/usr/lib/python3/dist-packages/pydap/responses/dods.py is in python3-pydap 3.2.2+ds1-1ubuntu1.
This file is owned by root:root, with mode 0o644.
The actual contents of the file can be viewed below.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 | """The DODS response.
This is the DAP response that carries data. The response comes with a DDS
header describing the structure of the data, followed by the data encoded as
XDR.
Even though Python has a library for XDR encoding/decoding, the DODS response
uses Numpy directly since it's faster.
"""
import copy
import numpy as np
from six import string_types
from ..model import (BaseType,
SequenceType, StructureType)
from ..lib import (walk, START_OF_SEQUENCE, END_OF_SEQUENCE, __version__,
NUMPY_TO_DAP2_TYPEMAP,
DAP2_TO_NUMPY_RESPONSE_TYPEMAP,
DAP2_ARRAY_LENGTH_NUMPY_TYPE)
from .lib import BaseResponse
from .dds import dds
try:
from functools import singledispatch
except ImportError:
from singledispatch import singledispatch
def DAP2_response_dtypemap(dtype):
"""
This function takes a numpy dtype object
and returns a dtype object that is compatible with
the DAP2 specification.
"""
dtype_str = DAP2_TO_NUMPY_RESPONSE_TYPEMAP[
NUMPY_TO_DAP2_TYPEMAP[
dtype.char]]
return np.dtype(dtype_str)
def tostring_with_byteorder(x, dtype):
return (x
.astype(dtype.str)
.newbyteorder(dtype.byteorder)
.tostring())
class DODSResponse(BaseResponse):
"""The DODS response."""
__version__ = __version__
def __init__(self, dataset):
BaseResponse.__init__(self, dataset)
self.headers.extend([('Content-description', 'dods_data'),
('Content-type', 'application/octet-stream')])
length = calculate_size(dataset)
if length is not None:
self.headers.append(('Content-length', str(length)))
def __iter__(self):
# generate DDS
for line in dds(self.dataset):
yield line.encode('ascii')
yield b'Data:\n'
for block in dods(self.dataset):
yield block
@singledispatch
def dods(var):
"""Single dispatcher for generating the DODS response."""
raise StopIteration
@dods.register(StructureType)
def _structuretype(var):
for child in var.children():
for block in dods(child):
yield block
@dods.register(SequenceType)
def _sequencetype(var):
# a flat array can be processed one record (or more?) at a time
if all(isinstance(child, BaseType) for child in var.children()):
DAP2_types = []
position = 0
for child in var.children():
if DAP2_response_dtypemap(child.dtype).char == 'S':
(DAP2_types
.append(DAP2_ARRAY_LENGTH_NUMPY_TYPE)) # string length
DAP2_types.append('|S{%s}' % position) # string padded to 4n
position += 1
else:
# Convert any numpy dtypes to numpy dtypes compatible
# with the DAP2 standard:
DAP2_types.append(DAP2_response_dtypemap(child.dtype).str)
DAP2_dtype_str = ','.join(DAP2_types)
strings = position > 0
# array initializations is costy, so we keep a cache here; this will
# be inneficient if there are many strings of different length only
cache = {}
for record in var.iterdata():
yield START_OF_SEQUENCE
if strings:
out = []
padded = []
for value in record:
if isinstance(value, string_types):
length = len(value) or 1
out.append(length)
padded.append(length + (-length % 4))
out.append(value)
record = out
DAP2_dtype_str = ','.join(DAP2_types).format(*padded)
if DAP2_dtype_str not in cache:
