bzip2-ruby-rb20 0.2.7
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +7 -0
- data/.gitignore +10 -0
- data/.rspec +2 -0
- data/.travis.yml +17 -0
- data/.yardopts +1 -0
- data/CHANGELOG.md +94 -0
- data/Gemfile +3 -0
- data/README.md +69 -0
- data/Rakefile +35 -0
- data/bzip2-ruby.gemspec +21 -0
- data/ext/bzip2/bzip2.c +250 -0
- data/ext/bzip2/common.c +55 -0
- data/ext/bzip2/common.h +76 -0
- data/ext/bzip2/extconf.rb +21 -0
- data/ext/bzip2/reader.c +1032 -0
- data/ext/bzip2/reader.h +35 -0
- data/ext/bzip2/writer.c +453 -0
- data/ext/bzip2/writer.h +22 -0
- data/lib/bzip2.rb +13 -0
- data/lib/bzip2/internals.rb +13 -0
- data/lib/bzip2/reader.rb +27 -0
- data/lib/bzip2/version.rb +3 -0
- data/lib/bzip2/writer.rb +64 -0
- data/spec/reader_spec.rb +261 -0
- data/spec/spec_helper.rb +19 -0
- data/spec/writer_spec.rb +134 -0
- metadata +103 -0
data/ext/bzip2/common.c
ADDED
@@ -0,0 +1,55 @@
|
|
1
|
+
#include <ruby.h>
|
2
|
+
#include <bzlib.h>
|
3
|
+
|
4
|
+
#include "common.h"
|
5
|
+
|
6
|
+
void bz_file_mark(struct bz_file * bzf) {
|
7
|
+
rb_gc_mark(bzf->io);
|
8
|
+
rb_gc_mark(bzf->in);
|
9
|
+
}
|
10
|
+
|
11
|
+
void * bz_malloc(void *opaque, int m, int n) {
|
12
|
+
return malloc(m * n);
|
13
|
+
}
|
14
|
+
|
15
|
+
void bz_free(void *opaque, void *p) {
|
16
|
+
free(p);
|
17
|
+
}
|
18
|
+
|
19
|
+
VALUE bz_raise(int error) {
|
20
|
+
VALUE exc;
|
21
|
+
const char *msg;
|
22
|
+
|
23
|
+
exc = bz_eError;
|
24
|
+
switch (error) {
|
25
|
+
case BZ_SEQUENCE_ERROR:
|
26
|
+
msg = "incorrect sequence";
|
27
|
+
break;
|
28
|
+
case BZ_PARAM_ERROR:
|
29
|
+
msg = "parameter out of range";
|
30
|
+
break;
|
31
|
+
case BZ_MEM_ERROR:
|
32
|
+
msg = "not enough memory is available";
|
33
|
+
break;
|
34
|
+
case BZ_DATA_ERROR:
|
35
|
+
msg = "data integrity error is detected";
|
36
|
+
break;
|
37
|
+
case BZ_DATA_ERROR_MAGIC:
|
38
|
+
msg = "compressed stream does not start with the correct magic bytes";
|
39
|
+
break;
|
40
|
+
case BZ_IO_ERROR:
|
41
|
+
msg = "error reading or writing";
|
42
|
+
break;
|
43
|
+
case BZ_UNEXPECTED_EOF:
|
44
|
+
exc = bz_eEOZError;
|
45
|
+
msg = "compressed file finishes before the logical end of stream is detected";
|
46
|
+
break;
|
47
|
+
case BZ_OUTBUFF_FULL:
|
48
|
+
msg = "output buffer full";
|
49
|
+
break;
|
50
|
+
default:
|
51
|
+
msg = "unknown error";
|
52
|
+
exc = bz_eError;
|
53
|
+
}
|
54
|
+
rb_raise(exc, "%s", msg);
|
55
|
+
}
|
data/ext/bzip2/common.h
ADDED
@@ -0,0 +1,76 @@
|
|
1
|
+
#ifndef _RB_BZIP2_COMMON_H_
|
2
|
+
#define _RB_BZIP2_COMMON_H_
|
3
|
+
|
4
|
+
#include <ruby.h>
|
5
|
+
#include <bzlib.h>
|
6
|
+
|
7
|
+
#ifndef RUBY_19_COMPATIBILITY
|
8
|
+
# include <rubyio.h>
|
9
|
+
# include <version.h>
|
10
|
+
#else
|
11
|
+
# include <ruby/io.h>
|
12
|
+
#endif
|
13
|
+
|
14
|
+
#define BZ2_RB_CLOSE 1
|
15
|
+
#define BZ2_RB_INTERNAL 2
|
16
|
+
|
17
|
+
#define BZ_RB_BLOCKSIZE 4096
|
18
|
+
#define DEFAULT_BLOCKS 9
|
19
|
+
#define ASIZE (1 << CHAR_BIT)
|
20
|
+
|
21
|
+
/* Older versions of Ruby (< 1.8.6) need these */
|
22
|
+
#ifndef RSTRING_PTR
|
23
|
+
# define RSTRING_PTR(s) (RSTRING(s)->ptr)
|
24
|
+
#endif
|
25
|
+
#ifndef RSTRING_LEN
|
26
|
+
# define RSTRING_LEN(s) (RSTRING(s)->len)
|
27
|
+
#endif
|
28
|
+
#ifndef RARRAY_PTR
|
29
|
+
# define RARRAY_PTR(s) (RARRAY(s)->ptr)
|
30
|
+
#endif
|
31
|
+
#ifndef RARRAY_LEN
|
32
|
+
# define RARRAY_LEN(s) (RARRAY(s)->len)
|
33
|
+
#endif
|
34
|
+
|
35
|
+
struct bz_file {
|
36
|
+
bz_stream bzs;
|
37
|
+
VALUE in, io;
|
38
|
+
char *buf;
|
39
|
+
unsigned int buflen;
|
40
|
+
int blocks, work, small;
|
41
|
+
int flags, lineno, state;
|
42
|
+
};
|
43
|
+
|
44
|
+
struct bz_str {
|
45
|
+
VALUE str;
|
46
|
+
int pos;
|
47
|
+
};
|
48
|
+
|
49
|
+
struct bz_iv {
|
50
|
+
VALUE bz2, io;
|
51
|
+
void (*finalize)();
|
52
|
+
};
|
53
|
+
|
54
|
+
#define Get_BZ2(obj, bzf) \
|
55
|
+
rb_io_taint_check(obj); \
|
56
|
+
Data_Get_Struct(obj, struct bz_file, bzf); \
|
57
|
+
if (!RTEST(bzf->io)) { \
|
58
|
+
rb_raise(rb_eIOError, "closed IO"); \
|
59
|
+
}
|
60
|
+
|
61
|
+
#ifndef ASDFasdf
|
62
|
+
extern VALUE bz_cWriter, bz_cReader, bz_cInternal;
|
63
|
+
extern VALUE bz_eError, bz_eEOZError;
|
64
|
+
|
65
|
+
extern VALUE bz_internal_ary;
|
66
|
+
|
67
|
+
extern ID id_new, id_write, id_open, id_flush, id_read;
|
68
|
+
extern ID id_closed, id_close, id_str;
|
69
|
+
#endif
|
70
|
+
|
71
|
+
void bz_file_mark(struct bz_file * bzf);
|
72
|
+
void* bz_malloc(void *opaque, int m, int n);
|
73
|
+
void bz_free(void *opaque, void *p);
|
74
|
+
VALUE bz_raise(int err);
|
75
|
+
|
76
|
+
#endif
|
@@ -0,0 +1,21 @@
|
|
1
|
+
# encoding: UTF-8
|
2
|
+
require 'mkmf'
|
3
|
+
dir_config('bz2')
|
4
|
+
have_header('bzlib.h')
|
5
|
+
|
6
|
+
$CFLAGS << ' -Wall -Wextra -Wno-unused -funroll-loops '
|
7
|
+
# $CFLAGS << ' -O0 -ggdb -Wextra'
|
8
|
+
|
9
|
+
if have_library("bz2", "BZ2_bzWriteOpen")
|
10
|
+
if enable_config("shared", true)
|
11
|
+
$static = nil
|
12
|
+
end
|
13
|
+
|
14
|
+
if RUBY_VERSION.to_f >= 1.9
|
15
|
+
$CFLAGS << ' -DRUBY_19_COMPATIBILITY'
|
16
|
+
end
|
17
|
+
|
18
|
+
create_makefile('bzip2/bzip2')
|
19
|
+
else
|
20
|
+
puts "libbz2 not found, maybe try manually specifying --with-bz2-dir to find it?"
