zscan 0.1
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +7 -0
- data/ext/extconf.rb +3 -0
- data/ext/zscan.c +242 -0
- data/lib/zscan.rb +47 -0
- data/readme.md +52 -0
- data/zscan.gemspec +16 -0
- metadata +48 -0
checksums.yaml
ADDED
@@ -0,0 +1,7 @@
|
|
1
|
+
---
|
2
|
+
SHA1:
|
3
|
+
metadata.gz: 708c4bbc6710c4d02ed3def05702611c16cd187a
|
4
|
+
data.tar.gz: aaf8f91e2a98fcfb18d589d4e6bb6687aeb46ca9
|
5
|
+
SHA512:
|
6
|
+
metadata.gz: a322da85ae70a1b35f33d02a9d5862250f59058c37109aad9d59b12b91dbdf66068f9740f69570d40c85290bdebf291b108e8b1daa26669e90af748bffa52d45
|
7
|
+
data.tar.gz: 5489bccc470ae3d9870cc7887f41bc43887159c1666a093cb4273709f58afcf83ada19dacf74fd74edad918fb2269e131bb1b3e2c6f757af7fc3ad8e92b53837
|
data/ext/extconf.rb
ADDED
data/ext/zscan.c
ADDED
@@ -0,0 +1,242 @@
|
|
1
|
+
#include <ruby/ruby.h>
|
2
|
+
#include <ruby/re.h>
|
3
|
+
#include <ruby/encoding.h>
|
4
|
+
|
5
|
+
typedef struct {
|
6
|
+
size_t pos;
|
7
|
+
size_t bytepos;
|
8
|
+
} Pos;
|
9
|
+
|
10
|
+
typedef struct {
|
11
|
+
size_t pos;
|
12
|
+
size_t bytepos;
|
13
|
+
VALUE s;
|
14
|
+
struct re_registers regs;
|
15
|
+
size_t stack_i;
|
16
|
+
size_t stack_cap;
|
17
|
+
Pos* stack;
|
18
|
+
} ZScan;
|
19
|
+
|
20
|
+
#define P ZScan* p = rb_check_typeddata(self, &zscan_type)
|
21
|
+
|
22
|
+
static void zscan_mark(void* pp) {
|
23
|
+
ZScan* p = pp;
|
24
|
+
rb_gc_mark(p->s);
|
25
|
+
}
|
26
|
+
|
27
|
+
static void zscan_free(void* pp) {
|
28
|
+
ZScan* p = pp;
|
29
|
+
onig_region_free(&(p->regs), 0);
|
30
|
+
free(p->stack);
|
31
|
+
ruby_xfree(p);
|
32
|
+
}
|
33
|
+
|
34
|
+
extern size_t onig_region_memsize P_((const struct re_registers *regs));
|
35
|
+
static size_t zscan_memsize(const void* pp) {
|
36
|
+
const ZScan* p = pp;
|
37
|
+
return p ? sizeof(*p) - sizeof(p->regs) + onig_region_memsize(&p->regs) : 0;
|
38
|
+
}
|
39
|
+
|
40
|
+
static const rb_data_type_t zscan_type = {
|
41
|
+
"ZScan",
|
42
|
+
{zscan_mark, zscan_free, zscan_memsize}
|
43
|
+
};
|
44
|
+
|
45
|
+
static VALUE zscan_alloc(VALUE klass) {
|
46
|
+
ZScan* p = ALLOC(ZScan);
|
47
|
+
MEMZERO(p, ZScan, 1);
|
48
|
+
onig_region_init(&(p->regs));
|
49
|
+
p->s = Qnil;
|
50
|
+
p->stack_cap = 5;
|
51
|
+
p->stack = (Pos*)malloc(sizeof(Pos) * 5);
|
52
|
+
return TypedData_Wrap_Struct(klass, &zscan_type, p);
|
53
|
+
}
|
54
|
+
|
55
|
+
static VALUE zscan_internal_init(VALUE self, VALUE v_s) {
|
56
|
+
P;
|
57
|
+
p->s = v_s;
|
58
|
+
return self;
|
59
|
+
}
|
60
|
+
|
61
|
+
static VALUE zscan_internal_string(VALUE self) {
|
62
|
+
P;
|
63
|
+
return p->s;
|
64
|
+
}
|
65
|
+
|
66
|
+
static VALUE zscan_pos(VALUE self) {
|
67
|
+
P;
|
68
|
+
