Mercurial > jwe > octave
annotate liboctave/util/lo-regexp.cc @ 31121:7d3bda173b63
regexp: Be more thorough in detecting named patterns.
* liboctave/util/lo-regexp.cc (regexp::compile_internal): Check more thoroughly
if named pattern expressions are complete before adapting syntax.
* libinterp/corefcn/regexp.cc (Fregexp): Add tests.
author | Markus Mützel <markus.muetzel@gmx.de> |
---|---|
date | Mon, 04 Jul 2022 21:38:15 +0200 |
parents | 5cf18ef0377c |
children |
rev | line source |
---|---|
27928
bd51beb6205e
update formatting of copyright notices
John W. Eaton <jwe@octave.org>
parents:
27924
diff
changeset
|
1 //////////////////////////////////////////////////////////////////////// |
bd51beb6205e
update formatting of copyright notices
John W. Eaton <jwe@octave.org>
parents:
27924
diff
changeset
|
2 // |
30569
796f54d4ddbf
update Octave Project Developers copyright for the new year
John W. Eaton <jwe@octave.org>
parents:
30399
diff
changeset
|
3 // Copyright (C) 2002-2022 The Octave Project Developers |
27928
bd51beb6205e
update formatting of copyright notices
John W. Eaton <jwe@octave.org>
parents:
27924
diff
changeset
|
4 // |
bd51beb6205e
update formatting of copyright notices
John W. Eaton <jwe@octave.org>
parents:
27924
diff
changeset
|
5 // See the file COPYRIGHT.md in the top-level directory of this |
bd51beb6205e
update formatting of copyright notices
John W. Eaton <jwe@octave.org>
parents:
27924
diff
changeset
|
6 // distribution or <https://octave.org/copyright/>. |
bd51beb6205e
update formatting of copyright notices
John W. Eaton <jwe@octave.org>
parents:
27924
diff
changeset
|
7 // |
bd51beb6205e
update formatting of copyright notices
John W. Eaton <jwe@octave.org>
parents:
27924
diff
changeset
|
8 // This file is part of Octave. |
bd51beb6205e
update formatting of copyright notices
John W. Eaton <jwe@octave.org>
parents:
27924
diff
changeset
|
9 // |
bd51beb6205e
update formatting of copyright notices
John W. Eaton <jwe@octave.org>
parents:
27924
diff
changeset
|
10 // Octave is free software: you can redistribute it and/or modify it |
bd51beb6205e
update formatting of copyright notices
John W. Eaton <jwe@octave.org>
parents:
27924
diff
changeset
|
11 // under the terms of the GNU General Public License as published by |
bd51beb6205e
update formatting of copyright notices
John W. Eaton <jwe@octave.org>
parents:
27924
diff
changeset
|
12 // the Free Software Foundation, either version 3 of the License, or |
bd51beb6205e
update formatting of copyright notices
John W. Eaton <jwe@octave.org>
parents:
27924
diff
changeset
|
13 // (at your option) any later version. |
bd51beb6205e
update formatting of copyright notices
John W. Eaton <jwe@octave.org>
parents:
27924
diff
changeset
|
14 // |
bd51beb6205e
update formatting of copyright notices
John W. Eaton <jwe@octave.org>
parents:
27924
diff
changeset
|
15 // Octave is distributed in the hope that it will be useful, but |
bd51beb6205e
update formatting of copyright notices
John W. Eaton <jwe@octave.org>
parents:
27924
diff
changeset
|
16 // WITHOUT ANY WARRANTY; without even the implied warranty of |
bd51beb6205e
update formatting of copyright notices
John W. Eaton <jwe@octave.org>
parents:
27924
diff
changeset
|
17 // MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
bd51beb6205e
update formatting of copyright notices
John W. Eaton <jwe@octave.org>
parents:
27924
diff
changeset
|
18 // GNU General Public License for more details. |
bd51beb6205e
update formatting of copyright notices
John W. Eaton <jwe@octave.org>
parents:
27924
diff
changeset
|
19 // |
bd51beb6205e
update formatting of copyright notices
John W. Eaton <jwe@octave.org>
parents:
27924
diff
changeset
|
20 // You should have received a copy of the GNU General Public License |
bd51beb6205e
update formatting of copyright notices
John W. Eaton <jwe@octave.org>
parents:
27924
diff
changeset
|
21 // along with Octave; see the file COPYING. If not, see |
bd51beb6205e
update formatting of copyright notices
John W. Eaton <jwe@octave.org>
parents:
27924
diff
changeset
|
22 // <https://www.gnu.org/licenses/>. |
bd51beb6205e
update formatting of copyright notices
John W. Eaton <jwe@octave.org>
parents:
27924
diff
changeset
|
23 // |
bd51beb6205e
update formatting of copyright notices
John W. Eaton <jwe@octave.org>
parents:
27924
diff
changeset
|
24 //////////////////////////////////////////////////////////////////////// |
7779
791231dac333
Add regexp matching to Fwho and Fclear
David Bateman <dbateman@free.fr>
parents:
diff
changeset
|
25 |
21724
aba2e6293dd8
use "#if ..." consistently instead of "#ifdef" and "#ifndef"
John W. Eaton <jwe@octave.org>
parents:
21301
diff
changeset
|
26 #if defined (HAVE_CONFIG_H) |
21301
40de9f8f23a6
Use '#include "config.h"' rather than <config.h>.
Rik <rik@octave.org>
parents:
21202
diff
changeset
|
27 # include "config.h" |
7779
791231dac333
Add regexp matching to Fwho and Fclear
David Bateman <dbateman@free.fr>
parents:
diff
changeset
|
28 #endif |
791231dac333
Add regexp matching to Fwho and Fclear
David Bateman <dbateman@free.fr>
parents:
diff
changeset
|
29 |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
30 #include <list> |
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
31 #include <sstream> |
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
32 #include <string> |
7779
791231dac333
Add regexp matching to Fwho and Fclear
David Bateman <dbateman@free.fr>
parents:
diff
changeset
|
33 #include <vector> |
791231dac333
Add regexp matching to Fwho and Fclear
David Bateman <dbateman@free.fr>
parents:
diff
changeset
|
34 |
14025
9867be070ee1
use pcre/pcre.h if it is present
John W. Eaton <jwe@octave.org>
parents:
14024
diff
changeset
|
35 #if defined (HAVE_PCRE_H) |
21202
f7121e111991
maint: indent #ifdef blocks in liboctave and src directories.
Rik <rik@octave.org>
parents:
21136
diff
changeset
|
36 # include <pcre.h> |
14025
9867be070ee1
use pcre/pcre.h if it is present
John W. Eaton <jwe@octave.org>
parents:
14024
diff
changeset
|
37 #elif defined (HAVE_PCRE_PCRE_H) |
21202
f7121e111991
maint: indent #ifdef blocks in liboctave and src directories.
Rik <rik@octave.org>
parents:
21136
diff
changeset
|
38 # include <pcre/pcre.h> |
14025
9867be070ee1
use pcre/pcre.h if it is present
John W. Eaton <jwe@octave.org>
parents:
14024
diff
changeset
|
39 #endif |
7779
791231dac333
Add regexp matching to Fwho and Fclear
David Bateman <dbateman@free.fr>
parents:
diff
changeset
|
40 |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
41 #include "Matrix.h" |
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
42 #include "base-list.h" |
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
43 #include "lo-error.h" |
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
44 #include "oct-locbuf.h" |
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
45 #include "quit.h" |
16940
99122191d3dd
maint: Rename regexp.h to lo-regexp.h, regexp.cc to lo-regexp.cc in liboctave.
