Mercurial > public > mercurial-scm > hg

comparison contrib/python-zstandard/c-ext/compressionparams.c @ 30895:c32454d69b85

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
zstd: vendor python-zstandard 0.7.0 Commit 3054ae3a66112970a091d3939fee32c2d0c1a23e from https://github.com/indygreg/python-zstandard is imported without modifications (other than removing unwanted files). The vendored zstd library within has been upgraded from 1.1.2 to 1.1.3. This version introduced new APIs for threads, thread pools, multi-threaded compression, and a new dictionary builder (COVER). These features are not yet used by python-zstandard (or Mercurial for that matter). However, that will likely change in the next python-zstandard release (and I think there are opportunities for Mercurial to take advantage of the multi-threaded APIs). Relevant to Mercurial, the CFFI bindings are now fully implemented. This means zstd should "just work" with PyPy (although I haven't tried). The python-zstandard test suite also runs all tests against both the C extension and CFFI bindings to ensure feature parity. There is also a "decompress_content_dict_chain()" API. This was derived from discussions with Yann Collet on list about alternate ways of encoding delta chains. The change most relevant to Mercurial is a performance enhancement in the simple decompression API to reuse a data structure across operations. This makes decompression of multiple inputs significantly faster. (This scenario occurs when reading revlog delta chains, for example.) Using python-zstandard's bench.py to measure the performance difference... On changelog chunks in the mozilla-unified repo: decompress discrete decompress() reuse zctx 1.262243 wall; 1.260000 CPU; 1.260000 user; 0.000000 sys 170.43 MB/s (best of 3) 0.949106 wall; 0.950000 CPU; 0.950000 user; 0.000000 sys 226.66 MB/s (best of 4) decompress discrete dict decompress() reuse zctx 0.692170 wall; 0.690000 CPU; 0.690000 user; 0.000000 sys 310.80 MB/s (best of 5) 0.437088 wall; 0.440000 CPU; 0.440000 user; 0.000000 sys 492.17 MB/s (best of 7) On manifest chunks in the mozilla-unified repo: decompress discrete decompress() reuse zctx 1.367284 wall; 1.370000 CPU; 1.370000 user; 0.000000 sys 274.01 MB/s (best of 3) 1.086831 wall; 1.080000 CPU; 1.080000 user; 0.000000 sys 344.72 MB/s (best of 3) decompress discrete dict decompress() reuse zctx 0.993272 wall; 0.990000 CPU; 0.990000 user; 0.000000 sys 377.19 MB/s (best of 3) 0.678651 wall; 0.680000 CPU; 0.680000 user; 0.000000 sys 552.06 MB/s (best of 5) That should make reads on zstd revlogs a bit faster ;) # no-check-commit
author Gregory Szorc <gregory.szorc@gmail.com>
date Tue, 07 Feb 2017 23:24:47 -0800
parents d7a05d8351d0
children e0dc40530c5a
comparison
equal deleted inserted replaced
30894:5b60464efbde 30895:c32454d69b85
23 unsigned PY_LONG_LONG sourceSize = 0; 23 unsigned PY_LONG_LONG sourceSize = 0;
24 Py_ssize_t dictSize = 0; 24 Py_ssize_t dictSize = 0;
25 ZSTD_compressionParameters params; 25 ZSTD_compressionParameters params;
26 CompressionParametersObject* result; 26 CompressionParametersObject* result;
27 27
28 if (!PyArg_ParseTuple(args, "i|Kn", &compressionLevel, &sourceSize, &dictSize)) { 28 if (!PyArg_ParseTuple(args, "i|Kn:get_compression_parameters",
29 &compressionLevel, &sourceSize, &dictSize)) {
29 return NULL; 30 return NULL;
