) + { + found: + if (end.pattern == NULL) + return 0; + break; + } + if (end.pattern != NULL) + { + p = end.pattern; + n = end.string; + no_leading_period = end.no_leading_period; + continue; + } + } + } + + /* If we come here no match is possible with the wildcard. */ + return FNM_NOMATCH; + + case L_('['): + { + /* Nonzero if the sense of the character class is inverted. */ + const CHAR *p_init = p; + const CHAR *n_init = n; + bool not; + CHAR cold; + UCHAR fn; + + if (posixly_correct == 0) + posixly_correct = getenv ("POSIXLY_CORRECT") != NULL ? 1 : -1; + + if (n == string_end) + return FNM_NOMATCH; + + if (*n == L_('.') && no_leading_period) + return FNM_NOMATCH; + + if (*n == L_('/') && (flags & FNM_FILE_NAME)) + /* '/' cannot be matched. */ + return FNM_NOMATCH; + + not = (*p == L_('!') || (posixly_correct < 0 && *p == L_('^'))); + if (not) + ++p; + + fn = FOLD ((UCHAR) *n); + + c = *p++; + for (;;) + { + if (!(flags & FNM_NOESCAPE) && c == L_('\\')) + { + if (*p == L_('\0')) + return FNM_NOMATCH; + c = FOLD ((UCHAR) *p); + ++p; + + goto normal_bracket; + } + else if (c == L_('[') && *p == L_(':')) + { + /* Leave room for the null. */ + CHAR str[CHAR_CLASS_MAX_LENGTH + 1]; + size_t c1 = 0; + wctype_t wt; + const CHAR *startp = p; + + for (;;) + { + if (c1 == CHAR_CLASS_MAX_LENGTH) + /* The name is too long and therefore the pattern + is ill-formed. */ + return FNM_NOMATCH; + + c = *++p; + if (c == L_(':') && p[1] == L_(']')) + { + p += 2; + break; + } + if (c < L_('a') || c >= L_('z')) + { + /* This cannot possibly be a character class name. + Match it as a normal range. */ + p = startp; + c = L_('['); + goto normal_bracket; + } + str[c1++] = c; + } + str[c1] = L_('\0'); + + wt = IS_CHAR_CLASS (str); + if (wt == 0) + /* Invalid character class name. */ + return FNM_NOMATCH; + +#if defined _LIBC && ! WIDE_CHAR_VERSION + /* The following code is glibc specific but does + there a good job in speeding up the code since + we can avoid the btowc() call. */ + if (_ISCTYPE ((UCHAR) *n, wt)) + goto matched; #else - if ((STREQ (str, L("alnum")) && ISALNUM ((UCHAR) *n)) - || (STREQ (str, L("alpha")) && ISALPHA ((UCHAR) *n)) - || (STREQ (str, L("blank")) && ISBLANK ((UCHAR) *n)) - || (STREQ (str, L("cntrl")) && ISCNTRL ((UCHAR) *n)) - || (STREQ (str, L("digit")) && ISDIGIT ((UCHAR) *n)) - || (STREQ (str, L("graph")) && ISGRAPH ((UCHAR) *n)) - || (STREQ (str, L("lower")) && ISLOWER ((UCHAR) *n)) - || (STREQ (str, L("print")) && ISPRINT ((UCHAR) *n)) - || (STREQ (str, L("punct")) && ISPUNCT ((UCHAR) *n)) - || (STREQ (str, L("space")) && ISSPACE ((UCHAR) *n)) - || (STREQ (str, L("upper")) && ISUPPER ((UCHAR) *n)) - || (STREQ (str, L("xdigit")) && ISXDIGIT ((UCHAR) *n))) - goto matched; + if (iswctype (BTOWC ((UCHAR) *n), wt)) + goto matched; #endif - c = *p++; - } + c = *p++; + } #ifdef _LIBC - else if (c == L('[') && *p == L('=')) - { - /* It's important that STR be a scalar variable rather - than a one-element array, because GCC (at least 4.9.2 - -O2 on x86-64) can be confused by the array and - diagnose a "used initialized" in a dead branch in the - findidx function. */ - UCHAR str; - uint32_t nrules = - _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES); - const CHAR *startp = p; - - c = *++p; - if (c == L('\0')) - { - p = startp; - c = L('['); - goto normal_bracket; - } - str = c; - - c = *++p; - if (c != L('=') || p[1] != L(']')) - { - p = startp; - c = L('['); - goto normal_bracket; - } - p += 2; - - if (nrules == 0) - { - if ((UCHAR) *n == str) - goto matched; - } - else - { - const int32_t *table; + else if (c == L_('[') && *p == L_('=')) + { + /* It's important that STR be a scalar variable rather + than a one-element array, because GCC (at least 4.9.2 + -O2 on x86-64) can be confused by the array and + diagnose a "used initialized" in a dead branch in the + findidx function. */ + UCHAR str; + uint32_t nrules = + _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES); + const CHAR *startp = p; + + c = *++p; + if (c == L_('\0')) + { + p = startp; + c = L_('['); + goto normal_bracket; + } + str = c; + + c = *++p; + if (c != L_('=') || p[1] != L_(']')) + { + p = startp; + c = L_('['); + goto normal_bracket; + } + p += 2; + + if (nrules == 0) + { + if ((UCHAR) *n == str) + goto matched; + } + else + { + const int32_t *table; # if WIDE_CHAR_VERSION - const int32_t *weights; - const wint_t *extra; + const int32_t *weights; + const wint_t *extra; # else - const unsigned char *weights; - const unsigned char *extra; + const unsigned char *weights; + const unsigned char *extra; # endif - const int32_t *indirect; - int32_t idx; - const UCHAR *cp = (const UCHAR *) &str; + const int32_t *indirect; + int32_t idx; + const UCHAR *cp = (const UCHAR *) &str; # if WIDE_CHAR_VERSION - table = (const int32_t *) - _NL_CURRENT (LC_COLLATE, _NL_COLLATE_TABLEWC); - weights = (const int32_t *) - _NL_CURRENT (LC_COLLATE, _NL_COLLATE_WEIGHTWC); - extra = (const wint_t *) - _NL_CURRENT (LC_COLLATE, _NL_COLLATE_EXTRAWC); - indirect = (const int32_t *) - _NL_CURRENT (LC_COLLATE, _NL_COLLATE_INDIRECTWC); + table = (const int32_t *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_TABLEWC); + weights = (const int32_t *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_WEIGHTWC); + extra = (const wint_t *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_EXTRAWC); + indirect = (const int32_t *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_INDIRECTWC); # else - table = (const int32_t *) - _NL_CURRENT (LC_COLLATE, _NL_COLLATE_TABLEMB); - weights = (const unsigned char *) - _NL_CURRENT (LC_COLLATE, _NL_COLLATE_WEIGHTMB); - extra = (const unsigned char *) - _NL_CURRENT (LC_COLLATE, _NL_COLLATE_EXTRAMB); - indirect = (const int32_t *) - _NL_CURRENT (LC_COLLATE, _NL_COLLATE_INDIRECTMB); + table = (const int32_t *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_TABLEMB); + weights = (const unsigned char *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_WEIGHTMB); + extra = (const unsigned char *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_EXTRAMB); + indirect = (const int32_t *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_INDIRECTMB); # endif - idx = FINDIDX (table, indirect, extra, &cp, 1); - if (idx != 0) - { - /* We found a table entry. Now see whether the - character we are currently at has the same - equivalance class value. */ - int len = weights[idx & 0xffffff]; - int32_t idx2; - const UCHAR *np = (const UCHAR *) n; - - idx2 = FINDIDX (table, indirect, extra, - &np, string_end - n); - if (idx2 != 0 - && (idx >> 24) == (idx2 >> 24) - && len == weights[idx2 & 0xffffff]) - { - int cnt = 0; - - idx &= 0xffffff; - idx2 &= 0xffffff; - - while (cnt < len - && (weights[idx + 1 + cnt] - == weights[idx2 + 1 + cnt])) - ++cnt; - - if (cnt == len) - goto matched; - } - } - } - - c = *p++; - } + idx = FINDIDX (table, indirect, extra, &cp, 1); + if (idx != 0) + { + /* We found a table entry. Now see whether the + character we are currently at has the same + equivalence class value. */ + int len = weights[idx & 0xffffff]; + int32_t idx2; + const UCHAR *np = (const UCHAR *) n; + + idx2 = FINDIDX (table, indirect, extra, + &np, string_end - n); + if (idx2 != 0 + && (idx >> 24) == (idx2 >> 24) + && len == weights[idx2 & 0xffffff]) + { + int cnt = 0; + + idx &= 0xffffff; + idx2 &= 0xffffff; + + while (cnt < len + && (weights[idx + 1 + cnt] + == weights[idx2 + 1 + cnt])) + ++cnt; + + if (cnt == len) + goto matched; + } + } + } + + c = *p++; + } #endif - else if (c == L('\0')) - { - /* [ unterminated, treat as normal character. */ - p = p_init; - n = n_init; - c = L('['); - goto normal_match; - } - else - { - int is_range = 0; + else if (c == L_('\0')) + { + /* [ unterminated, treat as normal character. */ + p = p_init; + n = n_init; + c = L_('['); + goto normal_match; + } + else + { + bool is_range = false; #ifdef _LIBC - int is_seqval = 0; - - if (c == L('[') && *p == L('.')) - { - uint32_t nrules = - _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES); - const CHAR *startp = p; - size_t c1 = 0; - - while (1) - { - c = *++p; - if (c == L('.') && p[1] == L(']')) - { - p += 2; - break; - } - if (c == '\0') - return FNM_NOMATCH; - ++c1; - } - - /* We have to handling the symbols differently in - ranges since then the collation sequence is - important. */ - is_range = *p == L('-') && p[1] != L('\0'); - - if (nrules == 0) - { - /* There are no names defined in the collation - data. Therefore we only accept the trivial - names consisting of the character itself. */ - if (c1 != 1) - return FNM_NOMATCH; - - if (!is_range && *n == startp[1]) - goto matched; - - cold = startp[1]; - c = *p++; - } - else - { - int32_t table_size; - const int32_t *symb_table; - const unsigned char *extra; - int32_t idx; - int32_t elem; + bool is_seqval = false; + + if (c == L_('[') && *p == L_('.')) + { + uint32_t nrules = + _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES); + const CHAR *startp = p; + size_t c1 = 0; + + while (1) + { + c = *++p; + if (c == L_('.') && p[1] == L_(']')) + { + p += 2; + break; + } + if (c == '\0') + return FNM_NOMATCH; + ++c1; + } + + /* We have to handling the symbols differently in + ranges since then the collation sequence is + important. */ + is_range = *p == L_('-') && p[1] != L_('\0'); + + if (nrules == 0) + { + /* There are no names defined in the collation + data. Therefore we only accept the trivial + names consisting of the character itself. */ + if (c1 != 1) + return FNM_NOMATCH; + + if (!is_range && *n == startp[1]) + goto matched; + + cold = startp[1]; + c = *p++; + } + else + { + int32_t table_size; + const int32_t *symb_table; + const unsigned char *extra; + int32_t idx; + int32_t elem; # if WIDE_CHAR_VERSION - CHAR *wextra; + CHAR *wextra; # endif - table_size = - _NL_CURRENT_WORD (LC_COLLATE, - _NL_COLLATE_SYMB_HASH_SIZEMB); - symb_table = (const int32_t *) - _NL_CURRENT (LC_COLLATE, - _NL_COLLATE_SYMB_TABLEMB); - extra = (const unsigned char *) - _NL_CURRENT (LC_COLLATE, - _NL_COLLATE_SYMB_EXTRAMB); - - for (elem = 0; elem < table_size; elem++) - if (symb_table[2 * elem] != 0) - { - idx = symb_table[2 * elem + 1]; - /* Skip the name of collating element. */ - idx += 1 + extra[idx]; + table_size = + _NL_CURRENT_WORD (LC_COLLATE, + _NL_COLLATE_SYMB_HASH_SIZEMB); + symb_table = (const int32_t *) + _NL_CURRENT (LC_COLLATE, + _NL_COLLATE_SYMB_TABLEMB); + extra = (const unsigned char *) + _NL_CURRENT (LC_COLLATE, + _NL_COLLATE_SYMB_EXTRAMB); + + for (elem = 0; elem < table_size; elem++) + if (symb_table[2 * elem] != 0) + { + idx = symb_table[2 * elem + 1]; + /* Skip the name of collating element. */ + idx += 1 + extra[idx]; # if WIDE_CHAR_VERSION - /* Skip the byte sequence of the - collating element. */ - idx += 1 + extra[idx]; - /* Adjust for the alignment. */ - idx = (idx + 3) & ~3; - - wextra = (CHAR *) &extra[idx + 4]; - - if (/* Compare the length of the sequence. */ - c1 == wextra[0] - /* Compare the wide char sequence. */ - && WMEMCMP (startp + 1, &wextra[1], - c1) == 0) - /* Yep, this is the entry. */ - break; + /* Skip the byte sequence of the + collating element. */ + idx += 1 + extra[idx]; + /* Adjust for the alignment. */ + idx = (idx + 3) & ~3; + + wextra = (CHAR *) &extra[idx + 4]; + + if (/* Compare the length of the sequence. */ + c1 == wextra[0] + /* Compare the wide char sequence. */ + && (__wmemcmp (startp + 1, &wextra[1], + c1) + == 0)) + /* Yep, this is the entry. */ + break; # else - if (/* Compare the length of the sequence. */ - c1 == extra[idx] - /* Compare the byte sequence. */ - && memcmp (startp + 1, - &extra[idx + 1], c1) == 0) - /* Yep, this is the entry. */ - break; + if (/* Compare the length of the sequence. */ + c1 == extra[idx] + /* Compare the byte sequence. */ + && memcmp (startp + 1, + &extra[idx + 1], c1) == 0) + /* Yep, this is the entry. */ + break; # endif - } + } - if (elem < table_size) - { - /* Compare the byte sequence but only if - this is not part of a range. */ - if (! is_range + if (elem < table_size) + { + /* Compare the byte sequence but only if + this is not part of a range. */ + if (! is_range # if WIDE_CHAR_VERSION - && WMEMCMP (n, &wextra[1], c1) == 0 + && __wmemcmp (n, &wextra[1], c1) == 0 # else - && memcmp (n, &extra[idx + 1], c1) == 0 + && memcmp (n, &extra[idx + 1], c1) == 0 # endif - ) - { - n += c1 - 1; - goto matched; - } - - /* Get the collation sequence value. */ - is_seqval = 1; + ) + { + n += c1 - 1; + goto matched; + } + + /* Get the collation sequence value. */ + is_seqval = true; # if WIDE_CHAR_VERSION - cold = wextra[1 + wextra[0]]; + cold = wextra[1 + wextra[0]]; # else - idx += 1 + extra[idx]; - /* Adjust for the alignment. */ - idx = (idx + 3) & ~3; - cold = *((int32_t *) &extra[idx]); + idx += 1 + extra[idx]; + /* Adjust for the alignment. */ + idx = (idx + 3) & ~3; + cold = *((int32_t *) &extra[idx]); # endif - c = *p++; - } - else if (c1 == 1) - { - /* No valid character. Match it as a - single byte. */ - if (!is_range && *n == startp[1]) - goto matched; - - cold = startp[1]; - c = *p++; - } - else - return FNM_NOMATCH; - } - } - else + c = *p++; + } + else if (c1 == 1) + { + /* No valid character. Match it as a + single byte. */ + if (!is_range && *n == startp[1]) + goto matched; + + cold = startp[1]; + c = *p++; + } + else + return FNM_NOMATCH; + } + } + else #endif - { - c = FOLD (c); - normal_bracket: - - /* We have to handling the symbols differently in - ranges since then the collation sequence is - important. */ - is_range = (*p == L('-') && p[1] != L('\0') - && p[1] != L(']')); - - if (!is_range && c == fn) - goto matched; - - /* This is needed if we goto normal_bracket; from - outside of is_seqval's scope. */ - is_seqval = 0; - cold = c; - c = *p++; - } - - if (c == L('-') && *p != L(']')) - { + { + c = FOLD (c); + normal_bracket: + + /* We have to handling the symbols differently in + ranges since then the collation sequence is + important. */ + is_range = (*p == L_('-') && p[1] != L_('\0') + && p[1] != L_(']')); + + if (!is_range && c == fn) + goto matched; + +#if _LIBC + /* This is needed if we goto normal_bracket; from + outside of is_seqval's scope. */ + is_seqval = false; +#endif + cold = c; + c = *p++; + } + + if (c == L_('-') && *p != L_(']')) + { #if _LIBC - /* We have