Rik <rik@octave.org>
parents:
15819
diff
changeset
|
46 #include "lo-regexp.h" |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
47 #include "str-vec.h" |
27528
19ad9150dd69
Check if input to regexp is valid UTF-8 is faster than PCRE's checks (bug #57064).
Markus Mützel <markus.muetzel@gmx.de>
parents:
27384
diff
changeset
|
48 #include "unistr-wrappers.h" |
7779
791231dac333
Add regexp matching to Fwho and Fclear
David Bateman <dbateman@free.fr>
parents:
diff
changeset
|
49 |
27105
097774bed4ed
move some static functions inside octave namespace
John W. Eaton <jwe@octave.org>
parents:
26379
diff
changeset
|
50 namespace octave |
097774bed4ed
move some static functions inside octave namespace
John W. Eaton <jwe@octave.org>
parents:
26379
diff
changeset
|
51 { |
097774bed4ed
move some static functions inside octave namespace
John W. Eaton <jwe@octave.org>
parents:
26379
diff
changeset
|
52 // Define the maximum number of retries for a pattern |
097774bed4ed
move some static functions inside octave namespace
John W. Eaton <jwe@octave.org>
parents:
26379
diff
changeset
|
53 // that possibly results in an infinite recursion. |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
54 #define PCRE_MATCHLIMIT_MAX 10 |
7779
791231dac333
Add regexp matching to Fwho and Fclear
David Bateman <dbateman@free.fr>
parents:
diff
changeset
|
55 |
27105
097774bed4ed
move some static functions inside octave namespace
John W. Eaton <jwe@octave.org>
parents:
26379
diff
changeset
|
56 // FIXME: should this be configurable? |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
57 #define MAXLOOKBEHIND 10 |
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
58 |
27105
097774bed4ed
move some static functions inside octave namespace
John W. Eaton <jwe@octave.org>
parents:
26379
diff
changeset
|
59 static bool lookbehind_warned = false; |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
60 |
27105
097774bed4ed
move some static functions inside octave namespace
John W. Eaton <jwe@octave.org>
parents:
26379
diff
changeset
|
61 // FIXME: don't bother collecting and composing return values |
097774bed4ed
move some static functions inside octave namespace
John W. Eaton <jwe@octave.org>
parents:
26379
diff
changeset
|
62 // the user doesn't want. |
7779
791231dac333
Add regexp matching to Fwho and Fclear
David Bateman <dbateman@free.fr>
parents:
diff
changeset
|
63 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
64 void |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
65 regexp::free (void) |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
66 { |
27384
3db033e86376
use m_ prefix for data members in most liboctave/util classes
John W. Eaton <jwe@octave.org>
parents:
27372
diff
changeset
|
67 if (m_data) |
3db033e86376
use m_ prefix for data members in most liboctave/util classes
John W. Eaton <jwe@octave.org>
parents:
27372
diff
changeset
|
68 pcre_free (static_cast<pcre *> (m_data)); |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
69 } |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
70 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
71 void |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
72 regexp::compile_internal (void) |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
73 { |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
74 // If we had a previously compiled pattern, release it. |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
75 free (); |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
76 |
29659
d13d090cb03a
use std::size_t and std::ptrdiff_t in C++ code (bug #60471)
John W. Eaton <jwe@octave.org>
parents:
29363
diff
changeset
|
77 std::size_t max_length = MAXLOOKBEHIND; |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
78 |
29659
d13d090cb03a
use std::size_t and std::ptrdiff_t in C++ code (bug #60471)
John W. Eaton <jwe@octave.org>
parents:
29363
diff
changeset
|
79 std::size_t pos = 0; |
d13d090cb03a
use std::size_t and std::ptrdiff_t in C++ code (bug #60471)
John W. Eaton <jwe@octave.org>
parents:
29363
diff
changeset
|
80 std::size_t new_pos; |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
81 int inames = 0; |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
82 std::ostringstream buf; |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
83 |
27384
3db033e86376
use m_ prefix for data members in most liboctave/util classes
John W. Eaton <jwe@octave.org>
parents:
27372
diff
changeset
|
84 while ((new_pos = m_pattern.find ("(?", pos)) != std::string::npos) |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
85 { |
31121
7d3bda173b63
regexp: Be more thorough in detecting named patterns.
Markus Mützel <markus.muetzel@gmx.de>
parents:
31119
diff
changeset
|
86 std::size_t tmp_pos; |
31119
5cf18ef0377c
regexp: Check pattern length before accessing it (bug #62704).
Markus Mützel <markus.muetzel@gmx.de>
parents:
30569
diff
changeset
|
87 if (m_pattern.size () > new_pos + 2 |
5cf18ef0377c
regexp: Check pattern length before accessing it (bug #62704).
Markus Mützel <markus.muetzel@gmx.de>
parents:
30569
diff
changeset
|
88 && m_pattern.at (new_pos + 2) == '<' |
5cf18ef0377c
regexp: Check pattern length before accessing it (bug #62704).
Markus Mützel <markus.muetzel@gmx.de>
parents:
30569
diff
changeset
|
89 && ! (m_pattern.size () > new_pos + 3 |
5cf18ef0377c
regexp: Check pattern length before accessing it (bug #62704).
Markus Mützel <markus.muetzel@gmx.de>
parents:
30569
diff
changeset
|
90 && (m_pattern.at (new_pos + 3) == '=' |
31121
7d3bda173b63
regexp: Be more thorough in detecting named patterns.
Markus Mützel <markus.muetzel@gmx.de>
parents:
31119
diff
changeset
|
91 || m_pattern.at (new_pos + 3) == '!')) |
7d3bda173b63
regexp: Be more thorough in detecting named patterns.
Markus Mützel <markus.muetzel@gmx.de>
parents:
31119
diff
changeset
|
92 && (tmp_pos = m_pattern.find_first_of ('>', new_pos)) |
7d3bda173b63
regexp: Be more thorough in detecting named patterns.
Markus Mützel <markus.muetzel@gmx.de>
parents:
31119
diff
changeset
|
93 != std::string::npos |
7d3bda173b63
regexp: Be more thorough in detecting named patterns.