# Remember that DAP2_dtype is a (possibly composite)
# numpy dtype that is compatible with the DAP2
# data model. This means that all dtypes in
# DAP2_dtype are representable in DAP2 -- AND --
# the data in var can all be upconverted
# in a lossless manner to the dtypes in DAP2_dtype.
cache[DAP2_dtype_str] = np.zeros((1,), dtype=DAP2_dtype_str)
# By assigning record to ``cache`` the upconversion
# occurs naturally:
cache[DAP2_dtype_str][:] = tuple(record)
# byteorder was taken care of during the upconversion:
yield cache[DAP2_dtype_str].tostring()
yield END_OF_SEQUENCE
# nested array, need to process individually
else:
# create a template structure
struct = StructureType(var.name)
for name in var.keys():
struct[name] = copy.copy(var[name])
for record in var.iterdata():
yield START_OF_SEQUENCE
struct.data = record
for block in dods(struct):
yield block
yield END_OF_SEQUENCE
@dods.register(BaseType)
def _basetype(var):
data = var.data
if not hasattr(data, "shape"):
data = np.asarray(data)
# Convert any numpy dtypes to numpy dtypes compatible
# with the DAP2 standard:
DAP2_dtype = DAP2_response_dtypemap(data.dtype)
if data.shape:
# pack length for arrays
length = np.prod(data.shape).astype(np.int)
# send length twice at the begining of an array...
factor = 2
if DAP2_dtype.char == 'S':
# ... expcept for strings:
factor = 1
yield tostring_with_byteorder(
length,
np.dtype(DAP2_ARRAY_LENGTH_NUMPY_TYPE)) * factor
# make data iterable; 1D arrays must be converted to 2D, since
# iteration over 1D yields scalars which are not properly cast to big
# endian
# This line was removed because endianness is now treated explicitly
# in tostring_with_byteorder()
# if len(data.shape) < 2:
# data = data.reshape(1, -1)
# Only ensure that 0d arrays are iterable:
if len(data.shape) == 0:
data = data[np.newaxis]
# unsigned bytes are padded up to 4n
if DAP2_dtype == np.ubyte:
length = np.prod(data.shape).astype(np.int)
for block in data:
yield tostring_with_byteorder(block, DAP2_dtype)
yield (-length % 4) * b'\0'
# regular data
else:
# strings are also zero padded and preceeded by their length
if DAP2_dtype.char == 'S':
for block in data:
for word in block.flat:
length = len(word)
yield tostring_with_byteorder(
np.array(length),
np.dtype(DAP2_ARRAY_LENGTH_NUMPY_TYPE))
# byteorder is not important for strings:
if hasattr(word, 'encode'):
yield word.encode('ascii')
elif hasattr(word, 'tostring'):
yield word.tostring()
else:
raise TypeError("Could not convert word '{0}' to bytes"
.format(word))
yield (-length % 4) * b'\0'
else:
for block in data:
# Remember that DAP2_dtype is a
# numpy dtype that is compatible with the DAP2
# data model. This means that the dtype in
# DAP2_dtype is representable in DAP2 -- AND --
# the data in var can all be upconverted
# in a lossless manner to the dtype in DAP2_dtype.
yield tostring_with_byteorder(block, DAP2_dtype)
def calculate_size(dataset):
"""Calculate the size of the response. Returns the size in bytes."""
length = 0
for var in walk(dataset):
# Pydap can't calculate the size of sequences since the data is
# streamed directly from the source. Also, strings are encoded
# individually, so it's not possible to get their size unless we read
# everything.
if (isinstance(var, SequenceType) or
(isinstance(var, BaseType) and
DAP2_response_dtypemap(var.dtype).char == 'S')):
return None
elif isinstance(var, BaseType):
if var.shape:
length += 8 # account for array size marker
size = int(np.prod(var.shape))
# Convert any numpy dtype to numpy dtype compatible
# with the DAP2 standard:
DAP2_dtype = DAP2_response_dtypemap(var.data.dtype)
if DAP2_dtype == np.ubyte:
length += size + (-size % 4)
else:
# Remember that numpy dtypes are upconverted to
# DAP2_dtype when sent in the response.
# Their length must thus be modified accordingly:
length += size * DAP2_dtype.itemsize
# account for DDS
length += len(''.join(dds(dataset))) + len(b'Data:\n')
return length
|