tags/pcre-4.5/maketables.c

/*************************************************
*      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 <ph10@cam.ac.uk>

           Copyright (c) 1997-2003 University of Cambridge

-----------------------------------------------------------------------------
Permission is granted to anyone to use this software for any purpose on any
computer system, and to redistribute it freely, subject to the following
restrictions:

1. This software is distributed in the hope that it will be useful,
   but WITHOUT ANY WARRANTY; without even the implied warranty of
   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.

2. The origin of this software must not be misrepresented, either by
   explicit claim or by omission.

3. Altered versions must be plainly marked as such, and must not be
   misrepresented as being the original software.

4. If PCRE is embedded in any software that is released under the GNU
   General Purpose Licence (GPL), then the terms of that licence shall
   supersede any condition above with which it is incompatible.
-----------------------------------------------------------------------------

See the file Tech.Notes for some information on the internals.
*/


/* This file is compiled on its own as part of the PCRE library. However,
it is also included in the compilation of dftables.c, in which case the macro
DFTABLES is defined. */

#ifndef DFTABLES
#include "internal.h"
#endif


/*************************************************
*           Create PCRE character tables         *
*************************************************/

/* This function builds a set of character tables for use by PCRE and returns
a pointer to them. They are build using the ctype functions, and consequently
their contents will depend upon the current locale setting. When compiled as
part of the library, the store is obtained via pcre_malloc(), but when compiled
inside dftables, use malloc().

Arguments:   none
Returns:     pointer to the contiguous block of data
*/

const unsigned char *
pcre_maketables(void)
{
unsigned char *yield, *p;
int i;

#ifndef DFTABLES
yield = (unsigned char*)(pcre_malloc)(tables_length);
#else
yield = (unsigned char*)malloc(tables_length);
#endif

if (yield == NULL) return NULL;
p = yield;

/* First comes the lower casing table */

for (i = 0; i < 256; i++) *p++ = tolower(i);

/* Next the case-flipping table */

for (i = 0; i < 256; i++) *p++ = islower(i)? toupper(i) : tolower(i);

/* Then the character class tables. Don't try to be clever and save effort
on exclusive ones - in some locales things may be different. Note that the
table for "space" includes everything "isspace" gives, including VT in the
default locale. This makes it work for the POSIX class [:space:]. */

memset(p, 0, cbit_length);
for (i = 0; i < 256; i++)
  {
  if (isdigit(i))
    {
    p[cbit_digit  + i/8] |= 1 << (i&7);
    p[cbit_word   + i/8] |= 1 << (i&7);
    }
  if (isupper(i))
    {
    p[cbit_upper  + i/8] |= 1 << (i&7);
    p[cbit_word   + i/8] |= 1 << (i&7);
    }
  if (islower(i))
    {
    p[cbit_lower  + i/8] |= 1 << (i&7);
    p[cbit_word   + i/8] |= 1 << (i&7);
    }
  if (i == '_')   p[cbit_word   + i/8] |= 1 << (i&7);
  if (isspace(i)) p[cbit_space  + i/8] |= 1 << (i&7);
  if (isxdigit(i))p[cbit_xdigit + i/8] |= 1 << (i&7);
  if (isgraph(i)) p[cbit_graph  + i/8] |= 1 << (i&7);
  if (isprint(i)) p[cbit_print  + i/8] |= 1 << (i&7);
  if (ispunct(i)) p[cbit_punct  + i/8] |= 1 << (i&7);
  if (iscntrl(i)) p[cbit_cntrl  + i/8] |= 1 << (i&7);
  }
p += cbit_length;

/* Finally, the character type table. In this, we exclude VT from the white
space chars, because Perl doesn't recognize it as such for \s and for comments
within regexes. */

for (i = 0; i < 256; i++)
  {
  int x = 0;
  if (i != 0x0b && isspace(i)) x += ctype_space;
  if (isalpha(i)) x += ctype_letter;
  if (isdigit(i)) x += ctype_digit;
  if (isxdigit(i)) x += ctype_xdigit;
  if (isalnum(i) || i == '_') x += ctype_word;

  /* Note: strchr includes the terminating zero in the characters it considers.
  In this instance, that is ok because we want binary zero to be flagged as a
  meta-character, which in this sense is any character that terminates a run
  of data characters. */