Markus Mützel <markus.muetzel@gmx.de>
parents:
31119
diff
changeset
|
94 && m_pattern.find_first_of (')', tmp_pos) != std::string::npos) |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
95 { |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
96 // The syntax of named tokens in pcre is "(?P<name>...)" while |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
97 // we need a syntax "(?<name>...)", so fix that here. Also an |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
98 // expression like |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
99 // "(?<first>\w+)\s+(?<last>\w+)|(?<last>\w+),\s+(?<first>\w+)" |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
100 // should be perfectly legal, while pcre does not allow the same |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
101 // named token name on both sides of the alternative. Also fix |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
102 // that here by replacing name tokens by dummy names, and dealing |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
103 // with the dummy names later. |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
104 |
27281
db687716fed6
style fixes: generally aim to break long lines before operators, not after
John W. Eaton <jwe@octave.org>
parents:
27105
diff
changeset
|
105 std::string tmp_name |
27384
3db033e86376
use m_ prefix for data members in most liboctave/util classes
John W. Eaton <jwe@octave.org>
parents:
27372
diff
changeset
|
106 = m_pattern.substr (new_pos+3, tmp_pos-new_pos-3); |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
107 |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
108 bool found = false; |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
109 |
27384
3db033e86376
use m_ prefix for data members in most liboctave/util classes
John W. Eaton <jwe@octave.org>
parents:
27372
diff
changeset
|
110 for (int i = 0; i < m_names; i++) |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
111 { |
27384
3db033e86376
use m_ prefix for data members in most liboctave/util classes
John W. Eaton <jwe@octave.org>
parents:
27372
diff
changeset
|
112 if (m_named_pats(i) == tmp_name) |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
113 { |
27384
3db033e86376
use m_ prefix for data members in most liboctave/util classes
John W. Eaton <jwe@octave.org>
parents:
27372
diff
changeset
|
114 m_named_idx.resize (dim_vector (inames+1, 1)); |
3db033e86376
use m_ prefix for data members in most liboctave/util classes
John W. Eaton <jwe@octave.org>
parents:
27372
diff
changeset
|
115 m_named_idx(inames) = i; |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
116 found = true; |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
117 break; |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
118 } |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
119 } |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
120 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
121 if (! found) |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
122 { |
27384
3db033e86376
use m_ prefix for data members in most liboctave/util classes
John W. Eaton <jwe@octave.org>
parents:
27372
diff
changeset
|
123 m_named_idx.resize (dim_vector (inames+1, 1)); |
3db033e86376
use m_ prefix for data members in most liboctave/util classes
John W. Eaton <jwe@octave.org>
parents:
27372
diff
changeset
|
124 m_named_idx(inames) = m_names; |
3db033e86376
use m_ prefix for data members in most liboctave/util classes
John W. Eaton <jwe@octave.org>
parents:
27372
diff
changeset
|
125 m_named_pats.append (tmp_name); |
3db033e86376
use m_ prefix for data members in most liboctave/util classes
John W. Eaton <jwe@octave.org>
parents:
27372
diff
changeset
|
126 m_names++; |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
127 } |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
128 |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
129 if (new_pos - pos > 0) |
27384
3db033e86376
use m_ prefix for data members in most liboctave/util classes
John W. Eaton <jwe@octave.org>
parents:
27372
diff
changeset
|
130 buf << m_pattern.substr (pos, new_pos-pos); |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
131 if (inames < 10) |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
132 buf << "(?P<n00" << inames++; |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
133 else if (inames < 100) |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
134 buf << "(?P<n0" << inames++; |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
135 else |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
136 buf << "(?P<n" << inames++; |
7779
791231dac333
Add regexp matching to Fwho and Fclear
David Bateman <dbateman@free.fr>
parents:
diff
changeset
|
137 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
138 pos = tmp_pos; |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
139 } |
31119
5cf18ef0377c
regexp: Check pattern length before accessing it (bug #62704).
Markus Mützel <markus.muetzel@gmx.de>
parents:
30569
diff
changeset
|
140 else if (m_pattern.size () > new_pos + 2 |
5cf18ef0377c
regexp: Check pattern length before accessing it (bug #62704).
Markus Mützel <markus.muetzel@gmx.de>
parents:
30569
diff
changeset
|
141 && m_pattern.at (new_pos + 2) == '<') |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
142 { |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
143 // Find lookbehind operators of arbitrary length (ie like |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
144 // "(?<=[a-z]*)") and replace with a maximum length operator |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
145 // as PCRE can not yet handle arbitrary length lookahead |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
146 // operators. Use the string length as the maximum length to |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
147 // avoid issues. |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
148 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
149 int brackets = 1; |
29659
d13d090cb03a
use std::size_t and std::ptrdiff_t in C++ code (bug #60471)
John W. Eaton <jwe@octave.org>
parents:
29363
diff
changeset
|
150 std::size_t tmp_pos1 = new_pos + 2; |
d13d090cb03a
use std::size_t and std::ptrdiff_t in C++ code (bug #60471)
John W. Eaton <jwe@octave.org>
parents:
29363
diff
changeset
|
151 std::size_t tmp_pos2 = tmp_pos1; |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
152 |
27384
3db033e86376
use m_ prefix for data members in most liboctave/util classes
John W. Eaton <jwe@octave.org>
parents:
27372
diff
changeset
|
153 while (tmp_pos1 < m_pattern.length () && brackets > 0) |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
154 { |
27384
3db033e86376
use m_ prefix for data members in most liboctave/util classes
John W. Eaton <jwe@octave.org>
parents:
27372
diff
changeset
|
155 char ch = m_pattern.at (tmp_pos1); |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
156 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
157 if (ch == '(') |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
158 brackets++; |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
159 else if (ch == ')') |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
160 { |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
161 if (brackets > 1) |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
162 tmp_pos2 = tmp_pos1; |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
163 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
164 brackets--; |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
165 } |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
166 |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
167 tmp_pos1++; |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
168 } |
7779
791231dac333
Add regexp matching to Fwho and Fclear
David Bateman <dbateman@free.fr>
parents:
diff
changeset
|
169 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
170 if (brackets != 0) |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
171 { |
27384
3db033e86376
use m_ prefix for data members in most liboctave/util classes
John W. Eaton <jwe@octave.org>
parents:
27372
diff
changeset
|
172 buf << m_pattern.substr (pos, new_pos - pos) << "(?"; |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
173 pos = new_pos + 2; |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
174 } |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
175 else |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
176 { |
29659
d13d090cb03a
use std::size_t and std::ptrdiff_t in C++ code (bug #60471)
John W. Eaton <jwe@octave.org>
parents:
29363
diff
changeset
|
177 std::size_t tmp_pos3 = m_pattern.find_first_of ("*+", tmp_pos2); |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
178 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
179 if (tmp_pos3 != std::string::npos && tmp_pos3 < tmp_pos1) |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
180 { |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
181 if (! lookbehind_warned) |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
182 { |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
183 lookbehind_warned = true; |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
184 (*current_liboctave_warning_with_id_handler) |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
185 ("Octave:regexp-lookbehind-limit", |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
186 "%s: arbitrary length lookbehind patterns are only supported up to length %d", |
27384
3db033e86376
use m_ prefix for data members in most liboctave/util classes
John W. Eaton <jwe@octave.org>
parents:
27372
diff
changeset
|
187 m_who.c_str (), MAXLOOKBEHIND); |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
188 } |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
189 |
27384
3db033e86376
use m_ prefix for data members in most liboctave/util classes
John W. Eaton <jwe@octave.org>
parents:
27372
diff
changeset
|
190 buf << m_pattern.substr (pos, new_pos - pos) << '('; |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
191 |
29659
d13d090cb03a
use std::size_t and std::ptrdiff_t in C++ code (bug #60471)
John W. Eaton <jwe@octave.org>
parents:
29363
diff
changeset
|
192 std::size_t i; |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
193 |
27384
3db033e86376
use m_ prefix for data members in most liboctave/util classes
John W. Eaton <jwe@octave.org>
parents:
27372
diff
changeset
|
194 if (m_pattern.at (tmp_pos3) == '*') |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
195 i = 0; |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
196 else |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
197 i = 1; |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
198 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
199 for (; i < max_length + 1; i++) |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
200 { |
27384
3db033e86376
use m_ prefix for data members in most liboctave/util classes
John W. Eaton <jwe@octave.org>
parents:
27372
diff
changeset
|
201 buf << m_pattern.substr (new_pos, tmp_pos3 - new_pos) |
23807
336f89b6208b
Use character literals 'c' rather than string literals "c" when possible.