30 } 31 }
31 32
32 params = ZSTD_getCParams(compressionLevel, sourceSize, dictSize); 33 params = ZSTD_getCParams(compressionLevel, sourceSize, dictSize);
33 34
45 result->strategy = params.strategy; 46 result->strategy = params.strategy;
46 47
47 return result; 48 return result;
48 } 49 }
49 50
50 PyObject* estimate_compression_context_size(PyObject* self, PyObject* args) { 51 static int CompressionParameters_init(CompressionParametersObject* self, PyObject* args, PyObject* kwargs) {
51 CompressionParametersObject* params; 52 static char* kwlist[] = {
52 ZSTD_compressionParameters zparams; 53 "window_log",
53 PyObject* result; 54 "chain_log",
54 55 "hash_log",
55 if (!PyArg_ParseTuple(args, "O!", &CompressionParametersType, &params)) { 56 "search_log",
56 return NULL; 57 "search_length",
57 } 58 "target_length",
58 59 "strategy",
59 ztopy_compression_parameters(params, &zparams); 60 NULL
60 result = PyLong_FromSize_t(ZSTD_estimateCCtxSize(zparams)); 61 };
61 return result; 62
62 }
63
64 PyDoc_STRVAR(CompressionParameters__doc__,
65 "CompressionParameters: low-level control over zstd compression");
66
67 static PyObject* CompressionParameters_new(PyTypeObject* subtype, PyObject* args, PyObject* kwargs) {
68 CompressionParametersObject* self;
69 unsigned windowLog; 63 unsigned windowLog;
70 unsigned chainLog; 64 unsigned chainLog;
71 unsigned hashLog; 65 unsigned hashLog;
72 unsigned searchLog; 66 unsigned searchLog;
73 unsigned searchLength; 67 unsigned searchLength;
74 unsigned targetLength; 68 unsigned targetLength;
75 unsigned strategy; 69 unsigned strategy;
76 70
77 if (!PyArg_ParseTuple(args, "IIIIIII", &windowLog, &chainLog, &hashLog, &searchLog, 71 if (!PyArg_ParseTupleAndKeywords(args, kwargs, "IIIIIII:CompressionParameters",
78 &searchLength, &targetLength, &strategy)) { 72 kwlist, &windowLog, &chainLog, &hashLog, &searchLog, &searchLength,
79 return NULL; 73 &targetLength, &strategy)) {
74 return -1;
80 } 75 }
81 76
82 if (windowLog < ZSTD_WINDOWLOG_MIN || windowLog > ZSTD_WINDOWLOG_MAX) { 77 if (windowLog < ZSTD_WINDOWLOG_MIN || windowLog > ZSTD_WINDOWLOG_MAX) {
83 PyErr_SetString(PyExc_ValueError, "invalid window log value"); 78 PyErr_SetString(PyExc_ValueError, "invalid window log value");
84 return NULL; 79 return -1;
85 } 80 }
86 81
87 if (chainLog < ZSTD_CHAINLOG_MIN || chainLog > ZSTD_CHAINLOG_MAX) { 82 if (chainLog < ZSTD_CHAINLOG_MIN || chainLog > ZSTD_CHAINLOG_MAX) {
88 PyErr_SetString(PyExc_ValueError, "invalid chain log value"); 83 PyErr_SetString(PyExc_ValueError, "invalid chain log value");
89 return NULL; 84 return -1;
90 } 85 }
91 86
92 if (hashLog < ZSTD_HASHLOG_MIN || hashLog > ZSTD_HASHLOG_MAX) { 87 if (hashLog < ZSTD_HASHLOG_MIN || hashLog > ZSTD_HASHLOG_MAX) {
93 PyErr_SetString(PyExc_ValueError, "invalid hash log value"); 88 PyErr_SetString(PyExc_ValueError, "invalid hash log value");
94 return NULL; 89 return -1;
95 } 90 }
96 91
97 if (searchLog < ZSTD_SEARCHLOG_MIN || searchLog > ZSTD_SEARCHLOG_MAX) { 92 if (searchLog < ZSTD_SEARCHLOG_MIN || searchLog > ZSTD_SEARCHLOG_MAX) {
98 PyErr_SetString(PyExc_ValueError, "invalid search log value"); 93 PyErr_SetString(PyExc_ValueError, "invalid search log value");
99 return NULL; 94 return -1;
100 } 95 }
101 96
102 if (searchLength < ZSTD_SEARCHLENGTH_MIN || searchLength > ZSTD_SEARCHLENGTH_MAX) { 97 if (searchLength < ZSTD_SEARCHLENGTH_MIN || searchLength > ZSTD_SEARCHLENGTH_MAX) {
