Rooba
/
cpython
mirror of https://github.com/python/cpython.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

New pcre version from AMK

This commit is contained in:
Guido van Rossum 1997-12-08 17:15:20 +00:00
parent dfa6790bd6
commit 5070060d40
4 changed files with 1815 additions and 1227 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

146
Modules/pcre-internal.h
View File

@ -3,7 +3,7 @@
*************************************************/
#define PCRE_VERSION "0.95 23-Sep-1997"
#define PCRE_VERSION "1.01 19-Nov-1997"
/* This is a library of functions to support regular expressions whose syntax
@ -34,38 +34,54 @@ computer system, and to redistribute it freely, subject to the following
/* This header contains definitions that are shared between the different
modules, but which are not relevant to the outside. */
/* To cope with SunOS4 and other systems that lack memmove() but have bcopy(),
define a macro for memmove() if USE_BCOPY is defined. */
#ifdef USE_BCOPY
#define memmove(a, b, c) bcopy(b, a, c)
#endif
/* Standard C headers plus the external interface definition */
#include <ctype.h>
#include <limits.h>
#include <setjmp.h>
#include <stddef.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include "pcre.h"
/* Private options flags start at the most significant end of the byte. The
public options defined in pcre.h start at the least significant end. Make sure
they don't overlap! */
/* Private options flags start at the most significant end of the two bytes.
The public options defined in pcre.h start at the least significant end. Make
sure they don't overlap! */
#define PCRE_FIRSTSET 0x80 /* first_char is set */
#define PCRE_STARTLINE 0x40 /* start after \n for multiline */
#define PCRE_FIRSTSET 0x8000 /* first_char is set */
#define PCRE_STARTLINE 0x4000 /* start after \n for multiline */
#define PCRE_COMPILED_CASELESS 0x2000 /* like it says */
/* Options for the "extra" block produced by pcre_study(). */
#define PCRE_STUDY_CASELESS 0x01 /* study was caseless */
#define PCRE_STUDY_MAPPED 0x20 /* a map of starting chars exists */
#define PCRE_STUDY_MAPPED 0x02 /* a map of starting chars exists */
/* Masks for identifying the public options: all permitted at compile time,
only some permitted at run or study time. */
#ifdef FOR_PYTHON
#define PUBLIC_OPTIONS \
(PCRE_CASELESS|PCRE_EXTENDED|PCRE_ANCHORED|PCRE_MULTILINE|PCRE_DOTALL)
(PCRE_CASELESS|PCRE_EXTENDED|PCRE_ANCHORED|PCRE_MULTILINE| \
PCRE_DOTALL|PCRE_DOLLAR_ENDONLY|PCRE_EXTRA|PCRE_LOCALE)
#else
#define PUBLIC_OPTIONS \
(PCRE_CASELESS|PCRE_EXTENDED|PCRE_ANCHORED|PCRE_MULTILINE)
(PCRE_CASELESS|PCRE_EXTENDED|PCRE_ANCHORED|PCRE_MULTILINE| \
PCRE_DOTALL|PCRE_DOLLAR_ENDONLY|PCRE_EXTRA)
#endif
#define PUBLIC_EXEC_OPTIONS (PCRE_CASELESS|PCRE_ANCHORED|PCRE_MULTILINE)
#define PUBLIC_EXEC_OPTIONS \
(PCRE_CASELESS|PCRE_ANCHORED|PCRE_MULTILINE|PCRE_NOTBOL|PCRE_NOTEOL| \
PCRE_DOTALL|PCRE_DOLLAR_ENDONLY)
#define PUBLIC_STUDY_OPTIONS (PCRE_CASELESS)
/* Magic number to provide a small check against being handed junk. */
@ -79,26 +95,22 @@ typedef int BOOL;
#define FALSE 0
#define TRUE 1
/* Flags for character classes - see also class_ops table below. */
#define CLASS_DIGITS 0x01
#define CLASS_NOT_DIGITS 0x02
#define CLASS_WHITESPACE 0x04
#define CLASS_NOT_WHITESPACE 0x08
#define CLASS_WORD 0x10
#define CLASS_NOT_WORD 0x20
/* 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. The final one must be ESC_REF as subsequent
values are used for \1, \2, \3, etc. There is a test in the code for an escape
greater than ESC_b and less than ESC_Z to detect the types that may be
greater than ESC_b and less than ESC_X to detect the types that may be
repeated. 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_B, ESC_b, ESC_D, ESC_d, ESC_S, ESC_s, ESC_W, ESC_w,
ESC_Z, ESC_REF };
/* These are not Perl escapes, so can't appear in the */
ESC_X, /* simple table-lookup because they must be conditional */
/* on PCRE_EXTRA. */
ESC_Z,
ESC_REF };
/* Opcode table: OP_BRA must be last, as all values >= it are used for brackets
