diff options
Diffstat (limited to 'erts/emulator/pcre/pcre_internal.h')
| -rw-r--r-- | erts/emulator/pcre/pcre_internal.h | 1136 |
1 files changed, 1136 insertions, 0 deletions
diff --git a/erts/emulator/pcre/pcre_internal.h b/erts/emulator/pcre/pcre_internal.h new file mode 100644 index 0000000000..6aafabb0c9 --- /dev/null +++ b/erts/emulator/pcre/pcre_internal.h @@ -0,0 +1,1136 @@ +/************************************************* +* Perl-Compatible Regular Expressions * +*************************************************/ + + +/* PCRE is a library of functions to support regular expressions whose syntax +and semantics are as close as possible to those of the Perl 5 language. + + Written by Philip Hazel + Copyright (c) 1997-2008 University of Cambridge + +----------------------------------------------------------------------------- +Redistribution and use in source and binary forms, with or without +modification, are permitted provided that the following conditions are met: + + * Redistributions of source code must retain the above copyright notice, + this list of conditions and the following disclaimer. + + * 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. + + * Neither the name of the University of Cambridge nor the names of its + contributors may be used to endorse or promote products derived from + this software without specific prior written permission. + +THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "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 COPYRIGHT OWNER OR CONTRIBUTORS 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. +----------------------------------------------------------------------------- +*/ + +/* This header contains definitions that are shared between the different +modules, but which are not relevant to the exported API. This includes some +functions whose names all begin with "_erts_pcre_". */ + +/* %ExternalCopyright% */ + +#ifndef PCRE_INTERNAL_H +#define PCRE_INTERNAL_H + +#ifdef ERLANG_INTEGRATION +#include "local_config.h" +#endif + +/* Define DEBUG to get debugging output on stdout. */ + +#if 0 +#define DEBUG +#endif + +/* Use a macro for debugging printing, 'cause that eliminates the use of #ifdef +inline, and there are *still* stupid compilers about that don't like indented +pre-processor statements, or at least there were when I first wrote this. After +all, it had only been about 10 years then... + +It turns out that the Mac Debugging.h header also defines the macro DPRINTF, so +be absolutely sure we get our version. */ + +#undef DPRINTF +#ifdef DEBUG +#define DPRINTF(p) printf p +#else +#define DPRINTF(p) /* Nothing */ +#endif + + +/* Standard C headers plus the external interface definition. The only time +setjmp and stdarg are used is when NO_RECURSE is set. */ + +#include <ctype.h> +#include <limits.h> +#include <setjmp.h> +#include <stdarg.h> +#include <stddef.h> +#include <stdio.h> +#include <stdlib.h> +#include <string.h> + +/* When compiling a DLL for Windows, the exported symbols have to be declared +using some MS magic. I found some useful information on this web page: +http://msdn2.microsoft.com/en-us/library/y4h7bcy6(VS.80).aspx. According to the +information there, using __declspec(dllexport) without "extern" we have a +definition; with "extern" we have a declaration. The settings here override the +setting in pcre.h (which is included below); it defines only PCRE_EXP_DECL, +which is all that is needed for applications (they just import the symbols). We +use: + + PCRE_EXP_DECL for declarations + PCRE_EXP_DEFN for definitions of exported functions + PCRE_EXP_DATA_DEFN for definitions of exported variables + +The reason for the two DEFN macros is that in non-Windows environments, one +does not want to have "extern" before variable definitions because it leads to +compiler warnings. So we distinguish between functions and variables. In +Windows, the two should always be the same. + +The reason for wrapping this in #ifndef PCRE_EXP_DECL is so that pcretest, +which is an application, but needs to import this file in order to "peek" at +internals, can #include pcre.h first to get an application's-eye view. + +In principle, people compiling for non-Windows, non-Unix-like (i.e. uncommon, +special-purpose environments) might want to stick other stuff in front of +exported symbols. That's why, in the non-Windows case, we set PCRE_EXP_DEFN and +PCRE_EXP_DATA_DEFN only if they are not already set. */ + +#ifndef PCRE_EXP_DECL +# ifdef _WIN32 +# ifndef PCRE_STATIC +# define PCRE_EXP_DECL extern __declspec(dllexport) +# define PCRE_EXP_DEFN __declspec(dllexport) +# define PCRE_EXP_DATA_DEFN __declspec(dllexport) +# else +# define PCRE_EXP_DECL extern +# define PCRE_EXP_DEFN +# define PCRE_EXP_DATA_DEFN +# endif +# else +# ifdef __cplusplus +# define PCRE_EXP_DECL extern "C" +# else +# define PCRE_EXP_DECL extern +# endif +# ifndef PCRE_EXP_DEFN +# define PCRE_EXP_DEFN PCRE_EXP_DECL +# endif +# ifndef PCRE_EXP_DATA_DEFN +# define PCRE_EXP_DATA_DEFN +# endif +# endif +#endif + +/* We need to have types that specify unsigned 16-bit and 32-bit integers. We +cannot determine these outside the compilation (e.g. by running a program as +part