zstd-ruby 1.5.6.1 → 1.5.6.3
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +4 -4
- data/README.md +40 -9
- data/ext/zstdruby/common.h +110 -1
- data/ext/zstdruby/extconf.rb +1 -1
- data/ext/zstdruby/main.c +2 -1
- data/ext/zstdruby/skippable_frame.c +1 -1
- data/ext/zstdruby/streaming_compress.c +52 -20
- data/ext/zstdruby/streaming_decompress.c +18 -37
- data/ext/zstdruby/zstdruby.c +59 -48
- data/lib/zstd-ruby/stream_reader.rb +22 -0
- data/lib/zstd-ruby/stream_writer.rb +23 -0
- data/lib/zstd-ruby/version.rb +1 -1
- data/zstd-ruby.gemspec +2 -1
- metadata +18 -3
- data/ext/zstdruby/streaming_compress.h +0 -5
checksums.yaml
CHANGED
@@ -1,7 +1,7 @@
|
|
1
1
|
---
|
2
2
|
SHA256:
|
3
|
-
metadata.gz:
|
4
|
-
data.tar.gz:
|
3
|
+
metadata.gz: 729474c9a3f3196fe69a06dfdd280c5879bc368b1884b8d508b75164aa7e1dcd
|
4
|
+
data.tar.gz: 1cdf1b0554dd89aa1f7b6450be8eb183773a0f3b895024715ae7b3243cec598b
|
5
5
|
SHA512:
|
6
|
-
metadata.gz:
|
7
|
-
data.tar.gz:
|
6
|
+
metadata.gz: d69bc35190b5dbc8ce36f9e5acf6ef26a2c85a899ddd409f8eee4a83b005166b20d22e2659cae1b101fb1b97f6a7bf12cf4066e3da94b2a1991d41bf255cd1ce
|
7
|
+
data.tar.gz: 2ee85e56a8ae268da1e4b69bf35a016a5d9bc677f240435725e082b5c5279e60d7f8be32dbe57283e618b7cf59b3937b57aa859f6cbb0aafa64458a903747fa2
|
data/README.md
CHANGED
@@ -34,20 +34,22 @@ Or install it yourself as:
|
|
34
34
|
require 'zstd-ruby'
|
35
35
|
```
|
36
36
|
|
37
|
-
###
|
37
|
+
### Compression
|
38
|
+
|
39
|
+
#### Simple Compression
|
38
40
|
|
39
41
|
```ruby
|
40
42
|
compressed_data = Zstd.compress(data)
|
41
|
-
compressed_data = Zstd.compress(data, complession_level) # default compression_level is
|
43
|
+
compressed_data = Zstd.compress(data, level: complession_level) # default compression_level is 3
|
42
44
|
```
|
43
45
|
|
44
|
-
|
46
|
+
#### Compression with Dictionary
|
45
47
|
```ruby
|
46
48
|
# dictionary is supposed to have been created using `zstd --train`
|
47
|
-
compressed_using_dict = Zstd.
|
49
|
+
compressed_using_dict = Zstd.compress("", dict: File.read('dictionary_file'))
|
48
50
|
```
|
49
51
|
|
50
|
-
|
52
|
+
#### Streaming Compression
|
51
53
|
```ruby
|
52
54
|
stream = Zstd::StreamingCompress.new
|
53
55
|
stream << "abc" << "def"
|
@@ -66,19 +68,39 @@ res << stream.compress("def")
|
|
66
68
|
res << stream.finish
|
67
69
|
```
|
68
70
|
|
69
|
-
|
71
|
+
#### Streaming Compression with Dictionary
|
72
|
+
```ruby
|
73
|
+
stream = Zstd::StreamingCompress.new(dict: File.read('dictionary_file'))
|
74
|
+
stream << "abc" << "def"
|
75
|
+
res = stream.flush
|
76
|
+
stream << "ghi"
|
77
|
+
res << stream.finish
|
78
|
+
```
|
79
|
+
|
80
|
+
#### Streaming Compression with level and Dictionary
|
81
|
+
```ruby
|
82
|
+
stream = Zstd::StreamingCompress.new(level: 5, dict: File.read('dictionary_file'))
|
83
|
+
stream << "abc" << "def"
|
84
|
+
res = stream.flush
|
85
|
+
stream << "ghi"
|
86
|
+
res << stream.finish
|
87
|
+
```
|
88
|
+
|
89
|
+
### Decompression
|
90
|
+
|
91
|
+
#### Simple Decompression
|
70
92
|
|
71
93
|
```ruby
|
72
94
|
data = Zstd.decompress(compressed_data)
|
73
95
|
```
|
74
96
|
|
75
|
-
|
97
|
+
#### Decompression with Dictionary
|
76
98
|
```ruby
|
77
99
|
# dictionary is supposed to have been created using `zstd --train`
|
78
|
-
Zstd.