that extract substrings. Starting from 1 (i.e. after OP_END), the values up to
@ -110,21 +122,28 @@ enum {
/* Values corresponding to backslashed metacharacters */
OP_SOD, /* Start of data: \A */
OP_NOT_WORD_BOUNDARY, /* \W */
OP_WORD_BOUNDARY, /* \w */
OP_NOT_WORD_BOUNDARY, /* \B */
OP_WORD_BOUNDARY, /* \b */
OP_NOT_DIGIT, /* \D */
OP_DIGIT, /* \d */
OP_NOT_WHITESPACE, /* \S */
OP_WHITESPACE, /* \s */
OP_NOT_WORDCHAR, /* \W */
OP_WORDCHAR, /* \w */
OP_CUT, /* The analogue of Prolog's "cut" operation (extension) */
OP_EOD, /* End of data: or \Z. This must always be the last
of the backslashed meta values. */
OP_NOT_WORD_BOUNDARY_L, /* localized \B */
OP_WORD_BOUNDARY_L, /* localized \b */
OP_NOT_WORDCHAR_L, /* localized \W */
OP_WORDCHAR_L, /* localized \w */
OP_CIRC, /* Start of line - varies with multiline switch */
OP_DOLL, /* End of line - varies with multiline switch */
OP_ANY, /* Match any character */
OP_CHARS, /* Match string of characters */
OP_NOT, /* Match anything but the following char */
OP_STAR, /* The maximizing and minimizing versions of */
OP_MINSTAR, /* all these opcodes must come in pairs, with */
@ -132,9 +151,19 @@ enum {
OP_MINPLUS, /* This first set applies to single characters */
OP_QUERY,
OP_MINQUERY,
OP_UPTO, /* From 0 to n matches. */
OP_UPTO, /* From 0 to n matches */
OP_MINUPTO,
OP_EXACT, /* Exactly n matches. */
OP_EXACT, /* Exactly n matches */
OP_NOTSTAR, /* The maximizing and minimizing versions of */
OP_NOTMINSTAR, /* all these opcodes must come in pairs, with */
OP_NOTPLUS, /* the minimizing one second. */
OP_NOTMINPLUS, /* This first set applies to "not" single characters */
OP_NOTQUERY,
OP_NOTMINQUERY,
OP_NOTUPTO, /* From 0 to n matches */
OP_NOTMINUPTO,
OP_NOTEXACT, /* Exactly n matches */
OP_TYPESTAR, /* The maximizing and minimizing versions of */
OP_TYPEMINSTAR, /* all these opcodes must come in pairs, with */
@ -142,9 +171,9 @@ enum {
OP_TYPEMINPLUS, /* be in exactly the same order as those above. */
OP_TYPEQUERY, /* This set applies to character types such as \d */
OP_TYPEMINQUERY,
OP_TYPEUPTO,
OP_TYPEUPTO, /* From 0 to n matches */
OP_TYPEMINUPTO,
OP_TYPEEXACT,
OP_TYPEEXACT, /* Exactly n matches */
OP_CRSTAR, /* The maximizing and minimizing versions of */
OP_CRMINSTAR, /* all these opcodes must come in pairs, with */
@ -152,11 +181,11 @@ enum {
OP_CRMINPLUS, /* be in exactly the same order as those above. */
OP_CRQUERY, /* These are for character classes and back refs */
OP_CRMINQUERY,
OP_CRRANGE, /* These are different to the two seta above. */
OP_CRRANGE, /* These are different to the three seta above. */
OP_CRMINRANGE,
OP_CLASS, /* Match a character class */
OP_NEGCLASS, /* Don't match a character class */
OP_CLASS_L, /* Match a character class */
OP_REF, /* Match a back reference */
OP_ALT, /* Start of alternation */
@ -166,6 +195,7 @@ enum {
OP_ASSERT,
OP_ASSERT_NOT,
OP_ONCE, /* Once matched, don't back up into the subpattern */
OP_BRAZERO, /* These two must remain together and in this */
OP_BRAMINZERO, /* order. */
@ -179,6 +209,35 @@ left after OP_BRA, i.e. 255 - OP_BRA. We actually set it somewhat lower. */
#define EXTRACT_MAX 99
/* The texts of compile-time error messages are defined as macros here so that
they can be accessed by the POSIX wrapper and converted into error codes. Yes,
I could have used error codes in the first place, but didn't feel like changing
just to accommodate the POSIX wrapper. */
#define ERR1 "\\ at end of pattern"
#define ERR2 "\\c at end of pattern"
#define ERR3 "unrecognized character follows \\"
#define ERR4 "numbers out of order in {} quantifier"
#define ERR5 "number too big in {} quantifier"
#define ERR6 "missing terminating ] for character class"
#define ERR7 "invalid escape sequence in character class"
#define ERR8 "range out of order in character class"
#define ERR9 "nothing to repeat"
#define ERR10 "operand of unlimited repeat could match the empty string"
#define ERR11 "internal error: unexpected repeat"
#define ERR12 "unrecognized character after (?"
