Thanks to visit codestin.com
Credit goes to github.com

Skip to content

Commit 51b3aa3

Browse files
gvanrossumgvanrossum
committed
New "re" regular expression support.
This code is written by Philip Hazel and Andrew Kuchling. It requires a new "re.py" module, too.
1 parent 04ac894 commit 51b3aa3

4 files changed

Lines changed: 5135 additions & 0 deletions

File tree

‎Modules/pcre-internal.h‎

Lines changed: 226 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,226 @@
1+
/*************************************************
2+
* Perl-Compatible Regular Expressions *
3+
*************************************************/
4+
5+
6+
#define PCRE_VERSION "0.95 23-Sep-1997"
7+
8+
9+
/* This is a library of functions to support regular expressions whose syntax
10+
and semantics are as close as possible to those of the Perl 5 language. See
11+
the file Tech.Notes for some information on the internals.
12+
13+
Written by: Philip Hazel <[email protected]>
14+
15+
Copyright (c) 1997 University of Cambridge
16+
17+
-----------------------------------------------------------------------------
18+
Permission is granted to anyone to use this software for any purpose on any
19+
computer system, and to redistribute it freely, subject to the following
20+
restrictions:
21+
22+
1. This software is distributed in the hope that it will be useful,
23+
but WITHOUT ANY WARRANTY; without even the implied warranty of
24+
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
25+
26+
2. The origin of this software must not be misrepresented, either by
27+
explicit claim or by omission.
28+
29+
3. Altered versions must be plainly marked as such, and must not be
30+
misrepresented as being the original software.
31+
-----------------------------------------------------------------------------
32+
*/
33+
34+
/* This header contains definitions that are shared between the different
35+
modules, but which are not relevant to the outside. */
36+
37+
/* Standard C headers plus the external interface definition */
38+
39+
#include <ctype.h>
40+
#include <limits.h>
41+
#include <stdio.h>
42+
#include <stdlib.h>
43+
#include <string.h>
44+
#include "pcre.h"
45+
46+
/* Private options flags start at the most significant end of the byte. The
47+
public options defined in pcre.h start at the least significant end. Make sure
48+
they don't overlap! */
49+
50+
#define PCRE_FIRSTSET 0x80 /* first_char is set */
51+
#define PCRE_STARTLINE 0x40 /* start after \n for multiline */
52+
53+
/* Options for the "extra" block produced by pcre_study(). */
54+
55+
#define PCRE_STUDY_CASELESS 0x01 /* study was caseless */
56+
#define PCRE_STUDY_MAPPED 0x20 /* a map of starting chars exists */
57+
58+
/* Masks for identifying the public options: all permitted at compile time,
59+
only some permitted at run or study time. */
60+
61+
#ifdef FOR_PYTHON
62+
#define PUBLIC_OPTIONS \
63+
(PCRE_CASELESS|PCRE_EXTENDED|PCRE_ANCHORED|PCRE_MULTILINE|PCRE_DOTALL)
64+
#else
65+
#define PUBLIC_OPTIONS \
66+
(PCRE_CASELESS|PCRE_EXTENDED|PCRE_ANCHORED|PCRE_MULTILINE)
67+
#endif
68+
#define PUBLIC_EXEC_OPTIONS (PCRE_CASELESS|PCRE_ANCHORED|PCRE_MULTILINE)
69+
#define PUBLIC_STUDY_OPTIONS (PCRE_CASELESS)
70+
71+
/* Magic number to provide a small check against being handed junk. */
72+
73+
#define MAGIC_NUMBER 0x50435245 /* 'PCRE' */
74+
75+
/* Miscellaneous definitions */
76+
77+
typedef int BOOL;
78+
79+
#define FALSE 0
80+
#define TRUE 1
81+
82+
/* Flags for character classes - see also class_ops table below. */
83+
84+
#define CLASS_DIGITS 0x01
85+
#define CLASS_NOT_DIGITS 0x02
86+
#define CLASS_WHITESPACE 0x04
87+
#define CLASS_NOT_WHITESPACE 0x08
88+
#define CLASS_WORD 0x10
89+
#define CLASS_NOT_WORD 0x20
90+
91+
/* These are escaped items that aren't just an encoding of a particular data
92+
value such as \n. They must have non-zero values, as check_escape() returns
93+
their negation. Also, they must appear in the same order as in the opcode
94+
definitions below, up to ESC_Z. The final one must be ESC_REF as subsequent
95+
values are used for \1, \2, \3, etc. There is a test in the code for an escape
96+
greater than ESC_b and less than ESC_Z to detect the types that may be
97+
repeated. If any new escapes are put in-between that don't consume a character,
98+
that code will have to change. */
99+
100+
enum { ESC_A = 1, ESC_B, ESC_b, ESC_D, ESC_d, ESC_S, ESC_s, ESC_W, ESC_w,
101+
ESC_Z, ESC_REF };
102+
103+
/* Opcode table: OP_BRA must be last, as all values >= it are used for brackets
104+
that extract substrings. Starting from 1 (i.e. after OP_END), the values up to