return ULONG2NUM(p->pos);
|
69
|
+
}
|
70
|
+
|
71
|
+
static VALUE zscan_advance(VALUE self, VALUE v_diff) {
|
72
|
+
P;
|
73
|
+
long signed_n = p->pos + NUM2LONG(v_diff);
|
74
|
+
if (signed_n < 0) {
|
75
|
+
p->pos = 0;
|
76
|
+
p->bytepos = 0;
|
77
|
+
return self;
|
78
|
+
}
|
79
|
+
size_t n = signed_n;
|
80
|
+
|
81
|
+
// because there's no "reverse scan" API, we have a O(n) routine :(
|
82
|
+
if (n < p->pos) {
|
83
|
+
p->pos = 0;
|
84
|
+
p->bytepos = 0;
|
85
|
+
}
|
86
|
+
|
87
|
+
if (n > p->pos) {
|
88
|
+
rb_encoding* enc = rb_enc_get(p->s);
|
89
|
+
size_t byteend = RSTRING_LEN(p->s);
|
90
|
+
char* ptr = RSTRING_PTR(p->s);
|
91
|
+
for (; p->pos < n && p->bytepos < byteend;) {
|
92
|
+
int n = rb_enc_mbclen(ptr + p->bytepos, ptr + byteend, enc);
|
93
|
+
if (n) {
|
94
|
+
p->pos++;
|
95
|
+
p->bytepos += n;
|
96
|
+
} else {
|
97
|
+
break;
|
98
|
+
}
|
99
|
+
}
|
100
|
+
}
|
101
|
+
return self;
|
102
|
+
}
|
103
|
+
|
104
|
+
static VALUE zscan_bytepos(VALUE self) {
|
105
|
+
P;
|
106
|
+
return ULONG2NUM(p->bytepos);
|
107
|
+
}
|
108
|
+
|
109
|
+
static VALUE zscan_bytepos_eq(VALUE self, VALUE v_bytepos) {
|
110
|
+
P;
|
111
|
+
size_t bytepos = NUM2ULONG(v_bytepos);
|
112
|
+
size_t from, to;
|
113
|
+
|
114
|
+
if (bytepos > p->bytepos) {
|
115
|
+
from = p->bytepos;
|
116
|
+
to = bytepos;
|
117
|
+
} else if (bytepos < p->bytepos) {
|
118
|
+
from = bytepos;
|
119
|
+
to = p->bytepos;
|
120
|
+
} else {
|
121
|
+
return v_bytepos;
|
122
|
+
}
|
123
|
+
|
124
|
+
rb_encoding* enc = rb_enc_get(p->s);
|
125
|
+
char* ptr = RSTRING_PTR(p->s);
|
126
|
+
size_t diff = 0;
|
127
|
+
for (; from < to;) {
|
128
|
+
int n = rb_enc_mbclen(ptr + from, ptr + to, enc);
|
129
|
+
if (n) {
|
130
|
+
diff++;
|
131
|
+
from += n;
|
132
|
+
} else {
|
133
|
+
if (from < to) {
|
134
|
+
rb_raise(rb_eRuntimeError, "the given bytepos splits character");
|
135
|
+
return v_bytepos;
|
136
|
+
}
|
137
|
+
break;
|
138
|
+
}
|
139
|
+
}
|
140
|
+
|
141
|
+
if (bytepos > p->bytepos) {
|
142
|
+
p->pos += diff;
|
143
|
+
} else if (bytepos < p->bytepos) {
|
144
|
+
p->pos -= diff;
|
145
|
+
}
|
146
|
+
p->bytepos = bytepos;
|
147
|
+
|
148
|
+
return v_bytepos;
|
149
|
+
}
|
150
|
+
|
151
|
+
static VALUE zscan_eos_p(VALUE self) {
|
152
|
+
P;
|
153
|
+
return (p->bytepos == (size_t)RSTRING_LEN(p->s) ? Qtrue : Qfalse);
|
154
|
+
}
|
155
|
+
|
156
|
+
regex_t *rb_reg_prepare_re(VALUE re, VALUE str);
|
157
|
+
static VALUE zscan_bmatch_p(VALUE self, VALUE pattern) {
|
158
|
+
P;
|
159
|
+
if (TYPE(pattern) == T_STRING) {
|
160
|
+
// todo
|
161
|
+
} else if (TYPE(pattern) == T_REGEXP) {
|
162
|
+