Rik <rik@octave.org>
parents:
23795
diff
changeset
|
202 << '{' << i << '}'; |
27384
3db033e86376
use m_ prefix for data members in most liboctave/util classes
John W. Eaton <jwe@octave.org>
parents:
27372
diff
changeset
|
203 buf << m_pattern.substr (tmp_pos3 + 1, |
27938
863ae57eee69
maint: Use Octave coding conventions in liboctave/
Rik <rik@octave.org>
parents:
27928
diff
changeset
|
204 tmp_pos1 - tmp_pos3 - 1); |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
205 if (i != max_length) |
23807
336f89b6208b
Use character literals 'c' rather than string literals "c" when possible.
Rik <rik@octave.org>
parents:
23795
diff
changeset
|
206 buf << '|'; |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
207 } |
23807
336f89b6208b
Use character literals 'c' rather than string literals "c" when possible.
Rik <rik@octave.org>
parents:
23795
diff
changeset
|
208 buf << ')'; |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
209 } |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
210 else |
27384
3db033e86376
use m_ prefix for data members in most liboctave/util classes
John W. Eaton <jwe@octave.org>
parents:
27372
diff
changeset
|
211 buf << m_pattern.substr (pos, tmp_pos1 - pos); |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
212 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
213 pos = tmp_pos1; |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
214 } |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
215 } |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
216 else |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
217 { |
27384
3db033e86376
use m_ prefix for data members in most liboctave/util classes
John W. Eaton <jwe@octave.org>
parents:
27372
diff
changeset
|
218 buf << m_pattern.substr (pos, new_pos - pos) << "(?"; |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
219 pos = new_pos + 2; |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
220 } |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
221 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
222 } |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
223 |
27384
3db033e86376
use m_ prefix for data members in most liboctave/util classes
John W. Eaton <jwe@octave.org>
parents:
27372
diff
changeset
|
224 buf << m_pattern.substr (pos); |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
225 |
25103
078b795c5219
maint: style check C++ ahead of 4.4 release.
Rik <rik@octave.org>
parents:
25054
diff
changeset
|
226 // Replace NULLs with escape sequence because conversion function c_str() |
24741
00dfa167c1fe
Fix handling of NULL character in regular expressions for Matlab compatibility.
Rik <rik@octave.org>
parents:
24607
diff
changeset
|
227 // will terminate string early at embedded NULLs. |
00dfa167c1fe
Fix handling of NULL character in regular expressions for Matlab compatibility.
Rik <rik@octave.org>
parents:
24607
diff
changeset
|
228 std::string buf_str = buf.str (); |
00dfa167c1fe
Fix handling of NULL character in regular expressions for Matlab compatibility.
Rik <rik@octave.org>
parents:
24607
diff
changeset
|
229 while ((pos = buf_str.find ('\0')) != std::string::npos) |
00dfa167c1fe
Fix handling of NULL character in regular expressions for Matlab compatibility.
Rik <rik@octave.org>
parents:
24607
diff
changeset
|
230 buf_str.replace (pos, 1, "\\000"); |
00dfa167c1fe
Fix handling of NULL character in regular expressions for Matlab compatibility.
Rik <rik@octave.org>
parents:
24607
diff
changeset
|
231 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
232 const char *err; |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
233 int erroffset; |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
234 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
235 int pcre_options |
27384
3db033e86376
use m_ prefix for data members in most liboctave/util classes
John W. Eaton <jwe@octave.org>
parents:
27372
diff
changeset
|
236 = ( (m_options.case_insensitive () ? PCRE_CASELESS : 0) |
3db033e86376
use m_ prefix for data members in most liboctave/util classes
John W. Eaton <jwe@octave.org>
parents:
27372
diff
changeset
|
237 | (m_options.dotexceptnewline () ? 0 : PCRE_DOTALL) |
3db033e86376
use m_ prefix for data members in most liboctave/util classes
John W. Eaton <jwe@octave.org>
parents:
27372
diff
changeset
|
238 | (m_options.lineanchors () ? PCRE_MULTILINE : 0) |
27570
74173f04d2a3
Use syntax for UTF-8 mode with PCRE that is used in the configure test (bug #35910).
Markus Mützel <markus.muetzel@gmx.de>
parents:
27528
diff
changeset
|
239 | (m_options.freespacing () ? PCRE_EXTENDED : 0) |
74173f04d2a3
Use syntax for UTF-8 mode with PCRE that is used in the configure test (bug #35910).
Markus Mützel <markus.muetzel@gmx.de>
parents:
27528
diff
changeset
|
240 | PCRE_UTF8); |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
241 |
27384
3db033e86376
use m_ prefix for data members in most liboctave/util classes
John W. Eaton <jwe@octave.org>
parents:
27372
diff
changeset
|
242 m_data = pcre_compile (buf_str.c_str (), pcre_options, |
27938
863ae57eee69
maint: Use Octave coding conventions in liboctave/
Rik <rik@octave.org>
parents:
27928
diff
changeset
|
243 &err, &erroffset, nullptr); |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
244 |
27384
3db033e86376
use m_ prefix for data members in most liboctave/util classes
John W. Eaton <jwe@octave.org>
parents:
27372
diff
changeset
|
245 if (! m_data) |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
246 (*current_liboctave_error_handler) |
27384
3db033e86376
use m_ prefix for data members in most liboctave/util classes
John W. Eaton <jwe@octave.org>
parents:
27372
diff
changeset
|
247 ("%s: %s at position %d of expression", m_who.c_str (), err, erroffset); |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
248 } |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
249 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
250 regexp::match_data |
29218
397d043bb2ff
Add changes missing in cf059093ffbc.
Markus Mützel <markus.muetzel@gmx.de>
parents:
29217
diff
changeset
|
251 regexp::match (const std::string& buffer) const |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
252 { |
27528
19ad9150dd69
Check if input to regexp is valid UTF-8 is faster than PCRE's checks (bug #57064).
Markus Mützel <markus.muetzel@gmx.de>
parents:
27384
diff
changeset
|
253 // check if input is valid utf-8 |
19ad9150dd69
Check if input to regexp is valid UTF-8 is faster than PCRE's checks (bug #57064).
Markus Mützel <markus.muetzel@gmx.de>
parents:
27384
diff
changeset
|
254 const uint8_t *buf_str = reinterpret_cast<const uint8_t *> (buffer.c_str ()); |
19ad9150dd69
Check if input to regexp is valid UTF-8 is faster than PCRE's checks (bug #57064).
Markus Mützel <markus.muetzel@gmx.de>
parents:
27384
diff
changeset
|
255 if (octave_u8_check_wrapper (buf_str, buffer.length ())) |
19ad9150dd69
Check if input to regexp is valid UTF-8 is faster than PCRE's checks (bug #57064).
Markus Mützel <markus.muetzel@gmx.de>
parents:
27384
diff
changeset
|
256 (*current_liboctave_error_handler) |
19ad9150dd69
Check if input to regexp is valid UTF-8 is faster than PCRE's checks (bug #57064).
Markus Mützel <markus.muetzel@gmx.de>
parents:
27384
diff
changeset
|
257 ("%s: the input string is invalid UTF-8", m_who.c_str ()); |
19ad9150dd69
Check if input to regexp is valid UTF-8 is faster than PCRE's checks (bug #57064).
Markus Mützel <markus.muetzel@gmx.de>
parents:
27384
diff
changeset
|
258 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
259 regexp::match_data retval; |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
260 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
261 std::list<regexp::match_element> lst; |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
262 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
263 int subpatterns; |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
264 int namecount; |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
265 int nameentrysize; |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
266 char *nametable; |
29659
d13d090cb03a
use std::size_t and std::ptrdiff_t in C++ code (bug #60471)
John W. Eaton <jwe@octave.org>
parents:
29363
diff
changeset
|
267 std::size_t idx = 0; |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
268 |
27384
3db033e86376
use m_ prefix for data members in most liboctave/util classes
John W. Eaton <jwe@octave.org>
parents:
27372
diff
changeset
|
269 pcre *re = static_cast<pcre *> (m_data); |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
270 |
23795
980f39c3ab90
Use C++11 nullptr rather than 0 in code (bug #51565).