|
21
|
+
end
|
data/ext/bzip2/reader.c
ADDED
@@ -0,0 +1,1032 @@
|
|
1
|
+
#include <bzlib.h>
|
2
|
+
#include <ruby.h>
|
3
|
+
|
4
|
+
#include "reader.h"
|
5
|
+
#include "common.h"
|
6
|
+
|
7
|
+
void bz_str_mark(struct bz_str *bzs) {
|
8
|
+
rb_gc_mark(bzs->str);
|
9
|
+
}
|
10
|
+
|
11
|
+
struct bz_file * bz_get_bzf(VALUE obj) {
|
12
|
+
struct bz_file *bzf;
|
13
|
+
|
14
|
+
Get_BZ2(obj, bzf);
|
15
|
+
if (!bzf->buf) {
|
16
|
+
if (bzf->state != BZ_OK) {
|
17
|
+
bz_raise(bzf->state);
|
18
|
+
}
|
19
|
+
bzf->state = BZ2_bzDecompressInit(&(bzf->bzs), 0, bzf->small);
|
20
|
+
if (bzf->state != BZ_OK) {
|
21
|
+
BZ2_bzDecompressEnd(&(bzf->bzs));
|
22
|
+
bz_raise(bzf->state);
|
23
|
+
}
|
24
|
+
bzf->buf = ALLOC_N(char, BZ_RB_BLOCKSIZE + 1);
|
25
|
+
bzf->buflen = BZ_RB_BLOCKSIZE;
|
26
|
+
bzf->buf[0] = bzf->buf[bzf->buflen] = '\0';
|
27
|
+
bzf->bzs.total_out_hi32 = bzf->bzs.total_out_lo32 = 0;
|
28
|
+
bzf->bzs.next_out = bzf->buf;
|
29
|
+
bzf->bzs.avail_out = 0;
|
30
|
+
}
|
31
|
+
if (bzf->state == BZ_STREAM_END && !bzf->bzs.avail_out) {
|
32
|
+
return 0;
|
33
|
+
}
|
34
|
+
return bzf;
|
35
|
+
}
|
36
|
+
|
37
|
+
int bz_next_available(struct bz_file *bzf, int in){
|
38
|
+
bzf->bzs.next_out = bzf->buf;
|
39
|
+
bzf->bzs.avail_out = 0;
|
40
|
+
if (bzf->state == BZ_STREAM_END) {
|
41
|
+
return BZ_STREAM_END;
|
42
|
+
}
|
43
|
+
if (!bzf->bzs.avail_in) {
|
44
|
+
bzf->in = rb_funcall(bzf->io, id_read, 1, INT2FIX(1024));
|
45
|
+
if (TYPE(bzf->in) != T_STRING || RSTRING_LEN(bzf->in) == 0) {
|
46
|
+
BZ2_bzDecompressEnd(&(bzf->bzs));
|
47
|
+
bzf->bzs.avail_out = 0;
|
48
|
+
bzf->state = BZ_UNEXPECTED_EOF;
|
49
|
+
bz_raise(bzf->state);
|
50
|
+
}
|
51
|
+
bzf->bzs.next_in = RSTRING_PTR(bzf->in);
|
52
|
+
bzf->bzs.avail_in = (int) RSTRING_LEN(bzf->in);
|
53
|
+
}
|
54
|
+
if ((bzf->buflen - in) < (BZ_RB_BLOCKSIZE / 2)) {
|
55
|
+
bzf->buf = REALLOC_N(bzf->buf, char, bzf->buflen+BZ_RB_BLOCKSIZE+1);
|
56
|
+
bzf->buflen += BZ_RB_BLOCKSIZE;
|
57
|
+
bzf->buf[bzf->buflen] = '\0';
|
58
|
+
}
|
59
|
+
bzf->bzs.avail_out = bzf->buflen - in;
|
60
|
+
bzf->bzs.next_out = bzf->buf + in;
|
61
|
+
bzf->state = BZ2_bzDecompress(&(bzf->bzs));
|
62
|
+
if (bzf->state != BZ_OK) {
|
63
|
+
BZ2_bzDecompressEnd(&(bzf->bzs));
|
64
|
+
if (bzf->state != BZ_STREAM_END) {
|
65
|
+
bzf->bzs.avail_out = 0;
|
66
|
+
bz_raise(bzf->state);
|
67
|
+
}
|
68
|
+
}
|
69
|
+
bzf->bzs.avail_out = bzf->buflen - bzf->bzs.avail_out;
|
70
|
+
bzf->bzs.next_out = bzf->buf;
|
71
|
+
return 0;
|
72
|
+
}
|
73
|
+
|
74
|
+
VALUE bz_read_until(struct bz_file *bzf, const char *str, int len, int *td1) {
|
75
|
+
VALUE res;
|
76
|
+
int total, i, nex = 0;
|
77
|
+
char *p, *t, *tx, *end, *pend = ((char*) str) + len;
|
78
|
+
|
79
|
+
res = rb_str_new(0, 0);
|
80
|
+
while (1) {
|
81
|
+
total = bzf->bzs.avail_out;
|
82
|
+
if (len == 1) {
|
83
|
+
tx = memchr(bzf->bzs.next_out, *str, bzf->bzs.avail_out);
|
84
|
+
if (tx) {
|
85
|
+
i = (int)(tx - bzf->bzs.next_out + len);
|
86
|
+
res = rb_str_cat(res, bzf->bzs.next_out, i);
|
87
|
+
bzf->bzs.next_out += i;
|
88
|
+
bzf->bzs.avail_out -= i;
|
89
|
+
return res;
|
90
|
+
}
|
91
|
+
} else {
|
92
|
+
tx = bzf->bzs.next_out;
|
93
|
+
end = bzf->bzs.next_out + bzf->bzs.avail_out;
|
94
|
+
while (tx + len <= end) {
|
95
|
+
for (p = (char*) str, t = tx; p != pend; ++p, ++t) {
|
96
|
+
if (*p != *t) break;
|
97
|
+
}
|
98
|
+
if (p == pend) {
|
99
|
+
i = (int)(tx - bzf->bzs.next_out + len);
|
100
|
+
res = rb_str_cat(res, bzf->bzs.next_out, i);
|
101
|
+
bzf->bzs.next_out += i;
|
102
|
+
bzf->bzs.avail_out -= i;
|
103
|
+
return res;
|
104
|
+
}
|
105
|
+
if (td1) {
|
106
|
+
tx += td1[(int)*(tx + len)];
|
107
|
+
} else {
|
108
|
+
tx += 1;
|
109
|
+
}
|
110
|
+
}
|
111
|
+
}
|
112
|
+
nex = 0;
|
113
|
+
if (total) {
|
114
|
+
nex = len - 1;
|
115
|
+
res = rb_str_cat(res, bzf->bzs.next_out, total - nex);
|
116
|
+
if (nex) {
|
117
|
+
MEMMOVE(bzf->buf, bzf->bzs.next_out + total - nex, char, nex);
|
118
|
+
}
|
119
|
+
}
|
120
|
+
if (bz_next_available(bzf, nex) == BZ_STREAM_END) {
|
121
|
+
if (nex) {
|
122
|
+
res = rb_str_cat(res, bzf->buf, nex);
|
123
|
+
}
|
124
|
+
if (RSTRING_LEN(res)) {
|
125
|
+
return res;
|
126
|
+
}
|
127
|
+
return Qnil;
|
128
|
+
}
|
129
|
+
}
|
130
|
+
return Qnil;
|
131
|
+
}
|
132
|
+
|
133
|
+
/**
|
134
|
+
* Reads a stream as long as the next character is equal to the specified
|
135
|
+
* character
|
136
|
+
*
|
137
|
+
* Returns the next character in the sequence that's not the same as the one
|
138
|
+
* given or EOF if it's there until the end of the file.
