Coverage for /pythoncovmergedfiles/medio/medio/usr/local/lib/python3.8/site-packages/bitarray/util.py: 23%
213 statements
« prev ^ index » next coverage.py v7.3.1, created at 2023-09-25 06:15 +0000
1# Copyright (c) 2019 - 2023, Ilan Schnell; All Rights Reserved
2# bitarray is published under the PSF license.
3#
4# Author: Ilan Schnell
5"""
6Useful utilities for working with bitarrays.
7"""
8from __future__ import absolute_import
10import os
11import sys
13from bitarray import bitarray, bits2bytes
15from bitarray._util import (
16 zeros, count_n, rindex, parity,
17 count_and, count_or, count_xor, any_and, subset,
18 _correspond_all,
19 serialize, deserialize,
20 ba2hex, hex2ba,
21 ba2base, base2ba,
22 sc_encode, sc_decode,
23 vl_encode, vl_decode,
24 canonical_decode,
25)
27__all__ = [
28 'zeros', 'urandom', 'pprint', 'make_endian', 'rindex', 'strip', 'count_n',
29 'parity', 'count_and', 'count_or', 'count_xor', 'any_and', 'subset',
30 'intervals',
31 'ba2hex', 'hex2ba',
32 'ba2base', 'base2ba',
33 'ba2int', 'int2ba',
34 'serialize', 'deserialize',
35 'sc_encode', 'sc_decode',
36 'vl_encode', 'vl_decode',
37 'huffman_code', 'canonical_huffman', 'canonical_decode',
38]
41_is_py2 = bool(sys.version_info[0] == 2)
44def urandom(__length, endian=None):
45 """urandom(length, /, endian=None) -> bitarray
47Return a bitarray of `length` random bits (uses `os.urandom`).
48"""
49 a = bitarray(0, endian)
50 a.frombytes(os.urandom(bits2bytes(__length)))
51 del a[__length:]
52 return a
55def pprint(__a, stream=None, group=8, indent=4, width=80):
56 """pprint(bitarray, /, stream=None, group=8, indent=4, width=80)
58Prints the formatted representation of object on `stream` (which defaults
59to `sys.stdout`). By default, elements are grouped in bytes (8 elements),
60and 8 bytes (64 elements) per line.
61Non-bitarray objects are printed by the standard library
62function `pprint.pprint()`.
63"""
64 if stream is None:
65 stream = sys.stdout
67 if not isinstance(__a, bitarray):
68 import pprint as _pprint
69 _pprint.pprint(__a, stream=stream, indent=indent, width=width)
70 return
72 group = int(group)
73 if group < 1:
74 raise ValueError('group must be >= 1')
75 indent = int(indent)
76 if indent < 0:
77 raise ValueError('indent must be >= 0')
78 width = int(width)
79 if width <= indent:
80 raise ValueError('width must be > %d (indent)' % indent)
82 gpl = (width - indent) // (group + 1) # groups per line
83 epl = group * gpl # elements per line
84 if epl == 0:
85 epl = width - indent - 2
86 type_name = type(__a).__name__
87 # here 4 is len("'()'")
88 multiline = len(type_name) + 4 + len(__a) + len(__a) // group >= width
89 if multiline:
90 quotes = "'''"
91 elif __a:
92 quotes = "'"
93 else:
94 quotes = ""
96 stream.write("%s(%s" % (type_name, quotes))
97 for i, b in enumerate(__a):
98 if multiline and i % epl == 0:
99 stream.write('\n%s' % (indent * ' '))
100 if i % group == 0 and i % epl != 0:
101 stream.write(' ')
102 stream.write(str(b))
104 if multiline:
105 stream.write('\n')
107 stream.write("%s)\n" % quotes)
108 stream.flush()
111def make_endian(__a, endian):
112 """make_endian(bitarray, /, endian) -> bitarray
114When the endianness of the given bitarray is different from `endian`,
115return a new bitarray, with endianness `endian` and the same elements
116as the original bitarray.
117Otherwise (endianness is already `endian`) the original bitarray is returned
118unchanged.
119"""
120 if not isinstance(__a, bitarray):
121 raise TypeError("bitarray expected, got '%s'" % type(__a).__name__)
123 if __a.endian() == endian:
124 return __a
126 return bitarray(__a, endian)
129def strip(__a, mode='right'):
130 """strip(bitarray, /, mode='right') -> bitarray
132Return a new bitarray with zeros stripped from left, right or both ends.