of "configure") because PCRE is often cross-compiled for use on other +systems. Instead we make use of the maximum sizes that are available at +preprocessor time in standard C environments. */ + +#if USHRT_MAX == 65535 + typedef unsigned short pcre_uint16; +#elif UINT_MAX == 65535 + typedef unsigned int pcre_uint16; +#else + #error Cannot determine a type for 16-bit unsigned integers +#endif + +#if UINT_MAX == 4294967295 + typedef unsigned int pcre_uint32; +#elif ULONG_MAX == 4294967295 + typedef unsigned long int pcre_uint32; +#else + #error Cannot determine a type for 32-bit unsigned integers +#endif + +/* All character handling must be done as unsigned characters. Otherwise there +are problems with top-bit-set characters and functions such as isspace(). +However, we leave the interface to the outside world as char *, because that +should make things easier for callers. We define a short type for unsigned char +to save lots of typing. I tried "uchar", but it causes problems on Digital +Unix, where it is defined in sys/types, so use "uschar" instead. */ + +typedef unsigned char uschar; + +/* This is an unsigned int value that no character can ever have. UTF-8 +characters only go up to 0x7fffffff (though Unicode doesn't go beyond +0x0010ffff). */ + +#define NOTACHAR 0xffffffff + +/* PCRE is able to support several different kinds of newline (CR, LF, CRLF, +"any" and "anycrlf" at present). The following macros are used to package up +testing for newlines. NLBLOCK, PSSTART, and PSEND are defined in the various +modules to indicate in which datablock the parameters exist, and what the +start/end of string field names are. */ + +#define NLTYPE_FIXED 0 /* Newline is a fixed length string */ +#define NLTYPE_ANY 1 /* Newline is any Unicode line ending */ +#define NLTYPE_ANYCRLF 2 /* Newline is CR, LF, or CRLF */ + +/* This macro checks for a newline at the given position */ + +#define IS_NEWLINE(p) \ + ((NLBLOCK->nltype != NLTYPE_FIXED)? \ + ((p) < NLBLOCK->PSEND && \ + _erts_pcre_is_newline((p), NLBLOCK->nltype, NLBLOCK->PSEND, &(NLBLOCK->nllen),\ + utf8)) \ + : \ + ((p) <= NLBLOCK->PSEND - NLBLOCK->nllen && \ + (p)[0] == NLBLOCK->nl[0] && \ + (NLBLOCK->nllen == 1 || (p)[1] == NLBLOCK->nl[1]) \ + ) \ + ) + +/* This macro checks for a newline immediately preceding the given position */ + +#define WAS_NEWLINE(p) \ + ((NLBLOCK->nltype != NLTYPE_FIXED)? \ + ((p) > NLBLOCK->PSSTART && \ + _erts_pcre_was_newline((p), NLBLOCK->nltype, NLBLOCK->PSSTART, \ + &(NLBLOCK->nllen), utf8)) \ + : \ + ((p) >= NLBLOCK->PSSTART + NLBLOCK->nllen && \ + (p)[-NLBLOCK->nllen] == NLBLOCK->nl[0] && \ + (NLBLOCK->nllen == 1 || (p)[-NLBLOCK->nllen+1] == NLBLOCK->nl[1]) \ + ) \ + ) + +/* When PCRE is compiled as a C++ library, the subject pointer can be replaced +with a custom type. This makes it possible, for example, to allow pcre_exec() +to process subject strings that are discontinuous by using a smart pointer +class. It must always be possible to inspect all of the subject string in +pcre_exec() because of the way it backtracks. Two macros are required in the +normal case, for sign-unspecified and unsigned char pointers. The former is +used for the external interface and appears in pcre.h, which is why its name +must begin with PCRE_. */ + +#ifdef CUSTOM_SUBJECT_PTR +#define PCRE_SPTR CUSTOM_SUBJECT_PTR +#define USPTR CUSTOM_SUBJECT_PTR +#else +#define PCRE_SPTR const char * +#define USPTR const unsigned char * +#endif + + + +/* Include the public PCRE header and the definitions of UCP character property +values. */ + +#include "pcre.h" +#include "ucp.h" + +/* When compiling for use with the Virtual Pascal compiler, these functions +need to have their names changed. PCRE must be compiled with the -DVPCOMPAT +option on the command line. */ + +#ifdef VPCOMPAT +#define strlen(s) _strlen(s) +#define strncmp(s1,s2,m) _strncmp(s1,s2,m) +#define memcmp(s,c,n) _memcmp(s,c,n) +#define memcpy(d,s,n) _memcpy(d,s,n) +#define memmove(d,s,n) _memmove(d,s,n) +#define memset(s,c,n) _memset(s,c,n) +#else /* VPCOMPAT */ + +/* To cope with SunOS4 and other systems that lack memmove() but have bcopy(), +define a macro for memmove() if HAVE_MEMMOVE is false, provided that HAVE_BCOPY +is set. Otherwise, include an emulating function for those systems that have +neither (there some non-Unix environments where this is the case). */ + +#ifndef HAVE_MEMMOVE +#undef memmove /* some systems may have a macro */ +#ifdef HAVE_BCOPY +#define memmove(a, b, c) bcopy(b, a, c) +#else /* HAVE_BCOPY */ +static void * +pcre_memmove(void *d, const void *s, size_t n) +{ +size_t i; +unsigned char *dest = (unsigned char *)d; +const unsigned char *src = (const unsigned char *)s; +if (dest > src) + { + dest += n; + src += n; + for (i = 0; i < n; ++i) *(--dest) = *(--src); + return (void *)dest; + } +else + { + for (i = 0; i < n; ++i) *dest++ = *src++; + return (void *)(dest - n); + } +} +#define memmove(a, b, c) pcre_memmove(a, b, c) +#endif /* not HAVE_BCOPY */ +#endif /* not HAVE_MEMMOVE */ +#endif /* not VPCOMPAT */ + + +/* PCRE keeps offsets in its compiled code as 2-byte quantities (always stored +in big-endian order) by default. These are used, for example, to