|
139
|
+
*/
|
140
|
+
int bz_read_while(struct bz_file *bzf, char c) {
|
141
|
+
char *end;
|
142
|
+
|
143
|
+
while (1) {
|
144
|
+
end = bzf->bzs.next_out + bzf->bzs.avail_out;
|
145
|
+
while (bzf->bzs.next_out < end) {
|
146
|
+
if (c != *bzf->bzs.next_out) {
|
147
|
+
return *bzf->bzs.next_out;
|
148
|
+
}
|
149
|
+
++bzf->bzs.next_out;
|
150
|
+
--bzf->bzs.avail_out;
|
151
|
+
}
|
152
|
+
if (bz_next_available(bzf, 0) == BZ_STREAM_END) {
|
153
|
+
return EOF;
|
154
|
+
}
|
155
|
+
}
|
156
|
+
return EOF;
|
157
|
+
}
|
158
|
+
|
159
|
+
/*
|
160
|
+
* Internally allocates data for a new Reader
|
161
|
+
* @private
|
162
|
+
*/
|
163
|
+
VALUE bz_reader_s_alloc(VALUE obj) {
|
164
|
+
struct bz_file *bzf;
|
165
|
+
VALUE res;
|
166
|
+
res = Data_Make_Struct(obj, struct bz_file, bz_file_mark, free, bzf);
|
167
|
+
bzf->bzs.bzalloc = bz_malloc;
|
168
|
+
bzf->bzs.bzfree = bz_free;
|
169
|
+
bzf->blocks = DEFAULT_BLOCKS;
|
170
|
+
bzf->state = BZ_OK;
|
171
|
+
return res;
|
172
|
+
}
|
173
|
+
|
174
|
+
VALUE bz_reader_close __((VALUE));
|
175
|
+
|
176
|
+
/*
|
177
|
+
* call-seq:
|
178
|
+
* open(filename, &block=nil) -> Bzip2::Reader
|
179
|
+
*
|
180
|
+
* @param [String] filename the name of the file to read from
|
181
|
+
* @yieldparam [Bzip2::Reader] reader the Bzip2::Reader instance
|
182
|
+
*
|
183
|
+
* If a block is given, the created Bzip2::Reader instance is yielded to the
|
184
|
+
* block and will be closed when the block completes. It is guaranteed via
|
185
|
+
* +ensure+ that the reader is closed
|
186
|
+
*
|
187
|
+
* If a block is not given, a Bzip2::Reader instance will be returned
|
188
|
+
*
|
189
|
+
* Bzip2::Reader.open('file') { |f| puts f.gets }
|
190
|
+
*
|
191
|
+
* reader = Bzip2::Reader.open('file')
|
192
|
+
* puts reader.gets
|
193
|
+
* reader.close
|
194
|
+
*
|
195
|
+
* @return [Bzip2::Reader, nil]
|
196
|
+
*/
|
197
|
+
VALUE bz_reader_s_open(int argc, VALUE *argv, VALUE obj) {
|
198
|
+
VALUE res;
|
199
|
+
struct bz_file *bzf;
|
200
|
+
|
201
|
+
if (argc < 1) {
|
202
|
+
rb_raise(rb_eArgError, "invalid number of arguments");
|
203
|
+
}
|
204
|
+
argv[0] = rb_funcall2(rb_mKernel, id_open, 1, argv);
|
205
|
+
if (NIL_P(argv[0])) {
|
206
|
+
return Qnil;
|
207
|
+
}
|
208
|
+
res = rb_funcall2(obj, id_new, argc, argv);
|
209
|
+
Data_Get_Struct(res, struct bz_file, bzf);
|
210
|
+
bzf->flags |= BZ2_RB_CLOSE;
|
211
|
+
if (rb_block_given_p()) {
|
212
|
+
return rb_ensure(rb_yield, res, bz_reader_close, res);
|
213
|
+
}
|
214
|
+
return res;
|
215
|
+
}
|
216
|
+
|
217
|
+
/*
|
218
|
+
* call-seq:
|
219
|
+
* initialize(io)
|
220
|
+
*
|
221
|
+
* Creates a new stream for reading a bzip file or string
|
222
|
+
*
|
223
|
+
* @param [File, string, #read] io the source for input data. If the source is
|
224
|
+
* a file or something responding to #read, then data will be read via #read,
|
225
|
+
* otherwise if the input is a string it will be taken as the literal data
|
226
|
+
* to decompress
|
227
|
+
*/
|
228
|
+
VALUE bz_reader_init(int argc, VALUE *argv, VALUE obj) {
|
229
|
+
struct bz_file *bzf;
|
230
|
+
int small = 0;
|
231
|
+
VALUE a, b;
|
232
|
+
int internal = 0;
|
233
|
+
|
234
|
+
if (rb_scan_args(argc, argv, "11", &a, &b) == 2) {
|
235
|
+
small = RTEST(b);
|
236
|
+
}
|
237
|
+
rb_io_taint_check(a);
|
238
|
+
if (OBJ_TAINTED(a)) {
|
239
|
+
OBJ_TAINT(obj);
|
240
|
+
}
|
241
|
+
if (rb_respond_to(a, id_read)) {
|
242
|
+
if (TYPE(a) == T_FILE) {
|
243
|
+
#ifndef RUBY_19_COMPATIBILITY
|
244
|
+
OpenFile *fptr;
|
245
|
+
#else
|
246
|
+
rb_io_t *fptr;
|
247
|
+
#endif
|
248
|
+
|
249
|
+
GetOpenFile(a, fptr);
|
250
|
+
rb_io_check_readable(fptr);
|
251
|
+
} else if (rb_respond_to(a, id_closed)) {
|
252
|
+
VALUE iv = rb_funcall2(a, id_closed, 0, 0);
|
253
|
+
if (RTEST(iv)) {
|
254
|
+
rb_raise(rb_eArgError, "closed object");
|
255
|
+
}
|
256
|
+
}
|
257
|
+
} else {
|
258
|
+
struct bz_str *bzs;
|
259
|
+
VALUE res;
|
260
|
+
|
261
|
+
if (!rb_respond_to(a, id_str)) {
|
262
|
+
rb_raise(rb_eArgError, "first argument must respond to #read");
|
263
|
+
}
|
264
|
+
a = rb_funcall2(a, id_str, 0, 0);
|
265
|
+
if (TYPE(a) != T_STRING) {
|
266
|
+
rb_raise(rb_eArgError, "#to_str must return a String");
|
267
|
+
}
|
268
|
+
res = Data_Make_Struct(bz_cInternal, struct bz_str,
|
269
|
+
bz_str_mark, free, bzs);
|
270
|
+
bzs->str = a;
|
271
|
+
a = res;
|
272
|
+
internal = BZ2_RB_INTERNAL;
|
273
|
+
}
|
274
|
+
Data_Get_Struct(obj, struct bz_file, bzf);
|
275
|
+
bzf->io = a;
|
276
|
+
bzf->small = small;
|
277
|
+
bzf->flags |= internal;
|
278
|
+
return obj;
|
279
|
+
}
|
280
|
+
|
281
|
+
/*
|
282
|
+
* call-seq:
|
283
|
+
* read(len = nil)
|
284
|
+
*
|
285
|
+
* Read decompressed data from the stream.