to find the collation sequence - value for C. Collation sequence is nothing - we can regularly access. The sequence - value is defined by the order in which the - definitions of the collation values for the - various characters appear in the source - file. A strange concept, nowhere - documented. */ - uint32_t fcollseq; - uint32_t lcollseq; - UCHAR cend = *p++; + /* We have to find the collation sequence + value for C. Collation sequence is nothing + we can regularly access. The sequence + value is defined by the order in which the + definitions of the collation values for the + various characters appear in the source + file. A strange concept, nowhere + documented. */ + uint32_t fcollseq; + uint32_t lcollseq; + UCHAR cend = *p++; # if WIDE_CHAR_VERSION - /* Search in the `names' array for the characters. */ - fcollseq = __collseq_table_lookup (collseq, fn); - if (fcollseq == ~((uint32_t) 0)) - /* XXX We don't know anything about the character - we are supposed to match. This means we are - failing. */ - goto range_not_matched; - - if (is_seqval) - lcollseq = cold; - else - lcollseq = __collseq_table_lookup (collseq, cold); + /* Search in the 'names' array for the characters. */ + fcollseq = __collseq_table_lookup (collseq, fn); + if (fcollseq == ~((uint32_t) 0)) + /* XXX We don't know anything about the character + we are supposed to match. This means we are + failing. */ + goto range_not_matched; + + if (is_seqval) + lcollseq = cold; + else + lcollseq = __collseq_table_lookup (collseq, cold); # else - fcollseq = collseq[fn]; - lcollseq = is_seqval ? cold : collseq[(UCHAR) cold]; + fcollseq = collseq[fn]; + lcollseq = is_seqval ? cold : collseq[(UCHAR) cold]; # endif - is_seqval = 0; - if (cend == L('[') && *p == L('.')) - { - uint32_t nrules = - _NL_CURRENT_WORD (LC_COLLATE, - _NL_COLLATE_NRULES); - const CHAR *startp = p; - size_t c1 = 0; - - while (1) - { - c = *++p; - if (c == L('.') && p[1] == L(']')) - { - p += 2; - break; - } - if (c == '\0') - return FNM_NOMATCH; - ++c1; - } - - if (nrules == 0) - { - /* There are no names defined in the - collation data. Therefore we only - accept the trivial names consisting - of the character itself. */ - if (c1 != 1) - return FNM_NOMATCH; - - cend = startp[1]; - } - else - { - int32_t table_size; - const int32_t *symb_table; - const unsigned char *extra; - int32_t idx; - int32_t elem; + is_seqval = false; + if (cend == L_('[') && *p == L_('.')) + { + uint32_t nrules = + _NL_CURRENT_WORD (LC_COLLATE, + _NL_COLLATE_NRULES); + const CHAR *startp = p; + size_t c1 = 0; + + while (1) + { + c = *++p; + if (c == L_('.') && p[1] == L_(']')) + { + p += 2; + break; + } + if (c == '\0') + return FNM_NOMATCH; + ++c1; + } + + if (nrules == 0) + { + /* There are no names defined in the + collation data. Therefore we only + accept the trivial names consisting + of the character itself. */ + if (c1 != 1) + return FNM_NOMATCH; + + cend = startp[1]; + } + else + { + int32_t table_size; + const int32_t *symb_table; + const unsigned char *extra; + int32_t idx; + int32_t elem; # if WIDE_CHAR_VERSION - CHAR *wextra; + CHAR *wextra; # endif - table_size = - _NL_CURRENT_WORD (LC_COLLATE, - _NL_COLLATE_SYMB_HASH_SIZEMB); - symb_table = (const int32_t *) - _NL_CURRENT (LC_COLLATE, - _NL_COLLATE_SYMB_TABLEMB); - extra = (const unsigned char *) - _NL_CURRENT (LC_COLLATE, - _NL_COLLATE_SYMB_EXTRAMB); - - for (elem = 0; elem < table_size; elem++) - if (symb_table[2 * elem] != 0) - { - idx = symb_table[2 * elem + 1]; - /* Skip the name of collating - element. */ - idx += 1 + extra[idx]; + table_size = + _NL_CURRENT_WORD (LC_COLLATE, + _NL_COLLATE_SYMB_HASH_SIZEMB); + symb_table = (const int32_t *) + _NL_CURRENT (LC_COLLATE, + _NL_COLLATE_SYMB_TABLEMB); + extra = (const unsigned char *) + _NL_CURRENT (LC_COLLATE, + _NL_COLLATE_SYMB_EXTRAMB); + + for (elem = 0; elem < table_size; elem++) + if (symb_table[2 * elem] != 0) + { + idx = symb_table[2 * elem + 1]; + /* Skip the name of collating + element. */ + idx += 1 + extra[idx]; # if WIDE_CHAR_VERSION - /* Skip the byte sequence of the - collating element. */ - idx += 1 + extra[idx]; - /* Adjust for the alignment. */ - idx = (idx + 3) & ~3; - - wextra = (CHAR *) &extra[idx + 4]; - - if (/* Compare the length of the - sequence. */ - c1 == wextra[0] - /* Compare the wide char sequence. */ - && WMEMCMP (startp + 1, &wextra[1], - c1) == 0) - /* Yep, this is the entry. */ - break; + /* Skip the byte sequence of the + collating element. */ + idx += 1 + extra[idx]; + /* Adjust for the alignment. */ + idx = (idx + 3) & ~3; + + wextra = (CHAR *) &extra[idx + 4]; + + if (/* Compare the length of the + sequence. */ + c1 == wextra[0] + /* Compare the wide char sequence. */ + && (__wmemcmp (startp + 1, + &wextra[1], c1) + == 