133Allowed values for mode are the strings: `left`, `right`, `both`
134"""
135 if not isinstance(mode, str):
136 raise TypeError("str expected for mode, got '%s'" % type(__a).__name__)
137 if mode not in ('left', 'right', 'both'):
138 raise ValueError("mode must be 'left', 'right' or 'both', got %r" %
139 mode)
140 if mode == 'right':
141 start = None
142 else:
143 try:
144 start = __a.index(1)
145 except ValueError:
146 return __a[:0]
148 if mode == 'left':
149 stop = None
150 else:
151 try:
152 stop = rindex(__a) + 1
153 except ValueError:
154 return __a[:0]
156 return __a[start:stop]
159def intervals(__a):
160 """intervals(bitarray, /) -> iterator
162Compute all uninterrupted intervals of 1s and 0s, and return an
163iterator over tuples `(value, start, stop)`. The intervals are guaranteed
164to be in order, and their size is always non-zero (`stop - start > 0`).
165"""
166 try:
167 value = __a[0] # value of current interval
168 except IndexError:
169 return
170 n = len(__a)
171 stop = 0 # "previous" stop - becomes next start
173 while stop < n:
174 start = stop
175 # assert __a[start] == value
176 try: # find next occurrence of opposite value
177 stop = __a.index(not value, start)
178 except ValueError:
179 stop = n
180 yield int(value), start, stop
181 value = not value # next interval has opposite value
184def ba2int(__a, signed=False):
185 """ba2int(bitarray, /, signed=False) -> int
187Convert the given bitarray to an integer.
188The bit-endianness of the bitarray is respected.
189`signed` indicates whether two's complement is used to represent the integer.
190"""
191 if not isinstance(__a, bitarray):
192 raise TypeError("bitarray expected, got '%s'" % type(__a).__name__)
193 length = len(__a)
194 if length == 0:
195 raise ValueError("non-empty bitarray expected")
197 le = bool(__a.endian() == 'little')
198 if __a.padbits:
199 pad = zeros(__a.padbits, __a.endian())
200 __a = __a + pad if le else pad + __a
202 if _is_py2:
203 a = bitarray(__a, 'big')
204 if le:
205 a.reverse()
206 res = int(ba2hex(a), 16)
207 else: # py3
208 res = int.from_bytes(__a.tobytes(), byteorder=__a.endian())
210 if signed and res >= 1 << (length - 1):
211 res -= 1 << length
212 return res
215def int2ba(__i, length=None, endian=None, signed=False):
216 """int2ba(int, /, length=None, endian=None, signed=False) -> bitarray
218Convert the given integer to a bitarray (with given endianness,
219and no leading (big-endian) / trailing (little-endian) zeros), unless
220the `length` of the bitarray is provided. An `OverflowError` is raised
221if the integer is not representable with the given number of bits.
222`signed` determines whether two's complement is used to represent the integer,
223and requires `length` to be provided.
224"""
225 if not isinstance(__i, (int, long) if _is_py2 else int):
226 raise TypeError("int expected, got '%s'" % type(__i).__name__)
227 if length is not None:
228 if not isinstance(length, int):
229 raise TypeError("int expected for length")
230 if length <= 0:
231 raise ValueError("length must be > 0")
232 if signed and length is None:
233 raise TypeError("signed requires length")
235 if __i == 0:
236 # there are special cases for 0 which we'd rather not deal with below
237 return zeros(length or 1, endian)
239 if signed:
240 m = 1 << (length - 1)
241 if not (-m <= __i < m):
242 raise OverflowError("signed integer not in range(%d, %d), "
243 "got %d" % (-m, m, __i))
244 if __i < 0:
245 __i += 1 << length
246 else: # unsigned
247 if __i < 0:
248 raise OverflowError("unsigned integer not positive, got %d" % __i)
249 if length and __i >= (1 << length):
250 raise OverflowError("unsigned integer not in range(0, %d), "
251 "got %d" % (1 << length, __i))
253 a = bitarray(0, endian)
254 le = bool(a.endian() == 'little')
255 if _is_py2:
256 s = hex(__i)[2:].rstrip('L')
257 a.extend(hex2ba(s, 'big'))
258 if le:
259 a.reverse()
260 else: # py3
261 b = __i.to_bytes(bits2bytes(__i.bit_length()), byteorder=a.endian())
262 a.frombytes(b)
264 if length is None:
265 return strip(a, 'right' if le else 'left')
267 la = len(a)
268 if la > length:
269 a = a[:length] if le else a[-length:]
270 if la < length:
271 pad = zeros(length - la, a.endian())
272 a = a + pad if le else pad + a
273 assert len(a) == length
274 return a
276# ------------------------------ Huffman coding -----------------------------
278def _huffman_tree(__freq_map):
279 """_huffman_tree(dict, /) -> Node
281Given a dict mapping symbols to their frequency, construct a Huffman tree
282and return its root node.