103 PyErr_SetString(PyExc_ValueError, "invalid search length value"); 98 PyErr_SetString(PyExc_ValueError, "invalid search length value");
104 return NULL; 99 return -1;
105 } 100 }
106 101
107 if (targetLength < ZSTD_TARGETLENGTH_MIN || targetLength > ZSTD_TARGETLENGTH_MAX) { 102 if (targetLength < ZSTD_TARGETLENGTH_MIN || targetLength > ZSTD_TARGETLENGTH_MAX) {
108 PyErr_SetString(PyExc_ValueError, "invalid target length value"); 103 PyErr_SetString(PyExc_ValueError, "invalid target length value");
109 return NULL; 104 return -1;
110 } 105 }
111 106
112 if (strategy < ZSTD_fast || strategy > ZSTD_btopt) { 107 if (strategy < ZSTD_fast || strategy > ZSTD_btopt) {
113 PyErr_SetString(PyExc_ValueError, "invalid strategy value"); 108 PyErr_SetString(PyExc_ValueError, "invalid strategy value");
114 return NULL; 109 return -1;
115 }
116
117 self = (CompressionParametersObject*)subtype->tp_alloc(subtype, 1);
118 if (!self) {
119 return NULL;
120 } 110 }
121 111
122 self->windowLog = windowLog; 112 self->windowLog = windowLog;
123 self->chainLog = chainLog; 113 self->chainLog = chainLog;
124 self->hashLog = hashLog; 114 self->hashLog = hashLog;
125 self->searchLog = searchLog; 115 self->searchLog = searchLog;
126 self->searchLength = searchLength; 116 self->searchLength = searchLength;
127 self->targetLength = targetLength; 117 self->targetLength = targetLength;
128 self->strategy = strategy; 118 self->strategy = strategy;
129 119
130 return (PyObject*)self; 120 return 0;
131 } 121 }
122
123 PyObject* estimate_compression_context_size(PyObject* self, PyObject* args) {
124 CompressionParametersObject* params;
125 ZSTD_compressionParameters zparams;
126 PyObject* result;
127
128 if (!PyArg_ParseTuple(args, "O!:estimate_compression_context_size",
129 &CompressionParametersType, &params)) {
130 return NULL;
131 }
132
133 ztopy_compression_parameters(params, &zparams);
134 result = PyLong_FromSize_t(ZSTD_estimateCCtxSize(zparams));
135 return result;
136 }
137
138 PyDoc_STRVAR(CompressionParameters__doc__,
139 "CompressionParameters: low-level control over zstd compression");
132 140
133 static void CompressionParameters_dealloc(PyObject* self) { 141 static void CompressionParameters_dealloc(PyObject* self) {
134 PyObject_Del(self); 142 PyObject_Del(self);
135 } 143 }
136 144
137 static Py_ssize_t CompressionParameters_length(PyObject* self) { 145 static PyMemberDef CompressionParameters_members[] = {
138 return 7; 146 { "window_log", T_UINT,
139 } 147 offsetof(CompressionParametersObject, windowLog), READONLY,
140 148 "window log" },
141 static PyObject* CompressionParameters_item(PyObject* o, Py_ssize_t i) { 149 { "chain_log", T_UINT,
142 CompressionParametersObject* self = (CompressionParametersObject*)o; 150 offsetof(CompressionParametersObject, chainLog), READONLY,
143 151 "chain log" },
144 switch (i) { 152 { "hash_log", T_UINT,
145 case 0: 153 offsetof(CompressionParametersObject, hashLog), READONLY,
146 return PyLong_FromLong(self->windowLog); 154 "hash log" },
147 case 1: 155 { "search_log", T_UINT,
148 return PyLong_FromLong(self->chainLog); 156 offsetof(CompressionParametersObject, searchLog), READONLY,
149 case 2: 157 "search log" },
150 return PyLong_FromLong(self->hashLog); 158 { "search_length", T_UINT,