Rik <rik@octave.org>
parents:
23302
diff
changeset
|
271 pcre_fullinfo (re, nullptr, PCRE_INFO_CAPTURECOUNT, &subpatterns); |
980f39c3ab90
Use C++11 nullptr rather than 0 in code (bug #51565).
Rik <rik@octave.org>
parents:
23302
diff
changeset
|
272 pcre_fullinfo (re, nullptr, PCRE_INFO_NAMECOUNT, &namecount); |
980f39c3ab90
Use C++11 nullptr rather than 0 in code (bug #51565).
Rik <rik@octave.org>
parents:
23302
diff
changeset
|
273 pcre_fullinfo (re, nullptr, PCRE_INFO_NAMEENTRYSIZE, &nameentrysize); |
980f39c3ab90
Use C++11 nullptr rather than 0 in code (bug #51565).
Rik <rik@octave.org>
parents:
23302
diff
changeset
|
274 pcre_fullinfo (re, nullptr, PCRE_INFO_NAMETABLE, &nametable); |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
275 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
276 OCTAVE_LOCAL_BUFFER (int, ovector, (subpatterns+1)*3); |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
277 OCTAVE_LOCAL_BUFFER (int, nidx, namecount); |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
278 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
279 for (int i = 0; i < namecount; i++) |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
280 { |
23302
78cf75ffd382
maint: tweak readability of code in lo-regexp.cc.
Rik <rik@octave.org>
parents:
23084
diff
changeset
|
281 // Index of subpattern in first two bytes of name (MSB first). |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
282 // Extract index. |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
283 nidx[i] = (static_cast<int> (nametable[i*nameentrysize])) << 8 |
22402
4caa7b28d183
maint: Style check C++ code in liboctave/
Rik <rik@octave.org>
parents:
22333
diff
changeset
|
284 | static_cast<int> (nametable[i*nameentrysize+1]); |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
285 } |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
286 |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
287 while (true) |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
288 { |
22860
0b1e25cc4457
eliminate use of OCTAVE_QUIT macro in C++ sources
John W. Eaton <jwe@octave.org>
parents:
22755
diff
changeset
|
289 octave_quit (); |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
290 |
23795
980f39c3ab90
Use C++11 nullptr rather than 0 in code (bug #51565).
Rik <rik@octave.org>
parents:
23302
diff
changeset
|
291 int matches = pcre_exec (re, nullptr, buffer.c_str (), |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
292 buffer.length (), idx, |
27528
19ad9150dd69
Check if input to regexp is valid UTF-8 is faster than PCRE's checks (bug #57064).
Markus Mützel <markus.muetzel@gmx.de>
parents:
27384
diff
changeset
|
293 PCRE_NO_UTF8_CHECK | (idx ? PCRE_NOTBOL : 0), |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
294 ovector, (subpatterns+1)*3); |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
295 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
296 if (matches == PCRE_ERROR_MATCHLIMIT) |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
297 { |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
298 // Try harder; start with default value for MATCH_LIMIT |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
299 // and increase it. |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
300 (*current_liboctave_warning_with_id_handler) |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
301 ("Octave:regexp-match-limit", |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
302 "your pattern caused PCRE to hit its MATCH_LIMIT; trying harder now, but this will be slow"); |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
303 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
304 pcre_extra pe; |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
305 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
306 pcre_config (PCRE_CONFIG_MATCH_LIMIT, |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
307 static_cast<void *> (&pe.match_limit)); |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
308 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
309 pe.flags = PCRE_EXTRA_MATCH_LIMIT; |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
310 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
311 int i = 0; |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
312 while (matches == PCRE_ERROR_MATCHLIMIT |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
313 && i++ < PCRE_MATCHLIMIT_MAX) |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
314 { |
22860
0b1e25cc4457
eliminate use of OCTAVE_QUIT macro in C++ sources
John W. Eaton <jwe@octave.org>
parents:
22755
diff
changeset
|
315 octave_quit (); |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
316 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
317 pe.match_limit *= 10; |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
318 matches = pcre_exec (re, &pe, buffer.c_str (), |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
319 buffer.length (), idx, |
27528
19ad9150dd69
Check if input to regexp is valid UTF-8 is faster than PCRE's checks (bug #57064).
Markus Mützel <markus.muetzel@gmx.de>
parents:
27384
diff
changeset
|
320 PCRE_NO_UTF8_CHECK |
19ad9150dd69
Check if input to regexp is valid UTF-8 is faster than PCRE's checks (bug #57064).
Markus Mützel <markus.muetzel@gmx.de>
parents:
27384
diff
changeset
|
321 | (idx ? PCRE_NOTBOL : 0), |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
322 ovector, (subpatterns+1)*3); |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
323 } |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
324 } |
7779
791231dac333
Add regexp matching to Fwho and Fclear
David Bateman <dbateman@free.fr>
parents:
diff
changeset
|
325 |
27528
19ad9150dd69
Check if input to regexp is valid UTF-8 is faster than PCRE's checks (bug #57064).
Markus Mützel <markus.muetzel@gmx.de>
parents:
27384
diff
changeset
|
326 if (matches < 0 && matches != PCRE_ERROR_NOMATCH) |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
327 (*current_liboctave_error_handler) |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
328 ("%s: internal error calling pcre_exec; " |
27384
3db033e86376
use m_ prefix for data members in most liboctave/util classes
John W. Eaton <jwe@octave.org>
parents:
27372
diff
changeset
|
329 "error code from pcre_exec is %i", m_who.c_str (), matches); |
21136
7cac4e7458f2
maint: clean up code around calls to current_liboctave_error_handler.
Rik <rik@octave.org>
parents:
20955
diff
changeset
|
330 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
331 if (matches == PCRE_ERROR_NOMATCH) |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
332 break; |
27384
3db033e86376
use m_ prefix for data members in most liboctave/util classes
John W. Eaton <jwe@octave.org>
parents:
27372
diff
changeset
|
333 else if (ovector[0] >= ovector[1] && ! m_options.emptymatch ()) |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
334 { |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
335 // Zero length match. Skip to next char. |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
336 idx = ovector[0] + 1; |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
337 if (idx < buffer.length ()) |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
338 continue; |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
339 else |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
340 break; |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
341 } |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
342 else |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
343 { |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
344 int pos_match = 0; |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
345 Matrix token_extents (matches-1, 2); |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
346 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
347 for (int i = 1; i < matches; i++) |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
348 { |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
349 if (ovector[2*i] >= 0 && ovector[2*i+1] > 0 |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
350 && (i == 1 || ovector[2*i] != ovector[2*i-2] |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
351 || ovector[2*i-1] != ovector[2*i+1])) |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
352 { |
30399
f3f3e3793fb5
maint: style check C++ files in liboctave/ ahead of 7.1 release.
Rik <rik@octave.org>
parents:
30183
diff
changeset
|
353 token_extents(pos_match, 0) = double (ovector[2*i]+1); |
f3f3e3793fb5
maint: style check C++ files in liboctave/ ahead of 7.1 release.