#define ERR13 "too many capturing parenthesized sub-patterns"
#define ERR14 "missing )"
#define ERR15 "back reference to non-existent subpattern"
#define ERR16 "erroffset passed as NULL"
#define ERR17 "unknown option bit(s) set"
#define ERR18 "missing ) after comment"
#define ERR19 "too many sets of parentheses"
#define ERR20 "regular expression too large"
#define ERR21 "failed to get memory"
#define ERR22 "unmatched brackets"
#define ERR23 "internal error: code overflow"
/* 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
@ -193,8 +252,9 @@ runs on as long as necessary after the end. */
typedef struct real_pcre {
unsigned int magic_number;
unsigned char options;
unsigned short int options;
unsigned char top_bracket;
unsigned char top_backref;
unsigned char first_char;
unsigned char code[1];
} real_pcre;
@ -206,21 +266,29 @@ typedef struct real_pcre_extra {
unsigned char start_bits[32];
} real_pcre_extra;
/* Global tables from pcre-chartables.c */
/* Global tables from chartables.c */
extern uschar pcre_lcc[];
extern uschar pcre_ucc[];
extern uschar pcre_fcc[];
extern uschar pcre_cbits[];
extern uschar pcre_ctypes[];
/* Bit definitions for entries in pcre_ctypes[]. */
#define ctype_space 0x01
#define ctype_digit 0x02
#define ctype_xdigit 0x04
#define ctype_word 0x08 /* alphameric or '_' */
#ifdef FOR_PYTHON
#define ctype_odigit 0x10 /* Octal digits */
#endif
#define ctype_letter 0x02
#define ctype_digit 0x04
#define ctype_xdigit 0x08
#define ctype_word 0x10 /* alphameric or '_' */
#define ctype_odigit 0x20 /* octal digit */
#define ctype_meta 0x80 /* regexp meta char or zero (end pattern) */
/* End of pcre-internal.h */
/* Offsets for the bitmap tables */
#define cbit_digit 0
#define cbit_letter 32
#define cbit_word 64
#define cbit_space 96
#define cbit_length 128 /* Length of the cbits table */
/* End of internal.h */

45
Modules/pcre.h
View File

@ -4,24 +4,32 @@
/* Copyright (c) 1997 University of Cambridge */
/* Have to include stdlib.h in order to ensure that size_t is defined;
it is needed in there for malloc. */
#ifndef _PCRE_H
#define _PCRE_H
#ifndef PCRE_H
#define PCRE_H
#include <stdlib.h>
#ifdef FOR_PYTHON
#include "Python.h"
#endif
/* Have to include stdlib.h in order to ensure that size_t is defined;
it is needed here for malloc. */
#include <stdlib.h>
/* Options */
#define PCRE_CASELESS 0x01
#define PCRE_EXTENDED 0x02
#define PCRE_ANCHORED 0x04
#define PCRE_MULTILINE 0x08
#define PCRE_DOTALL 0x10
#define PCRE_CASELESS 0x0001
#define PCRE_EXTENDED 0x0002
#define PCRE_ANCHORED 0x0004
#define PCRE_MULTILINE 0x0008
#define PCRE_DOTALL 0x0010
#define PCRE_DOLLAR_ENDONLY 0x0020
#define PCRE_EXTRA 0x0040
#define PCRE_NOTBOL 0x0080
#define PCRE_NOTEOL 0x0100
#ifdef FOR_PYTHON
#define PCRE_LOCALE 0x0200
#endif
/* Exec-time error codes */
@ -31,6 +39,7 @@ it is needed in there for malloc. */
#define PCRE_ERROR_BADOPTION (-4)
#define PCRE_ERROR_BADMAGIC (-5)
#define PCRE_ERROR_UNKNOWN_NODE (-6)
#define PCRE_ERROR_NOMEMORY (-7)
/* Types */
@ -46,14 +55,14 @@ extern void (*pcre_free)(void *);
/* Functions */
#ifdef FOR_PYTHON
extern pcre *pcre_compile(char *, int, char **, int *, PyObject *);
extern pcre *pcre_compile(const char *, int, char **, int *, PyObject *);
#else
extern pcre *pcre_compile(char *, int, char **, int *);
extern pcre *pcre_compile(const char *, int, char **, int *);
#endif
extern int pcre_exec(pcre *, pcre_extra *, char *, int, int, int *, int);
extern int pcre_info(pcre *, int *, int *);
extern pcre_extra *pcre_study(pcre *, int, char **);
extern int pcre_exec(const pcre *, const pcre_extra *, const char *,
int, int, int *, int);
extern int pcre_info(const pcre *, int *, int *);
extern pcre_extra *pcre_study(const pcre *, int, char **);
extern char *pcre_version(void);
#endif /* ifndef PCRE_H */
/* End of pcre.h */
#endif /* End of pcre.h */

814
Modules/pcremodule.c
View File

@ -105,39 +105,47 @@ PyPcre_exec(self, args)
PcreObject *self;
PyObject *args;
{
unsigned char *string;
int stringlen, pos=0, options=0, i, count;
int offsets[100*2]; /* XXX must this be fixed? */
char *string;