|
100
|
+
Zstd.decompress(compressed_using_dict, dict: File.read('dictionary_file'))
|
79
101
|
```
|
80
102
|
|
81
|
-
|
103
|
+
#### Streaming Decompression
|
82
104
|
```ruby
|
83
105
|
cstr = "" # Compressed data
|
84
106
|
stream = Zstd::StreamingDecompress.new
|
@@ -87,6 +109,15 @@ result << stream.decompress(cstr[0, 10])
|
|
87
109
|
result << stream.decompress(cstr[10..-1])
|
88
110
|
```
|
89
111
|
|
112
|
+
#### Streaming Decompression with dictionary
|
113
|
+
```ruby
|
114
|
+
cstr = "" # Compressed data
|
115
|
+
stream = Zstd::StreamingDecompress.new(dict: File.read('dictionary_file'))
|
116
|
+
result = ''
|
117
|
+
result << stream.decompress(cstr[0, 10])
|
118
|
+
result << stream.decompress(cstr[10..-1])
|
119
|
+
```
|
120
|
+
|
90
121
|
### Skippable frame
|
91
122
|
|
92
123
|
```ruby
|
data/ext/zstdruby/common.h
CHANGED
@@ -1,7 +1,11 @@
|
|
1
1
|
#ifndef ZSTD_RUBY_H
|
2
2
|
#define ZSTD_RUBY_H 1
|
3
3
|
|
4
|
-
#include
|
4
|
+
#include <ruby.h>
|
5
|
+
#ifdef HAVE_RUBY_THREAD_H
|
6
|
+
#include <ruby/thread.h>
|
7
|
+
#endif
|
8
|
+
#include <stdbool.h>
|
5
9
|
#include "./libzstd/zstd.h"
|
6
10
|
|
7
11
|
static int convert_compression_level(VALUE compression_level_value)
|
@@ -12,4 +16,109 @@ static int convert_compression_level(VALUE compression_level_value)
|
|
12
16
|
return NUM2INT(compression_level_value);
|
13
17
|
}
|
14
18
|
|
19
|
+
static void set_compress_params(ZSTD_CCtx* const ctx, VALUE level_from_args, VALUE kwargs)
|
20
|
+
{
|
21
|
+
ID kwargs_keys[2];
|
22
|
+
kwargs_keys[0] = rb_intern("level");
|
23
|
+
kwargs_keys[1] = rb_intern("dict");
|
24
|
+
VALUE kwargs_values[2];
|
25
|
+
rb_get_kwargs(kwargs, kwargs_keys, 0, 2, kwargs_values);
|
26
|
+
|
27
|
+
int compression_level = ZSTD_CLEVEL_DEFAULT;
|
28
|
+
if (kwargs_values[0] != Qundef && kwargs_values[0] != Qnil) {
|
29
|
+
compression_level = convert_compression_level(kwargs_values[0]);
|
30
|
+
} else if (!NIL_P(level_from_args)) {
|
31
|
+
rb_warn("`level` in args is deprecated; use keyword args `level:` instead.");
|
32
|
+
compression_level = convert_compression_level(level_from_args);
|
33
|
+
}
|
34
|
+
ZSTD_CCtx_setParameter(ctx, ZSTD_c_compressionLevel, compression_level);
|
35
|
+
|
36
|
+
if (kwargs_values[1] != Qundef && kwargs_values[1] != Qnil) {
|
37
|
+
char* dict_buffer = RSTRING_PTR(kwargs_values[1]);
|
38
|
+
size_t dict_size = RSTRING_LEN(kwargs_values[1]);
|
39
|
+
size_t load_dict_ret = ZSTD_CCtx_loadDictionary(ctx, dict_buffer, dict_size);
|
40
|
+
if (ZSTD_isError(load_dict_ret)) {
|
41
|
+
ZSTD_freeCCtx(ctx);
|
42
|
+
rb_raise(rb_eRuntimeError, "%s", "ZSTD_CCtx_loadDictionary failed");
|
43
|
+
}
|
44
|
+
}
|
45
|
+
}
|
46
|
+
|
47
|
+
struct compress_params {
|
48
|
+
ZSTD_CCtx* ctx;
|
49
|
+
ZSTD_outBuffer* output;
|
50
|
+
ZSTD_inBuffer* input;
|
51
|
+
ZSTD_EndDirective endOp;
|
52
|
+
size_t ret;
|
53
|
+
};
|
54
|
+
|
55
|
+
static void* compress_wrapper(void* args)
|
56
|
+
{
|
57
|
+
struct compress_params* params = args;
|
58
|
+
params->ret = ZSTD_compressStream2(params->ctx, params->output, params->input, params->endOp);
|
59
|
+
return NULL;
|
60
|
+
}
|
61
|
+
|
62
|
+
static size_t zstd_compress(ZSTD_CCtx* const ctx, ZSTD_outBuffer* output, ZSTD_inBuffer* input, ZSTD_EndDirective endOp, bool gvl)
|
63
|
+
{
|
64
|
+
#ifdef HAVE_RUBY_THREAD_H
|
65
|
+
if (gvl) {
|
66
|
+
return ZSTD_compressStream2(ctx, output, input, endOp);
|
67
|
+
} else {
|
68
|
+
struct compress_params params = { ctx, output, input, endOp };
|
69
|
+
rb_thread_call_without_gvl(compress_wrapper, ¶ms, NULL, NULL);
|
70
|
+
return params.ret;
|
71
|
+
}
|
72
|
+
#else
|
73
|
+
return ZSTD_compressStream2(ctx, output, input, endOp);
|
74
|
+
#endif
|
75
|
+
}
|
76
|
+
|
77
|
+
static void set_decompress_params(ZSTD_DCtx* const dctx, VALUE kwargs)
|
78
|
+
{
|
79
|
+
ID kwargs_keys[1];
|
80
|
+
kwargs_keys[0] = rb_intern("dict");
|
81
|
+
VALUE kwargs_values[1];
|
82
|
+
rb_get_kwargs(kwargs, kwargs_keys, 0, 1, kwargs_values);
|
83
|
+
|
84
|
+
if (kwargs_values[0] != Qundef && kwargs_values[0] != Qnil) {
|
85
|
+
char* dict_buffer = RSTRING_PTR(kwargs_values[0]);
|
86
|
+