|
286
|
+
*
|
287
|
+
* Bzip2::Reader.new(Bzip2.compress('ab')).read # => "ab"
|
288
|
+
* Bzip2::Reader.new(Bzip2.compress('ab')).read(1) # => "a"
|
289
|
+
*
|
290
|
+
* @return [String, nil] the decompressed data read or +nil+ if eoz has been
|
291
|
+
* reached
|
292
|
+
* @param [Integer] len the number of decompressed bytes which should be read.
|
293
|
+
* If nothing is specified, the entire stream is read
|
294
|
+
*/
|
295
|
+
VALUE bz_reader_read(int argc, VALUE *argv, VALUE obj) {
|
296
|
+
struct bz_file *bzf;
|
297
|
+
VALUE res, length;
|
298
|
+
int total;
|
299
|
+
int n;
|
300
|
+
|
301
|
+
rb_scan_args(argc, argv, "01", &length);
|
302
|
+
if (NIL_P(length)) {
|
303
|
+
n = -1;
|
304
|
+
} else {
|
305
|
+
n = NUM2INT(length);
|
306
|
+
if (n < 0) {
|
307
|
+
rb_raise(rb_eArgError, "negative length %d given", n);
|
308
|
+
}
|
309
|
+
}
|
310
|
+
bzf = bz_get_bzf(obj);
|
311
|
+
if (!bzf) {
|
312
|
+
return Qnil;
|
313
|
+
}
|
314
|
+
res = rb_str_new(0, 0);
|
315
|
+
if (OBJ_TAINTED(obj)) {
|
316
|
+
OBJ_TAINT(res);
|
317
|
+
}
|
318
|
+
if (n == 0) {
|
319
|
+
free(bzf->buf);
|
320
|
+
return res;
|
321
|
+
}
|
322
|
+
while (1) {
|
323
|
+
total = bzf->bzs.avail_out;
|
324
|
+
if (n != -1 && (RSTRING_LEN(res) + total) >= n) {
|
325
|
+
n -= (int) RSTRING_LEN(res);
|
326
|
+
res = rb_str_cat(res, bzf->bzs.next_out, n);
|
327
|
+
bzf->bzs.next_out += n;
|
328
|
+
bzf->bzs.avail_out -= n;
|
329
|
+
free(bzf->buf);
|
330
|
+
return res;
|
331
|
+
}
|
332
|
+
if (total) {
|
333
|
+
res = rb_str_cat(res, bzf->bzs.next_out, total);
|
334
|
+
}
|
335
|
+
if (bz_next_available(bzf, 0) == BZ_STREAM_END) {
|
336
|
+
free(bzf->buf);
|
337
|
+
return res;
|
338
|
+
}
|
339
|
+
}
|
340
|
+
return Qnil;
|
341
|
+
}
|
342
|
+
|
343
|
+
int bz_getc(VALUE obj) {
|
344
|
+
VALUE length = INT2FIX(1);
|
345
|
+
VALUE res = bz_reader_read(1, &length, obj);
|
346
|
+
if (NIL_P(res) || RSTRING_LEN(res) == 0) {
|
347
|
+
return EOF;
|
348
|
+
}
|
349
|
+
return RSTRING_PTR(res)[0];
|
350
|
+
}
|
351
|
+
|
352
|
+
/*
|
353
|
+
* call-seq:
|
354
|
+
* ungetc(byte)
|
355
|
+
*
|
356
|
+
* "Ungets" a character/byte. This rewinds the stream by 1 character and inserts
|
357
|
+
* the given character into that position. The next read will return the given
|
358
|
+
* character as the first one read
|
359
|
+
*
|
360
|
+
* reader = Bzip2::Reader.new Bzip2.compress('abc')
|
361
|
+
* reader.getc # => 97
|
362
|
+
* reader.ungetc 97 # => nil
|
363
|
+
* reader.getc # => 97
|
364
|
+
* reader.ungetc 42 # => nil
|
365
|
+
* reader.getc # => 42
|
366
|
+
* reader.getc # => 98
|
367
|
+
* reader.getc # => 99
|
368
|
+
* reader.ungetc 100 # => nil
|
369
|
+
* reader.getc # => 100
|
370
|
+
*
|
371
|
+
* @param [Integer] byte the byte to 'unget'
|
372
|
+
* @return [nil] always
|
373
|
+
*/
|
374
|
+
VALUE bz_reader_ungetc(VALUE obj, VALUE a) {
|
375
|
+
struct bz_file *bzf;
|
376
|
+
int c = NUM2INT(a);
|
377
|
+
|
378
|
+
Get_BZ2(obj, bzf);
|
379
|
+
if (!bzf->buf) {
|
380
|
+
bz_raise(BZ_SEQUENCE_ERROR);
|
381
|
+
}
|
382
|
+
if (bzf->bzs.avail_out < bzf->buflen) {
|
383
|
+
bzf->bzs.next_out -= 1;
|
384
|
+
bzf->bzs.next_out[0] = c;
|
385
|
+
bzf->bzs.avail_out += 1;
|
386
|
+
} else {
|
387
|
+
bzf->buf = REALLOC_N(bzf->buf, char, bzf->buflen + 2);
|
388
|
+
bzf->buf[bzf->buflen++] = c;
|
389
|
+
bzf->buf[bzf->buflen] = '\0';
|
390
|
+
bzf->bzs.next_out = bzf->buf;
|
391
|
+
bzf->bzs.avail_out = bzf->buflen;
|
392
|
+
}
|
393
|
+
return Qnil;
|
394
|
+
}
|
395
|
+
|
396
|
+
/*
|
397
|
+
* call-seq:
|
398
|
+
* ungets(str)
|
399
|
+
*
|
400
|
+
* Equivalently "unget" a string. When called on a string that was just read
|
401
|
+
* from the stream, this inserts the string back into the stream to br read
|
402
|
+
* again.
|
403
|
+
*
|
404
|
+
* When called with a string which hasn't been read from the stream, it does
|
405
|
+
* the same thing, and the next read line/data will start from the beginning
|
406
|
+
* of the given data and the continue on with the rest of the stream.