int stringlen, pos = 0, options=0, endpos = -1, i, count;
int offsets[100*2];
PyObject *list;
if (!PyArg_ParseTuple(args, "s#|ii", &string, &stringlen, &pos, &options))
if (!PyArg_ParseTuple(args, "s#|iiii", &string, &stringlen, &pos, &endpos, &options))
return NULL;
if (endpos == -1) {endpos = stringlen;}
count = pcre_exec(self->regex, self->regex_extra,
string+pos, stringlen-pos, options,
(char *)string+pos, endpos - pos, options,
offsets, sizeof(offsets)/sizeof(int) );
/* If an error occurred during the match, and an exception was raised,
just return NULL and leave the exception alone. The most likely
problem to cause this would be running out of memory for
the failure stack. */
if (PyErr_Occurred())
{
return NULL;
}
if (count==PCRE_ERROR_NOMATCH) {Py_INCREF(Py_None); return Py_None;}
if (count<0)
{
PyErr_SetObject(ErrorObject, Py_BuildValue("si", "Regex error", count));
return NULL;
}
{
PyErr_SetObject(ErrorObject, Py_BuildValue("si", "Regex error", count));
return NULL;
}
list=PyList_New(self->num_groups+1);
if (list==NULL) return NULL;
/* XXX count can be >size_offset! */
for(i=0; i<=self->num_groups; i++)
{
PyObject *v;
int start=offsets[i*2], end=offsets[i*2+1];
/* If the group wasn't affected by the match, return -1, -1 */
if (start<0 || count<=i)
{start=end=-1;}
else
{start += pos; end +=pos;}
v=Py_BuildValue("ii", start, end);
if (v==NULL) {Py_DECREF(list); return NULL;}
PyList_SetItem(list, i, v);
}
{
PyObject *v;
int start=offsets[i*2], end=offsets[i*2+1];
/* If the group wasn't affected by the match, return -1, -1 */
if (start<0 || count<=i)
{start=end=-1;}
else
{start += pos; end +=pos;}
v=Py_BuildValue("ii", start, end);
if (v==NULL) {Py_DECREF(list); return NULL;}
PyList_SetItem(list, i, v);
}
return list;
}
@ -182,7 +190,7 @@ PyPcre_compile(self, args)
{
PcreObject *rv;
PyObject *dictionary;
unsigned char *pattern, *newpattern;
char *pattern, *newpattern;
char *error;
int num_zeros, i, j;
@ -192,282 +200,274 @@ PyPcre_compile(self, args)
return NULL;
rv = newPcreObject(args);
if ( rv == NULL )
return NULL;
return NULL;
/* PCRE doesn't like having null bytes in its pattern, so we have to replace
any zeros in the string with the characters '\0'. */
num_zeros=1;
for(i=0; i<patternlen; i++) {
if (pattern[i]==0) num_zeros++;
if (pattern[i]==0) num_zeros++;
}
newpattern=malloc(patternlen+num_zeros);
if (newpattern==NULL) {
PyErr_SetString(PyExc_MemoryError, "can't allocate memory for new pattern");
return NULL;
PyErr_SetString(PyExc_MemoryError, "can't allocate memory for new pattern");
return NULL;
}
for (i=j=0; i<patternlen; i++, j++)
{
if (pattern[i]!=0) newpattern[j]=pattern[i];
else {
newpattern[j++]='\\';
newpattern[j] ='0';
}
}
{
if (pattern[i]!=0) newpattern[j]=pattern[i];
else {
newpattern[j++]='\\';
newpattern[j] ='0';
}
}
newpattern[j]='\0';
rv->regex = pcre_compile(newpattern, options,
rv->regex = pcre_compile((char*)newpattern, options,
&error, &erroroffset, dictionary);
free(newpattern);
if (rv->regex==NULL)
{
PyMem_DEL(rv);
if (!PyErr_Occurred())
{
PyErr_SetObject(ErrorObject, Py_BuildValue("si", error, erroroffset));
}
return NULL;
}
{
PyMem_DEL(rv);
if (!PyErr_Occurred())
{
PyErr_SetObject(ErrorObject, Py_BuildValue("si", error, erroroffset));
}
return NULL;
}
rv->regex_extra=pcre_study(rv->regex, 0, &error);
if (rv->regex_extra==NULL && error!=NULL)
{
PyMem_DEL(rv);
PyErr_SetObject(ErrorObject, Py_BuildValue("si", error, 0));
return NULL;
}
{
PyMem_DEL(rv);
PyErr_SetObject(ErrorObject, Py_BuildValue("si", error, 0));
return NULL;
}
rv->num_groups = pcre_info(rv->regex, NULL, NULL);
if (rv->num_groups<0)
{
PyErr_SetObject(ErrorObject, Py_BuildValue("si", "Regex error", rv->num_groups));
PyMem_DEL(rv);
return NULL;
}
{
PyErr_SetObject(ErrorObject, Py_BuildValue("si", "Regex error", rv->num_groups));
PyMem_DEL(rv);
return NULL;
}
return (PyObject *)rv;
}
static PyObject *
PyPcre_expand_escape(pattern, pattern_len, indexptr, typeptr)
unsigned char *pattern;
int pattern_len, *indexptr, *typeptr;
unsigned char *pattern;
int pattern_len, *indexptr, *typeptr;
{
unsigned char c;
int index = *indexptr;
unsigned char c;
int index = *indexptr;