0)) + /* Yep, this is the entry. */ + break; # else - if (/* Compare the length of the - sequence. */ - c1 == extra[idx] - /* Compare the byte sequence. */ - && memcmp (startp + 1, - &extra[idx + 1], c1) == 0) - /* Yep, this is the entry. */ - break; + if (/* Compare the length of the + sequence. */ + c1 == extra[idx] + /* Compare the byte sequence. */ + && memcmp (startp + 1, + &extra[idx + 1], c1) == 0) + /* Yep, this is the entry. */ + break; # endif - } + } - if (elem < table_size) - { - /* Get the collation sequence value. */ - is_seqval = 1; + if (elem < table_size) + { + /* Get the collation sequence value. */ + is_seqval = true; # if WIDE_CHAR_VERSION - cend = wextra[1 + wextra[0]]; + cend = wextra[1 + wextra[0]]; # else - idx += 1 + extra[idx]; - /* Adjust for the alignment. */ - idx = (idx + 3) & ~3; - cend = *((int32_t *) &extra[idx]); + idx += 1 + extra[idx]; + /* Adjust for the alignment. */ + idx = (idx + 3) & ~3; + cend = *((int32_t *) &extra[idx]); # endif - } - else if (c1 == 1) - { - cend = startp[1]; - c = *p++; - } - else - return FNM_NOMATCH; - } - } - else - { - if (!(flags & FNM_NOESCAPE) && cend == L('\\')) - cend = *p++; - if (cend == L('\0')) - return FNM_NOMATCH; - cend = FOLD (cend); - } - - /* XXX It is not entirely clear to me how to handle - characters which are not mentioned in the - collation specification. */ - if ( + } + else if (c1 == 1) + { + cend = startp[1]; + c = *p++; + } + else + return FNM_NOMATCH; + } + } + else + { + if (!(flags & FNM_NOESCAPE) && cend == L_('\\')) + cend = *p++; + if (cend == L_('\0')) + return FNM_NOMATCH; + cend = FOLD (cend); + } + + /* XXX It is not entirely clear to me how to handle + characters which are not mentioned in the + collation specification. */ + if ( # if WIDE_CHAR_VERSION - lcollseq == 0xffffffff || + lcollseq == 0xffffffff || # endif - lcollseq <= fcollseq) - { - /* We have to look at the upper bound. */ - uint32_t hcollseq; - - if (is_seqval) - hcollseq = cend; - else - { + lcollseq <= fcollseq) + { + /* We have to look at the upper bound. */ + uint32_t hcollseq; + + if (is_seqval) + hcollseq = cend; + else + { # if WIDE_CHAR_VERSION - hcollseq = - __collseq_table_lookup (collseq, cend); - if (hcollseq == ~((uint32_t) 0)) - { - /* Hum, no information about the upper - bound. The matching succeeds if the - lower bound is matched exactly. */ - if (lcollseq != fcollseq) - goto range_not_matched; - - goto matched; - } + hcollseq = + __collseq_table_lookup (collseq, cend); + if (hcollseq == ~((uint32_t) 0)) + { + /* Hum, no information about the upper + bound. The matching succeeds if the + lower bound is matched exactly. */ + if (lcollseq != fcollseq) + goto range_not_matched; + + goto matched; + } # else - hcollseq = collseq[cend]; + hcollseq = collseq[cend]; # endif - } + } - if (lcollseq <= hcollseq && fcollseq <= hcollseq) - goto matched; - } + if (lcollseq <= hcollseq && fcollseq <= hcollseq) + goto matched; + } # if WIDE_CHAR_VERSION - range_not_matched: + range_not_matched: # endif #else - /* We use a boring value comparison of the character - values. This is better than comparing using - `strcoll' since the latter would have surprising - and sometimes fatal consequences. */ - UCHAR cend = *p++; - - if (!(flags & FNM_NOESCAPE) && cend == L('\\')) - cend = *p++; - if (cend == L('\0')) - return FNM_NOMATCH; - - /* It is a range. */ - if (cold <= fn && fn <= cend) - goto matched; + /* We use a boring value comparison of the character + values. This is better than comparing using + 'strcoll' since the latter would have surprising + and sometimes fatal consequences. */ + UCHAR cend = *p++; + + if (!(flags & FNM_NOESCAPE) && cend == L_('\\')) + cend = *p++; + if (cend == L_('\0')) + return FNM_NOMATCH; + + /* It is a range. */ + if ((UCHAR) cold <= fn && fn <= cend) + goto matched; #endif - c = *p++; - } - } - - if (c == L(']')) - break; - } - - if (!not) - return FNM_NOMATCH; - break; - - matched: - /* Skip the rest of the [...] that already matched. */ - while ((c = *p++) != L (']')) - { - if (c == L('\0')) - /* [... (unterminated) loses. */ - return FNM_NOMATCH; - - if (!