size_t dict_size = RSTRING_LEN(kwargs_values[0]);
|
87
|
+
size_t load_dict_ret = ZSTD_DCtx_loadDictionary(dctx, dict_buffer, dict_size);
|
88
|
+
if (ZSTD_isError(load_dict_ret)) {
|
89
|
+
ZSTD_freeDCtx(dctx);
|
90
|
+
rb_raise(rb_eRuntimeError, "%s", "ZSTD_CCtx_loadDictionary failed");
|
91
|
+
}
|
92
|
+
}
|
93
|
+
}
|
94
|
+
|
95
|
+
struct decompress_params {
|
96
|
+
ZSTD_DCtx* dctx;
|
97
|
+
ZSTD_outBuffer* output;
|
98
|
+
ZSTD_inBuffer* input;
|
99
|
+
size_t ret;
|
100
|
+
};
|
101
|
+
|
102
|
+
static void* decompress_wrapper(void* args)
|
103
|
+
{
|
104
|
+
struct decompress_params* params = args;
|
105
|
+
params->ret = ZSTD_decompressStream(params->dctx, params->output, params->input);
|
106
|
+
return NULL;
|
107
|
+
}
|
108
|
+
|
109
|
+
static size_t zstd_decompress(ZSTD_DCtx* const dctx, ZSTD_outBuffer* output, ZSTD_inBuffer* input, bool gvl)
|
110
|
+
{
|
111
|
+
#ifdef HAVE_RUBY_THREAD_H
|
112
|
+
if (gvl) {
|
113
|
+
return ZSTD_decompressStream(dctx, output, input);
|
114
|
+
} else {
|
115
|
+
struct decompress_params params = { dctx, output, input };
|
116
|
+
rb_thread_call_without_gvl(decompress_wrapper, ¶ms, NULL, NULL);
|
117
|
+
return params.ret;
|
118
|
+
}
|
119
|
+
#else
|
120
|
+
return ZSTD_decompressStream(dctx, output, input);
|
121
|
+
#endif
|
122
|
+
}
|
123
|
+
|
15
124
|
#endif /* ZSTD_RUBY_H */
|
data/ext/zstdruby/extconf.rb
CHANGED
@@ -2,7 +2,7 @@ require "mkmf"
|
|
2
2
|
|
3
3
|
have_func('rb_gc_mark_movable')
|
4
4
|
|
5
|
-
$CFLAGS = '-I. -O3 -std=c99 -DZSTD_STATIC_LINKING_ONLY'
|
5
|
+
$CFLAGS = '-I. -O3 -std=c99 -DZSTD_STATIC_LINKING_ONLY -DZSTD_MULTITHREAD -pthread -DDEBUGLEVEL=0'
|
6
6
|
$CPPFLAGS += " -fdeclspec" if CONFIG['CXX'] =~ /clang/
|
7
7
|
|
8
8
|
Dir.chdir File.expand_path('..', __FILE__) do
|
data/ext/zstdruby/main.c
CHANGED
@@ -1,5 +1,4 @@
|
|
1
|
-
#include
|
2
|
-
#include <streaming_compress.h>
|
1
|
+
#include "common.h"
|
3
2
|
|
4
3
|
struct streaming_compress_t {
|
5
4
|
ZSTD_CCtx* ctx;
|
@@ -71,9 +70,9 @@ rb_streaming_compress_allocate(VALUE klass)
|
|
71
70
|
static VALUE
|
72
71
|
rb_streaming_compress_initialize(int argc, VALUE *argv, VALUE obj)
|
73
72
|
{
|
73
|
+
VALUE kwargs;
|
74
74
|
VALUE compression_level_value;
|
75
|
-
rb_scan_args(argc, argv, "01", &compression_level_value);
|
76
|
-
int compression_level = convert_compression_level(compression_level_value);
|
75
|
+
rb_scan_args(argc, argv, "01:", &compression_level_value, &kwargs);
|
77
76
|
|
78
77
|
struct streaming_compress_t* sc;
|
79
78
|
TypedData_Get_Struct(obj, struct streaming_compress_t, &streaming_compress_type, sc);
|
@@ -83,7 +82,8 @@ rb_streaming_compress_initialize(int argc, VALUE *argv, VALUE obj)
|
|
83
82
|
if (ctx == NULL) {
|
84
83
|
rb_raise(rb_eRuntimeError, "%s", "ZSTD_createCCtx error");
|
85
84
|
}
|
86
|
-
|
85
|
+
set_compress_params(ctx, compression_level_value, kwargs);
|
86
|
+
|
87
87
|
sc->ctx = ctx;
|
88
88
|
sc->buf = rb_str_new(NULL, buffOutSize);
|
89
89
|
sc->buf_size = buffOutSize;
|
@@ -106,7 +106,7 @@ no_compress(struct streaming_compress_t* sc, ZSTD_EndDirective endOp)
|
|
106
106
|
do {
|
107
107
|
ZSTD_outBuffer output = { (void*)output_data, sc->buf_size, 0 };
|
108
108
|
|
109
|
-
size_t const ret =
|
109
|
+
size_t const ret = zstd_compress(sc->ctx, &output, &input, endOp, false);
|
110
110
|
if (ZSTD_isError(ret)) {
|
111
111
|
rb_raise(rb_eRuntimeError, "flush error error code: %s", ZSTD_getErrorName(ret));
|
112
112
|
}
|
@@ -125,11 +125,12 @@ rb_streaming_compress_compress(VALUE obj, VALUE src)
|
|
125
125
|
|
126
126
|
struct streaming_compress_t* sc;
|
127
127
|
TypedData_Get_Struct(obj, struct streaming_compress_t, &streaming_compress_type, sc);
|
128
|
+
|
128
129
|
const char* output_data = RSTRING_PTR(sc->buf);
|
129
130
|
VALUE result = rb_str_new(0, 0);
|
130
131
|
while (input.pos < input.size) {
|
131
132
|
ZSTD_outBuffer output = { (void*)output_data, sc->buf_size, 0 };
|
132
|
-
size_t const ret =
|
133
|
+
size_t const ret = zstd_compress(sc->ctx, &output, &input, ZSTD_e_continue, false);
|
133
134
|
if (ZSTD_isError(ret)) {
|
134
135
|
rb_raise(rb_eRuntimeError, "compress error error code: %s", ZSTD_getErrorName(ret));