if (pattern_len<=index)
{
PyErr_SetString(ErrorObject, "escape ends too soon");
return NULL;
}
c=pattern[index]; index++;
*typeptr=CHAR;
switch (c)
{
case('t'):
*indexptr=index;
return Py_BuildValue("c", (char)9);
break;
case('n'):
*indexptr = index;
return Py_BuildValue("c", (char)10);
break;
case('v'):
*indexptr = index;
return Py_BuildValue("c", (char)11);
break;
case('r'):
*indexptr = index;
return Py_BuildValue("c", (char)13);
break;
case('f'):
*indexptr = index;
return Py_BuildValue("c", (char)12);
break;
case('a'):
*indexptr = index;
return Py_BuildValue("c", (char)7);
break;
case('b'):
*indexptr=index;
return Py_BuildValue("c", (char)8);
break;
case('x'):
{
int end, length;
unsigned char *string;
PyObject *v;
end=index;
while (end<pattern_len &&
( pcre_ctypes[ pattern[end] ] & ctype_xdigit ) )
end++;
if (end==index)
{
PyErr_SetString(ErrorObject, "\\x must be followed by hex digits");
return NULL;
}
length=end-index;
string=malloc(length+4+1);
if (string==NULL)
{
PyErr_SetString(PyExc_MemoryError, "can't allocate memory for \\x string");
return NULL;
}
/* Create a string containing "\x<hexdigits>", which will be
passed to eval() */
string[0]=string[length+3]='"';
string[1]='\\';
string[length+4]='\0';
memcpy(string+2, pattern+index-1, length+1);
v=PyRun_String((char *)string, Py_eval_input,
PyEval_GetGlobals(), PyEval_GetLocals());
free(string);
/* The evaluation raised an exception */
if (v==NULL) return NULL;
*indexptr = end;
return v;
}
break;
case('E'): case('G'): case('L'): case('Q'):
case('U'): case('l'): case('u'):
{
char message[50];
sprintf(message, "\\%c is not allowed", c);
PyErr_SetString(ErrorObject, message);
return NULL;
}
case('g'):
{
int end, valid, i;
if (pattern_len<=index)
{
PyErr_SetString(ErrorObject, "unfinished symbolic reference");
return NULL;
}
if (pattern[index]!='<')
{
PyErr_SetString(ErrorObject, "missing < in symbolic reference");
return NULL;
}
index++;
end=index;
while (end<pattern_len && pattern[end]!='>')
end++;
if (end==pattern_len)
{
PyErr_SetString(ErrorObject, "unfinished symbolic reference");
return NULL;
}
valid=1;
if (index==end /* Zero-length name */
|| !(pcre_ctypes[pattern[index]] & ctype_word) /* First char. not alphanumeric */
|| (pcre_ctypes[pattern[index]] & ctype_digit) ) /* First char. a digit */
valid=0;
{
PyErr_SetString(ErrorObject, "escape ends too soon");
return NULL;
}
c=pattern[index]; index++;
*typeptr=CHAR;
for(i=index+1; i<end; i++)
{
if (!(pcre_ctypes[pattern[i]] & ctype_word) )
valid=0;
}
if (!valid)
{
/* XXX should include the text of the reference */
PyErr_SetString(ErrorObject, "illegal symbolic reference");
return NULL;
}
*typeptr = MEMORY_REFERENCE;
*indexptr = end+1;
return Py_BuildValue("s#", pattern+index, end-index);
}
break;
switch (c)
{
case('t'):
*indexptr=index;
return Py_BuildValue("c", (char)9);
break;
case('n'):
*indexptr = index;
return Py_BuildValue("c", (char)10);
break;
case('v'):
*indexptr = index;
return Py_BuildValue("c", (char)11);
break;
case('r'):
*indexptr = index;
return Py_BuildValue("c", (char)13);
break;
case('f'):
*indexptr = index;
return Py_BuildValue("c", (char)12);
break;
case('a'):
*indexptr = index;
return Py_BuildValue("c", (char)7);
break;
case('b'):
*indexptr=index;
return Py_BuildValue("c", (char)8);
break;
case('0'):
{
/* \0 always indicates an octal escape, so we consume up to 3
characters, as long as they're all octal digits */
int octval=0, i;
index--;
for(i=index;
i<=index+2 && i<pattern_len
&& (pcre_ctypes[ pattern[i] ] & ctype_odigit );
i++)
{
octval = octval * 8 + pattern[i] - '0';
}
if (octval>255)
{
PyErr_SetString(ErrorObject, "octal value out of range");
return NULL;
}
*indexptr = i;
return Py_BuildValue("c", (unsigned char)octval);
}
break;
case('1'): case('2'): case('3'): case('4'):
case('5'): case('6'): case('7'): case('8'):
case('9'):
{
/* Handle \?, where ? is from 1 through 9 */
int value=0;
index--;
/* If it's at least a two-digit reference, like \34, it might
either be a 3-digit octal escape (\123) or a 2-digit
decimal memory reference (\34) */
case('x'):
{
int x, ch, end;
if ( (index+1) <pattern_len &&