|
407
|
+
*
|
408
|
+
* reader = Bzip2::Reader.new Bzip2.compress("a\nb")
|
409
|
+
* reader.gets # => "a\n"
|
410
|
+
* reader.ungets "a\n" # => nil
|
411
|
+
* reader.gets # => "a\n"
|
412
|
+
* reader.ungets "foo" # => nil
|
413
|
+
* reader.gets # => "foob"
|
414
|
+
*
|
415
|
+
* @param [String] str the string to insert back into the stream
|
416
|
+
* @return [nil] always
|
417
|
+
*/
|
418
|
+
VALUE bz_reader_ungets(VALUE obj, VALUE a) {
|
419
|
+
struct bz_file *bzf;
|
420
|
+
|
421
|
+
Check_Type(a, T_STRING);
|
422
|
+
Get_BZ2(obj, bzf);
|
423
|
+
if (!bzf->buf) {
|
424
|
+
bz_raise(BZ_SEQUENCE_ERROR);
|
425
|
+
}
|
426
|
+
if ((bzf->bzs.avail_out + RSTRING_LEN(a)) < bzf->buflen) {
|
427
|
+
bzf->bzs.next_out -= RSTRING_LEN(a);
|
428
|
+
MEMCPY(bzf->bzs.next_out, RSTRING_PTR(a), char, RSTRING_LEN(a));
|
429
|
+
bzf->bzs.avail_out += (int) RSTRING_LEN(a);
|
430
|
+
} else {
|
431
|
+
bzf->buf = REALLOC_N(bzf->buf, char, bzf->buflen + RSTRING_LEN(a) + 1);
|
432
|
+
MEMCPY(bzf->buf + bzf->buflen, RSTRING_PTR(a), char,RSTRING_LEN(a));
|
433
|
+
bzf->buflen += (int) RSTRING_LEN(a);
|
434
|
+
bzf->buf[bzf->buflen] = '\0';
|
435
|
+
bzf->bzs.next_out = bzf->buf;
|
436
|
+
bzf->bzs.avail_out = bzf->buflen;
|
437
|
+
}
|
438
|
+
return Qnil;
|
439
|
+
}
|
440
|
+
|
441
|
+
VALUE bz_reader_gets(VALUE obj) {
|
442
|
+
struct bz_file *bzf;
|
443
|
+
VALUE str = Qnil;
|
444
|
+
|
445
|
+
bzf = bz_get_bzf(obj);
|
446
|
+
if (bzf) {
|
447
|
+
str = bz_read_until(bzf, "\n", 1, 0);
|
448
|
+
if (!NIL_P(str)) {
|
449
|
+
bzf->lineno++;
|
450
|
+
OBJ_TAINT(str);
|
451
|
+
}
|
452
|
+
}
|
453
|
+
return str;
|
454
|
+
}
|
455
|
+
|
456
|
+
VALUE bz_reader_gets_internal(int argc, VALUE *argv, VALUE obj, int *td, int init) {
|
457
|
+
struct bz_file *bzf;
|
458
|
+
VALUE rs, res;
|
459
|
+
const char *rsptr;
|
460
|
+
int rslen, rspara, *td1;
|
461
|
+
|
462
|
+
rs = rb_rs;
|
463
|
+
if (argc) {
|
464
|
+
rb_scan_args(argc, argv, "1", &rs);
|
465
|
+
if (!NIL_P(rs)) {
|
466
|
+
Check_Type(rs, T_STRING);
|
467
|
+
}
|
468
|
+
}
|
469
|
+
if (NIL_P(rs)) {
|
470
|
+
return bz_reader_read(1, &rs, obj);
|
471
|
+
}
|
472
|
+
rslen = (int) RSTRING_LEN(rs);
|
473
|
+
if (rs == rb_default_rs || (rslen == 1 && RSTRING_PTR(rs)[0] == '\n')) {
|
474
|
+
return bz_reader_gets(obj);
|
475
|
+
}
|
476
|
+
|
477
|
+
if (rslen == 0) {
|
478
|
+
rsptr = "\n\n";
|
479
|
+
rslen = 2;
|
480
|
+
rspara = 1;
|
481
|
+
} else {
|
482
|
+
rsptr = RSTRING_PTR(rs);
|
483
|
+
rspara = 0;
|
484
|
+
}
|
485
|
+
|
486
|
+
bzf = bz_get_bzf(obj);
|
487
|
+
if (!bzf) {
|
488
|
+
return Qnil;
|
489
|
+
}
|
490
|
+
if (rspara) {
|
491
|
+
bz_read_while(bzf, '\n');
|
492
|
+
}
|
493
|
+
td1 = 0;
|
494
|
+
if (rslen != 1) {
|
495
|
+
if (init) {
|
496
|
+
int i;
|
497
|
+
|
498
|
+
for (i = 0; i < ASIZE; i++) {
|
499
|
+
td[i] = rslen + 1;
|
500
|
+
}
|
501
|
+
for (i = 0; i < rslen; i++) {
|
502
|
+
td[(int)*(rsptr + i)] = rslen - i;
|
503
|
+
}
|
504
|
+
}
|
505
|
+
td1 = td;
|
506
|
+
}
|
507
|
+
|
508
|
+
res = bz_read_until(bzf, rsptr, rslen, td1);
|
509
|
+
if (rspara) {
|
510
|
+
bz_read_while(bzf, '\n');
|
511
|
+
}
|
512
|
+
|
513
|
+
if (!NIL_P(res)) {
|
514
|
+
bzf->lineno++;
|
515
|
+
OBJ_TAINT(res);
|
516
|
+
}
|
517
|
+
return res;
|
518
|
+
}
|
519
|
+
|
520
|
+
/*
|
521
|
+
* Specs were missing for this method originally and playing around with it
|
522
|
+
* gave some very odd results, so unless you know what you're doing, I wouldn't
|
523
|
+
* mess around with this...
|
524
|
+
*/
|
525
|
+
VALUE bz_reader_set_unused(VALUE obj, VALUE a) {
|
526
|
+
struct bz_file *bzf;
|
527
|
+
|
528
|
+
Check_Type(a, T_STRING);
|
529
|
+
Get_BZ2(obj, bzf);
|
530
|
+
if (!bzf->in) {
|
531
|
+
bzf->in = rb_str_new(RSTRING_PTR(a), RSTRING_LEN(a));
|
532
|
+
} else {
|
533
|
+
bzf->in = rb_str_cat(bzf->in, RSTRING_PTR(a), RSTRING_LEN(a));
|
534
|
+
}
|
535
|
+
bzf->bzs.next_in = RSTRING_PTR(bzf->in);
|
536
|
+
bzf->bzs.avail_in = (int) RSTRING_LEN(bzf->in);
|
537
|
+
return Qnil;
|
538
|
+
}
|
539
|
+
|
540
|
+
/*
|
541
|
+
* Reads one character from the stream, returning the byte read.