|
135
136
|
}
|
@@ -139,27 +140,54 @@ rb_streaming_compress_compress(VALUE obj, VALUE src)
|
|
139
140
|
}
|
140
141
|
|
141
142
|
static VALUE
|
142
|
-
|
143
|
+
rb_streaming_compress_write(int argc, VALUE *argv, VALUE obj)
|
143
144
|
{
|
144
|
-
|
145
|
-
|
146
|
-
size_t input_size = RSTRING_LEN(src);
|
147
|
-
ZSTD_inBuffer input = { input_data, input_size, 0 };
|
148
|
-
|
145
|
+
size_t total = 0;
|
146
|
+
VALUE result = rb_str_new(0, 0);
|
149
147
|
struct streaming_compress_t* sc;
|
150
148
|
TypedData_Get_Struct(obj, struct streaming_compress_t, &streaming_compress_type, sc);
|
151
149
|
const char* output_data = RSTRING_PTR(sc->buf);
|
152
150
|
|
153
|
-
while (
|
154
|
-
|
155
|
-
|
156
|
-
|
157
|
-
|
151
|
+
while (argc-- > 0) {
|
152
|
+
VALUE str = *argv++;
|
153
|
+
StringValue(str);
|
154
|
+
const char* input_data = RSTRING_PTR(str);
|
155
|
+
size_t input_size = RSTRING_LEN(str);
|
156
|
+
ZSTD_inBuffer input = { input_data, input_size, 0 };
|
157
|
+
|
158
|
+
while (input.pos < input.size) {
|
159
|
+
ZSTD_outBuffer output = { (void*)output_data, sc->buf_size, 0 };
|
160
|
+
size_t const ret = zstd_compress(sc->ctx, &output, &input, ZSTD_e_continue, false);
|
161
|
+
if (ZSTD_isError(ret)) {
|
162
|
+
rb_raise(rb_eRuntimeError, "compress error error code: %s", ZSTD_getErrorName(ret));
|
163
|
+
}
|
164
|
+
total += RSTRING_LEN(str);
|
158
165
|
}
|
159
166
|
}
|
160
|
-
return
|
167
|
+
return SIZET2NUM(total);
|
161
168
|
}
|
162
169
|
|
170
|
+
/*
|
171
|
+
* Document-method: <<
|
172
|
+
* Same as IO.
|
173
|
+
*/
|
174
|
+
#define rb_streaming_compress_addstr rb_io_addstr
|
175
|
+
/*
|
176
|
+
* Document-method: printf
|
177
|
+
* Same as IO.
|
178
|
+
*/
|
179
|
+
#define rb_streaming_compress_printf rb_io_printf
|
180
|
+
/*
|
181
|
+
* Document-method: print
|
182
|
+
* Same as IO.
|
183
|
+
*/
|
184
|
+
#define rb_streaming_compress_print rb_io_print
|
185
|
+
/*
|
186
|
+
* Document-method: puts
|
187
|
+
* Same as IO.
|
188
|
+
*/
|
189
|
+
#define rb_streaming_compress_puts rb_io_puts
|
190
|
+
|
163
191
|
static VALUE
|
164
192
|
rb_streaming_compress_flush(VALUE obj)
|
165
193
|
{
|
@@ -186,7 +214,12 @@ zstd_ruby_streaming_compress_init(void)
|
|
186
214
|
rb_define_alloc_func(cStreamingCompress, rb_streaming_compress_allocate);
|
187
215
|
rb_define_method(cStreamingCompress, "initialize", rb_streaming_compress_initialize, -1);
|
188
216
|
rb_define_method(cStreamingCompress, "compress", rb_streaming_compress_compress, 1);
|
217
|
+
rb_define_method(cStreamingCompress, "write", rb_streaming_compress_write, -1);
|
189
218
|
rb_define_method(cStreamingCompress, "<<", rb_streaming_compress_addstr, 1);
|
219
|
+
rb_define_method(cStreamingCompress, "printf", rb_streaming_compress_printf, -1);
|
220
|
+
rb_define_method(cStreamingCompress, "print", rb_streaming_compress_print, -1);
|
221
|
+
rb_define_method(cStreamingCompress, "puts", rb_streaming_compress_puts, -1);
|
222
|
+
|
190
223
|
rb_define_method(cStreamingCompress, "flush", rb_streaming_compress_flush, 0);
|
191
224
|
rb_define_method(cStreamingCompress, "finish", rb_streaming_compress_finish, 0);
|
192
225
|
|
@@ -194,4 +227,3 @@ zstd_ruby_streaming_compress_init(void)
|
|
194
227
|
rb_define_const(cStreamingCompress, "FLUSH", INT2FIX(ZSTD_e_flush));
|
195
228
|
rb_define_const(cStreamingCompress, "END", INT2FIX(ZSTD_e_end));
|
196
229
|
}
|
197
|
-
|
@@ -1,7 +1,7 @@
|
|
1
|
-
#include
|
1
|
+
#include "common.h"
|
2
2
|
|
3
3
|
struct streaming_decompress_t {
|
4
|
-
ZSTD_DCtx*
|
4
|
+
ZSTD_DCtx* dctx;
|
5
5
|
VALUE buf;
|
6
6
|
size_t buf_size;
|
7
7
|
};
|
@@ -21,9 +21,9 @@ static void
|
|
21
21
|
streaming_decompress_free(void *p)
|
22
22
|
{
|
23
23
|
struct streaming_decompress_t *sd = p;
|
24
|
-
ZSTD_DCtx*
|
25
|
-
if (
|
26
|
-
ZSTD_freeDCtx(
|
24
|
+
ZSTD_DCtx* dctx = sd->dctx;
|
25
|
+
if (dctx != NULL) {
|
26
|
+
ZSTD_freeDCtx(dctx);
|
27
27
|
}
|
28
28
|
xfree(sd);
|
29
29
|
}
|
@@ -61,24 +61,29 @@ rb_streaming_decompress_allocate(VALUE klass)
|
|
61
61
|
{
|
62
62
|
struct streaming_decompress_t* sd;
|
63
63
|