283"""
284 from heapq import heappush, heappop
286 class Node(object):
287 """
288 A Node instance will either have a 'symbol' (leaf node) or
289 a 'child' (a tuple with both children) attribute.
290 The 'freq' attribute will always be present.
291 """
292 def __lt__(self, other):
293 # heapq needs to be able to compare the nodes
294 return self.freq < other.freq
296 minheap = []
297 # create all leaf nodes and push them onto the queue
298 for sym, f in __freq_map.items():
299 leaf = Node()
300 leaf.symbol = sym
301 leaf.freq = f
302 heappush(minheap, leaf)
304 # repeat the process until only one node remains
305 while len(minheap) > 1:
306 # take the two nodes with lowest frequencies from the queue
307 # to construct a new node and push it onto the queue
308 parent = Node()
309 parent.child = heappop(minheap), heappop(minheap)
310 parent.freq = parent.child[0].freq + parent.child[1].freq
311 heappush(minheap, parent)
313 # the single remaining node is the root of the Huffman tree
314 return minheap[0]
317def huffman_code(__freq_map, endian=None):
318 """huffman_code(dict, /, endian=None) -> dict
320Given a frequency map, a dictionary mapping symbols to their frequency,
321calculate the Huffman code, i.e. a dict mapping those symbols to
322bitarrays (with given endianness). Note that the symbols are not limited
323to being strings. Symbols may may be any hashable object (such as `None`).
324"""
325 if not isinstance(__freq_map, dict):
326 raise TypeError("dict expected, got '%s'" % type(__freq_map).__name__)
328 b0 = bitarray('0', endian)
329 b1 = bitarray('1', endian)
331 if len(__freq_map) < 2:
332 if len(__freq_map) == 0:
333 raise ValueError("cannot create Huffman code with no symbols")
334 # Only one symbol: Normally if only one symbol is given, the code
335 # could be represented with zero bits. However here, the code should
336 # be at least one bit for the .encode() and .decode() methods to work.
337 # So we represent the symbol by a single code of length one, in
338 # particular one 0 bit. This is an incomplete code, since if a 1 bit
339 # is received, it has no meaning and will result in an error.
340 return {list(__freq_map)[0]: b0}
342 result = {}
344 def traverse(nd, prefix=bitarray(0, endian)):
345 try: # leaf
346 result[nd.symbol] = prefix
347 except AttributeError: # parent, so traverse each of the children
348 traverse(nd.child[0], prefix + b0)
349 traverse(nd.child[1], prefix + b1)
351 traverse(_huffman_tree(__freq_map))
352 return result
355def canonical_huffman(__freq_map):
356 """canonical_huffman(dict, /) -> tuple
358Given a frequency map, a dictionary mapping symbols to their frequency,
359calculate the canonical Huffman code. Returns a tuple containing:
3610. the canonical Huffman code as a dict mapping symbols to bitarrays
3621. a list containing the number of symbols of each code length
3632. a list of symbols in canonical order
365Note: the two lists may be used as input for `canonical_decode()`.
366"""
367 if not isinstance(__freq_map, dict):
368 raise TypeError("dict expected, got '%s'" % type(__freq_map).__name__)
370 if len(__freq_map) < 2:
371 if len(__freq_map) == 0:
372 raise ValueError("cannot create Huffman code with no symbols")
373 # Only one symbol: see note above in huffman_code()
374 sym = list(__freq_map)[0]
375 return {sym: bitarray('0', 'big')}, [0, 1], [sym]
377 code_length = {} # map symbols to their code length
379 def traverse(nd, length=0):
380 # traverse the Huffman tree, but (unlike in huffman_code() above) we
381 # now just simply record the length for reaching each symbol
382 try: # leaf
383 code_length[nd.symbol] = length
384 except AttributeError: # parent, so traverse each of the children
385 traverse(nd.child[0], length + 1)
386 traverse(nd.child[1], length + 1)
388 traverse(_huffman_tree(__freq_map))
390 # we now have a mapping of symbols to their code length,
391 # which is all we need
393 table = sorted(code_length.items(), key=lambda item: (item[1], item[0]))
395 maxbits = max(item[1] for item in table)
396 codedict = {}
397 count = (maxbits + 1) * [0]
399 code = 0
400 for i, (sym, length) in enumerate(table):
401 codedict[sym] = int2ba(code, length, 'big')
402 count[length] += 1
403 if i + 1 < len(table):
404 code += 1
405 code <<= table[i + 1][1] - length
407 return codedict, count, [item[0] for item in table]