link from the +start of a subpattern to its alternatives and its end. The use of 2 bytes per +offset limits the size of the compiled regex to around 64K, which is big enough +for almost everybody. However, I received a request for an even bigger limit. +For this reason, and also to make the code easier to maintain, the storing and +loading of offsets from the byte string is now handled by the macros that are +defined here. + +The macros are controlled by the value of LINK_SIZE. This defaults to 2 in +the config.h file, but can be overridden by using -D on the command line. This +is automated on Unix systems via the "configure" command. */ + +#if LINK_SIZE == 2 + +#define PUT(a,n,d) \ + (a[n] = (d) >> 8), \ + (a[(n)+1] = (d) & 255) + +#define GET(a,n) \ + (((a)[n] << 8) | (a)[(n)+1]) + +#define MAX_PATTERN_SIZE (1 << 16) + + +#elif LINK_SIZE == 3 + +#define PUT(a,n,d) \ + (a[n] = (d) >> 16), \ + (a[(n)+1] = (d) >> 8), \ + (a[(n)+2] = (d) & 255) + +#define GET(a,n) \ + (((a)[n] << 16) | ((a)[(n)+1] << 8) | (a)[(n)+2]) + +#define MAX_PATTERN_SIZE (1 << 24) + + +#elif LINK_SIZE == 4 + +#define PUT(a,n,d) \ + (a[n] = (d) >> 24), \ + (a[(n)+1] = (d) >> 16), \ + (a[(n)+2] = (d) >> 8), \ + (a[(n)+3] = (d) & 255) + +#define GET(a,n) \ + (((a)[n] << 24) | ((a)[(n)+1] << 16) | ((a)[(n)+2] << 8) | (a)[(n)+3]) + +#define MAX_PATTERN_SIZE (1 << 30) /* Keep it positive */ + + +#else +#error LINK_SIZE must be either 2, 3, or 4 +#endif + + +/* Convenience macro defined in terms of the others */ + +#define PUTINC(a,n,d) PUT(a,n,d), a += LINK_SIZE + + +/* PCRE uses some other 2-byte quantities that do not change when the size of +offsets changes. There are used for repeat counts and for other things such as +capturing parenthesis numbers in back references. */ + +#define PUT2(a,n,d) \ + a[n] = (d) >> 8; \ + a[(n)+1] = (d) & 255 + +#define GET2(a,n) \ + (((a)[n] << 8) | (a)[(n)+1]) + +#define PUT2INC(a,n,d) PUT2(a,n,d), a += 2 + + +/* When UTF-8 encoding is being used, a character is no longer just a single +byte. The macros for character handling generate simple sequences when used in +byte-mode, and more complicated ones for UTF-8 characters. BACKCHAR should +never be called in byte mode. To make sure it can never even appear when UTF-8 +support is omitted, we don't even define it. */ + +#ifndef SUPPORT_UTF8 +#define NEXTCHAR(p,end) p++; +#define GETCHAR(c, eptr) c = *eptr; +#define GETCHARTEST(c, eptr) c = *eptr; +#define GETCHARINC(c, eptr) c = *eptr++; +#define GETCHARINCTEST(c, eptr) c = *eptr++; +#define GETCHARLEN(c, eptr, len) c = *eptr; +/* #define BACKCHAR(eptr) */ + +#else /* SUPPORT_UTF8 */ + +/* Advance a character pointer one byte in non-UTF-8 mode and by one character +in UTF-8 mode. */ + +#define NEXTCHAR(p,end) \ + p++; \ + if (utf8) { while(p < end && (*p & 0xc0) == 0x80) p++; } + +/* Get the next UTF-8 character, not advancing the pointer. This is called when +we know we are in UTF-8 mode. */ + +#define GETCHAR(c, eptr) \ + c = *eptr; \ + if (c >= 0xc0) \ + { \ + int gcii; \ + int gcaa = _erts_pcre_utf8_table4[c & 0x3f]; /* Number of additional bytes */ \ + int gcss = 6*gcaa; \ + c = (c & _erts_pcre_utf8_table3[gcaa]) << gcss; \ + for (gcii = 1; gcii <= gcaa; gcii++) \ + { \ + gcss -= 6; \ + c |= (eptr[gcii] & 0x3f) << gcss; \ + } \ + } + +/* Get the next UTF-8 character, testing for UTF-8 mode, and not advancing the +pointer. */ + +#define GETCHARTEST(c, eptr) \ + c = *eptr; \ + if (utf8 && c >= 0xc0) \ + { \ + int gcii; \ + int gcaa = _erts_pcre_utf8_table4[c & 0x3f]; /* Number of additional bytes */ \ + int gcss = 6*gcaa; \ + c = (c & _erts_pcre_utf8_table3[gcaa]) << gcss; \ + for (gcii = 1; gcii <= gcaa; gcii++) \ + { \ + gcss -= 6; \ + c |= (eptr[gcii] & 0x3f) << gcss; \ + } \ + } + +/* Get the next UTF-8 character, advancing the pointer. This is called when we +know we are in UTF-8 mode. */ + +#define GETCHARINC(c, eptr) \ + c = *eptr++; \ + if (c >= 0xc0) \ + { \ + int gcaa = _erts_pcre_utf8_table4[c & 0x3f]; /* Number of additional bytes */ \ + int gcss = 6*gcaa; \ + c = (c & _erts_pcre_utf8_table3[gcaa]) << gcss; \ + while (gcaa-- > 0) \ + { \ + gcss -= 6; \ + c |= (*eptr++ & 0x3f) << gcss; \ + } \ + } + +/* Get the next character, testing for UTF-8 mode, and advancing the pointer */ + +#define GETCHARINCTEST(c, eptr) \ + c = *eptr++; \ + if (utf8 && c >= 0xc0) \ + { \ + int gcaa = _erts_pcre_utf8_table4[c & 0x3f]; /* Number of additional bytes */ \ + int gcss = 6*gcaa; \ + c = (c & _erts_pcre_utf8_table3[gcaa]) << gcss; \ + while (gcaa-- > 0) \ + { \ + gcss -= 6; \ + c |= (*eptr++ & 0x3f) << gcss; \ + } \ + } + +/* Get the next UTF-8 character, not advancing the pointer, incrementing length +if there are extra bytes. This is called when we know we are in UTF-8 mode. */ + +#define GETCHARLEN(c, eptr, len) \ + c = *eptr; \ + if (c >= 0xc0) \ + { \ + int gcii; \ + int gcaa = _erts_pcre_utf8_table4[c & 0x3f]; /* Number of additional bytes */ \ + int gcss = 6*gcaa; \ + c = (c & _erts_pcre_utf8_table3[gcaa]) << gcss; \ + for (gcii = 1; gcii <= gcaa; gcii++) \ + { \ + gcss -= 6; \ + c |= (eptr[gcii] & 0x3f) << gcss; \ + } \ + len += gcaa; \ + } + +/* If the pointer is not at the start of a