VALUE obj = TypedData_Make_Struct(klass, struct streaming_decompress_t, &streaming_decompress_type, sd);
|
64
|
-
sd->
|
64
|
+
sd->dctx = NULL;
|
65
65
|
sd->buf = Qnil;
|
66
66
|
sd->buf_size = 0;
|
67
67
|
return obj;
|
68
68
|
}
|
69
69
|
|
70
70
|
static VALUE
|
71
|
-
rb_streaming_decompress_initialize(VALUE obj)
|
71
|
+
rb_streaming_decompress_initialize(int argc, VALUE *argv, VALUE obj)
|
72
72
|
{
|
73
|
+
VALUE kwargs;
|
74
|
+
rb_scan_args(argc, argv, "00:", &kwargs);
|
75
|
+
|
73
76
|
struct streaming_decompress_t* sd;
|
74
77
|
TypedData_Get_Struct(obj, struct streaming_decompress_t, &streaming_decompress_type, sd);
|
75
78
|
size_t const buffOutSize = ZSTD_DStreamOutSize();
|
76
79
|
|
77
|
-
ZSTD_DCtx*
|
78
|
-
if (
|
80
|
+
ZSTD_DCtx* dctx = ZSTD_createDCtx();
|
81
|
+
if (dctx == NULL) {
|
79
82
|
rb_raise(rb_eRuntimeError, "%s", "ZSTD_createDCtx error");
|
80
83
|
}
|
81
|
-
|
84
|
+
set_decompress_params(dctx, kwargs);
|
85
|
+
|
86
|
+
sd->dctx = dctx;
|
82
87
|
sd->buf = rb_str_new(NULL, buffOutSize);
|
83
88
|
sd->buf_size = buffOutSize;
|
84
89
|
|
@@ -99,45 +104,21 @@ rb_streaming_decompress_decompress(VALUE obj, VALUE src)
|
|
99
104
|
VALUE result = rb_str_new(0, 0);
|
100
105
|
while (input.pos < input.size) {
|
101
106
|
ZSTD_outBuffer output = { (void*)output_data, sd->buf_size, 0 };
|
102
|
-
size_t const ret =
|
107
|
+
size_t const ret = zstd_decompress(sd->dctx, &output, &input, false);
|
103
108
|
if (ZSTD_isError(ret)) {
|
104
|
-
rb_raise(rb_eRuntimeError, "
|
109
|
+
rb_raise(rb_eRuntimeError, "decompress error error code: %s", ZSTD_getErrorName(ret));
|
105
110
|
}
|
106
111
|
rb_str_cat(result, output.dst, output.pos);
|
107
112
|
}
|
108
113
|
return result;
|
109
114
|
}
|
110
115
|
|
111
|
-
static VALUE
|
112
|
-
rb_streaming_decompress_addstr(VALUE obj, VALUE src)
|
113
|
-
{
|
114
|
-
StringValue(src);
|
115
|
-
const char* input_data = RSTRING_PTR(src);
|
116
|
-
size_t input_size = RSTRING_LEN(src);
|
117
|
-
ZSTD_inBuffer input = { input_data, input_size, 0 };
|
118
|
-
|
119
|
-
struct streaming_decompress_t* sd;
|
120
|
-
TypedData_Get_Struct(obj, struct streaming_decompress_t, &streaming_decompress_type, sd);
|
121
|
-
const char* output_data = RSTRING_PTR(sd->buf);
|
122
|
-
|
123
|
-
while (input.pos < input.size) {
|
124
|
-
ZSTD_outBuffer output = { (void*)output_data, sd->buf_size, 0 };
|
125
|
-
size_t const result = ZSTD_decompressStream(sd->ctx, &output, &input);
|
126
|
-
if (ZSTD_isError(result)) {
|
127
|
-
rb_raise(rb_eRuntimeError, "compress error error code: %s", ZSTD_getErrorName(result));
|
128
|
-
}
|
129
|
-
}
|
130
|
-
return obj;
|
131
|
-
}
|
132
|
-
|
133
116
|
extern VALUE rb_mZstd, cStreamingDecompress;
|
134
117
|
void
|
135
118
|
zstd_ruby_streaming_decompress_init(void)
|
136
119
|
{
|
137
120
|
VALUE cStreamingDecompress = rb_define_class_under(rb_mZstd, "StreamingDecompress", rb_cObject);
|
138
121
|
rb_define_alloc_func(cStreamingDecompress, rb_streaming_decompress_allocate);
|
139
|
-
rb_define_method(cStreamingDecompress, "initialize", rb_streaming_decompress_initialize,
|
122
|
+
rb_define_method(cStreamingDecompress, "initialize", rb_streaming_decompress_initialize, -1);
|
140
123
|
rb_define_method(cStreamingDecompress, "decompress", rb_streaming_decompress_decompress, 1);
|
141
|
-
rb_define_method(cStreamingDecompress, "<<", rb_streaming_decompress_addstr, 1);
|
142
124
|
}
|
143
|
-
|
data/ext/zstdruby/zstdruby.c
CHANGED
@@ -12,28 +12,40 @@ static VALUE rb_compress(int argc, VALUE *argv, VALUE self)
|
|
12
12
|
{
|
13
13
|
VALUE input_value;
|
14
14
|
VALUE compression_level_value;
|
15
|
-
|
16
|
-
|
15
|
+
VALUE kwargs;
|
16
|
+
rb_scan_args(argc, argv, "11:", &input_value, &compression_level_value, &kwargs);
|
17
|
+
|
18
|
+
ZSTD_CCtx* const ctx = ZSTD_createCCtx();
|
19
|
+
if (ctx == NULL) {
|
20
|
+
rb_raise(rb_eRuntimeError, "%s", "ZSTD_createCCtx error");
|
21
|
+
}
|
22
|
+
|
23
|
+
set_compress_params(ctx, compression_level_value, kwargs);
|
17
24
|
|
18
25
|
StringValue(input_value);
|
19
26
|
char* input_data = RSTRING_PTR(input_value);
|
20
27
|
size_t input_size = RSTRING_LEN(input_value);
|
28
|
+