|
542
|
+
*
|
543
|
+
* reader = Bzip2::Reader.new Bzip2.compress('ab')
|
544
|
+
* reader.getc # => 97
|
545
|
+
* reader.getc # => 98
|
546
|
+
* reader.getc # => nil
|
547
|
+
*
|
548
|
+
* @return [Integer, nil] the byte value of the character read or +nil+ if eoz
|
549
|
+
* has been reached
|
550
|
+
*/
|
551
|
+
VALUE bz_reader_getc(VALUE obj) {
|
552
|
+
VALUE str;
|
553
|
+
VALUE len = INT2FIX(1);
|
554
|
+
|
555
|
+
str = bz_reader_read(1, &len, obj);
|
556
|
+
if (NIL_P(str) || RSTRING_LEN(str) == 0) {
|
557
|
+
return Qnil;
|
558
|
+
}
|
559
|
+
return INT2FIX(RSTRING_PTR(str)[0] & 0xff);
|
560
|
+
}
|
561
|
+
|
562
|
+
void bz_eoz_error() {
|
563
|
+
rb_raise(bz_eEOZError, "End of Zip component reached");
|
564
|
+
}
|
565
|
+
|
566
|
+
/*
|
567
|
+
* Performs the same as Bzip2::Reader#getc except Bzip2::EOZError is raised if
|
568
|
+
* eoz has been readhed
|
569
|
+
*
|
570
|
+
* @raise [Bzip2::EOZError] if eoz has been reached
|
571
|
+
*/
|
572
|
+
VALUE bz_reader_readchar(VALUE obj) {
|
573
|
+
VALUE res = bz_reader_getc(obj);
|
574
|
+
|
575
|
+
if (NIL_P(res)) {
|
576
|
+
bz_eoz_error();
|
577
|
+
}
|
578
|
+
return res;
|
579
|
+
}
|
580
|
+
|
581
|
+
/*
|
582
|
+
* call-seq:
|
583
|
+
* gets(sep = "\n")
|
584
|
+
*
|
585
|
+
* Reads a line from the stream until the separator is reached. This does not
|
586
|
+
* throw an exception, but rather returns nil if an eoz/eof error occurs
|
587
|
+
*
|
588
|
+
* reader = Bzip2::Reader.new Bzip2.compress("a\nb")
|
589
|
+
* reader.gets # => "a\n"
|
590
|
+
* reader.gets # => "b"
|
591
|
+
* reader.gets # => nil
|
592
|
+
*
|
593
|
+
* @return [String, nil] the read data or nil if eoz has been reached
|
594
|
+
* @see Bzip2::Reader#readline
|
595
|
+
*/
|
596
|
+
VALUE bz_reader_gets_m(int argc, VALUE *argv, VALUE obj) {
|
597
|
+
int td[ASIZE];
|
598
|
+
VALUE str = bz_reader_gets_internal(argc, argv, obj, td, Qtrue);
|
599
|
+
|
600
|
+
if (!NIL_P(str)) {
|
601
|
+
rb_lastline_set(str);
|
602
|
+
}
|
603
|
+
return str;
|
604
|
+
}
|
605
|
+
|
606
|
+
/*
|
607
|
+
* call-seq:
|
608
|
+
* readline(sep = "\n")
|
609
|
+
*
|
610
|
+
* Reads one line from the stream and returns it (including the separator)
|
611
|
+
*
|
612
|
+
* reader = Bzip2::Reader.new Bzip2.compress("a\nb")
|
613
|
+
* reader.readline # => "a\n"
|
614
|
+
* reader.readline # => "b"
|
615
|
+
* reader.readline # => raises Bzip2::EOZError
|
616
|
+
*
|
617
|
+
*
|
618
|
+
* @param [String] sep the newline separator character
|
619
|
+
* @return [String] the read line
|
620
|
+
* @see Bzip2::Reader.readlines
|
621
|
+
* @raise [Bzip2::EOZError] if the stream has reached its end
|
622
|
+
*/
|
623
|
+
VALUE bz_reader_readline(int argc, VALUE *argv, VALUE obj) {
|
624
|
+
VALUE res = bz_reader_gets_m(argc, argv, obj);
|
625
|
+
|
626
|
+
if (NIL_P(res)) {
|
627
|
+
bz_eoz_error();
|
628
|
+
}
|
629
|
+
return res;
|
630
|
+
}
|
631
|
+
|
632
|
+
/*
|
633
|
+
* call-seq:
|
634
|
+
* readlines(sep = "\n")
|
635
|
+
*
|
636
|
+
* Reads the lines of the files and returns the result as an array.
|
637
|
+
*
|
638
|
+
* If the stream has reached eoz, then an empty array is returned
|
639
|
+
*
|
640
|
+
* @param [String] sep the newline separator character
|
641
|
+
* @return [Array] an array of lines read
|
642
|
+
* @see Bzip2::Reader.readlines
|
643
|
+
*/
|
644
|
+
VALUE bz_reader_readlines(int argc, VALUE *argv, VALUE obj) {
|
645
|
+
VALUE line, ary;
|
646
|
+
int td[ASIZE], in;
|
647
|
+
|
648
|
+
in = Qtrue;
|
649
|
+
ary = rb_ary_new();
|
650
|
+
while (!NIL_P(line = bz_reader_gets_internal(argc, argv, obj, td, in))) {
|
651
|
+
in = Qfalse;
|
652
|
+
rb_ary_push(ary, line);
|
653
|
+
}
|
654
|
+
return ary;
|
655
|
+
}
|
656
|
+
|
657
|
+
/*
|
658
|
+
* call-seq:
|
659
|
+
* each(sep = "\n", &block)
|
660
|
+
*
|
661
|
+
* Iterates over the lines of the stream.
|
662
|
+
*
|
663
|
+
* @param [String] sep the byte which separates lines
|
664
|
+
* @yieldparam [String] line the next line of the file (including the separator
|
665
|
+
* character)
|
666
|
+
* @see Bzip2::Reader.foreach
|
667
|
+
*/
|
668
|
+
VALUE bz_reader_each_line(int argc, VALUE *argv, VALUE obj) {
|
669
|
+
VALUE line;
|
670
|
+
int td[ASIZE], in;
|
671
|
+
|
672
|
+
in = Qtrue;
|
673
|
+
while (!NIL_P(line = bz_reader_gets_internal(argc, argv, obj, td, in))) {
|
674
|
+
in = Qfalse;
|
675
|
+
rb_yield(line);
|
676
|
+
}
|
677
|
+
return obj;
|
678
|
+
}
|
679
|
+
|
680
|
+
/*
|
681
|
+
* call-seq:
|
682
|
+
* each_byte(&block)
|
683
|
+
*
|
684
|
+
* Iterates over the decompressed bytes of the file.
|
685
|
+
*
|
686
|
+
* Bzip2::Writer.open('file'){ |f| f << 'asdf' }
|
687
|
+
* reader = Bzip2::Reader.new File.open('file')
|
688
|
+
* reader.each_byte{ |b| puts "#{b} #{b.chr}" }
|
689
|
+
*
|
690
|
+
* # Output:
|
691
|
+
* # 97 a
|
692
|
+
* # 115 s
|
693
|
+
* # 100 d
|
694
|
+
* # 102 f
|
695
|
+
*
|
696
|
+
* @yieldparam [Integer] byte the decompressed bytes of the file
|
697
|
+
*/
|
698
|
+
VALUE bz_reader_each_byte(VALUE obj) {
|
699
|
+
int c;
|
700
|
+
|
701
|
+
while ((c = bz_getc(obj)) != EOF) {
|
702
|
+
rb_yield(INT2FIX(c & 0xff));
|
703
|
+
}
|
704
|
+
return obj;
|
705
|
+
}
|
706
|
+
|
707
|
+
/*
|
708
|
+
* Specs were missing for this method originally and playing around with it
|
709
|
+
* gave some very odd results, so unless you know what you're doing, I wouldn't
|
710
|
+
* mess around with this...
|
711
|
+
*/
|
712
|
+
VALUE bz_reader_unused(VALUE obj) {
|
713
|
+
struct bz_file *bzf;
|
714
|
+
VALUE res;
|
715
|
+
|
716
|
+
Get_BZ2(obj, bzf);
|
717
|
+
if (!bzf->in || bzf->state != BZ_STREAM_END) {
|
718
|
+
return Qnil;
|
719
|
+
}
|
720
|
+
if (bzf->bzs.avail_in) {
|
721
|
+
res = rb_tainted_str_new(bzf->bzs.next_in, bzf->bzs.avail_in);
|
722
|
+
bzf->bzs.avail_in = 0;
|
723
|
+
} else {
|
724
|
+
res = rb_tainted_str_new(0, 0);
|
725
|
+
}
|
726
|
+
return res;
|
727
|
+
}
|
728
|
+
|
729
|
+
/*
|
730
|
+
* Test whether the end of the bzip stream has been reached
|
731
|
+
*
|
732
|
+
* @return [Boolean] +true+ if the reader is at the end of the bz stream or
|
733
|
+
* +false+ otherwise
|
734
|
+
*/
|
735
|
+
VALUE bz_reader_eoz(VALUE obj) {
|
736
|
+
struct bz_file *bzf;
|
737
|
+
|
738
|
+
Get_BZ2(obj, bzf);
|
739
|
+
if (!bzf->in || !bzf->buf) {
|
740
|
+
return Qnil;
|
741
|
+
}
|
742
|
+
if (bzf->state == BZ_STREAM_END && !bzf->bzs.avail_out) {
|
743
|
+
return Qtrue;
|
744
|
+
}
|
745
|
+
return Qfalse;
|
746
|
+
}
|
747
|
+
|
748
|
+
/*
|
749
|
+
* Test whether the bzip stream has reached its end (see Bzip2::Reader#eoz?)