105+
OP_EOL must correspond in order to the list of escapes immediately above. */
106+
107+
enum {
108+
OP_END, /* End of pattern */
109+
110+
/* Values corresponding to backslashed metacharacters */
111+
112+
OP_SOD, /* Start of data: \A */
113+
OP_NOT_WORD_BOUNDARY, /* \W */
114+
OP_WORD_BOUNDARY, /* \w */
115+
OP_NOT_DIGIT, /* \D */
116+
OP_DIGIT, /* \d */
117+
OP_NOT_WHITESPACE, /* \S */
118+
OP_WHITESPACE, /* \s */
119+
OP_NOT_WORDCHAR, /* \W */
120+
OP_WORDCHAR, /* \w */
121+
OP_EOD, /* End of data: or \Z. This must always be the last
122+
of the backslashed meta values. */
123+
124+
OP_CIRC, /* Start of line - varies with multiline switch */
125+
OP_DOLL, /* End of line - varies with multiline switch */
126+
OP_ANY, /* Match any character */
127+
OP_CHARS, /* Match string of characters */
128+
129+
OP_STAR, /* The maximizing and minimizing versions of */
130+
OP_MINSTAR, /* all these opcodes must come in pairs, with */
131+
OP_PLUS, /* the minimizing one second. */
132+
OP_MINPLUS, /* This first set applies to single characters */
133+
OP_QUERY,
134+
OP_MINQUERY,
135+
OP_UPTO, /* From 0 to n matches. */
136+
OP_MINUPTO,
137+
OP_EXACT, /* Exactly n matches. */
138+
139+
OP_TYPESTAR, /* The maximizing and minimizing versions of */
140+
OP_TYPEMINSTAR, /* all these opcodes must come in pairs, with */
141+
OP_TYPEPLUS, /* the minimizing one second. These codes must */
142+
OP_TYPEMINPLUS, /* be in exactly the same order as those above. */
143+
OP_TYPEQUERY, /* This set applies to character types such as \d */
144+
OP_TYPEMINQUERY,
145+
OP_TYPEUPTO,
146+
OP_TYPEMINUPTO,
147+
OP_TYPEEXACT,
148+
149+
OP_CRSTAR, /* The maximizing and minimizing versions of */
150+
OP_CRMINSTAR, /* all these opcodes must come in pairs, with */
151+
OP_CRPLUS, /* the minimizing one second. These codes must */
152+
OP_CRMINPLUS, /* be in exactly the same order as those above. */
153+
OP_CRQUERY, /* These are for character classes and back refs */
154+
OP_CRMINQUERY,
155+
OP_CRRANGE, /* These are different to the two seta above. */
156+
OP_CRMINRANGE,
157+
158+
OP_CLASS, /* Match a character class */
159+
OP_NEGCLASS, /* Don't match a character class */
160+
OP_REF, /* Match a back reference */
161+
162+
OP_ALT, /* Start of alternation */
163+
OP_KET, /* End of group that doesn't have an unbounded repeat */
164+
OP_KETRMAX, /* These two must remain together and in this */
165+
OP_KETRMIN, /* order. They are for groups the repeat for ever. */
166+
167+
OP_ASSERT,
168+
OP_ASSERT_NOT,
169+
170+
OP_BRAZERO, /* These two must remain together and in this */
171+
OP_BRAMINZERO, /* order. */
172+
173+
OP_BRA /* This and greater values are used for brackets that
174+
extract substrings. */
175+
};
176+
177+
/* The highest extraction number. This is limited by the number of opcodes
178+
left after OP_BRA, i.e. 255 - OP_BRA. We actually set it somewhat lower. */
179+
180+
#define EXTRACT_MAX 99
181+
182+
/* All character handling must be done as unsigned characters. Otherwise there
183+
are problems with top-bit-set characters and functions such as isspace().
184+
However, we leave the interface to the outside world as char *, because that
185+
should make things easier for callers. We define a short type for unsigned char
186+
to save lots of typing. I tried "uchar", but it causes problems on Digital
187+
Unix, where it is defined in sys/types, so use "uschar" instead. */
188+
189+
typedef unsigned char uschar;
190+
191+
/* The real format of the start of the pcre block; the actual code vector
192+
runs on as long as necessary after the end. */
193+
194+
typedef struct real_pcre {
195+
unsigned int magic_number;
196+
unsigned char options;
197+
unsigned char top_bracket;
198+
unsigned char first_char;
199+
unsigned char code[1];
200+
} real_pcre;
201+
202+
/* The real format of the extra block returned by pcre_study(). */
203+
204+
typedef struct real_pcre_extra {
205+
unsigned char options;
206+
unsigned char start_bits[32];
207+
} real_pcre_extra;
208+
209+
/* Global tables from pcre-chartables.c */
210+
211+
extern uschar pcre_lcc[];
212+
extern uschar pcre_ucc[];
213+
extern uschar pcre_ctypes[];
214+
215+
/* Bit definitions for entries in pcre_ctypes[]. */
216+
217+
#define ctype_space 0x01
218+
#define ctype_digit 0x02
219+
#define ctype_xdigit 0x04
220+
#define ctype_word 0x08 /* alphameric or '_' */
221+
#ifdef FOR_PYTHON
222+
#define ctype_odigit 0x10 /* Octal digits */
223+
#endif
224+
#define ctype_meta 0x80 /* regexp meta char or zero (end pattern) */
225+
226+
/* End of pcre-internal.h */