ZSTD_inBuffer input = { input_data, input_size, 0 };
|
29
|
+
// ZSTD_compressBound causes SEGV under multi-thread
|
21
30
|
size_t max_compressed_size = ZSTD_compressBound(input_size);
|
31
|
+
VALUE buf = rb_str_new(NULL, max_compressed_size);
|
32
|
+
char* output_data = RSTRING_PTR(buf);
|
33
|
+
ZSTD_outBuffer output = { (void*)output_data, max_compressed_size, 0 };
|
22
34
|
|
23
|
-
|
24
|
-
|
25
|
-
|
26
|
-
|
27
|
-
if (ZSTD_isError(compressed_size)) {
|
28
|
-
rb_raise(rb_eRuntimeError, "%s: %s", "compress failed", ZSTD_getErrorName(compressed_size));
|
35
|
+
size_t const ret = zstd_compress(ctx, &output, &input, ZSTD_e_end, true);
|
36
|
+
if (ZSTD_isError(ret)) {
|
37
|
+
ZSTD_freeCCtx(ctx);
|
38
|
+
rb_raise(rb_eRuntimeError, "%s: %s", "compress failed", ZSTD_getErrorName(ret));
|
29
39
|
}
|
30
|
-
|
31
|
-
|
32
|
-
|
40
|
+
VALUE result = rb_str_new(0, 0);
|
41
|
+
rb_str_cat(result, output.dst, output.pos);
|
42
|
+
ZSTD_freeCCtx(ctx);
|
43
|
+
return result;
|
33
44
|
}
|
34
45
|
|
35
46
|
static VALUE rb_compress_using_dict(int argc, VALUE *argv, VALUE self)
|
36
47
|
{
|
48
|
+
rb_warn("Zstd.compress_using_dict is deprecated; use Zstd.compress with `dict:` instead.");
|
37
49
|
VALUE input_value;
|
38
50
|
VALUE dict;
|
39
51
|
VALUE compression_level_value;
|
@@ -76,61 +88,56 @@ static VALUE rb_compress_using_dict(int argc, VALUE *argv, VALUE self)
|
|
76
88
|
}
|
77
89
|
|
78
90
|
|
79
|
-
static VALUE decompress_buffered(const char* input_data, size_t input_size)
|
91
|
+
static VALUE decompress_buffered(ZSTD_DCtx* dctx, const char* input_data, size_t input_size)
|
80
92
|
{
|
81
|
-
const size_t outputBufferSize = 4096;
|
82
|
-
|
83
|
-
ZSTD_DStream* const dstream = ZSTD_createDStream();
|
84
|
-
if (dstream == NULL) {
|
85
|
-
rb_raise(rb_eRuntimeError, "%s", "ZSTD_createDStream failed");
|
86
|
-
}
|
87
|
-
|
88
|
-
size_t initResult = ZSTD_initDStream(dstream);
|
89
|
-
if (ZSTD_isError(initResult)) {
|
90
|
-
ZSTD_freeDStream(dstream);
|
91
|
-
rb_raise(rb_eRuntimeError, "%s: %s", "ZSTD_initDStream failed", ZSTD_getErrorName(initResult));
|
92
|
-
}
|
93
|
-
|
94
93
|
VALUE output_string = rb_str_new(NULL, 0);
|
95
94
|
ZSTD_outBuffer output = { NULL, 0, 0 };
|
96
95
|
|
97
96
|
ZSTD_inBuffer input = { input_data, input_size, 0 };
|
98
97
|
while (input.pos < input.size) {
|
99
|
-
output.size +=
|
98
|
+
output.size += ZSTD_DStreamOutSize();
|
100
99
|
rb_str_resize(output_string, output.size);
|
101
100
|
output.dst = RSTRING_PTR(output_string);
|
102
101
|
|
103
|
-
size_t
|
104
|
-
if (ZSTD_isError(
|
105
|
-
|
106
|
-
rb_raise(rb_eRuntimeError, "%s: %s", "ZSTD_decompressStream failed", ZSTD_getErrorName(
|
102
|
+
size_t ret = zstd_decompress(dctx, &output, &input, true);
|
103
|
+
if (ZSTD_isError(ret)) {
|
104
|
+
ZSTD_freeDCtx(dctx);
|
105
|
+
rb_raise(rb_eRuntimeError, "%s: %s", "ZSTD_decompressStream failed", ZSTD_getErrorName(ret));
|
107
106
|
}
|
108
107
|
}
|
109
|
-
|
110
|
-
ZSTD_freeDStream(dstream);
|
111
108
|
rb_str_resize(output_string, output.pos);
|
109
|
+
ZSTD_freeDCtx(dctx);
|
112
110
|
return output_string;
|
113
111
|
}
|
114
112
|
|
115
|
-
static VALUE rb_decompress(VALUE
|
113
|
+
static VALUE rb_decompress(int argc, VALUE *argv, VALUE self)
|
116
114
|
{
|
115
|
+
VALUE input_value;
|
116
|
+
VALUE kwargs;
|
117
|
+
rb_scan_args(argc, argv, "10:", &input_value, &kwargs);
|
117
118
|
StringValue(input_value);
|
118
119
|
char* input_data = RSTRING_PTR(input_value);
|
119
120
|
size_t input_size = RSTRING_LEN(input_value);
|
121
|
+
ZSTD_DCtx* const dctx = ZSTD_createDCtx();
|
122
|
+
if (dctx == NULL) {
|
123
|
+
rb_raise(rb_eRuntimeError, "%s", "ZSTD_createDCtx failed");
|
124
|
+
}
|
125
|
+
set_decompress_params(dctx, kwargs);
|
120
126
|
|
121
127
|
unsigned long long const uncompressed_size = ZSTD_getFrameContentSize(input_data, input_size);
|
122
128
|
if (uncompressed_size == ZSTD_CONTENTSIZE_ERROR) {
|
123
129
|
rb_raise(rb_eRuntimeError, "%s: %s", "not compressed by zstd", ZSTD_getErrorName(uncompressed_size));
|
124
130
|
}
|
131
|
+
// ZSTD_decompressStream may be called multiple times when ZSTD_CONTENTSIZE_UNKNOWN, causing slowness.