regex_t *re = rb_reg_prepare_re(pattern, p->s);
|
163
|
+
int tmpreg = re != RREGEXP(pattern)->ptr;
|
164
|
+
if (!tmpreg) RREGEXP(pattern)->usecnt++;
|
165
|
+
|
166
|
+
char* ptr = RSTRING_PTR(p->s);
|
167
|
+
UChar* ptr_end = (UChar*)(ptr + RSTRING_LEN(p->s));
|
168
|
+
UChar* ptr_match_from = (UChar*)(ptr + p->bytepos);
|
169
|
+
long ret = onig_match(re, (UChar*)ptr, ptr_end, ptr_match_from, &(p->regs), ONIG_OPTION_NONE);
|
170
|
+
|
171
|
+
if (!tmpreg) {
|
172
|
+
RREGEXP(pattern)->usecnt--;
|
173
|
+
}
|
174
|
+
if (tmpreg) {
|
175
|
+
if (RREGEXP(pattern)->usecnt) {
|
176
|
+
onig_free(re);
|
177
|
+
} else {
|
178
|
+
onig_free(RREGEXP(pattern)->ptr);
|
179
|
+
RREGEXP(pattern)->ptr = re;
|
180
|
+
}
|
181
|
+
}
|
182
|
+
|
183
|
+
if (ret == -2) {
|
184
|
+
rb_raise(rb_eRuntimeError, "regexp buffer overflow");
|
185
|
+
} else if (ret >= 0) {
|
186
|
+
return ULONG2NUM(p->regs.end[0]);
|
187
|
+
}
|
188
|
+
} else {
|
189
|
+
rb_raise(rb_eTypeError, "expect String or Regexp");
|
190
|
+
}
|
191
|
+
|
192
|
+
return Qnil;
|
193
|
+
}
|
194
|
+
|
195
|
+
static VALUE zscan_push_pos(VALUE self) {
|
196
|
+
P;
|
197
|
+
if (p->stack_i + 1 == p->stack_cap) {
|
198
|
+
p->stack_cap *= 2;
|
199
|
+
p->stack = (Pos*)realloc(p->stack, sizeof(Pos) * p->stack_cap);
|
200
|
+
}
|
201
|
+
Pos e = {p->pos, p->bytepos};
|
202
|
+
p->stack[++p->stack_i] = e;
|
203
|
+
return self;
|
204
|
+
}
|
205
|
+
|
206
|
+
static VALUE zscan_pop_pos(VALUE self) {
|
207
|
+
P;
|
208
|
+
if (p->stack_i) {
|
209
|
+
p->pos = p->stack[p->stack_i].pos;
|
210
|
+
p->bytepos = p->stack[p->stack_i].bytepos;
|
211
|
+
p->stack_i--;
|
212
|
+
} else {
|
213
|
+
p->pos = 0;
|
214
|
+
p->bytepos = 0;
|
215
|
+
}
|
216
|
+
return self;
|
217
|
+
}
|
218
|
+
|
219
|
+
static VALUE zscan_drop_top(VALUE self) {
|
220
|
+
P;
|
221
|
+
if (p->stack_i) {
|
222
|
+
p->stack_i--;
|
223
|
+
}
|
224
|
+
return self;
|
225
|
+
}
|
226
|
+
|
227
|
+
void Init_zscan() {
|
228
|
+
VALUE zscan = rb_define_class("ZScan", rb_cObject);
|
229
|
+
rb_define_alloc_func(zscan, zscan_alloc);
|
230
|
+
rb_define_method(zscan, "_internal_init", zscan_internal_init, 1);
|
231
|
+
rb_define_method(zscan, "_internal_string", zscan_internal_string, 0);
|
232
|
+
|
233
|
+
rb_define_method(zscan, "pos", zscan_pos, 0);
|
234
|
+
rb_define_method(zscan, "bytepos", zscan_bytepos, 0);
|
235
|
+
rb_define_method(zscan, "bytepos=", zscan_bytepos_eq, 1);
|
236
|
+
rb_define_method(zscan, "advance", zscan_advance, 1);
|
237
|
+
rb_define_method(zscan, "eos?", zscan_eos_p, 0);
|
238
|
+
rb_define_method(zscan, "bmatch?", zscan_bmatch_p, 1);
|
239
|
+
rb_define_method(zscan, "push_pos", zscan_push_pos, 0);