(flags & FNM_NOESCAPE) && c == L('\\')) - { - if (*p == L('\0')) - return FNM_NOMATCH; - /* XXX 1003.2d11 is unclear if this is right. */ - ++p; - } - else if (c == L('[') && *p == L(':')) - { - int c1 = 0; - const CHAR *startp = p; - - while (1) - { - c = *++p; - if (++c1 == CHAR_CLASS_MAX_LENGTH) - return FNM_NOMATCH; - - if (*p == L(':') && p[1] == L(']')) - break; - - if (c < L('a') || c >= L('z')) - { - p = startp - 2; - break; - } - } - p += 2; - } - else if (c == L('[') && *p == L('=')) - { - c = *++p; - if (c == L('\0')) - return FNM_NOMATCH; - c = *++p; - if (c != L('=') || p[1] != L(']')) - return FNM_NOMATCH; - p += 2; - } - else if (c == L('[') && *p == L('.')) - { - while (1) - { - c = *++p; - if (c == L('\0')) - return FNM_NOMATCH; - - if (c == L('.') && p[1] == L(']')) - break; - } - p += 2; - } - } - if (not) - return FNM_NOMATCH; - } - break; - - case L('+'): - case L('@'): - case L('!'): - if (__builtin_expect (flags & FNM_EXTMATCH, 0) && *p == '(') - { - int res = EXT (c, p, n, string_end, no_leading_period, flags, - alloca_used); - if (res != -1) - return res; - } - goto normal_match; - - case L('/'): - if (NO_LEADING_PERIOD (flags)) - { - if (n == string_end || c != (UCHAR) *n) - return FNM_NOMATCH; - - new_no_leading_period = 1; - break; - } - /* FALLTHROUGH */ - default: - normal_match: - if (n == string_end || c != FOLD ((UCHAR) *n)) - return FNM_NOMATCH; - } + c = *p++; + } + } + + if (c == L_(']')) + break; + } + + if (!not) + return FNM_NOMATCH; + break; + + matched: + /* Skip the rest of the [...] that already matched. */ + while ((c = *p++) != L_(']')) + { + if (c == L_('\0')) + /* [... (unterminated) loses. */ + return FNM_NOMATCH; + + if (!(flags & FNM_NOESCAPE) && c == L_('\\')) + { + if (*p == L_('\0')) + return FNM_NOMATCH; + /* XXX 1003.2d11 is unclear if this is right. */ + ++p; + } + else if (c == L_('[') && *p == L_(':')) + { + int c1 = 0; + const CHAR *startp = p; + + while (1) + { + c = *++p; + if (++c1 == CHAR_CLASS_MAX_LENGTH) + return FNM_NOMATCH; + + if (*p == L_(':') && p[1] == L_(']')) + break; + + if (c < L_('a') || c >= L_('z')) + { + p = startp - 2; + break; + } + } + p += 2; + } + else if (c == L_('[') && *p == L_('=')) + { + c = *++p; + if (c == L_('\0')) + return FNM_NOMATCH; + c = *++p; + if (c != L_('=') || p[1] != L_(']')) + return FNM_NOMATCH; + p += 2; + } + else if (c == L_('[') && *p == L_('.')) + { + while (1) + { + c = *++p; + if (c == L_('\0')) + return FNM_NOMATCH; + + if (c == L_('.') && p[1] == L_(']')) + break; + } + p += 2; + } + } + if (not) + return FNM_NOMATCH; + } + break; + + case L_('+'): + case L_('@'): + case L_('!'): + if (__glibc_unlikely (flags & FNM_EXTMATCH) && *p == '(') + { + int res = EXT (c, p, n, string_end, no_leading_period, flags, + alloca_used); + if (res != -1) + return res; + } + goto normal_match; + + case L_('/'): + if (NO_LEADING_PERIOD (flags)) + { + if (n == string_end || c != (UCHAR) *n) + return FNM_NOMATCH; + + new_no_leading_period = true; + break; + } + FALLTHROUGH; + default: + normal_match: + if (n == string_end || c != FOLD ((UCHAR) *n)) + return FNM_NOMATCH; + } no_leading_period = new_no_leading_period; ++n; @@ -934,7 +922,7 @@ FCT (const CHAR *pattern, const CHAR *string, const CHAR *string_end, if (n == string_end) return 0; - if ((flags & FNM_LEADING_DIR) && n != string_end && *n == L('/')) + if ((flags & FNM_LEADING_DIR) && n != string_end && *n == L_('/')) /* The FNM_LEADING_DIR flag says that "foo*" matches "foobar/frobozz". */ return 0; @@ -948,37 +936,37 @@ END (const CHAR *pattern) const CHAR *p = pattern; while (1) - if (*++p == L('\0')) + if (*++p == L_('\0')) /* This is an invalid pattern. */ return pattern; - else if (*p == L('[')) + else if (*p == L_('[')) { - /* Handle brackets special. */ - if (posixly_correct == 0) - posixly_correct = getenv ("POSIXLY_CORRECT") != NULL ? 1 : -1; - - /* Skip the not sign. We have to recognize it because of a possibly - following ']'. */ - if (*++p == L('!') || (posixly_correct < 0 && *p == L('^'))) - ++p; - /* A leading ']' is recognized as such. */ - if (*p == L(']')) - ++p; - /* Skip over all characters of the list. */ - while (*p != L(']')) - if (*p++ == L('\0')) - /* This is no valid pattern. */ - return pattern; + /* Handle brackets special. */ + if (posixly_correct == 0) + posixly_correct = getenv ("POSIXLY_CORRECT") != NULL ? 1 : -1; + + /* Skip the not sign. We have to recognize it because of a possibly + following ']'. */ + if (*++p == L_('!') || (posixly_correct < 0 && *p == L_('^'))) + ++p; + /* A leading ']' is recognized as such. */ + if (*p == L_(']')) + ++p; + /* Skip over all characters of the list. */ + while (*p != L_(']')) + if (*p++ == L_('\0')) + /* This is no valid pattern. */ + return pattern; } - else if ((*p == L('?') || *p == L('*') || *p == L('+') || *p == L('@') - || *p == L('!')) && p[1] == L('(')) + else if ((*p == L_('?') || *p == L_('*') || *p == L_('+') || *p == L_('@') + || *p == L_('!')) && p[1] == L_('(')) { - p = END (p + 1); - if (*p == L('\0')) - /* This is an invalid pattern. */ - return pattern; + p = END (p + 1); + if (*p == L_('\0')) + /* This is an invalid pattern. */ + return pattern; } - else if (*p == L(')')) + else if (*p == L_(')')) break; return p + 1; @@ -987,19 +975,19 @@ END (const CHAR *pattern) static int EXT (INT opt, const CHAR *pattern, const CHAR *string, const