(pcre_ctypes[ pattern[index+1] ] & ctype_digit) )
{
if ( (index+2) <pattern_len &&
(pcre_ctypes[ pattern[index+2] ] & ctype_odigit) &&
(pcre_ctypes[ pattern[index+1] ] & ctype_odigit) &&
(pcre_ctypes[ pattern[index ] ] & ctype_odigit)
)
{
/* 3 octal digits */
value= 8*8*(pattern[index ]-'0') +
8*(pattern[index+1]-'0') +
(pattern[index+2]-'0');
if (value>255)
{
PyErr_SetString(ErrorObject, "octal value out of range");
return NULL;
}
*indexptr = index+3;
return Py_BuildValue("c", (unsigned char)value);
}
else
{
/* 2-digit form, so it's a memory reference */
value= 10*(pattern[index ]-'0') +
(pattern[index+1]-'0');
if (value<1 || EXTRACT_MAX<=value)
{
PyErr_SetString(ErrorObject, "memory reference out of range");
return NULL;
}
*typeptr = MEMORY_REFERENCE;
*indexptr = index+2;
return Py_BuildValue("i", value);
}
}
else
{
/* Single-digit form, like \2, so it's a memory reference */
*typeptr = MEMORY_REFERENCE;
*indexptr = index+1;
return Py_BuildValue("i", pattern[index]-'0');
}
}
break;
default:
*indexptr = index;
return Py_BuildValue("c", c);
x = 0; end = index;
while ( (end<pattern_len && pcre_ctypes[ pattern[end] ] & ctype_xdigit) != 0)
{
ch = pattern[end];
x = x * 16 + pcre_lcc[ch] -
(((pcre_ctypes[ch] & ctype_digit) != 0)? '0' : 'W');
x &= 255;
end++;
}
if (end==index)
{
PyErr_SetString(ErrorObject, "\\x must be followed by hex digits");
return NULL;
}
*indexptr = end;
return Py_BuildValue("c", (char)x);
}
break;
}
case('E'): case('G'): case('L'): case('Q'):
case('U'): case('l'): case('u'):
{
char message[50];
sprintf(message, "\\%c is not allowed", c);
PyErr_SetString(ErrorObject, message);
return NULL;
}
case('g'):
{
int end, i;
if (pattern_len<=index)
{
PyErr_SetString(ErrorObject, "unfinished symbolic reference");
return NULL;
}
if (pattern[index]!='<')
{
PyErr_SetString(ErrorObject, "missing < in symbolic reference");
return NULL;
}
index++;
end=index;
while (end<pattern_len && pattern[end]!='>')
end++;
if (end==pattern_len)
{
PyErr_SetString(ErrorObject, "unfinished symbolic reference");
return NULL;
}
if (index==end) /* Zero-length name */
{
/* XXX should include the text of the reference */
PyErr_SetString(ErrorObject, "zero-length symbolic reference");
return NULL;
}
if (!(pcre_ctypes[pattern[index]] & ctype_word) /* First char. not alphanumeric */
|| (pcre_ctypes[pattern[index]] & ctype_digit) ) /* First char. a digit */
{
/* XXX should include the text of the reference */
PyErr_SetString(ErrorObject, "first character of symbolic reference not a letter or _");
return NULL;
}
for(i=index+1; i<end; i++)
{
if (!(pcre_ctypes[pattern[i]] & ctype_word) )
{
/* XXX should include the text of the reference */
PyErr_SetString(ErrorObject, "illegal symbolic reference");
return NULL;
}
}
*typeptr = MEMORY_REFERENCE;
*indexptr = end+1;
return Py_BuildValue("s#", pattern+index, end-index);
}
break;
case('0'):
{
/* \0 always indicates an octal escape, so we consume up to 3
characters, as long as they're all octal digits */
int octval=0, i;
index--;
for(i=index;
i<=index+2 && i<pattern_len
&& (pcre_ctypes[ pattern[i] ] & ctype_odigit );
i++)
{
octval = octval * 8 + pattern[i] - '0';
}
if (octval>255)
{
PyErr_SetString(ErrorObject, "octal value out of range");
return NULL;
}
*indexptr = i;
return Py_BuildValue("c", (unsigned char)octval);
}
break;
case('1'): case('2'): case('3'): case('4'):
case('5'): case('6'): case('7'): case('8'):
case('9'):
{
/* Handle \?, where ? is from 1 through 9 */
int value=0;
index--;
/* If it's at least a two-digit reference, like \34, it might
either be a 3-digit octal escape (\123) or a 2-digit
decimal memory reference (\34) */
if ( (index+1) <pattern_len &&
(pcre_ctypes[ pattern[index+1] ] & ctype_digit) )
{
if ( (index+2) <pattern_len &&
(pcre_ctypes[ pattern[index+2] ] & ctype_odigit) &&
(pcre_ctypes[ pattern[index+1] ] & ctype_odigit) &&
(pcre_ctypes[ pattern[index ] ] & ctype_odigit)
)
{
/* 3 octal digits */
value= 8*8*(pattern[index ]-'0') +
8*(pattern[index+1]-'0') +
(pattern[index+2]-'0');
if (value>255)
{
PyErr_SetString(ErrorObject, "octal value out of range");
return NULL;
}