151 case 3: 159 offsetof(CompressionParametersObject, searchLength), READONLY,
152 return PyLong_FromLong(self->searchLog); 160 "search length" },
153 case 4: 161 { "target_length", T_UINT,
154 return PyLong_FromLong(self->searchLength); 162 offsetof(CompressionParametersObject, targetLength), READONLY,
155 case 5: 163 "target length" },
156 return PyLong_FromLong(self->targetLength); 164 { "strategy", T_INT,
157 case 6: 165 offsetof(CompressionParametersObject, strategy), READONLY,
158 return PyLong_FromLong(self->strategy); 166 "strategy" },
159 default: 167 { NULL }
160 PyErr_SetString(PyExc_IndexError, "index out of range");
161 return NULL;
162 }
163 }
164
165 static PySequenceMethods CompressionParameters_sq = {
166 CompressionParameters_length, /* sq_length */
167 0, /* sq_concat */
168 0, /* sq_repeat */
169 CompressionParameters_item, /* sq_item */
170 0, /* sq_ass_item */
171 0, /* sq_contains */
172 0, /* sq_inplace_concat */
173 0 /* sq_inplace_repeat */
174 }; 168 };
175 169
176 PyTypeObject CompressionParametersType = { 170 PyTypeObject CompressionParametersType = {
177 PyVarObject_HEAD_INIT(NULL, 0) 171 PyVarObject_HEAD_INIT(NULL, 0)
178 "CompressionParameters", /* tp_name */ 172 "CompressionParameters", /* tp_name */
183 0, /* tp_getattr */ 177 0, /* tp_getattr */
184 0, /* tp_setattr */ 178 0, /* tp_setattr */
185 0, /* tp_compare */ 179 0, /* tp_compare */
186 0, /* tp_repr */ 180 0, /* tp_repr */
187 0, /* tp_as_number */ 181 0, /* tp_as_number */
188 &CompressionParameters_sq, /* tp_as_sequence */ 182 0, /* tp_as_sequence */
189 0, /* tp_as_mapping */ 183 0, /* tp_as_mapping */
190 0, /* tp_hash */ 184 0, /* tp_hash */
191 0, /* tp_call */ 185 0, /* tp_call */
192 0, /* tp_str */ 186 0, /* tp_str */
193 0, /* tp_getattro */ 187 0, /* tp_getattro */
194 0, /* tp_setattro */ 188 0, /* tp_setattro */
195 0, /* tp_as_buffer */ 189 0, /* tp_as_buffer */
196 Py_TPFLAGS_DEFAULT, /* tp_flags */ 190 Py_TPFLAGS_DEFAULT | Py_TPFLAGS_BASETYPE, /* tp_flags */
197 CompressionParameters__doc__, /* tp_doc */ 191 CompressionParameters__doc__, /* tp_doc */
198 0, /* tp_traverse */ 192 0, /* tp_traverse */
199 0, /* tp_clear */ 193 0, /* tp_clear */
200 0, /* tp_richcompare */ 194 0, /* tp_richcompare */
201 0, /* tp_weaklistoffset */ 195 0, /* tp_weaklistoffset */
202 0, /* tp_iter */ 196 0, /* tp_iter */
203 0, /* tp_iternext */ 197 0, /* tp_iternext */
204 0, /* tp_methods */ 198 0, /* tp_methods */
205 0, /* tp_members */ 199 CompressionParameters_members, /* tp_members */
206 0, /* tp_getset */ 200 0, /* tp_getset */
207 0, /* tp_base */ 201 0, /* tp_base */
208 0, /* tp_dict */ 202 0, /* tp_dict */
209 0, /* tp_descr_get */ 203 0, /* tp_descr_get */
210 0, /* tp_descr_set */ 204 0, /* tp_descr_set */
211 0, /* tp_dictoffset */ 205 0, /* tp_dictoffset */
212 0, /* tp_init */ 206 (initproc)CompressionParameters_init, /* tp_init */
213 0, /* tp_alloc */ 207 0, /* tp_alloc */
214 CompressionParameters_new, /* tp_new */ 208 PyType_GenericNew, /* tp_new */
215 }; 209 };
216 210
217 void compressionparams_module_init(PyObject* mod) { 211 void compressionparams_module_init(PyObject* mod) {
218 Py_TYPE(&CompressionParametersType) = &PyType_Type; 212 Py_TYPE(&CompressionParametersType) = &PyType_Type;
219 if (PyType_Ready(&CompressionParametersType) < 0) { 213 if (PyType_Ready(&CompressionParametersType) < 0) {