Rik <rik@octave.org>
parents:
30183
diff
changeset
|
354 token_extents(pos_match++, 1) = double (ovector[2*i+1]); |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
355 } |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
356 } |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
357 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
358 token_extents.resize (pos_match, 2); |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
359 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
360 double start = double (ovector[0]+1); |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
361 double end = double (ovector[1]); |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
362 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
363 const char **listptr; |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
364 int status = pcre_get_substring_list (buffer.c_str (), ovector, |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
365 matches, &listptr); |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
366 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
367 if (status == PCRE_ERROR_NOMEMORY) |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
368 (*current_liboctave_error_handler) |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
369 ("%s: cannot allocate memory in pcre_get_substring_list", |
27384
3db033e86376
use m_ prefix for data members in most liboctave/util classes
John W. Eaton <jwe@octave.org>
parents:
27372
diff
changeset
|
370 m_who.c_str ()); |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
371 |
24741
00dfa167c1fe
Fix handling of NULL character in regular expressions for Matlab compatibility.
Rik <rik@octave.org>
parents:
24607
diff
changeset
|
372 // Must use explicit length constructor as match can contain '\0'. |
00dfa167c1fe
Fix handling of NULL character in regular expressions for Matlab compatibility.
Rik <rik@octave.org>
parents:
24607
diff
changeset
|
373 std::string match_string = std::string (*listptr, end - start + 1); |
00dfa167c1fe
Fix handling of NULL character in regular expressions for Matlab compatibility.
Rik <rik@octave.org>
parents:
24607
diff
changeset
|
374 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
375 string_vector tokens (pos_match); |
27384
3db033e86376
use m_ prefix for data members in most liboctave/util classes
John W. Eaton <jwe@octave.org>
parents:
27372
diff
changeset
|
376 string_vector named_tokens (m_names); |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
377 int pos_offset = 0; |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
378 pos_match = 0; |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
379 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
380 for (int i = 1; i < matches; i++) |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
381 { |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
382 if (ovector[2*i] >= 0 && ovector[2*i+1] > 0) |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
383 { |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
384 if (i == 1 || ovector[2*i] != ovector[2*i-2] |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
385 || ovector[2*i-1] != ovector[2*i+1]) |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
386 { |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
387 if (namecount > 0) |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
388 { |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
389 // FIXME: Should probably do this with a map() |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
390 // rather than a linear search. However, |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
391 // the number of captured, named expressions |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
392 // is usually pretty small (< 4) |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
393 for (int j = 0; j < namecount; j++) |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
394 { |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
395 if (nidx[j] == i) |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
396 { |
29659
d13d090cb03a
use std::size_t and std::ptrdiff_t in C++ code (bug #60471)
John W. Eaton <jwe@octave.org>
parents:
29363
diff
changeset
|
397 std::size_t len = ovector[2*i+1] - ovector[2*i]; |
27384
3db033e86376
use m_ prefix for data members in most liboctave/util classes
John W. Eaton <jwe@octave.org>
parents:
27372
diff
changeset
|
398 named_tokens(m_named_idx(j)) |
27281
db687716fed6
style fixes: generally aim to break long lines before operators, not after
John W. Eaton <jwe@octave.org>
parents:
27105
diff
changeset
|
399 = std::string (*(listptr+i-pos_offset), |
db687716fed6
style fixes: generally aim to break long lines before operators, not after
John W. Eaton <jwe@octave.org>
parents:
27105
diff
changeset
|
400 len); |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
401 break; |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
402 } |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
403 } |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
404 } |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
405 |
29659
d13d090cb03a
use std::size_t and std::ptrdiff_t in C++ code (bug #60471)
John W. Eaton <jwe@octave.org>
parents:
29363
diff
changeset
|
406 std::size_t len = ovector[2*i+1] - ovector[2*i]; |
24741
00dfa167c1fe
Fix handling of NULL character in regular expressions for Matlab compatibility.
Rik <rik@octave.org>
parents:
24607
diff
changeset
|
407 tokens(pos_match++) = std::string (*(listptr+i), len); |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
408 } |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
409 else |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
410 pos_offset++; |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
411 } |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
412 } |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
413 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
414 pcre_free_substring_list (listptr); |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
415 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
416 regexp::match_element new_elem (named_tokens, tokens, match_string, |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
417 token_extents, start, end); |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
418 lst.push_back (new_elem); |
14536
6d5c951ec520
Add 'emptymatch', 'noemptymatch' options to regular expressions.
Rik <octave@nomad.inbox5.com>
parents:
14506
diff
changeset
|
419 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
420 if (ovector[1] <= ovector[0]) |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
421 { |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
422 // Zero length match. Skip to next char. |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
423 idx = ovector[0] + 1; |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
424 if (idx <= buffer.length ()) |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
425 continue; |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
426 } |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
427 else |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
428 idx = ovector[1]; |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
429 |
27384
3db033e86376
use m_ prefix for data members in most liboctave/util classes
John W. Eaton <jwe@octave.org>
parents:
27372
diff
changeset
|
430 if (m_options.once () || idx >= buffer.length ()) |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
431 break; |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
432 } |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
433 } |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
434 |
27384
3db033e86376
use m_ prefix for data members in most liboctave/util classes
John W. Eaton <jwe@octave.org>
parents:
27372
diff
changeset
|
435 retval = regexp::match_data (lst, m_named_pats); |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
436 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
437 return retval; |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
438 } |
7779
791231dac333
Add regexp matching to Fwho and Fclear
David Bateman <dbateman@free.fr>
parents:
diff
changeset
|
439 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
440 bool |
29217
cf059093ffbc
stat: Make regular expression for Windows a static const (bug #59706).
Markus Mützel <markus.muetzel@gmx.de>
parents:
27938
diff
changeset
|
441 regexp::is_match (const std::string& buffer) const |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
442 { |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
443 regexp::match_data rx_lst = match (buffer); |
7779
791231dac333
Add regexp matching to Fwho and Fclear
David Bateman <dbateman@free.fr>
parents:
diff
changeset
|
444 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
445 return rx_lst.size () > 0; |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
446 } |
7779
791231dac333
Add regexp matching to Fwho and Fclear
David Bateman <dbateman@free.fr>
parents:
diff
changeset
|
447 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
448 Array<bool> |
29217
cf059093ffbc
stat: Make regular expression for Windows a static const (bug #59706).
Markus Mützel <markus.muetzel@gmx.de>
parents:
27938
diff
changeset
|
449 regexp::is_match (const string_vector& buffer) const |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
450 { |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
451 octave_idx_type len = buffer.numel (); |
7779
791231dac333
Add regexp matching to Fwho and Fclear
David Bateman <dbateman@free.fr>
parents:
diff
changeset
|
452 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
453 Array<bool> retval (dim_vector (len, 1)); |
7779
791231dac333
Add regexp matching to Fwho and Fclear
David Bateman <dbateman@free.fr>
parents:
diff
changeset
|
454 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
455 for (octave_idx_type i = 0; i < buffer.numel (); i++) |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
456 retval(i) = is_match (buffer(i)); |
7779
791231dac333
Add regexp matching to Fwho and Fclear
David Bateman <dbateman@free.fr>
parents:
diff
changeset
|
457 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
458 return retval; |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
459 } |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
460 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
461 // Declare rep_token_t used in processing replacement string |
30183
d0184bad0c07
maint: lo-regexp.cc: Use C++ named struct declaration rather than C-style typedef.