*indexptr = index+3;
return Py_BuildValue("c", (unsigned char)value);
}
else
{
/* 2-digit form, so it's a memory reference */
value= 10*(pattern[index ]-'0') +
(pattern[index+1]-'0');
if (value<1 || EXTRACT_MAX<=value)
{
PyErr_SetString(ErrorObject, "memory reference out of range");
return NULL;
}
*typeptr = MEMORY_REFERENCE;
*indexptr = index+2;
return Py_BuildValue("i", value);
}
}
else
{
/* Single-digit form, like \2, so it's a memory reference */
*typeptr = MEMORY_REFERENCE;
*indexptr = index+1;
return Py_BuildValue("i", pattern[index]-'0');
}
}
break;
default:
*indexptr = index;
return Py_BuildValue("c", c);
break;
}
}
static PyObject *
@ -475,117 +475,127 @@ PyPcre_expand(self, args)
PyObject *self;
PyObject *args;
{
PyObject *results, *match_obj;
PyObject *repl_obj, *newstring;
unsigned char *repl;
int size, total_len, i, start, pos;
PyObject *results, *match_obj;
PyObject *repl_obj, *newstring;
unsigned char *repl;
int size, total_len, i, start, pos;
if (!PyArg_ParseTuple(args, "OS", &match_obj, &repl_obj))
return NULL;
if (!PyArg_ParseTuple(args, "OS", &match_obj, &repl_obj))
return NULL;
repl=(unsigned char *)PyString_AsString(repl_obj);
size=PyString_Size(repl_obj);
results=PyList_New(0);
if (results==NULL) return NULL;
for(start=total_len=i=0; i<size; i++)
{
if (repl[i]=='\\')
repl=(unsigned char *)PyString_AsString(repl_obj);
size=PyString_Size(repl_obj);
results=PyList_New(0);
if (results==NULL) return NULL;
for(start=total_len=i=0; i<size; i++)
{
PyObject *value;
int escape_type;
if (repl[i]=='\\')
{
PyObject *value;
int escape_type;
if (start!=i)
{
PyList_Append(results,
PyString_FromStringAndSize((char *)repl+start, i-start));
total_len += i-start;
}
i++;
value=PyPcre_expand_escape(repl, size, &i, &escape_type);
if (value==NULL)
{
/* PyPcre_expand_escape triggered an exception of some sort,
so just return */
Py_DECREF(results);
return NULL;
}
switch (escape_type)
{
case (CHAR):
PyList_Append(results, value);
total_len += PyString_Size(value);
break;
case(MEMORY_REFERENCE):
{
PyObject *r, *tuple, *result;
r=PyObject_GetAttrString(match_obj, "group");
tuple=PyTuple_New(1);
Py_INCREF(value);
PyTuple_SetItem(tuple, 0, value);
result=PyEval_CallObject(r, tuple);
Py_DECREF(r); Py_DECREF(tuple);
if (result==NULL)
{
/* The group() method trigged an exception of some sort */
Py_DECREF(results);
return NULL;
}
if (result==Py_None)
{
char message[50];
sprintf(message,
"group %li did not contribute to the match",
PyInt_AsLong(value));
PyErr_SetString(ErrorObject,
message);
Py_DECREF(result);
Py_DECREF(results);
return NULL;
}
/* XXX typecheck that it's a string! */
PyList_Append(results, result);
total_len += PyString_Size(result);
Py_DECREF(result);
}
break;
default:
Py_DECREF(results);
PyErr_SetString(ErrorObject,
"bad escape in replacement");
return NULL;
}
start=i;
i--; /* Decrement now, because the 'for' loop will increment it */
if (start!=i)
{
PyList_Append(results,
PyString_FromStringAndSize((char *)repl+start, i-start));
total_len += i-start;
}
i++;
value=PyPcre_expand_escape(repl, size, &i, &escape_type);
if (value==NULL)
{
/* PyPcre_expand_escape triggered an exception of some sort,
so just return */
Py_DECREF(results);
return NULL;
}
switch (escape_type)
{
case (CHAR):
PyList_Append(results, value);
total_len += PyString_Size(value);
break;
case(MEMORY_REFERENCE):
{
PyObject *r, *tuple, *result;
r=PyObject_GetAttrString(match_obj, "group");
tuple=PyTuple_New(1);
Py_INCREF(value);
PyTuple_SetItem(tuple, 0, value);
result=PyEval_CallObject(r, tuple);
Py_DECREF(r); Py_DECREF(tuple);
if (result==NULL)
{
/* The group() method trigged an exception of some sort */
Py_DECREF(results);
Py_DECREF(value);
return NULL;
}
if (result==Py_None)
{
char message[50];
sprintf(message,
"group did not contribute to the match");
PyErr_SetString(ErrorObject,
message);
Py_DECREF(result);
Py_DECREF(value);
Py_DECREF(results);
return NULL;
}
/* typecheck that it's a string! */
if (!PyString_Check(result))
{
Py_DECREF(results);
Py_DECREF(result);
PyErr_SetString(ErrorObject,
"group() must return a string value for replacement");
return NULL;
}