|
240
|
+
rb_define_method(zscan, "pop_pos", zscan_pop_pos, 0);
|
241
|
+
rb_define_method(zscan, "drop_top", zscan_drop_top, 0);
|
242
|
+
}
|
data/lib/zscan.rb
ADDED
@@ -0,0 +1,47 @@
|
|
1
|
+
require_relative "../ext/zscan"
|
2
|
+
|
3
|
+
class ZScan
|
4
|
+
VERSION = '0.1'
|
5
|
+
|
6
|
+
def initialize s, dup=false
|
7
|
+
_internal_init dup ? s.dup : s
|
8
|
+
end
|
9
|
+
|
10
|
+
def string
|
11
|
+
_internal_string.dup
|
12
|
+
end
|
13
|
+
|
14
|
+
def scan re_or_str
|
15
|
+
if sz = bmatch?(re_or_str)
|
16
|
+
r = _internal_string.byteslice bytepos, sz
|
17
|
+
self.bytepos += sz
|
18
|
+
r
|
19
|
+
end
|
20
|
+
end
|
21
|
+
|
22
|
+
def skip re_or_str
|
23
|
+
if sz = bmatch?(re_or_str)
|
24
|
+
self.bytepos += sz
|
25
|
+
end
|
26
|
+
end
|
27
|
+
|
28
|
+
def pos= new_pos
|
29
|
+
advance new_pos - pos
|
30
|
+
end
|
31
|
+
|
32
|
+
def rest
|
33
|
+
_internal_string.byteslice bytepos
|
34
|
+
end
|
35
|
+
|
36
|
+
private :_internal_init, :_internal_string
|
37
|
+
end
|
38
|
+
|
39
|
+
# coding: utf-8
|
40
|
+
if __FILE__ == $PROGRAM_NAME
|
41
|
+
z = ZScan.new 'ab你好'
|
42
|
+
z.push_pos
|
43
|
+
z.scan /ab你/
|
44
|
+
p z.pos
|
45
|
+
p z.bytepos
|
46
|
+
z.pop_pos
|
47
|
+
end
|
data/readme.md
ADDED
@@ -0,0 +1,52 @@
|
|
1
|
+
## Motivation
|
2
|
+
|
3
|
+
A simple string scanner. Provides... much less methods than `StringScanner`.
|
4
|
+
|
5
|
+
It supports either string or regexp as scan param.
|
6
|
+
|
7
|
+
`pos` is by codepoints instead of bytes, use `bytepos` to locate byte position.
|
8
|
+
|
9
|
+
It provides a position stack for you to efficiently manage scanning locations.
|
10
|
+
|
11
|
+
It correctly scans anchors. The following codes demonstrate the behavior:
|
12
|
+
|
13
|
+
```ruby
|
14
|
+
require 'zscan'
|
15
|
+
z = ZScan.new 'ab'
|
16
|
+
z.pos = 1
|
17
|
+
z.scan /(?<a)/ #=> ''
|
18
|
+
z.scan /^/ #=> nil
|
19
|
+
```
|
20
|
+
|
21
|
+
While with `StringScanner`:
|
22
|
+
|
23
|
+
```ruby
|
24
|
+
require 'strscan'
|
25
|
+
s = StringScanner.new 'ab'
|
26
|
+
s.pos = 1
|
27
|
+
s.scan /(?<a)/ #=> nil
|
28
|
+
s.scan /^/ #=> ''
|
29
|
+
```
|
30
|
+
|
31
|
+
See also https://bugs.ruby-lang.org/issues/7092
|
32
|
+
|
33
|
+
## Methods
|
34
|
+
|
35
|
+
- `ZScan.new string, dup=false`
|
36
|
+
- `scan regexp_or_string`
|
37
|
+
- `skip regexp_or_string`
|
38
|
+
- `bmatch? regexp_or_string` returns length of matched bytes or nil
|
39
|
+
- `eos?`
|
40
|
+
- `string` note: returns a COW dup
|
41
|
+
- `rest`
|
42
|
+
|
43
|
+
## Position management
|
44
|
+
|
45
|
+
- `pos`
|
46
|
+
- `pos= new_pos` note: complexity ~ `new_pos > pos ? new_pos - pos : new_pos`.