Rik <rik@octave.org>
parents:
29660
diff
changeset
|
462 struct rep_token_t |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
463 { |
29659
d13d090cb03a
use std::size_t and std::ptrdiff_t in C++ code (bug #60471)
John W. Eaton <jwe@octave.org>
parents:
29363
diff
changeset
|
464 std::size_t pos; |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
465 int num; |
30183
d0184bad0c07
maint: lo-regexp.cc: Use C++ named struct declaration rather than C-style typedef.
Rik <rik@octave.org>
parents:
29660
diff
changeset
|
466 }; |
15541
9db32cabeacf
Fix backslash handling in regexp pattern (Bug #37092)
Rik <rik@octave.org>
parents:
15271
diff
changeset
|
467 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
468 std::string |
29218
397d043bb2ff
Add changes missing in cf059093ffbc.
Markus Mützel <markus.muetzel@gmx.de>
parents:
29217
diff
changeset
|
469 regexp::replace (const std::string& buffer, |
397d043bb2ff
Add changes missing in cf059093ffbc.
Markus Mützel <markus.muetzel@gmx.de>
parents:
29217
diff
changeset
|
470 const std::string& replacement) const |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
471 { |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
472 std::string retval; |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
473 |
25342
416856765a55
be more careful with using auto in place of explicit const iterator decls
John W. Eaton <jwe@octave.org>
parents:
25337
diff
changeset
|
474 const regexp::match_data rx_lst = match (buffer); |
15541
9db32cabeacf
Fix backslash handling in regexp pattern (Bug #37092)
Rik <rik@octave.org>
parents:
15271
diff
changeset
|
475 |
29659
d13d090cb03a
use std::size_t and std::ptrdiff_t in C++ code (bug #60471)
John W. Eaton <jwe@octave.org>
parents:
29363
diff
changeset
|
476 std::size_t num_matches = rx_lst.size (); |
15541
9db32cabeacf
Fix backslash handling in regexp pattern (Bug #37092)
Rik <rik@octave.org>
parents:
15271
diff
changeset
|
477 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
478 if (num_matches == 0) |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
479 { |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
480 retval = buffer; |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
481 return retval; |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
482 } |
15541
9db32cabeacf
Fix backslash handling in regexp pattern (Bug #37092)
Rik <rik@octave.org>
parents:
15271
diff
changeset
|
483 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
484 // Identify replacement tokens; build a vector of group numbers in |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
485 // the replacement string so that we can quickly calculate the size |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
486 // of the replacement. |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
487 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
488 // FIXME: All code assumes that only 10 tokens ($0-$9) exist. |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
489 // $11 represents $1 followed by the character '1' rather than |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
490 // the eleventh capture buffer. |
15541
9db32cabeacf
Fix backslash handling in regexp pattern (Bug #37092)
Rik <rik@octave.org>
parents:
15271
diff
changeset
|
491 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
492 std::string repstr = replacement; |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
493 std::vector<rep_token_t> tokens; |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
494 tokens.reserve (5); // Reserve memory for 5 pattern replacements |
15541
9db32cabeacf
Fix backslash handling in regexp pattern (Bug #37092)
Rik <rik@octave.org>
parents:
15271
diff
changeset
|
495 |
29659
d13d090cb03a
use std::size_t and std::ptrdiff_t in C++ code (bug #60471)
John W. Eaton <jwe@octave.org>
parents:
29363
diff
changeset
|
496 for (std::size_t i=0; i < repstr.size (); i++) |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
497 { |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
498 if (repstr[i] == '\\') |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
499 { |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
500 if (i < repstr.size () - 1 && repstr[i+1] == '$') |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
501 { |
30399
f3f3e3793fb5
maint: style check C++ files in liboctave/ ahead of 7.1 release.
Rik <rik@octave.org>
parents:
30183
diff
changeset
|
502 repstr.erase (i, 1); // erase backslash |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
503 i++; // skip over '$' |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
504 continue; |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
505 } |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
506 if (i < repstr.size () - 1 && repstr[i+1] == '\\') |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
507 { |
30399
f3f3e3793fb5
maint: style check C++ files in liboctave/ ahead of 7.1 release.
Rik <rik@octave.org>
parents:
30183
diff
changeset
|
508 repstr.erase (i, 1); // erase 1st backslash |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
509 continue; |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
510 } |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
511 } |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
512 else if (repstr[i] == '$') |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
513 { |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
514 if (i < repstr.size () - 1 && isdigit (repstr[i+1])) |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
515 { |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
516 rep_token_t tmp_token; |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
517 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
518 tmp_token.pos = i; |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
519 tmp_token.num = repstr[i+1]-'0'; |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
520 tokens.push_back (tmp_token); |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
521 } |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
522 } |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
523 } |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
524 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
525 std::string rep; |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
526 int num_tokens = tokens.size (); |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
527 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
528 if (num_tokens > 0) |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
529 { |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
530 // Determine replacement length |
29659
d13d090cb03a
use std::size_t and std::ptrdiff_t in C++ code (bug #60471)
John W. Eaton <jwe@octave.org>
parents:
29363
diff
changeset
|
531 const std::size_t replen = repstr.size () - 2*num_tokens; |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
532 int delta = 0; |
25337
3ff9192b676e
use auto keyword to declare iterator variables where possible
John W. Eaton <jwe@octave.org>
parents:
25166
diff
changeset
|
533 auto p = rx_lst.begin (); |
29659
d13d090cb03a
use std::size_t and std::ptrdiff_t in C++ code (bug #60471)
John W. Eaton <jwe@octave.org>
parents:
29363
diff
changeset
|
534 for (std::size_t i = 0; i < num_matches; i++) |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
535 { |
22860
0b1e25cc4457
eliminate use of OCTAVE_QUIT macro in C++ sources
John W. Eaton <jwe@octave.org>
parents:
22755
diff
changeset
|
536 octave_quit (); |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
537 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
538 double start = p->start (); |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
539 double end = p->end (); |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
540 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
541 const Matrix pairs (p->token_extents ()); |
29659
d13d090cb03a
use std::size_t and std::ptrdiff_t in C++ code (bug #60471)
John W. Eaton <jwe@octave.org>
parents:
29363
diff
changeset
|
542 std::size_t pairlen = 0; |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
543 for (int j = 0; j < num_tokens; j++) |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
544 { |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
545 if (tokens[j].num == 0) |
29659
d13d090cb03a
use std::size_t and std::ptrdiff_t in C++ code (bug #60471)
John W. Eaton <jwe@octave.org>
parents:
29363
diff
changeset
|
546 pairlen += static_cast<std::size_t> (end - start + 1); |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
547 else if (tokens[j].num <= pairs.rows ()) |
29659
d13d090cb03a
use std::size_t and std::ptrdiff_t in C++ code (bug #60471)
John W. Eaton <jwe@octave.org>
parents:
29363
diff
changeset
|
548 pairlen += static_cast<std::size_t> (pairs(tokens[j].num-1,1) |
24605
3e17190dfaea
Fix incorrect regexprep on ARM platforms (bug #52810).
Rik <rik@octave.org>
parents:
23083
diff
changeset
|
549 - pairs(tokens[j].num-1,0) |
3e17190dfaea
Fix incorrect regexprep on ARM platforms (bug #52810).