character, move it back until +it is. This is called only in UTF-8 mode - we don't put a test within the macro +because almost all calls are already within a block of UTF-8 only code. */ + +#define BACKCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr-- + +#endif + + +/* In case there is no definition of offsetof() provided - though any proper +Standard C system should have one. */ + +#ifndef offsetof +#define offsetof(p_type,field) ((size_t)&(((p_type *)0)->field)) +#endif + + +/* These are the public options that can change during matching. */ + +#define PCRE_IMS (PCRE_CASELESS|PCRE_MULTILINE|PCRE_DOTALL) + +/* Private flags containing information about the compiled regex. They used to +live at the top end of the options word, but that got almost full, so now they +are in a 16-bit flags word. */ + +#define PCRE_NOPARTIAL 0x0001 /* can't use partial with this regex */ +#define PCRE_FIRSTSET 0x0002 /* first_byte is set */ +#define PCRE_REQCHSET 0x0004 /* req_byte is set */ +#define PCRE_STARTLINE 0x0008 /* start after \n for multiline */ +#define PCRE_JCHANGED 0x0010 /* j option used in regex */ +#define PCRE_HASCRORLF 0x0020 /* explicit \r or \n in pattern */ + +/* Options for the "extra" block produced by pcre_study(). */ + +#define PCRE_STUDY_MAPPED 0x01 /* a map of starting chars exists */ + +/* Masks for identifying the public options that are permitted at compile +time, run time, or study time, respectively. */ + +#define PCRE_NEWLINE_BITS (PCRE_NEWLINE_CR|PCRE_NEWLINE_LF|PCRE_NEWLINE_ANY| \ + PCRE_NEWLINE_ANYCRLF) + +#define PUBLIC_OPTIONS \ + (PCRE_CASELESS|PCRE_EXTENDED|PCRE_ANCHORED|PCRE_MULTILINE| \ + PCRE_DOTALL|PCRE_DOLLAR_ENDONLY|PCRE_EXTRA|PCRE_UNGREEDY|PCRE_UTF8| \ + PCRE_NO_AUTO_CAPTURE|PCRE_NO_UTF8_CHECK|PCRE_AUTO_CALLOUT|PCRE_FIRSTLINE| \ + PCRE_DUPNAMES|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE) + +#define PUBLIC_EXEC_OPTIONS \ + (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NO_UTF8_CHECK| \ + PCRE_PARTIAL|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE) + +#define PUBLIC_DFA_EXEC_OPTIONS \ + (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NO_UTF8_CHECK| \ + PCRE_PARTIAL|PCRE_DFA_SHORTEST|PCRE_DFA_RESTART|PCRE_NEWLINE_BITS| \ + PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE) + +#define PUBLIC_STUDY_OPTIONS 0 /* None defined */ + +/* Magic number to provide a small check against being handed junk. Also used +to detect whether a pattern was compiled on a host of different endianness. */ + +#define MAGIC_NUMBER 0x50435245UL /* 'PCRE' */ + +/* Negative values for the firstchar and reqchar variables */ + +#define REQ_UNSET (-2) +#define REQ_NONE (-1) + +/* The maximum remaining length of subject we are prepared to search for a +req_byte match. */ + +#define REQ_BYTE_MAX 1000 + +/* Flags added to firstbyte or reqbyte; a "non-literal" item is either a +variable-length repeat, or a anything other than literal characters. */ + +#define REQ_CASELESS 0x0100 /* indicates caselessness */ +#define REQ_VARY 0x0200 /* reqbyte followed non-literal item */ + +/* Miscellaneous definitions */ + +typedef int BOOL; + +#define FALSE 0 +#define TRUE 1 + +/* Escape items that are just an encoding of a particular data value. */ + +#ifndef ESC_e +#define ESC_e 27 +#endif + +#ifndef ESC_f +#define ESC_f '\f' +#endif + +#ifndef ESC_n +#define ESC_n '\n' +#endif + +#ifndef ESC_r +#define ESC_r '\r' +#endif + +/* We can't officially use ESC_t because it is a POSIX reserved identifier +(presumably because of all the others like size_t). */ + +#ifndef ESC_tee +#define ESC_tee '\t' +#endif + +/* Codes for different types of Unicode property */ + +#define PT_ANY 0 /* Any property - matches all chars */ +#define PT_LAMP 1 /* L& - the union of Lu, Ll, Lt */ +#define PT_GC 2 /* General characteristic (e.g. L) */ +#define PT_PC 3 /* Particular characteristic (e.g. Lu) */ +#define PT_SC 4 /* Script (e.g. Han) */ + +/* Flag bits and data types for the extended class (OP_XCLASS) for classes that +contain UTF-8 characters with values greater than 255. */ + +#define XCL_NOT 0x01 /* Flag: this is a negative class */ +#define XCL_MAP 0x02 /* Flag: a 32-byte map is present */ + +#define XCL_END 0 /* Marks end of individual items */ +#define XCL_SINGLE 1 /* Single item (one multibyte char) follows */ +#define XCL_RANGE 2 /* A range (two multibyte chars) follows */ +#define XCL_PROP 3 /* Unicode property (2-byte property code follows) */ +#define XCL_NOTPROP 4 /* Unicode inverted property (ditto) */ + +/* These are escaped items that aren't just an encoding of a particular data +value such as \n. They must have non-zero values, as check_escape() returns +their negation. Also, they must appear in the same order as in the opcode +definitions below, up to ESC_z. There's a dummy for OP_ANY because it +corresponds to "." rather than an escape sequence. The final one must be +ESC_REF as subsequent values are used for backreferences (\1, \2, \3, etc). +There are two tests in the code for an escape greater than ESC_b and less than +ESC_Z to detect the types that may be repeated. These are the types that +consume characters. If any new escapes are put in between that don't consume a +character, that