|
750
|
+
* and then tests that the undlerying IO has also reached an eof
|
751
|
+
*
|
752
|
+
* @return [Boolean] +true+ if the stream has reached or +false+ otherwise.
|
753
|
+
*/
|
754
|
+
VALUE bz_reader_eof(VALUE obj) {
|
755
|
+
struct bz_file *bzf;
|
756
|
+
VALUE res;
|
757
|
+
|
758
|
+
res = bz_reader_eoz(obj);
|
759
|
+
if (RTEST(res)) {
|
760
|
+
Get_BZ2(obj, bzf);
|
761
|
+
if (bzf->bzs.avail_in) {
|
762
|
+
res = Qfalse;
|
763
|
+
} else {
|
764
|
+
res = bz_reader_getc(obj);
|
765
|
+
if (NIL_P(res)) {
|
766
|
+
res = Qtrue;
|
767
|
+
} else {
|
768
|
+
bz_reader_ungetc(obj, res);
|
769
|
+
res = Qfalse;
|
770
|
+
}
|
771
|
+
}
|
772
|
+
}
|
773
|
+
return res;
|
774
|
+
}
|
775
|
+
|
776
|
+
/*
|
777
|
+
* Tests whether this reader has be closed.
|
778
|
+
*
|
779
|
+
* @return [Boolean] +true+ if it is or +false+ otherwise.
|
780
|
+
*/
|
781
|
+
VALUE bz_reader_closed(VALUE obj) {
|
782
|
+
struct bz_file *bzf;
|
783
|
+
|
784
|
+
Data_Get_Struct(obj, struct bz_file, bzf);
|
785
|
+
return RTEST(bzf->io)?Qfalse:Qtrue;
|
786
|
+
}
|
787
|
+
|
788
|
+
/*
|
789
|
+
* Closes this reader to disallow further reads.
|
790
|
+
*
|
791
|
+
* reader = Bzip2::Reader.new File.open('file')
|
792
|
+
* reader.close
|
793
|
+
*
|
794
|
+
* reader.closed? # => true
|
795
|
+
*
|
796
|
+
* @return [File] the io with which the reader was created.
|
797
|
+
* @raise [IOError] if the stream has already been closed
|
798
|
+
*/
|
799
|
+
VALUE bz_reader_close(VALUE obj) {
|
800
|
+
struct bz_file *bzf;
|
801
|
+
VALUE res;
|
802
|
+
|
803
|
+
Get_BZ2(obj, bzf);
|
804
|
+
if (bzf->buf) {
|
805
|
+
free(bzf->buf);
|
806
|
+
bzf->buf = 0;
|
807
|
+
}
|
808
|
+
if (bzf->state == BZ_OK) {
|
809
|
+
BZ2_bzDecompressEnd(&(bzf->bzs));
|
810
|
+
}
|
811
|
+
if (bzf->flags & BZ2_RB_CLOSE) {
|
812
|
+
int closed = 0;
|
813
|
+
if (rb_respond_to(bzf->io, id_closed)) {
|
814
|
+
VALUE iv = rb_funcall2(bzf->io, id_closed, 0, 0);
|
815
|
+
closed = RTEST(iv);
|
816
|
+
}
|
817
|
+
if (!closed && rb_respond_to(bzf->io, id_close)) {
|
818
|
+
rb_funcall2(bzf->io, id_close, 0, 0);
|
819
|
+
}
|
820
|
+
}
|
821
|
+
if (bzf->flags & (BZ2_RB_CLOSE|BZ2_RB_INTERNAL)) {
|
822
|
+
res = Qnil;
|
823
|
+
} else {
|
824
|
+
res = bzf->io;
|
825
|
+
}
|
826
|
+
bzf->io = 0;
|
827
|
+
return res;
|
828
|
+
}
|
829
|
+
|
830
|
+
/*
|
831
|
+
* Originally undocument and had no sepcs. Appears to call Bzip2::Reader#read
|
832
|
+
* and then mark the stream as finished, but this didn't work for me...
|
833
|
+
*/
|
834
|
+
VALUE bz_reader_finish(VALUE obj) {
|
835
|
+
struct bz_file *bzf;
|
836
|
+
|
837
|
+
Get_BZ2(obj, bzf);
|
838
|
+
if (bzf->buf) {
|
839
|
+
rb_funcall2(obj, id_read, 0, 0);
|
840
|
+
free(bzf->buf);
|
841
|
+
}
|
842
|
+
bzf->buf = 0;
|
843
|
+
bzf->state = BZ_OK;
|
844
|
+
return Qnil;
|
845
|
+
}
|
846
|
+
|
847
|
+
/*
|
848
|
+
* Originally undocument and had no sepcs. Appears to work nearly the same
|
849
|
+
* as Bzip2::Reader#close...