Rik <rik@octave.org>
parents:
23083
diff
changeset
|
550 + 1); |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
551 } |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
552 delta += (static_cast<int> (replen + pairlen) |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
553 - static_cast<int> (end - start + 1)); |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
554 p++; |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
555 } |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
556 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
557 // Build replacement string |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
558 rep.reserve (buffer.size () + delta); |
29659
d13d090cb03a
use std::size_t and std::ptrdiff_t in C++ code (bug #60471)
John W. Eaton <jwe@octave.org>
parents:
29363
diff
changeset
|
559 std::size_t from = 0; |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
560 p = rx_lst.begin (); |
29659
d13d090cb03a
use std::size_t and std::ptrdiff_t in C++ code (bug #60471)
John W. Eaton <jwe@octave.org>
parents:
29363
diff
changeset
|
561 for (std::size_t i = 0; i < num_matches; i++) |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
562 { |
22860
0b1e25cc4457
eliminate use of OCTAVE_QUIT macro in C++ sources
John W. Eaton <jwe@octave.org>
parents:
22755
diff
changeset
|
563 octave_quit (); |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
564 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
565 double start = p->start (); |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
566 double end = p->end (); |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
567 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
568 const Matrix pairs (p->token_extents ()); |
29659
d13d090cb03a
use std::size_t and std::ptrdiff_t in C++ code (bug #60471)
John W. Eaton <jwe@octave.org>
parents:
29363
diff
changeset
|
569 rep.append (&buffer[from], static_cast<std::size_t> (start - 1 - from)); |
d13d090cb03a
use std::size_t and std::ptrdiff_t in C++ code (bug #60471)
John W. Eaton <jwe@octave.org>
parents:
29363
diff
changeset
|
570 from = static_cast<std::size_t> (end); |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
571 |
29659
d13d090cb03a
use std::size_t and std::ptrdiff_t in C++ code (bug #60471)
John W. Eaton <jwe@octave.org>
parents:
29363
diff
changeset
|
572 std::size_t cur_pos = 0; |
15541
9db32cabeacf
Fix backslash handling in regexp pattern (Bug #37092)
Rik <rik@octave.org>
parents:
15271
diff
changeset
|
573 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
574 for (int j = 0; j < num_tokens; j++) |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
575 { |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
576 rep.append (&repstr[cur_pos], (tokens[j].pos) - cur_pos); |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
577 cur_pos = tokens[j].pos+2; |
15541
9db32cabeacf
Fix backslash handling in regexp pattern (Bug #37092)
Rik <rik@octave.org>
parents:
15271
diff
changeset
|
578 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
579 int k = tokens[j].num; |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
580 if (k == 0) |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
581 { |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
582 // replace with entire match |
29659
d13d090cb03a
use std::size_t and std::ptrdiff_t in C++ code (bug #60471)
John W. Eaton <jwe@octave.org>
parents:
29363
diff
changeset
|
583 rep.append (&buffer[static_cast<std::size_t> (end - 1)], |
d13d090cb03a
use std::size_t and std::ptrdiff_t in C++ code (bug #60471)
John W. Eaton <jwe@octave.org>
parents:
29363
diff
changeset
|
584 static_cast<std::size_t> (end - start + 1)); |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
585 } |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
586 else if (k <= pairs.rows ()) |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
587 { |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
588 // replace with group capture |
29659
d13d090cb03a
use std::size_t and std::ptrdiff_t in C++ code (bug #60471)
John W. Eaton <jwe@octave.org>
parents:
29363
diff
changeset
|
589 rep.append (&buffer[static_cast<std::size_t> (pairs(k-1,0)-1)], |
d13d090cb03a
use std::size_t and std::ptrdiff_t in C++ code (bug #60471)
John W. Eaton <jwe@octave.org>
parents:
29363
diff
changeset
|
590 static_cast<std::size_t> (pairs(k-1,1) |
24605
3e17190dfaea
Fix incorrect regexprep on ARM platforms (bug #52810).
Rik <rik@octave.org>
parents:
23083
diff
changeset
|
591 - pairs(k-1,0) + 1)); |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
592 } |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
593 else |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
594 { |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
595 // replace with nothing |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
596 } |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
597 } |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
598 if (cur_pos < repstr.size ()) |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
599 rep.append (&repstr[cur_pos], repstr.size () - cur_pos); |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
600 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
601 p++; |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
602 } |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
603 rep.append (&buffer[from], buffer.size () - from); |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
604 } |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
605 else |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
606 { |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
607 // Determine repstr length |
29659
d13d090cb03a
use std::size_t and std::ptrdiff_t in C++ code (bug #60471)
John W. Eaton <jwe@octave.org>
parents:
29363
diff
changeset
|
608 const std::size_t replen = repstr.size (); |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
609 int delta = 0; |
25337
3ff9192b676e
use auto keyword to declare iterator variables where possible
John W. Eaton <jwe@octave.org>
parents:
25166
diff
changeset
|
610 auto p = rx_lst.begin (); |
29659
d13d090cb03a
use std::size_t and std::ptrdiff_t in C++ code (bug #60471)
John W. Eaton <jwe@octave.org>
parents:
29363
diff
changeset
|
611 for (std::size_t i = 0; i < num_matches; i++) |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
612 { |
22860
0b1e25cc4457
eliminate use of OCTAVE_QUIT macro in C++ sources
John W. Eaton <jwe@octave.org>
parents:
22755
diff
changeset
|
613 octave_quit (); |
0b1e25cc4457
eliminate use of OCTAVE_QUIT macro in C++ sources
John W. Eaton <jwe@octave.org>
parents:
22755
diff
changeset
|
614 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
615 delta += static_cast<int> (replen) |
22402
4caa7b28d183
maint: Style check C++ code in liboctave/
Rik <rik@octave.org>
parents:
22333
diff
changeset
|
616 - static_cast<int> (p->end () - p->start () + 1); |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
617 p++; |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
618 } |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
619 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
620 // Build replacement string |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
621 rep.reserve (buffer.size () + delta); |
29659
d13d090cb03a
use std::size_t and std::ptrdiff_t in C++ code (bug #60471)
John W. Eaton <jwe@octave.org>
parents:
29363
diff
changeset
|
622 std::size_t from = 0; |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
623 p = rx_lst.begin (); |
29659
d13d090cb03a
use std::size_t and std::ptrdiff_t in C++ code (bug #60471)
John W. Eaton <jwe@octave.org>
parents:
29363
diff
changeset
|
624 for (std::size_t i = 0; i < num_matches; i++) |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
625 { |
22860
0b1e25cc4457
eliminate use of OCTAVE_QUIT macro in C++ sources
John W. Eaton <jwe@octave.org>
parents:
22755
diff
changeset
|
626 octave_quit (); |
0b1e25cc4457
eliminate use of OCTAVE_QUIT macro in C++ sources
John W. Eaton <jwe@octave.org>
parents:
22755
diff
changeset
|
627 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
628 rep.append (&buffer[from], |
29659
d13d090cb03a
use std::size_t and std::ptrdiff_t in C++ code (bug #60471)
John W. Eaton <jwe@octave.org>
parents:
29363
diff
changeset
|
629 static_cast<std::size_t> (p->start () - 1 - from)); |
d13d090cb03a
use std::size_t and std::ptrdiff_t in C++ code (bug #60471)
John W. Eaton <jwe@octave.org>
parents:
29363
diff
changeset
|
630 from = static_cast<std::size_t> (p->end ()); |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
631 rep.append (repstr); |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
632 p++; |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
633 } |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
634 rep.append (&buffer[from], buffer.size () - from); |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
635 } |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
636 |
22333
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
637 retval = rep; |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
638 return retval; |
2758af148ced
move base_list and regexp classes to octave namespace
John W. Eaton <jwe@octave.org>
parents:
22323
diff
changeset
|
639 } |
14024
fc9f204faea0
refactor regexp (bug #34440)
John W. Eaton <jwe@octave.org>
parents:
11586
diff
changeset
|
640 } |