code will have to change. */ + +enum { ESC_A = 1, ESC_G, ESC_K, ESC_B, ESC_b, ESC_D, ESC_d, ESC_S, ESC_s, + ESC_W, ESC_w, ESC_dum1, ESC_C, ESC_P, ESC_p, ESC_R, ESC_H, ESC_h, + ESC_V, ESC_v, ESC_X, ESC_Z, ESC_z, ESC_E, ESC_Q, ESC_k, ESC_REF }; + + +/* Opcode table: Starting from 1 (i.e. after OP_END), the values up to +OP_EOD must correspond in order to the list of escapes immediately above. + +*** NOTE NOTE NOTE *** Whenever this list is updated, the two macro definitions +that follow must also be updated to match. There is also a table called +"coptable" in pcre_dfa_exec.c that must be updated. */ + +enum { + OP_END, /* 0 End of pattern */ + + /* Values corresponding to backslashed metacharacters */ + + OP_SOD, /* 1 Start of data: \A */ + OP_SOM, /* 2 Start of match (subject + offset): \G */ + OP_SET_SOM, /* 3 Set start of match (\K) */ + OP_NOT_WORD_BOUNDARY, /* 4 \B */ + OP_WORD_BOUNDARY, /* 5 \b */ + OP_NOT_DIGIT, /* 6 \D */ + OP_DIGIT, /* 7 \d */ + OP_NOT_WHITESPACE, /* 8 \S */ + OP_WHITESPACE, /* 9 \s */ + OP_NOT_WORDCHAR, /* 10 \W */ + OP_WORDCHAR, /* 11 \w */ + OP_ANY, /* 12 Match any character */ + OP_ANYBYTE, /* 13 Match any byte (\C); different to OP_ANY for UTF-8 */ + OP_NOTPROP, /* 14 \P (not Unicode property) */ + OP_PROP, /* 15 \p (Unicode property) */ + OP_ANYNL, /* 16 \R (any newline sequence) */ + OP_NOT_HSPACE, /* 17 \H (not horizontal whitespace) */ + OP_HSPACE, /* 18 \h (horizontal whitespace) */ + OP_NOT_VSPACE, /* 19 \V (not vertical whitespace) */ + OP_VSPACE, /* 20 \v (vertical whitespace) */ + OP_EXTUNI, /* 21 \X (extended Unicode sequence */ + OP_EODN, /* 22 End of data or \n at end of data: \Z. */ + OP_EOD, /* 23 End of data: \z */ + + OP_OPT, /* 24 Set runtime options */ + OP_CIRC, /* 25 Start of line - varies with multiline switch */ + OP_DOLL, /* 26 End of line - varies with multiline switch */ + OP_CHAR, /* 27 Match one character, casefully */ + OP_CHARNC, /* 28 Match one character, caselessly */ + OP_NOT, /* 29 Match one character, not the following one */ + + OP_STAR, /* 30 The maximizing and minimizing versions of */ + OP_MINSTAR, /* 31 these six opcodes must come in pairs, with */ + OP_PLUS, /* 32 the minimizing one second. */ + OP_MINPLUS, /* 33 This first set applies to single characters.*/ + OP_QUERY, /* 34 */ + OP_MINQUERY, /* 35 */ + + OP_UPTO, /* 36 From 0 to n matches */ + OP_MINUPTO, /* 37 */ + OP_EXACT, /* 38 Exactly n matches */ + + OP_POSSTAR, /* 39 Possessified star */ + OP_POSPLUS, /* 40 Possessified plus */ + OP_POSQUERY, /* 41 Posesssified query */ + OP_POSUPTO, /* 42 Possessified upto */ + + OP_NOTSTAR, /* 43 The maximizing and minimizing versions of */ + OP_NOTMINSTAR, /* 44 these six opcodes must come in pairs, with */ + OP_NOTPLUS, /* 45 the minimizing one second. They must be in */ + OP_NOTMINPLUS, /* 46 exactly the same order as those above. */ + OP_NOTQUERY, /* 47 This set applies to "not" single characters. */ + OP_NOTMINQUERY, /* 48 */ + + OP_NOTUPTO, /* 49 From 0 to n matches */ + OP_NOTMINUPTO, /* 50 */ + OP_NOTEXACT, /* 51 Exactly n matches */ + + OP_NOTPOSSTAR, /* 52 Possessified versions */ + OP_NOTPOSPLUS, /* 53 */ + OP_NOTPOSQUERY, /* 54 */ + OP_NOTPOSUPTO, /* 55 */ + + OP_TYPESTAR, /* 56 The maximizing and minimizing versions of */ + OP_TYPEMINSTAR, /* 57 these six opcodes must come in pairs, with */ + OP_TYPEPLUS, /* 58 the minimizing one second. These codes must */ + OP_TYPEMINPLUS, /* 59 be in exactly the same order as those above. */ + OP_TYPEQUERY, /* 60 This set applies to character types such as \d */ + OP_TYPEMINQUERY, /* 61 */ + + OP_TYPEUPTO, /* 62 From 0 to n matches */ + OP_TYPEMINUPTO, /* 63 */ + OP_TYPEEXACT, /* 64 Exactly n matches */ + + OP_TYPEPOSSTAR, /* 65 Possessified versions */ + OP_TYPEPOSPLUS, /* 66 */ + OP_TYPEPOSQUERY, /* 67 */ + OP_TYPEPOSUPTO, /* 68 */ + + OP_CRSTAR, /* 69 The maximizing and minimizing versions of */ + OP_CRMINSTAR, /* 70 all these opcodes must come in pairs, with */ + OP_CRPLUS, /* 71 the minimizing one second. These codes must */ + OP_CRMINPLUS, /* 72 be in exactly the same order as those above. */ + OP_CRQUERY, /* 73 These are for character classes and back refs */ + OP_CRMINQUERY, /* 74 */ + OP_CRRANGE, /* 75 These are different to the three sets above. */ + OP_CRMINRANGE, /* 76 */ + + OP_CLASS, /* 77 Match a character class, chars < 256 only */ + OP_NCLASS, /* 78 Same, but the bitmap was created from a negative + class - the difference is relevant only when a UTF-8 + character > 255 is encountered. */ + + OP_XCLASS, /* 79 Extended class for handling UTF-8 chars within the + class. This does both positive and negative. */ + + OP_REF, /* 80 Match a back reference */ + OP_RECURSE, /* 81 Match a numbered subpattern (possibly recursive) */ + OP_CALLOUT, /* 82 Call out to external function if provided */ + + OP_ALT, /* 83 Start of alternation */ + OP_KET, /* 84 End of group that doesn't have an unbounded repeat */ + OP_KETRMAX, /* 85 These two must remain together and in this */ + OP_KETRMIN, /* 86 order. They are for groups the repeat for ever. */ + + /* The assertions must come before BRA, CBRA, ONCE, and COND.