CHAR *string_end, - int no_leading_period, int flags, size_t alloca_used) + bool no_leading_period, int flags, size_t alloca_used) { const CHAR *startp; - int level; + ssize_t level; struct patternlist { struct patternlist *next; CHAR malloced; - CHAR str[0]; + CHAR str[]; } *list = NULL; struct patternlist **lastp = &list; size_t pattern_len = STRLEN (pattern); - int any_malloced = 0; + bool any_malloced = false; const CHAR *p; const CHAR *rs; int retval = 0; @@ -1007,174 +995,180 @@ EXT (INT opt, const CHAR *pattern, const CHAR *string, const CHAR *string_end, /* Parse the pattern. Store the individual parts in the list. */ level = 0; for (startp = p = pattern + 1; level >= 0; ++p) - if (*p == L('\0')) + if (*p == L_('\0')) { - /* This is an invalid pattern. */ - retval = -1; - goto out; + /* This is an invalid pattern. */ + retval = -1; + goto out; } - else if (*p == L('[')) + else if (*p == L_('[')) { - /* Handle brackets special. */ - if (posixly_correct == 0) - posixly_correct = getenv ("POSIXLY_CORRECT") != NULL ? 1 : -1; - - /* Skip the not sign. We have to recognize it because of a possibly - following ']'. */ - if (*++p == L('!') || (posixly_correct < 0 && *p == L('^'))) - ++p; - /* A leading ']' is recognized as such. */ - if (*p == L(']')) - ++p; - /* Skip over all characters of the list. */ - while (*p != L(']')) - if (*p++ == L('\0')) - { - /* This is no valid pattern. */ - retval = -1; - goto out; - } + /* Handle brackets special. */ + if (posixly_correct == 0) + posixly_correct = getenv ("POSIXLY_CORRECT") != NULL ? 1 : -1; + + /* Skip the not sign. We have to recognize it because of a possibly + following ']'. */ + if (*++p == L_('!') || (posixly_correct < 0 && *p == L_('^'))) + ++p; + /* A leading ']' is recognized as such. */ + if (*p == L_(']')) + ++p; + /* Skip over all characters of the list. */ + while (*p != L_(']')) + if (*p++ == L_('\0')) + { + /* This is no valid pattern. */ + retval = -1; + goto out; + } } - else if ((*p == L('?') || *p == L('*') || *p == L('+') || *p == L('@') - || *p == L('!')) && p[1] == L('(')) + else if ((*p == L_('?') || *p == L_('*') || *p == L_('+') || *p == L_('@') + || *p == L_('!')) && p[1] == L_('(')) /* Remember the nesting level. */ ++level; - else if (*p == L(')')) + else if (*p == L_(')')) { - if (level-- == 0) - { - /* This means we found the end of the pattern. */ + if (level-- == 0) + { + /* This means we found the end of the pattern. */ #define NEW_PATTERN \ - struct patternlist *newp; \ - size_t slen = (opt == L('?') || opt == L('@') \ - ? pattern_len : (p - startp + 1)); \ - slen = sizeof (struct patternlist) + (slen * sizeof (CHAR)); \ - int malloced = ! __libc_use_alloca (alloca_used + slen); \ - if (__builtin_expect (malloced, 0)) \ - { \ - newp = malloc (slen); \ - if (newp == NULL) \ - { \ - retval = -2; \ - goto out; \ - } \ - any_malloced = 1; \ - } \ - else \ - newp = alloca_account (slen, alloca_used); \ - newp->next = NULL; \ - newp->malloced = malloced; \ - *((CHAR *) MEMPCPY (newp->str, startp, p - startp)) = L('\0'); \ - *lastp = newp; \ - lastp = &newp->next - NEW_PATTERN; - } + struct patternlist *newp; \ + size_t plen = (opt == L_('?') || opt == L_('@') \ + ? pattern_len : (p - startp + 1UL)); \ + idx_t slen = FLEXSIZEOF (struct patternlist, str, 0); \ + idx_t new_used = alloca_used + slen; \ + idx_t plensize; \ + if (INT_MULTIPLY_WRAPV (plen, sizeof (CHAR), &plensize) \ + || INT_ADD_WRAPV (new_used, plensize, &new_used)) \ + { \ + retval = -2; \ + goto out; \ + } \ + slen += plensize; \ + bool malloced = ! __libc_use_alloca (new_used); \ + if (__glibc_unlikely (malloced)) \ + { \ + newp = malloc (slen); \ + if (newp == NULL) \ + { \ + retval = -2; \ + goto out; \ + } \ + any_malloced = true; \ + } \ + else \ + newp = alloca_account (slen, alloca_used); \ + newp->next = NULL; \ + newp->malloced = malloced; \ + *((CHAR *) MEMPCPY (newp->str, startp, p - startp)) = L_('\0'); \ + *lastp = newp; \ + lastp = &newp->next + NEW_PATTERN; + } } - else if (*p == L('|')) + else if (*p == L_('|')) { - if (level == 0) - { - NEW_PATTERN; - startp = p + 1; - } + if (level == 0) + { + NEW_PATTERN; + startp = p + 1; + } } assert (list != NULL); - assert (p[-1] == L(')')); + assert (p[-1] == L_(')')); #undef NEW_PATTERN switch (opt) { - case L('*'): + case L_('*'): if (FCT (p, string, string_end, no_leading_period, flags, NULL, - alloca_used) == 0) - goto success; - /* FALLTHROUGH */ - - case L('+'): + alloca_used) == 0) + goto success; + FALLTHROUGH; + case L_('+'): do - { - for (rs = string; rs <= string_end; ++rs) - /* First match the prefix with the current pattern with the - current pattern. */ - if (FCT (list->str, string, rs, no_leading_period, - flags & FNM_FILE_NAME ? flags : flags & ~FNM_PERIOD, - NULL, alloca_used) == 0 - /* This was successful. Now match the rest with the rest - of the pattern. */ - && (FCT (p, rs, string_end, - rs == string - ? no_leading_period - : rs[-1] == '/' && NO_LEADING_PERIOD (flags) ? 