  if (strchr("*+?{^.$|()[", i) != 0) x += ctype_meta; *p++ = x; }

return yield;
}

/* End of maketables.c */
1	nigel	3	/*************************************************
2			* Perl-Compatible Regular Expressions *
3			*************************************************/
4
5			/*
6			PCRE is a library of functions to support regular expressions whose syntax
7			and semantics are as close as possible to those of the Perl 5 language.
8
9			Written by: Philip Hazel <ph10@cam.ac.uk>
10
11	nigel	63	Copyright (c) 1997-2003 University of Cambridge
12	nigel	3
13			-----------------------------------------------------------------------------
14			Permission is granted to anyone to use this software for any purpose on any
15			computer system, and to redistribute it freely, subject to the following
16			restrictions:
17
18			1. This software is distributed in the hope that it will be useful,
19			but WITHOUT ANY WARRANTY; without even the implied warranty of
20			MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
21
22			2. The origin of this software must not be misrepresented, either by
23			explicit claim or by omission.
24
25			3. Altered versions must be plainly marked as such, and must not be
26			misrepresented as being the original software.
27	nigel	29
28			4. If PCRE is embedded in any software that is released under the GNU
29			General Purpose Licence (GPL), then the terms of that licence shall
30			supersede any condition above with which it is incompatible.
31	nigel	3	-----------------------------------------------------------------------------
32
33			See the file Tech.Notes for some information on the internals.
34			*/
35
36
37	nigel	25	/* This file is compiled on its own as part of the PCRE library. However,
38	nigel	27	it is also included in the compilation of dftables.c, in which case the macro
39			DFTABLES is defined. */
40	nigel	3
41	nigel	27	#ifndef DFTABLES
42	nigel	3	#include "internal.h"
43	nigel	25	#endif
44	nigel	3
45	nigel	25
46
47			/*************************************************
48			* Create PCRE character tables *
49			*************************************************/
50
51			/* This function builds a set of character tables for use by PCRE and returns
52			a pointer to them. They are build using the ctype functions, and consequently
53			their contents will depend upon the current locale setting. When compiled as
54			part of the library, the store is obtained via pcre_malloc(), but when compiled
55	nigel	27	inside dftables, use malloc().
56	nigel	25
57			Arguments: none
58			Returns: pointer to the contiguous block of data
59			*/
60
61	nigel	53	const unsigned char *
62	nigel	25	pcre_maketables(void)
63	nigel	3	{
64	nigel	25	unsigned char yield, p;
65	nigel	3	int i;
66
67	nigel	27	#ifndef DFTABLES
68	nigel	37	yield = (unsigned char*)(pcre_malloc)(tables_length);
69	nigel	25	#else
70	nigel	37	yield = (unsigned char*)malloc(tables_length);
71	nigel	25	#endif
72	nigel	3
73	nigel	25	if (yield == NULL) return NULL;
74			p = yield;
75	nigel	3
76	nigel	25	/* First comes the lower casing table */
77	nigel	3
78	nigel	25	for (i = 0; i < 256; i++) *p++ = tolower(i);
79	nigel	3
80	nigel	25	/* Next the case-flipping table */
81	nigel	3
82	nigel	25	for (i = 0; i < 256; i++) *p++ = islower(i)? toupper(i) : tolower(i);
83	nigel	3
84	nigel	43	/* Then the character class tables. Don't try to be clever and save effort
85	nigel	63	on exclusive ones - in some locales things may be different. Note that the
86			table for "space" includes everything "isspace" gives, including VT in the
87			default locale. This makes it work for the POSIX class [:space:]. */
88	nigel	25
89			memset(p, 0, cbit_length);
90	nigel	3	for (i = 0; i < 256; i++)
91			{
92	nigel	43	if (isdigit(i))
93			{
94			p[cbit_digit + i/8] \|= 1 << (i&7);
95			p[cbit_word + i/8] \|= 1 << (i&7);
96			}
97			if (isupper(i))
98			{
99			p[cbit_upper + i/8] \|= 1 << (i&7);
100			p[cbit_word + i/8] \|= 1 << (i&7);
101			}
102			if (islower(i))
103			{
104			p[cbit_lower + i/8] \|= 1 << (i&7);
105			p[cbit_word + i/8] \|= 1 << (i&7);
106			}
107			if (i == '_') p[cbit_word + i/8] \|= 1 << (i&7);
108	nigel	25	if (isspace(i)) p[cbit_space + i/8] \|= 1 << (i&7);
109	nigel	43	if (isxdigit(i))p[cbit_xdigit + i/8] \|= 1 << (i&7);
110			if (isgraph(i)) p[cbit_graph + i/8] \|= 1 << (i&7);
111			if (isprint(i)) p[cbit_print + i/8] \|= 1 << (i&7);
112			if (ispunct(i)) p[cbit_punct + i/8] \|= 1 << (i&7);
113			if (iscntrl(i)) p[cbit_cntrl + i/8] \|= 1 << (i&7);
114	nigel	3	}
115	nigel	25	p += cbit_length;
116	nigel	3
117	nigel	63	/* Finally, the character type table. In this, we exclude VT from the white
118			space chars, because Perl doesn't recognize it as such for \s and for comments
119			within regexes. */
120	nigel	3
121			for (i = 0; i < 256; i++)
122			{
123			int x = 0;
124	nigel	63	if (i != 0x0b && isspace(i)) x += ctype_space;
125	nigel	3	if (isalpha(i)) x += ctype_letter;
126			if (isdigit(i)) x += ctype_digit;
127			if (isxdigit(i)) x += ctype_xdigit;
128			if (isalnum(i) \|\| i == '_') x += ctype_word;
129
130	nigel	73	/* Note: strchr includes the terminating zero in the characters it considers.
131			In this instance, that is ok because we want binary zero to be flagged as a
132			meta-character, which in this sense is any character that terminates a run
133			of data characters. */
134
135			if (strchr("+?{^.$\|()[", i) != 0) x += ctype_meta; p++ = x; }
136
137	nigel	25	return yield;
138	nigel	3	}
139
140			/* End of maketables.c */