*/ + + OP_ASSERT, /* 87 Positive lookahead */ + OP_ASSERT_NOT, /* 88 Negative lookahead */ + OP_ASSERTBACK, /* 89 Positive lookbehind */ + OP_ASSERTBACK_NOT, /* 90 Negative lookbehind */ + OP_REVERSE, /* 91 Move pointer back - used in lookbehind assertions */ + + /* ONCE, BRA, CBRA, and COND must come after the assertions, with ONCE first, + as there's a test for >= ONCE for a subpattern that isn't an assertion. */ + + OP_ONCE, /* 92 Atomic group */ + OP_BRA, /* 93 Start of non-capturing bracket */ + OP_CBRA, /* 94 Start of capturing bracket */ + OP_COND, /* 95 Conditional group */ + + /* These three must follow the previous three, in the same order. There's a + check for >= SBRA to distinguish the two sets. */ + + OP_SBRA, /* 96 Start of non-capturing bracket, check empty */ + OP_SCBRA, /* 97 Start of capturing bracket, check empty */ + OP_SCOND, /* 98 Conditional group, check empty */ + + OP_CREF, /* 99 Used to hold a capture number as condition */ + OP_RREF, /* 100 Used to hold a recursion number as condition */ + OP_DEF, /* 101 The DEFINE condition */ + + OP_BRAZERO, /* 102 These two must remain together and in this */ + OP_BRAMINZERO, /* 103 order. */ + + /* These are backtracking control verbs */ + + OP_PRUNE, /* 104 */ + OP_SKIP, /* 105 */ + OP_THEN, /* 106 */ + OP_COMMIT, /* 107 */ + + /* These are forced failure and success verbs */ + + OP_FAIL, /* 108 */ + OP_ACCEPT /* 109 */ +}; + + +/* This macro defines textual names for all the opcodes. These are used only +for debugging. The macro is referenced only in pcre_printint.c. */ + +#define OP_NAME_LIST \ + "End", "\\A", "\\G", "\\K", "\\B", "\\b", "\\D", "\\d", \ + "\\S", "\\s", "\\W", "\\w", "Any", "Anybyte", \ + "notprop", "prop", "\\R", "\\H", "\\h", "\\V", "\\v", \ + "extuni", "\\Z", "\\z", \ + "Opt", "^", "$", "char", "charnc", "not", \ + "*", "*?", "+", "+?", "?", "??", "{", "{", "{", \ + "*+","++", "?+", "{", \ + "*", "*?", "+", "+?", "?", "??", "{", "{", "{", \ + "*+","++", "?+", "{", \ + "*", "*?", "+", "+?", "?", "??", "{", "{", "{", \ + "*+","++", "?+", "{", \ + "*", "*?", "+", "+?", "?", "??", "{", "{", \ + "class", "nclass", "xclass", "Ref", "Recurse", "Callout", \ + "Alt", "Ket", "KetRmax", "KetRmin", "Assert", "Assert not", \ + "AssertB", "AssertB not", "Reverse", \ + "Once", "Bra", "CBra", "Cond", "SBra", "SCBra", "SCond", \ + "Cond ref", "Cond rec", "Cond def", "Brazero", "Braminzero", \ + "*PRUNE", "*SKIP", "*THEN", "*COMMIT", "*FAIL", "*ACCEPT" + + +/* This macro defines the length of fixed length operations in the compiled +regex. The lengths are used when searching for specific things, and also in the +debugging printing of a compiled regex. We use a macro so that it can be +defined close to the definitions of the opcodes themselves. + +As things have been extended, some of these are no longer fixed lenths, but are +minima instead. For example, the length of a single-character repeat may vary +in UTF-8 mode. The code that uses this table must know about such things. */ + +#define OP_LENGTHS \ + 1, /* End */ \ + 1, 1, 1, 1, 1, /* \A, \G, \K, \B, \b */ \ + 1, 1, 1, 1, 1, 1, /* \D, \d, \S, \s, \W, \w */ \ + 1, 1, /* Any, Anybyte */ \ + 3, 3, 1, /* NOTPROP, PROP, EXTUNI */ \ + 1, 1, 1, 1, 1, /* \R, \H, \h, \V, \v */ \ + 1, 1, 2, 1, 1, /* \Z, \z, Opt, ^, $ */ \ + 2, /* Char - the minimum length */ \ + 2, /* Charnc - the minimum length */ \ + 2, /* not */ \ + /* Positive single-char repeats ** These are */ \ + 2, 2, 2, 2, 2, 2, /* *, *?, +, +?, ?, ?? ** minima in */ \ + 4, 4, 4, /* upto, minupto, exact ** UTF-8 mode */ \ + 2, 2, 2, 4, /* *+, ++, ?+, upto+ */ \ + /* Negative single-char repeats - only for chars < 256 */ \ + 2, 2, 2, 2, 2, 2, /* NOT *, *?, +, +?, ?, ?? */ \ + 4, 4, 4, /* NOT upto, minupto, exact */ \ + 2, 2, 2, 4, /* Possessive *, +, ?, upto */ \ + /* Positive type repeats */ \ + 2, 2, 2, 2, 2, 2, /* Type *, *?, +, +?, ?, ?? */ \ + 4, 4, 4, /* Type upto, minupto, exact */ \ + 2, 2, 2, 4, /* Possessive *+, ++, ?+, upto+ */ \ + /* Character class & ref repeats */ \ + 1, 1, 1, 1, 1, 1, /* *, *?, +, +?, ?, ?? */ \ + 5, 5, /* CRRANGE, CRMINRANGE */ \ + 33, /* CLASS */ \ + 33, /* NCLASS */ \ + 0, /* XCLASS - variable length */ \ + 3, /* REF */ \ + 1+LINK_SIZE, /* RECURSE */ \ + 2+2*LINK_SIZE, /* CALLOUT */ \ + 1+LINK_SIZE, /* Alt */ \ + 1+LINK_SIZE, /* Ket */ \ + 1+LINK_SIZE, /* KetRmax */ \ + 1+LINK_SIZE, /* KetRmin */ \ + 1+LINK_SIZE, /* Assert */ \ + 1+LINK_SIZE, /* Assert not */ \ + 1+LINK_SIZE, /* Assert behind */ \ + 1+LINK_SIZE, /* Assert behind not */ \ + 1+LINK_SIZE, /* Reverse */ \ + 1+LINK_SIZE, /* ONCE */ \ + 1+LINK_SIZE, /* BRA */ \ + 3+LINK_SIZE, /* CBRA */ \ + 1+LINK_SIZE, /* COND */ \ + 1+LINK_SIZE, /* SBRA */ \ + 3+LINK_SIZE, /* SCBRA */ \ + 1+LINK_SIZE, /* SCOND */ \ + 3, /* CREF */ \ + 3, /* RREF */ \ + 1, /* DEF */ \ + 1, 1, /* BRAZERO, BRAMINZERO */ \ + 1, 1, 1, 1, /* PRUNE, SKIP, THEN, COMMIT, */ \ + 1, 1 /* FAIL, ACCEPT */ + + +/* A magic value for OP_RREF to indicate the "any recursion" condition. */ + +#define RREF_ANY 0xffff + +/* Error code numbers. They are given names so that they can more easily be +tracked. */ + +enum { ERR0, ERR1, ERR2, ERR3, ERR4, ERR5, ERR6, ERR7, ERR8, ERR9, + ERR10, ERR11, ERR12, ERR13, ERR14, ERR15, ERR16, ERR17, ERR18, ERR19, + ERR20, ERR21, ERR22, ERR23, ERR24, ERR25, ERR26, ERR27, ERR28, ERR29, + ERR30, ERR31, ERR32, ERR33, ERR34, ERR35, ERR36, ERR37, ERR38, ERR39, + ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49, + ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58, ERR59, + ERR60, ERR61, ERR62, ERR63 }; + +/* The real format of the start of the pcre block; the index of names and the +code vector run on as long as necessary after the end. We store an explicit +offset to the name table so that if a regex is compiled on one host, saved, and +then run on another where the size of pointers is different, all might still +be well. For the case of compiled-on-4 and run-on-8, we include an extra +pointer that is always NULL. For future-proofing, a few dummy fields were +originally included - even though you can never get this planning right - but +there is only one left now. + +NOTE NOTE NOTE: +Because people can now save and re-use compiled patterns, any additions to this +structure should be made at the end, and something earlier (e.g. a new +flag in the options or one of the dummy fields) should indicate that the new +fields are present. Currently PCRE always sets the dummy fields to zero. +NOTE NOTE NOTE: +*/ + +typedef struct real_pcre { + pcre_uint32 magic_number; + pcre_uint32 size; /* Total that was malloced */ + pcre_uint32 options; /* Public options */ + pcre_uint16 flags; /* Private flags */ + pcre_uint16 dummy1; /* For future use */ + pcre_uint16 top_bracket; + pcre_uint16 top_backref; + pcre_uint16 first_byte; + pcre_uint16 req_byte; + pcre_uint16 name_table_offset; /* Offset to name table that follows */ + pcre_uint16 name_entry_size; /* Size of any name items */ + pcre_uint16 name_count; /* Number of name items */ + pcre_uint16 ref_count; /* Reference count */ + + const unsigned char *tables; /* Pointer to tables or NULL for std */ + const unsigned char *nullpad; /* NULL padding */ +} real_pcre; + +/* The format of the block used to store data from pcre_study(). The same +remark (see NOTE above) about extending this structure applies. */ + +typedef struct pcre_study_data { + pcre_uint32 size; /* Total that was malloced */ + pcre_uint32 options; + uschar start_bits[32]; +} pcre_study_data; + +/* Structure for passing "static" information around between the functions +doing the compiling, so that they are thread-safe. */ + +typedef struct compile_data { + const uschar *lcc; /* Points to lower casing table */ + const uschar *fcc; /* Points to case-flipping table */ + const uschar *cbits; /* Points to character type table */ + const uschar *ctypes; /* Points to table of type maps */ + const uschar *start_workspace;/* The start of working space */ + const uschar *start_code; /* The start of the compiled code */ + const uschar *start_pattern; /* The start of the pattern */ + const uschar *end_pattern; /* The end of the pattern */ + uschar *hwm; /* High watermark of workspace */ + uschar *name_table; /* The name/number table */ + int names_found; /* Number of entries so far */ + int name_entry_size; /* Size of each entry */ + int bracount; /* Count of capturing parens as we compile */ + int final_bracount; /* Saved value after first pass */ + int top_backref; /* Maximum back reference */ + unsigned int backref_map; /* Bitmap of low back refs */ + int external_options; /* External (initial) options */ + int external_flags; /* External flag bits to be set */ + int req_varyopt; /* "After variable item" flag for reqbyte */ + BOOL had_accept; /* (*ACCEPT) encountered */ + int nltype; /* Newline type */ + int nllen; /* Newline string length */ + uschar nl[4]; /* Newline string when fixed length */ +} compile_data; + +/* Structure for maintaining a chain of pointers to the currently incomplete +branches, for testing for left recursion. */ + +typedef struct branch_chain { + struct branch_chain *outer; + uschar *current; +} branch_chain; + +/* Structure for items in a linked list that represents an explicit recursive +call within the pattern. */ + +typedef struct recursion_info { + struct recursion_info *prevrec; /* Previous recursion record (or NULL) */ + int group_num; /* Number of group that was called */ + const uschar *after_call; /* "Return value": points after the call in the expr */ + USPTR save_start; /* Old value of mstart */ + int *offset_save; /* Pointer to start of saved offsets */ + int saved_max; /* Number of saved offsets */ +} recursion_info; + +/* Structure for building a chain of data for holding the values of the subject +pointer at the start of each subpattern, so as to detect when an empty string +has been matched by a subpattern - to break infinite loops. */ + +typedef struct eptrblock { + struct eptrblock *epb_prev; + USPTR epb_saved_eptr; +} eptrblock; + + +/* Structure for passing "static" information around between the functions +doing traditional NFA matching, so that they are thread-safe. */ + +typedef struct match_data { +#ifdef ERLANG_INTEGRATION + unsigned long int loop_limit; + void *state_save; +#endif + unsigned long int match_call_count; /* As it says */ + unsigned long int match_limit; /* As it says */ + unsigned long int match_limit_recursion; /* As it says */ + int *offset_vector; /* Offset