|
132
|
+
// Therefore, we will not standardize on ZSTD_decompressStream
|
125
133
|
if (uncompressed_size == ZSTD_CONTENTSIZE_UNKNOWN) {
|
126
|
-
return decompress_buffered(input_data, input_size);
|
134
|
+
return decompress_buffered(dctx, input_data, input_size);
|
127
135
|
}
|
128
136
|
|
129
137
|
VALUE output = rb_str_new(NULL, uncompressed_size);
|
130
138
|
char* output_data = RSTRING_PTR(output);
|
131
|
-
size_t const decompress_size = ZSTD_decompress((void*)output_data, uncompressed_size,
|
132
|
-
(void*)input_data, input_size);
|
133
139
|
|
140
|
+
size_t const decompress_size = ZSTD_decompressDCtx(dctx, output_data, uncompressed_size, input_data, input_size);
|
134
141
|
if (ZSTD_isError(decompress_size)) {
|
135
142
|
rb_raise(rb_eRuntimeError, "%s: %s", "decompress error", ZSTD_getErrorName(decompress_size));
|
136
143
|
}
|
@@ -140,6 +147,7 @@ static VALUE rb_decompress(VALUE self, VALUE input_value)
|
|
140
147
|
|
141
148
|
static VALUE rb_decompress_using_dict(int argc, VALUE *argv, VALUE self)
|
142
149
|
{
|
150
|
+
rb_warn("Zstd.decompress_using_dict is deprecated; use Zstd.decompress with `dict:` instead.");
|
143
151
|
VALUE input_value;
|
144
152
|
VALUE dict;
|
145
153
|
rb_scan_args(argc, argv, "20", &input_value, &dict);
|
@@ -147,15 +155,6 @@ static VALUE rb_decompress_using_dict(int argc, VALUE *argv, VALUE self)
|
|
147
155
|
StringValue(input_value);
|
148
156
|
char* input_data = RSTRING_PTR(input_value);
|
149
157
|
size_t input_size = RSTRING_LEN(input_value);
|
150
|
-
unsigned long long const uncompressed_size = ZSTD_getFrameContentSize(input_data, input_size);
|
151
|
-
if (uncompressed_size == ZSTD_CONTENTSIZE_ERROR) {
|
152
|
-
rb_raise(rb_eRuntimeError, "%s: %s", "not compressed by zstd", ZSTD_getErrorName(uncompressed_size));
|
153
|
-
}
|
154
|
-
if (uncompressed_size == ZSTD_CONTENTSIZE_UNKNOWN) {
|
155
|
-
return decompress_buffered(input_data, input_size);
|
156
|
-
}
|
157
|
-
VALUE output = rb_str_new(NULL, uncompressed_size);
|
158
|
-
char* output_data = RSTRING_PTR(output);
|
159
158
|
|
160
159
|
char* dict_buffer = RSTRING_PTR(dict);
|
161
160
|
size_t dict_size = RSTRING_LEN(dict);
|
@@ -163,12 +162,11 @@ static VALUE rb_decompress_using_dict(int argc, VALUE *argv, VALUE self)
|
|
163
162
|
if (ddict == NULL) {
|
164
163
|
rb_raise(rb_eRuntimeError, "%s", "ZSTD_createDDict failed");
|
165
164
|
}
|
166
|
-
|
167
165
|
unsigned const expected_dict_id = ZSTD_getDictID_fromDDict(ddict);
|
168
166
|
unsigned const actual_dict_id = ZSTD_getDictID_fromFrame(input_data, input_size);
|
169
167
|
if (expected_dict_id != actual_dict_id) {
|
170
168
|
ZSTD_freeDDict(ddict);
|
171
|
-
rb_raise(rb_eRuntimeError, "
|
169
|
+
rb_raise(rb_eRuntimeError, "DictID mismatch");
|
172
170
|
}
|
173
171
|
|
174
172
|
ZSTD_DCtx* const ctx = ZSTD_createDCtx();
|
@@ -176,6 +174,19 @@ static VALUE rb_decompress_using_dict(int argc, VALUE *argv, VALUE self)
|
|
176
174
|
ZSTD_freeDDict(ddict);
|
177
175
|
rb_raise(rb_eRuntimeError, "%s", "ZSTD_createDCtx failed");
|
178
176
|
}
|
177
|
+
|
178
|
+
unsigned long long const uncompressed_size = ZSTD_getFrameContentSize(input_data, input_size);
|
179
|
+
if (uncompressed_size == ZSTD_CONTENTSIZE_ERROR) {
|
180
|
+
ZSTD_freeDDict(ddict);
|
181
|
+
ZSTD_freeDCtx(ctx);
|
182
|
+
rb_raise(rb_eRuntimeError, "%s: %s", "not compressed by zstd", ZSTD_getErrorName(uncompressed_size));
|
183
|
+
}
|
184
|
+
if (uncompressed_size == ZSTD_CONTENTSIZE_UNKNOWN) {
|
185
|
+
return decompress_buffered(ctx, input_data, input_size);
|
186
|
+
}
|
187
|
+
|
188
|
+
VALUE output = rb_str_new(NULL, uncompressed_size);
|
189
|
+
char* output_data = RSTRING_PTR(output);
|
179
190
|
size_t const decompress_size = ZSTD_decompress_usingDDict(ctx, output_data, uncompressed_size, input_data, input_size, ddict);
|
180
191
|
if (ZSTD_isError(decompress_size)) {
|
181
192
|
ZSTD_freeDDict(ddict);
|
@@ -193,6 +204,6 @@ zstd_ruby_init(void)
|
|
193
204
|
rb_define_module_function(rb_mZstd, "zstd_version", zstdVersion, 0);
|
194
205
|
rb_define_module_function(rb_mZstd, "compress", rb_compress, -1);
|
195
206
|
rb_define_module_function(rb_mZstd, "compress_using_dict", rb_compress_using_dict, -1);
|
196
|
-
rb_define_module_function(rb_mZstd, "decompress", rb_decompress, 1);
|
207
|
+
rb_define_module_function(rb_mZstd, "decompress", rb_decompress, -1);
|
197
208
|
rb_define_module_function(rb_mZstd, "decompress_using_dict", rb_decompress_using_dict, -1);
|
198
209
|
}
|
@@ -0,0 +1,22 @@
|
|
1
|
+
module Zstd
|
2
|
+
# @todo Exprimental
|
3
|
+
class StreamReader
|
4
|
+
def initialize(io)
|
5
|
+
@io = io
|
6
|
+
@stream = Zstd::StreamingDecompress.new
|
7
|
+
end
|
8
|
+
|
9
|
+
def read(length)
|
10
|
+
if @io.eof?