1 : 0, - flags & FNM_FILE_NAME - ? flags : flags & ~FNM_PERIOD, NULL, alloca_used) == 0 - /* This didn't work. Try the whole pattern. */ - || (rs != string - && FCT (pattern - 1, rs, string_end, - rs == string - ? no_leading_period - : (rs[-1] == '/' && NO_LEADING_PERIOD (flags) - ? 1 : 0), - flags & FNM_FILE_NAME - ? flags : flags & ~FNM_PERIOD, NULL, - alloca_used) == 0))) - /* It worked. Signal success. */ - goto success; - } + { + for (rs = string; rs <= string_end; ++rs) + /* First match the prefix with the current pattern with the + current pattern. */ + if (FCT (list->str, string, rs, no_leading_period, + flags & FNM_FILE_NAME ? flags : flags & ~FNM_PERIOD, + NULL, alloca_used) == 0 + /* This was successful. Now match the rest with the rest + of the pattern. */ + && (FCT (p, rs, string_end, + rs == string + ? no_leading_period + : rs[-1] == '/' && NO_LEADING_PERIOD (flags), + flags & FNM_FILE_NAME + ? flags : flags & ~FNM_PERIOD, NULL, alloca_used) == 0 + /* This didn't work. Try the whole pattern. */ + || (rs != string + && FCT (pattern - 1, rs, string_end, + rs == string + ? no_leading_period + : rs[-1] == '/' && NO_LEADING_PERIOD (flags), + flags & FNM_FILE_NAME + ? flags : flags & ~FNM_PERIOD, NULL, + alloca_used) == 0))) + /* It worked. Signal success. */ + goto success; + } while ((list = list->next) != NULL); /* None of the patterns lead to a match. */ retval = FNM_NOMATCH; break; - case L('?'): + case L_('?'): if (FCT (p, string, string_end, no_leading_period, flags, NULL, - alloca_used) == 0) - goto success; - /* FALLTHROUGH */ - - case L('@'): + alloca_used) == 0) + goto success; + FALLTHROUGH; + case L_('@'): do - /* I cannot believe it but `strcat' is actually acceptable - here. Match the entire string with the prefix from the - pattern list and the rest of the pattern following the - pattern list. */ - if (FCT (STRCAT (list->str, p), string, string_end, - no_leading_period, - flags & FNM_FILE_NAME ? flags : flags & ~FNM_PERIOD, - NULL, alloca_used) == 0) - /* It worked. Signal success. */ - goto success; + /* I cannot believe it but 'strcat' is actually acceptable + here. Match the entire string with the prefix from the + pattern list and the rest of the pattern following the + pattern list. */ + if (FCT (STRCAT (list->str, p), string, string_end, + no_leading_period, + flags & FNM_FILE_NAME ? flags : flags & ~FNM_PERIOD, + NULL, alloca_used) == 0) + /* It worked. Signal success. */ + goto success; while ((list = list->next) != NULL); /* None of the patterns lead to a match. */ retval = FNM_NOMATCH; break; - case L('!'): + case L_('!'): for (rs = string; rs <= string_end; ++rs) - { - struct patternlist *runp; - - for (runp = list; runp != NULL; runp = runp->next) - if (FCT (runp->str, string, rs, no_leading_period, - flags & FNM_FILE_NAME ? flags : flags & ~FNM_PERIOD, - NULL, alloca_used) == 0) - break; - - /* If none of the patterns matched see whether the rest does. */ - if (runp == NULL - && (FCT (p, rs, string_end, - rs == string - ? no_leading_period - : rs[-1] == '/' && NO_LEADING_PERIOD (flags) ? 1 : 0, - flags & FNM_FILE_NAME ? flags : flags & ~FNM_PERIOD, - NULL, alloca_used) == 0)) - /* This is successful. */ - goto success; - } + { + struct patternlist *runp; + + for (runp = list; runp != NULL; runp = runp->next) + if (FCT (runp->str, string, rs, no_leading_period, + flags & FNM_FILE_NAME ? flags : flags & ~FNM_PERIOD, + NULL, alloca_used) == 0) + break; + + /* If none of the patterns matched see whether the rest does. */ + if (runp == NULL + && (FCT (p, rs, string_end, + rs == string + ? no_leading_period + : rs[-1] == '/' && NO_LEADING_PERIOD (flags), + flags & FNM_FILE_NAME ? flags : flags & ~FNM_PERIOD, + NULL, alloca_used) == 0)) + /* This is successful. */ + goto success; + } /* None of the patterns together with the rest of the pattern - lead to a match. */ + lead to a match. */ retval = FNM_NOMATCH; break; @@ -1189,10 +1183,10 @@ EXT (INT opt, const CHAR *pattern, const CHAR *string, const CHAR *string_end, if (any_malloced) while (list != NULL) { - struct patternlist *old = list; - list = list->next; - if (old->malloced) - free (old); + struct patternlist *old = list; + list = list->next; + if (old->malloced) + free (old); } return retval; @@ -1209,10 +1203,9 @@ EXT (INT opt, const CHAR *pattern, const CHAR *string, const CHAR *string_end, #undef STRUCT #undef MEMPCPY #undef MEMCHR -#undef STRCOLL #undef STRLEN #undef STRCAT -#undef L +#undef L_ #undef BTOWC #undef WIDE_CHAR_VERSION #undef FINDIDX -- 2.25.1