|
47
|
+
- `bytepos`
|
48
|
+
- `bytepos= new_bytepos` note: complexity ~ `abs(new_bytepos - bytepos)`.
|
49
|
+
- `advance n` move forward `n` codepoints, if `n < 0`, move backward. Stops at beginning or end.
|
50
|
+
- `push_pos` efficiently pushes current pos into the stack.
|
51
|
+
- `pop_pos` efficiently sets current pos to top of the stack, and pops it.
|
52
|
+
- `drop_top` drops top of pos stack without changing current pos.
|
data/zscan.gemspec
ADDED
@@ -0,0 +1,16 @@
|
|
1
|
+
Gem::Specification.new do |s|
|
2
|
+
s.name = "zscan"
|
3
|
+
s.version = "0.1"
|
4
|
+
s.author = "Zete Lui"
|
5
|
+
s.homepage = "https://github.com/luikore/zscan"
|
6
|
+
s.platform = Gem::Platform::RUBY
|
7
|
+
s.summary = "improved string scanner"
|
8
|
+
s.description = "improved string scanner"
|
9
|
+
s.required_ruby_version = ">=1.9.2"
|
10
|
+
|
11
|
+
s.files = %w"readme.md lib/zscan.rb ext/zscan.c ext/extconf.rb zscan.gemspec"
|
12
|
+
s.require_paths = ["lib"]
|
13
|
+
s.extensions = ["ext/extconf.rb"]
|
14
|
+
s.rubygems_version = '1.8.24'
|
15
|
+
s.has_rdoc = false
|
16
|
+
end
|
metadata
ADDED
@@ -0,0 +1,48 @@
|
|
1
|
+
--- !ruby/object:Gem::Specification
|
2
|
+
name: zscan
|
3
|
+
version: !ruby/object:Gem::Version
|
4
|
+
version: '0.1'
|
5
|
+
platform: ruby
|
6
|
+
authors:
|
7
|
+
- Zete Lui
|
8
|
+
autorequire:
|
9
|
+
bindir: bin
|
10
|
+
cert_chain: []
|
11
|
+
date: 2013-05-05 00:00:00.000000000 Z
|
12
|
+
dependencies: []
|
13
|
+
description: improved string scanner
|
14
|
+
email:
|
15
|
+
executables: []
|
16
|
+
extensions:
|
17
|
+
- ext/extconf.rb
|
18
|
+
extra_rdoc_files: []
|
19
|
+
files:
|
20
|
+
- readme.md
|
21
|
+
- lib/zscan.rb
|
22
|
+
- ext/zscan.c
|
23
|
+
- ext/extconf.rb
|
24
|
+
- zscan.gemspec
|
25
|
+
homepage: https://github.com/luikore/zscan
|
26
|
+
licenses: []
|
27
|
+
metadata: {}
|
28
|
+
post_install_message:
|
29
|
+
rdoc_options: []
|
30
|
+
require_paths:
|
31
|
+
- lib
|
32
|
+
required_ruby_version: !ruby/object:Gem::Requirement
|
33
|
+
requirements:
|
34
|
+
- - '>='
|
35
|
+
- !ruby/object:Gem::Version
|
36
|
+
version: 1.9.2
|
37
|
+
required_rubygems_version: !ruby/object:Gem::Requirement
|
38
|
+
requirements:
|
39
|
+
- - '>='
|
40
|
+
- !ruby/object:Gem::Version
|
41
|
+
version: '0'
|
42
|
+
requirements: []
|
43
|
+
rubyforge_project:
|
44
|
+
rubygems_version: 2.0.3
|
45
|
+
signing_key:
|
46
|
+
specification_version: 4
|
47
|
+
summary: improved string scanner
|
48
|
+
test_files: []
|