|
11
|
+
raise StandardError, "EOF"
|
12
|
+
end
|
13
|
+
data = @io.read(length)
|
14
|
+
@stream.decompress(data)
|
15
|
+
end
|
16
|
+
|
17
|
+
def close
|
18
|
+
@io.write(@stream.finish)
|
19
|
+
@io.close
|
20
|
+
end
|
21
|
+
end
|
22
|
+
end
|
@@ -0,0 +1,23 @@
|
|
1
|
+
module Zstd
|
2
|
+
# @todo Exprimental
|
3
|
+
class StreamWriter
|
4
|
+
def initialize(io, level: nil)
|
5
|
+
@io = io
|
6
|
+
@stream = Zstd::StreamingCompress.new(level)
|
7
|
+
end
|
8
|
+
|
9
|
+
def write(*data)
|
10
|
+
@stream.write(*data)
|
11
|
+
@io.write(@stream.flush)
|
12
|
+
end
|
13
|
+
|
14
|
+
def finish
|
15
|
+
@io.write(@stream.finish)
|
16
|
+
end
|
17
|
+
|
18
|
+
def close
|
19
|
+
@io.write(@stream.finish)
|
20
|
+
@io.close
|
21
|
+
end
|
22
|
+
end
|
23
|
+
end
|
data/lib/zstd-ruby/version.rb
CHANGED
data/zstd-ruby.gemspec
CHANGED
@@ -24,7 +24,7 @@ Gem::Specification.new do |spec|
|
|
24
24
|
#end
|
25
25
|
|
26
26
|
spec.files = `git ls-files -z`.split("\x0").reject do |f|
|
27
|
-
f.match(%r{^(test|spec|features|benchmarks|zstd|.github)/})
|
27
|
+
f.match(%r{^(test|spec|features|benchmarks|zstd|.github|examples)/})
|
28
28
|
end
|
29
29
|
spec.bindir = "exe"
|
30
30
|
spec.executables = spec.files.grep(%r{^exe/}) { |f| File.basename(f) }
|
@@ -35,4 +35,5 @@ Gem::Specification.new do |spec|
|
|
35
35
|
spec.add_development_dependency "rake", "~> 13.0"
|
36
36
|
spec.add_development_dependency "rake-compiler", '~> 1'
|
37
37
|
spec.add_development_dependency "rspec", "~> 3.0"
|
38
|
+
spec.add_development_dependency "pry"
|
38
39
|
end
|
metadata
CHANGED
@@ -1,14 +1,14 @@
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
2
2
|
name: zstd-ruby
|
3
3
|
version: !ruby/object:Gem::Version
|
4
|
-
version: 1.5.6.
|
4
|
+
version: 1.5.6.3
|
5
5
|
platform: ruby
|
6
6
|
authors:
|
7
7
|
- SpringMT
|
8
8
|
autorequire:
|
9
9
|
bindir: exe
|
10
10
|
cert_chain: []
|
11
|
-
date: 2024-04-
|
11
|
+
date: 2024-04-13 00:00:00.000000000 Z
|
12
12
|
dependencies:
|
13
13
|
- !ruby/object:Gem::Dependency
|
14
14
|
name: bundler
|
@@ -66,6 +66,20 @@ dependencies:
|
|
66
66
|
- - "~>"
|
67
67
|
- !ruby/object:Gem::Version
|
68
68
|
version: '3.0'
|
69
|
+
- !ruby/object:Gem::Dependency
|
70
|
+
name: pry
|
71
|
+
requirement: !ruby/object:Gem::Requirement
|
72
|
+
requirements:
|
73
|
+
- - ">="
|
74
|
+
- !ruby/object:Gem::Version
|
75
|
+
version: '0'
|
76
|
+
type: :development
|
77
|
+
prerelease: false
|
78
|
+
version_requirements: !ruby/object:Gem::Requirement
|
79
|
+
requirements:
|
80
|
+
- - ">="
|
81
|
+
- !ruby/object:Gem::Version
|
82
|
+
version: '0'
|
69
83
|
description: Ruby binding for zstd(Zstandard - Fast real-time compression algorithm).
|
70
84
|
See https://github.com/facebook/zstd
|
71
85
|
email:
|
@@ -159,10 +173,11 @@ files:
|
|
159
173
|
- ext/zstdruby/main.c
|
160
174
|
- ext/zstdruby/skippable_frame.c
|
161
175
|
- ext/zstdruby/streaming_compress.c
|
162
|
-
- ext/zstdruby/streaming_compress.h
|
163
176
|
- ext/zstdruby/streaming_decompress.c
|
164
177
|
- ext/zstdruby/zstdruby.c
|
165
178
|
- lib/zstd-ruby.rb
|
179
|
+
- lib/zstd-ruby/stream_reader.rb
|
180
|
+
- lib/zstd-ruby/stream_writer.rb
|
166
181
|
- lib/zstd-ruby/version.rb
|
167
182
|
- renovate.json
|
168
183
|
- zstd-ruby.gemspec
|