vector */ + int offset_end; /* One past the end */ + int offset_max; /* The maximum usable for return data */ + int nltype; /* Newline type */ + int nllen; /* Newline string length */ + uschar nl[4]; /* Newline string when fixed */ + const uschar *lcc; /* Points to lower casing table */ + const uschar *ctypes; /* Points to table of type maps */ + BOOL offset_overflow; /* Set if too many extractions */ + BOOL notbol; /* NOTBOL flag */ + BOOL noteol; /* NOTEOL flag */ + BOOL utf8; /* UTF8 flag */ + BOOL endonly; /* Dollar not before final \n */ + BOOL notempty; /* Empty string match not wanted */ + BOOL partial; /* PARTIAL flag */ + BOOL hitend; /* Hit the end of the subject at some point */ + BOOL bsr_anycrlf; /* \R is just any CRLF, not full Unicode */ + const uschar *start_code; /* For use when recursing */ + USPTR start_subject; /* Start of the subject string */ + USPTR end_subject; /* End of the subject string */ + USPTR start_match_ptr; /* Start of matched string */ + USPTR end_match_ptr; /* Subject position at end match */ + int end_offset_top; /* Highwater mark at end of match */ + int capture_last; /* Most recent capture number */ + int start_offset; /* The start offset value */ + eptrblock *eptrchain; /* Chain of eptrblocks for tail recursions */ + int eptrn; /* Next free eptrblock */ + recursion_info *recursive; /* Linked list of recursion data */ + void *callout_data; /* To pass back to callouts */ +} match_data; + +/* A similar structure is used for the same purpose by the DFA matching +functions. */ + +typedef struct dfa_match_data { + const uschar *start_code; /* Start of the compiled pattern */ + const uschar *start_subject; /* Start of the subject string */ + const uschar *end_subject; /* End of subject string */ + const uschar *tables; /* Character tables */ + int moptions; /* Match options */ + int poptions; /* Pattern options */ + int nltype; /* Newline type */ + int nllen; /* Newline string length */ + uschar nl[4]; /* Newline string when fixed */ + void *callout_data; /* To pass back to callouts */ +} dfa_match_data; + +/* Bit definitions for entries in the pcre_ctypes table. */ + +#define ctype_space 0x01 +#define ctype_letter 0x02 +#define ctype_digit 0x04 +#define ctype_xdigit 0x08 +#define ctype_word 0x10 /* alphanumeric or '_' */ +#define ctype_meta 0x80 /* regexp meta char or zero (end pattern) */ + +/* Offsets for the bitmap tables in pcre_cbits. Each table contains a set +of bits for a class map. Some classes are built by combining these tables. */ + +#define cbit_space 0 /* [:space:] or \s */ +#define cbit_xdigit 32 /* [:xdigit:] */ +#define cbit_digit 64 /* [:digit:] or \d */ +#define cbit_upper 96 /* [:upper:] */ +#define cbit_lower 128 /* [:lower:] */ +#define cbit_word 160 /* [:word:] or \w */ +#define cbit_graph 192 /* [:graph:] */ +#define cbit_print 224 /* [:print:] */ +#define cbit_punct 256 /* [:punct:] */ +#define cbit_cntrl 288 /* [:cntrl:] */ +#define cbit_length 320 /* Length of the cbits table */ + +/* Offsets of the various tables from the base tables pointer, and +total length. */ + +#define lcc_offset 0 +#define fcc_offset 256 +#define cbits_offset 512 +#define ctypes_offset (cbits_offset + cbit_length) +#define tables_length (ctypes_offset + 256) + +/* Layout of the UCP type table that translates property names into types and +codes. Each entry used to point directly to a name, but to reduce the number of +relocations in shared libraries, it now has an offset into a single string +instead. */ + +typedef struct { + pcre_uint16 name_offset; + pcre_uint16 type; + pcre_uint16 value; +} ucp_type_table; + + +/* Internal shared data tables. These are tables that are used by more than one +of the exported public functions. They have to be "external" in the C sense, +but are not part of the PCRE public API. The data for these tables is in the +pcre_tables.c module. */ + +extern const int _erts_pcre_utf8_table1[]; +extern const int _erts_pcre_utf8_table2[]; +extern const int _erts_pcre_utf8_table3[]; +extern const uschar _erts_pcre_utf8_table4[]; + +extern const int _erts_pcre_utf8_table1_size; + +extern const char _erts_pcre_utt_names[]; +extern const ucp_type_table _erts_pcre_utt[]; +extern const int _erts_pcre_utt_size; + +extern const uschar _erts_pcre_default_tables[]; + +extern const uschar _erts_pcre_OP_lengths[]; + + +/* Internal shared functions. These are functions that are used by more than +one of the exported public functions. They have to be "external" in the C +sense, but are not part of the PCRE public API. */ + +extern BOOL _erts_pcre_is_newline(const uschar *, int, const uschar *, + int *, BOOL); +extern int _erts_pcre_ord2utf8(int, uschar *); +extern real_pcre *_erts_pcre_try_flipped(const real_pcre *, real_pcre *, + const pcre_study_data *, pcre_study_data *); +extern int _erts_pcre_ucp_findprop(const unsigned int, int *, int *); +extern unsigned int _erts_pcre_ucp_othercase(const unsigned int); +extern int _erts_pcre_valid_utf8(const uschar *, int); +extern BOOL _erts_pcre_was_newline(const uschar *, int, const uschar *, + int *, BOOL); +extern BOOL _erts_pcre_xclass(int, const uschar *); + +#endif + +/* End of pcre_internal.h */ |