|
850
|
+
*/
|
851
|
+
VALUE bz_reader_close_bang(VALUE obj) {
|
852
|
+
struct bz_file *bzf;
|
853
|
+
int closed;
|
854
|
+
|
855
|
+
Get_BZ2(obj, bzf);
|
856
|
+
closed = bzf->flags & (BZ2_RB_CLOSE|BZ2_RB_INTERNAL);
|
857
|
+
bz_reader_close(obj);
|
858
|
+
if (!closed && rb_respond_to(bzf->io, id_close)) {
|
859
|
+
if (rb_respond_to(bzf->io, id_closed)) {
|
860
|
+
closed = RTEST(rb_funcall2(bzf->io, id_closed, 0, 0));
|
861
|
+
}
|
862
|
+
if (!closed) {
|
863
|
+
rb_funcall2(bzf->io, id_close, 0, 0);
|
864
|
+
}
|
865
|
+
}
|
866
|
+
return Qnil;
|
867
|
+
}
|
868
|
+
|
869
|
+
struct foreach_arg {
|
870
|
+
int argc;
|
871
|
+
VALUE sep;
|
872
|
+
VALUE obj;
|
873
|
+
};
|
874
|
+
|
875
|
+
VALUE bz_reader_foreach_line(struct foreach_arg *arg) {
|
876
|
+
VALUE str;
|
877
|
+
int td[ASIZE], in;
|
878
|
+
|
879
|
+
in = Qtrue;
|
880
|
+
while (!NIL_P(str = bz_reader_gets_internal(arg->argc, &arg->sep, arg->obj, td, in))) {
|
881
|
+
in = Qfalse;
|
882
|
+
rb_yield(str);
|
883
|
+
}
|
884
|
+
return Qnil;
|
885
|
+
}
|
886
|
+
|
887
|
+
/*
|
888
|
+
* call-seq:
|
889
|
+
* foreach(filename, &block)
|
890
|
+
*
|
891
|
+
* Reads a bz2 compressed file and yields each line to the block
|
892
|
+
*
|
893
|
+
* Bzip2::Writer.open('file'){ |f| f << "a\n" << "b\n" << "c\n\nd" }
|
894
|
+
* Bzip2::Reader.foreach('file'){ |l| p l }
|
895
|
+
*
|
896
|
+
* # Output:
|
897
|
+
* # "a\n"
|
898
|
+
* # "b\n"
|
899
|
+
* # "c\n"
|
900
|
+
* # "\n"
|
901
|
+
* # "d"
|
902
|
+
*
|
903
|
+
* @param [String] filename the path to the file to open
|
904
|
+
* @yieldparam [String] each line of the file
|
905
|
+
*/
|
906
|
+
VALUE bz_reader_s_foreach(int argc, VALUE *argv, VALUE obj) {
|
907
|
+
VALUE fname, sep;
|
908
|
+
struct foreach_arg arg;
|
909
|
+
struct bz_file *bzf;
|
910
|
+
|
911
|
+
if (!rb_block_given_p()) {
|
912
|
+
rb_raise(rb_eArgError, "call out of a block");
|
913
|
+
}
|
914
|
+
rb_scan_args(argc, argv, "11", &fname, &sep);
|
915
|
+
#ifdef SafeStringValue
|
916
|
+
SafeStringValue(fname);
|
917
|
+
#else
|
918
|
+
Check_SafeStr(fname);
|
919
|
+
#endif
|
920
|
+
arg.argc = argc - 1;
|
921
|
+
arg.sep = sep;
|
922
|
+
arg.obj = rb_funcall2(rb_mKernel, id_open, 1, &fname);
|
923
|
+
if (NIL_P(arg.obj)) {
|
924
|
+
return Qnil;
|
925
|
+
}
|
926
|
+
arg.obj = rb_funcall2(obj, id_new, 1, &arg.obj);
|
927
|
+
Data_Get_Struct(arg.obj, struct bz_file, bzf);
|
928
|
+
bzf->flags |= BZ2_RB_CLOSE;
|
929
|
+
return rb_ensure(bz_reader_foreach_line, (VALUE)&arg, bz_reader_close, arg.obj);
|
930
|
+
}
|
931
|
+
|
932
|
+
VALUE bz_reader_i_readlines(struct foreach_arg *arg) {
|
933
|
+
VALUE str, res;
|
934
|
+
int td[ASIZE], in;
|
935
|
+
|
936
|
+
in = Qtrue;
|
937
|
+
res = rb_ary_new();
|
938
|
+
while (!NIL_P(str = bz_reader_gets_internal(arg->argc, &arg->sep, arg->obj, td, in))) {
|
939
|
+
in = Qfalse;
|
940
|
+
rb_ary_push(res, str);
|
941
|
+
}
|
942
|
+
return res;
|
943
|
+
}
|
944
|
+
|
945
|
+
/*
|
946
|
+
* call-seq:
|
947
|
+
* readlines(filename, separator="\n")
|
948
|
+
*
|
949
|
+
* Opens the given bz2 compressed file for reading and decompresses the file,
|
950
|
+
* returning an array of the lines of the file. A line is denoted by the
|
951
|
+
* separator argument.
|
952
|
+
*
|
953
|
+
* Bzip2::Writer.open('file'){ |f| f << "a\n" << "b\n" << "c\n\nd" }
|
954
|
+
*
|
955
|
+
* Bzip2::Reader.readlines('file') # => ["a\n", "b\n", "c\n", "\n", "d"]
|
956
|
+
* Bzip2::Reader.readlines('file', 'c') # => ["a\nb\nc", "\n\nd"]
|
957
|
+
*
|
958
|
+
* @param [String] filename the path to the file to read
|
959
|
+
* @param [String] separator the character to denote a newline in the file
|
960
|
+
* @see Bzip2::Reader#readlines
|
961
|
+
* @return [Array] an array of lines for the file
|
962
|
+
* @raise [Bzip2::Error] if the file is not a valid bz2 compressed file
|
963
|
+
*/
|
964
|
+
VALUE bz_reader_s_readlines(int argc, VALUE *argv, VALUE obj) {
|
965
|
+
VALUE fname, sep;
|
966
|
+
struct foreach_arg arg;
|
967
|
+
struct bz_file *bzf;
|
968
|
+
|
969
|
+
rb_scan_args(argc, argv, "11", &fname, &sep);
|
970
|
+
#ifdef SafeStringValue
|
971
|
+
SafeStringValue(fname);
|
972
|
+
#else
|
973
|
+
Check_SafeStr(fname);
|
974
|
+
#endif
|
975
|
+
arg.argc = argc - 1;
|
976
|
+
arg.sep = sep;
|
977
|
+
arg.obj = rb_funcall2(rb_mKernel, id_open, 1, &fname);
|
978
|
+
if (NIL_P(arg.obj)) {
|
979
|
+
return Qnil;
|
980
|
+
}
|
981
|
+
arg.obj = rb_funcall2(obj, id_new, 1, &arg.obj);
|
982
|
+
Data_Get_Struct(arg.obj, struct bz_file, bzf);
|
983
|
+
bzf->flags |= BZ2_RB_CLOSE;
|
984
|
+
return rb_ensure(bz_reader_i_readlines, (VALUE)&arg, bz_reader_close, arg.obj);
|
985
|
+
}
|
986
|
+
|
987
|
+
/*
|
988
|
+
* Returns the current line number that the stream is at. This number is based
|
989
|
+
* on the newline separator being "\n"
|
990
|
+
*
|
991
|
+
* reader = Bzip2::Reader.new Bzip2.compress("a\nb")
|
992
|
+
* reader.lineno # => 0
|
993
|
+
* reader.readline # => "a\n"
|
994
|
+
* reader.lineno # => 1
|
995
|
+
* reader.readline # => "b"
|
996
|
+
* reader.lineno # => 2
|
997
|
+
|
998
|
+
* @return [Integer] the current line number
|
999
|
+
*/
|
1000
|
+
VALUE bz_reader_lineno(VALUE obj) {
|
1001
|
+
struct bz_file *bzf;
|
1002
|
+
|
1003
|
+
Get_BZ2(obj, bzf);
|
1004
|
+
return INT2NUM(bzf->lineno);
|
1005
|
+
}
|
1006
|
+
|
1007
|
+
/*
|
1008
|
+
* call-seq:
|
1009
|
+
* lineno=(num)
|
1010
|
+
*
|
1011
|
+
* Sets the internal line number count that this stream should be set at
|
1012
|
+
*
|
1013
|
+
* reader = Bzip2::Reader.new Bzip2.compress("a\nb")
|
1014
|
+
* reader.lineno # => 0
|
1015
|
+
* reader.readline # => "a\n"
|
1016
|
+
* reader.lineno # => 1
|
1017
|
+
* reader.lineno = 0
|
1018
|
+
* reader.readline # => "b"
|
1019
|
+
* reader.lineno # => 1
|
1020
|
+
*
|
1021
|
+
* @note This does not actually rewind or move the stream forward
|
1022
|
+
* @param [Integer] lineno the line number which the stream should consider
|
1023
|
+
* being set at
|
1024
|
+
* @return [Integer] the line number provided
|
1025
|
+
*/
|
1026
|
+
VALUE bz_reader_set_lineno(VALUE obj, VALUE lineno) {
|
1027
|
+
struct bz_file *bzf;
|
1028
|
+
|
1029
|
+
Get_BZ2(obj, bzf);
|
1030
|
+
bzf->lineno = NUM2INT(lineno);
|
1031
|
+
return lineno;
|
1032
|
+
}
|