‎Modules/pcre.h‎

Lines changed: 59 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,59 @@
1+
/*************************************************
2+
* Perl-Compatible Regular Expressions *
3+
*************************************************/
4+
5+
/* Copyright (c) 1997 University of Cambridge */
6+
7+
/* Have to include stdlib.h in order to ensure that size_t is defined;
8+
it is needed in there for malloc. */
9+
10+
#ifndef PCRE_H
11+
#define PCRE_H
12+
13+
#include <stdlib.h>
14+
#ifdef FOR_PYTHON
15+
#include "Python.h"
16+
#endif
17+
18+
/* Options */
19+
20+
#define PCRE_CASELESS 0x01
21+
#define PCRE_EXTENDED 0x02
22+
#define PCRE_ANCHORED 0x04
23+
#define PCRE_MULTILINE 0x08
24+
#define PCRE_DOTALL 0x10
25+
26+
/* Exec-time error codes */
27+
28+
#define PCRE_ERROR_NOMATCH (-1)
29+
#define PCRE_ERROR_BADREF (-2)
30+
#define PCRE_ERROR_NULL (-3)
31+
#define PCRE_ERROR_BADOPTION (-4)
32+
#define PCRE_ERROR_BADMAGIC (-5)
33+
#define PCRE_ERROR_UNKNOWN_NODE (-6)
34+
35+
/* Types */
36+
37+
typedef void pcre;
38+
typedef void pcre_extra;
39+
40+
/* Store get and free functions. These can be set to alternative malloc/free
41+
functions if required. */
42+
43+
extern void *(*pcre_malloc)(size_t);
44+
extern void (*pcre_free)(void *);
45+
46+
/* Functions */
47+
48+
#ifdef FOR_PYTHON
49+
extern pcre *pcre_compile(char *, int, char **, int *, PyObject *);
50+
#else
51+
extern pcre *pcre_compile(char *, int, char **, int *);
52+
#endif
53+
extern int pcre_exec(pcre *, pcre_extra *, char *, int, int, int *, int);
54+
extern int pcre_info(pcre *, int *, int *);
55+
extern pcre_extra *pcre_study(pcre *, int, char **);
56+
extern char *pcre_version(void);
57+
58+
#endif /* ifndef PCRE_H */
59+
/* End of pcre.h */

0 commit comments

Comments
 (0)