PyList_Append(results, result);
total_len += PyString_Size(result);
Py_DECREF(result);
}
break;
default:
Py_DECREF(results);
PyErr_SetString(ErrorObject,
"bad escape in replacement");
return NULL;
}
Py_DECREF(value);
start=i;
i--; /* Decrement now, because the 'for' loop will increment it */
}
} /* endif repl[i]!='\\' */
if (start!=i)
{
PyList_Append(results, PyString_FromStringAndSize((char *)repl+start, i-start));
total_len += i-start;
}
} /* endif repl[i]!='\\' */
if (start!=i)
{
PyList_Append(results, PyString_FromStringAndSize((char *)repl+start, i-start));
total_len += i-start;
}
/* Whew! Now we've constructed a list containing various pieces of
strings that will make up our final result. So, iterate over
the list concatenating them. A new string measuring total_len
bytes is allocated and filled in. */
/* Whew! Now we've constructed a list containing various pieces of
strings that will make up our final result. So, iterate over
the list concatenating them. A new string measuring total_len
bytes is allocated and filled in. */
newstring=PyString_FromStringAndSize(NULL, total_len);
if (newstring==NULL)
{
Py_DECREF(results);
return NULL;
}
newstring=PyString_FromStringAndSize(NULL, total_len);
if (newstring==NULL)
{
Py_DECREF(results);
return NULL;
}
repl=(unsigned char *)PyString_AsString(newstring);
for (pos=i=0; i<PyList_Size(results); i++)
{
PyObject *item=PyList_GetItem(results, i);
memcpy(repl+pos, PyString_AsString(item), PyString_Size(item) );
pos += PyString_Size(item);
}
Py_DECREF(results);
return newstring;
repl=(unsigned char *)PyString_AsString(newstring);
for (pos=i=0; i<PyList_Size(results); i++)
{
PyObject *item=PyList_GetItem(results, i);
memcpy(repl+pos, PyString_AsString(item), PyString_Size(item) );
pos += PyString_Size(item);
}
Py_DECREF(results);
return newstring;
}
@ -642,67 +652,7 @@ initpcre()
insint(d, "MULTILINE", PCRE_MULTILINE);
insint(d, "DOTALL", PCRE_DOTALL);
insint(d, "VERBOSE", PCRE_EXTENDED);
/* Insert the opcodes */
insint(d, "OP_END", OP_END);
insint(d, "OP_SOD", OP_SOD);
insint(d, "OP_NOT_WORD_BOUNDARY", OP_NOT_WORD_BOUNDARY);
insint(d, "OP_WORD_BOUNDARY", OP_WORD_BOUNDARY);
insint(d, "OP_NOT_DIGIT", OP_NOT_DIGIT);
insint(d, "OP_NOT_WHITESPACE", OP_NOT_WHITESPACE);
insint(d, "OP_WHITESPACE", OP_WHITESPACE);
insint(d, "OP_NOT_WORDCHAR", OP_NOT_WORDCHAR);
insint(d, "OP_WORDCHAR", OP_WORDCHAR);
insint(d, "OP_EOD", OP_EOD);
insint(d, "OP_CIRC", OP_CIRC);
insint(d, "OP_DOLL", OP_DOLL);
insint(d, "OP_ANY", OP_ANY);
insint(d, "OP_CHARS", OP_CHARS);
insint(d, "OP_STAR", OP_STAR);
insint(d, "OP_MINSTAR", OP_MINSTAR);
insint(d, "OP_PLUS", OP_PLUS);
insint(d, "OP_MINPLUS", OP_MINPLUS);
insint(d, "OP_QUERY", OP_QUERY);
insint(d, "OP_MINQUERY", OP_MINQUERY);
insint(d, "OP_UPTO", OP_UPTO);
insint(d, "OP_MINUPTO", OP_MINUPTO);
insint(d, "OP_EXACT", OP_EXACT);
insint(d, "OP_TYPESTAR", OP_TYPESTAR);
insint(d, "OP_TYPEMINSTAR", OP_TYPEMINSTAR);
insint(d, "OP_TYPEPLUS", OP_TYPEPLUS);
insint(d, "OP_TYPEMINPLUS", OP_TYPEMINPLUS);
insint(d, "OP_TYPEQUERY", OP_TYPEQUERY);
insint(d, "OP_TYPEMINQUERY", OP_TYPEMINQUERY);
insint(d, "OP_TYPEUPTO", OP_TYPEUPTO);
insint(d, "OP_TYPEMINUPTO", OP_TYPEMINUPTO);
insint(d, "OP_TYPEEXACT", OP_TYPEEXACT);
insint(d, "OP_CRSTAR", OP_CRSTAR);
insint(d, "OP_CRMINSTAR", OP_CRMINSTAR);
insint(d, "OP_CRPLUS", OP_CRPLUS);
insint(d, "OP_CRMINPLUS", OP_CRMINPLUS);
insint(d, "OP_CRQUERY", OP_CRQUERY);
insint(d, "OP_CRMINQUERY", OP_CRMINQUERY);
insint(d, "OP_CRRANGE", OP_CRRANGE);
insint(d, "OP_CRMINRANGE", OP_CRMINRANGE);
insint(d, "OP_CLASS", OP_CLASS);
insint(d, "OP_NEGCLASS", OP_NEGCLASS);
insint(d, "OP_REF", OP_REF);
insint(d, "OP_ALT", OP_ALT);
insint(d, "OP_KET", OP_KET);
insint(d, "OP_KETRMAX", OP_KETRMAX);
insint(d, "OP_KETRMIN", OP_KETRMIN);
insint(d, "OP_ASSERT", OP_ASSERT);
insint(d, "OP_ASSERT_NOT", OP_ASSERT_NOT);
insint(d, "OP_BRAZERO", OP_BRAZERO);
insint(d, "OP_BRAMINZERO", OP_BRAMINZERO);
insint(d, "OP_BRA", OP_BRA);
insint(d, "LOCALE", PCRE_LOCALE);
/* Check for errors */
if (PyErr_Occurred())

2037
Modules/pypcre.c
View File

File diff suppressed because it is too large Load Diff